JP2008031929A - Air-fuel ratio control device for internal combustion engine and fuel injection amount control device for internal combustion engine - Google Patents

Abstract

An air-fuel ratio control apparatus for an internal combustion engine, which can quickly match an actual air-fuel ratio with a target air-fuel ratio.
In this apparatus, during the transition period Tk following the fuel increase period Trc, a first time-integrated output deviation based on the current output deviation (difference between the output value of the air-fuel ratio sensor downstream of the catalyst and the target value). The time integration value SDVoxs1 is updated and the downstream FB correction amount is updated based on the updated first time integration value, and the second time integration value SDVoxs2 and the downstream FB learning correction amount are maintained constant and downstream. The fuel injection amount is determined based on the side FB correction amount and the downstream FB learning correction amount. After the transition period, the apparatus continues to update the first time integral value and the downstream FB correction amount, and updates the second time integral value based on the current output deviation and updates the second time integral value. Based on this, the downstream FB learning correction amount is updated. Thereby, the downstream FB correction amount and the downstream FB learning correction amount can be appropriately calculated.
[Selection] Figure 5

Description

  The present invention is applied to an internal combustion engine provided with an air-fuel ratio sensor disposed in an exhaust passage, and performs an air-fuel ratio feedback control on an air-fuel ratio of a mixed gas supplied to the internal combustion engine based on an output value of the air-fuel ratio sensor. The present invention relates to a fuel ratio control device. Further, the present invention is applied to the internal combustion engine, and the fuel injection amount of the internal combustion engine that feedback-controls the fuel injection amount that is the amount of fuel injected to be supplied to the internal combustion engine based on the output value of the air-fuel ratio sensor The present invention relates to a control device.

  2. Description of the Related Art Conventionally, an air-fuel ratio control apparatus that performs feedback (FB) control of an air-fuel ratio (air-fuel ratio of an internal combustion engine) of a mixed gas supplied to an internal combustion engine (formed in a combustion chamber of the internal combustion engine) is known. One such air-fuel ratio control device is disposed in the exhaust passage upstream and downstream of an exhaust purification catalyst (catalyst) disposed in the exhaust passage so as to purify harmful substances in the exhaust gas. Air-fuel ratio sensors (upstream air-fuel ratio sensor and downstream air-fuel ratio sensor).

  When the target air-fuel ratio of the internal combustion engine is set to the stoichiometric air-fuel ratio, this control apparatus determines the difference between the actual output value of the downstream air-fuel ratio sensor and the downstream target value (hereinafter simply referred to as “output deviation”). The downstream FB correction amount Vsb is calculated based on This downstream target value is set to a value corresponding to the output of the downstream air-fuel ratio sensor when the gas whose air-fuel ratio is the stoichiometric air-fuel ratio has reached the downstream air-fuel ratio sensor.

  Further, when the target air-fuel ratio of the internal combustion engine is set to the stoichiometric air-fuel ratio, the control device is based on the downstream FB correction amount Vsb, the actual output value of the upstream air-fuel ratio sensor, and the target air-fuel ratio of the internal combustion engine. To determine the upstream FB correction amount. On the other hand, when the target air-fuel ratio of the internal combustion engine is set to an air-fuel ratio other than the stoichiometric air-fuel ratio, the control device detects the actual output value of the upstream air-fuel ratio sensor and its output without using the downstream FB correction amount Vsb. The upstream FB correction amount is determined based on the target air-fuel ratio. Then, the control device determines the fuel injection amount based on the upstream FB correction amount determined as described above.

  This downstream FB correction amount Vsb is an output deviation (in this case, “steady deviation”) in which the FB control of the air-fuel ratio using the downstream FB correction amount Vsb described above has been performed for a sufficiently long time (steady state). In other words, an integral term calculated by multiplying a time integral value of the output deviation by a predetermined gain is included. Therefore, the correspondence relationship between the output value of the upstream air-fuel ratio sensor and the actual air-fuel ratio upstream of the catalyst (upstream air-fuel ratio) due to reasons such as deterioration of the upstream air-fuel ratio sensor has been planned. Even if the output is different from the output (the output deviation of the upstream air-fuel ratio sensor is occurring), according to this control device, the air-fuel ratio downstream of the catalyst (downstream air-fuel ratio) is reduced downstream in the steady state. The air-fuel ratio corresponding to the side target value (downstream target air-fuel ratio) can be matched. In other words, the output deviation of the upstream air-fuel ratio sensor can be compensated.

  Incidentally, as described above, when the target air-fuel ratio of the internal combustion engine is set to an air-fuel ratio other than the stoichiometric air-fuel ratio, the downstream FB correction amount Vsb is not used when determining the fuel injection amount. The output deviation of the air-fuel ratio sensor is not compensated. Therefore, even if the output value of the upstream air-fuel ratio sensor is controlled to match the upstream target value corresponding to the target air-fuel ratio, the actual air-fuel ratio of the internal combustion engine may not match the target air-fuel ratio. There is.

Therefore, the other control device obtains the downstream FB learning correction amount Vsg based on the downstream FB correction amount Vsb calculated in the steady state. Then, when the target air-fuel ratio of the internal combustion engine is set to an air-fuel ratio other than the stoichiometric air-fuel ratio, the control device determines the obtained downstream FB learning correction amount Vsg, the actual output value of the upstream air-fuel ratio sensor, and the internal combustion engine. The upstream FB correction amount is determined based on the target air-fuel ratio of the engine. Thereby, even when the target air-fuel ratio of the internal combustion engine is set to an air-fuel ratio other than the stoichiometric air-fuel ratio, the output deviation of the upstream air-fuel ratio sensor is compensated. As a result, the actual air-fuel ratio of the internal combustion engine can be matched with the target air-fuel ratio (see Patent Document 1).
JP 2005-61356 A

  However, according to the control apparatus, there arises a problem as described below with reference to FIG. 1 which is a time chart.

  First, it is assumed that the control device sets the target air-fuel ratio of the internal combustion engine to the stoichiometric air-fuel ratio in the period A before time t0 shown in FIG. At this time, the control device time-integrates the difference (output deviation) Δ between the actual output value of the downstream air-fuel ratio sensor shown in FIG. 1B and the downstream target value Vref (in FIG. The time integration value is updated as shown by the curve CS shown in F). Then, the control device calculates the integral term of the downstream FB correction amount Vsb based on the updated time integration value, updates the downstream FB correction amount Vsb, and also updates the downstream FB correction amount Vsb (accordingly, the time integration). Value), the downstream FB learning correction amount Vsg shown in the curve CG in FIG. 1F is updated. Further, the control device, based on the calculated downstream FB correction amount Vsb, the calculated downstream FB learning correction amount Vsg, the actual output value V of the upstream air-fuel ratio sensor, and the target air-fuel ratio. The upstream FB correction amount DFi is determined. Then, the control device determines the fuel injection amount Fi based on the basic fuel injection amount Fis corresponding to the stoichiometric air-fuel ratio and the upstream FB correction amount DFi. As a result, the air-fuel ratio of the internal combustion engine is controlled to the stoichiometric air-fuel ratio that is the target air-fuel ratio.

  Next, it is assumed that the fuel cut execution condition is satisfied at time t0. In a fuel cut condition satisfaction period (time t0 to time t1) Tfc, which is a period during which the fuel cut execution condition is satisfied, the control device stops fuel injection (performs fuel cut control). As a result, the amount of oxygen stored by the catalyst (oxygen storage amount) increases. Further, during the fuel cut condition establishment period Tfc, the control device stops updating the time integration value (maintains the time integration value constant), and stops updating the downstream FB learning correction amount Vsg to thereby reduce the downstream FB. The learning correction amount Vsg is kept constant. Naturally, the control device does not perform FB control of the air-fuel ratio.

  When the fuel cut execution condition is not established at time t1 (no longer established), the oxygen storage amount is excessively larger than an appropriate amount (for example, half the maximum value of the oxygen storage amount). Therefore, in order to quickly reduce the oxygen storage amount, the control device is in the fuel increase period (time point t1 to time point t2) Trc, which is a period from when the fuel cut condition satisfaction period Tfc ends until a predetermined time elapses. Then, the target air-fuel ratio of the internal combustion engine is set to an air-fuel ratio (fuel increase air-fuel ratio) richer than the stoichiometric air-fuel ratio (fuel increase control is performed).

  In this fuel increase period Trc, the control device stops updating the time integral value (maintains the time integral value constant) and updates the downstream FB learning correction amount Vsg, as in the fuel cut condition establishment period Tfc. Stop and keep the downstream FB learning correction amount Vsg constant. Then, the control device does not depend on the downstream FB correction amount Vsb, the downstream FB learning correction amount Vsg maintained at a constant value, the actual output value V of the upstream air-fuel ratio sensor, the target air-fuel ratio, Based on the above, the upstream FB correction amount is determined. Then, the control device determines the fuel injection amount Fi based on the basic fuel injection amount Fir corresponding to the fuel increasing air-fuel ratio and the upstream FB correction amount DFi. Thereby, the air fuel ratio of the internal combustion engine is controlled to the fuel increasing air fuel ratio which is the target air fuel ratio.

  Next, in period B after time t2, since the oxygen storage amount of the catalyst is substantially appropriate, the control device sets the target air-fuel ratio of the internal combustion engine to the stoichiometric air-fuel ratio. Further, the control device restarts the update of the time integration value, and calculates the downstream FB correction amount Vsb and the downstream FB learning correction amount Vsg based on the time integration value. In addition, the control device, based on the calculated downstream FB correction amount Vsb, the calculated downstream FB learning correction amount Vsg, the actual output value of the upstream air-fuel ratio sensor, and the target air-fuel ratio. The upstream FB correction amount is determined. Then, the control device determines the fuel injection amount Fi based on the basic fuel injection amount Fis corresponding to the stoichiometric air-fuel ratio and the upstream FB correction amount DFi. Thereby, the air-fuel ratio of the internal combustion engine is controlled to the stoichiometric air-fuel ratio.

  Thus, according to the conventional control device, the update of the time integral value is stopped and the update of the downstream FB learning correction amount Vsg is also stopped during the fuel increase period Trc, so that the state of the internal combustion engine is in the steady state. The downstream FB learning correction amount Vsg is not calculated based on the downstream FB correction amount Vsb affected by the output deviation in the fuel increase period Trc that is not in the state. Therefore, it is possible to prevent the downstream FB learning correction amount Vsg from becoming an inappropriate value as a value for compensating for the output deviation of the upstream air-fuel ratio sensor.

  By the way, in the transition period Tk (time t2 to time t3) from the end of the fuel increase period Trc to the elapse of a predetermined time, it is immediately after the target air-fuel ratio is changed, so the state of the internal combustion engine is the steady state. It is not. Therefore, when the downstream FB learning correction amount Vsg is calculated based on the downstream FB correction amount Vsb affected by the output deviation in the transition period Tk, the calculated downstream FB learning correction amount Vsg is calculated as the upstream air-fuel ratio. There is a possibility that it becomes an inappropriate value as a value for compensating the output deviation of the sensor. That is, it is desirable to stop updating the downstream FB learning correction amount Vsg during the transition period Tk. Furthermore, in order to set the downstream FB learning correction amount Vsg calculated after the transition period Tk to an appropriate value, it is desirable to stop updating the downstream FB correction amount Vsb during the transition period Tk. Therefore, it is desirable that the update of the time integral value that is the basis for calculating the downstream FB correction amount Vsb is also stopped in the transient period Tk.

  On the other hand, during this transition period Tk, the target air-fuel ratio of the internal combustion engine is set to the stoichiometric air-fuel ratio. Therefore, in order to quickly match the downstream air-fuel ratio with the downstream target air-fuel ratio, as described above, the time integration value is updated in the transient period Tk to calculate the downstream FB correction amount Vsb and the calculated downstream side It is also desirable to control the air-fuel ratio of the internal combustion engine by the FB correction amount Vsb.

  Thus, in the above control device, if the time integration value is updated in order to update the downstream FB correction amount Vsb in the transition period Tk, the downstream FB learning correction amount Vsg is an inappropriate value after the transition period Tk ends. There is a risk of being updated. On the contrary, if the update of the time integration value is stopped in the transition period Tk in order to set the downstream FB learning correction amount Vsg after the transition period Tk to an appropriate value, the integral term of the downstream FB correction amount Vsb in the transition period Tk. There is a problem that the emission cannot be improved.

  The present invention has been made to address the above-described problems, and one of its purposes is an air-fuel ratio control apparatus for an internal combustion engine that can quickly match the actual air-fuel ratio with the target air-fuel ratio. Is to provide.

In order to achieve this object, the air-fuel ratio control apparatus for an internal combustion engine according to the present invention is applied to an internal combustion engine in which an exhaust purification catalyst is disposed in an exhaust passage.
Furthermore, an air-fuel ratio control apparatus for an internal combustion engine according to the present invention includes:
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
An upstream feedback correction amount is calculated based on the output value of the upstream air-fuel ratio sensor and the output value of the downstream air-fuel ratio sensor, and is supplied to the internal combustion engine based on the calculated upstream feedback correction amount. Air-fuel ratio feedback control means for controlling the air-fuel ratio of the mixed gas.

  In addition, the air-fuel ratio feedback control means includes a downstream feedback correction amount calculation means, a downstream feedback learning correction amount calculation means, a time integral value setting means, and an upstream feedback correction amount calculation means.

  The downstream feedback correction amount calculating means calculates a difference between an output value of the downstream air-fuel ratio sensor and a predetermined downstream target value in a downstream feedback condition satisfaction period in which a predetermined downstream feedback execution condition is satisfied. A first time integration value obtained by time-integrating a certain output deviation is updated based on the first time integration value obtained at the present time and the current output deviation, and the updated first time integration value is obtained. This is means for calculating a downstream feedback correction amount based on the value.

  The downstream feedback learning correction amount calculating means is a second time obtained by time-integrating the output deviation in a learning condition satisfaction period in which a predetermined learning execution condition in which a condition is further added to the downstream feedback execution condition is satisfied. The integral value is updated based on the second time integral value obtained at the current time and the current output deviation, and the downstream feedback learning correction amount is calculated based on the updated second time integral value. It is a means for calculating and maintaining the downstream feedback learning correction amount constant during a learning condition non-satisfied period in which the learning execution condition is not satisfied.

  When the learning execution condition is not satisfied at the first time point and is satisfied at the subsequent second time point, the time integral value setting means sets the second time integral value at the second time point from the first time point to the same time point. It is a means for setting to a predetermined value unrelated to the output deviation up to two time points.

  The upstream feedback correction amount calculation means is configured to determine the upstream feedback correction amount based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor during the downstream feedback condition establishment period. The downstream feedback learning correction amount and the upstream air-fuel ratio sensor are calculated without calculating the correction amount and based on the downstream feedback correction amount in the downstream feedback condition non-fulfilling period in which the downstream feedback execution condition is not satisfied. The upstream feedback correction amount is calculated based on the output value.

  According to this, since the learning execution condition is a condition in which a condition is further added to the downstream feedback execution condition, there is a case where the learning execution condition is not satisfied even if the downstream feedback execution condition is satisfied. . That is, an overlapping period in which the downstream feedback condition establishment period overlaps with the learning condition establishment period occurs. For example, this overlap period is a fuel increase period in which the target air-fuel ratio of the internal combustion engine is set to a richer air-fuel ratio than the stoichiometric air-fuel ratio, and a subsequent period in which the target air-fuel ratio of the internal combustion engine becomes the stoichiometric air-fuel ratio. In the case where there is a set normal period, this is a transient period from the end of the fuel increase period in the normal period until a predetermined transient time elapses.

  In the overlap period, the first time integral value is updated based on the output deviation in the overlap period, and the downstream feedback correction amount is calculated based on the updated first time integral value. Thereby, it is possible to calculate the downstream feedback correction amount during the overlap period more appropriately than when calculating based on the time integral value in which the output deviation in the overlap period is not reflected. As a result, the air-fuel ratio (downstream air-fuel ratio) downstream of the exhaust purification catalyst (catalyst) can be quickly brought close to the air-fuel ratio (downstream target air-fuel ratio) corresponding to the downstream target value in the overlap period.

  On the other hand, during the learning condition failure period, the downstream feedback learning correction amount is maintained constant regardless of whether or not it is an overlapping period. Further, at the end of the learning condition failure period (second time point), the second time integral value is a predetermined value that is unrelated to the output deviation in the learning condition failure period (period from the first time point to the second time point). (That is, a value based on the time integral value of the output deviation calculated by time-integrating the output deviation only in the learning condition satisfaction period, for example, the second time integral value at the first time point).

  As a result, it is possible to avoid the output deviation in the learning condition non-satisfied period (particularly, the overlap period) being reflected in the downstream feedback learning correction amount, thereby preventing the downstream feedback learning correction amount from becoming an inappropriate value. be able to. As a result, a deviation of the correspondence relationship between the output value of the upstream air-fuel ratio sensor and the actual air-fuel ratio upstream of the catalyst (upstream air-fuel ratio) with respect to the planned correspondence relationship (the output deviation of the upstream air-fuel ratio sensor). ) Can be appropriately compensated, and the actual air-fuel ratio of the internal combustion engine can be reliably matched with the target air-fuel ratio.

Further, an air-fuel ratio control device for another internal combustion engine according to the present invention is applied to the above-described internal combustion engine,
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
An upstream feedback correction amount is calculated based on the output value of the upstream air-fuel ratio sensor and the output value of the downstream air-fuel ratio sensor, and is supplied to the internal combustion engine based on the calculated upstream feedback correction amount. Air-fuel ratio feedback control means for controlling the air-fuel ratio of the mixed gas.

  Further, the air-fuel ratio feedback control means includes a time integral value update means, a downstream feedback correction amount calculation means, a downstream feedback learning correction amount calculation means, a time integral value setting means, and an upstream feedback correction amount calculation means. And comprising.

  The time integral value updating means is the same time integral value currently obtained as a time integral value obtained by time-integrating an output deviation that is a difference between the output value of the downstream air-fuel ratio sensor and a predetermined downstream target value. And updating based on the current output deviation.

  The downstream feedback correction amount calculating means is a means for calculating a downstream feedback correction amount based on the time integral value in a downstream feedback condition satisfaction period in which a predetermined downstream feedback execution condition is satisfied.

  The downstream-side feedback learning correction amount calculating means is configured to provide downstream-side feedback learning based on the time integral value in a learning condition establishment period in which a predetermined learning execution condition in which a condition is further added to the downstream feedback execution condition is satisfied. It is a means for calculating a correction amount and maintaining the downstream feedback learning correction amount constant during a learning condition non-satisfied period in which the learning execution condition is not satisfied.

  The time integration value setting means sets the time integration value at the second time point from the first time point to the second time point when the learning execution condition is not satisfied at the first time point and is satisfied at the subsequent second time point. Means for setting a predetermined value irrelevant to the output deviation.

  The upstream feedback correction amount calculation means is configured to determine the upstream feedback correction amount based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor during the downstream feedback condition establishment period. The downstream feedback learning correction amount and the upstream air-fuel ratio sensor are calculated without calculating the correction amount and based on the downstream feedback correction amount in the downstream feedback condition non-fulfilling period in which the downstream feedback execution condition is not satisfied. The upstream feedback correction amount is calculated based on the output value.

  According to this, similar to the above-described air-fuel ratio control apparatus for an internal combustion engine, an overlapping period in which the learning condition non-establishment period and the downstream feedback condition satisfaction period overlap is generated. In the overlap period, the time integral value is updated based on the output deviation in the overlap period, and the downstream feedback correction amount is calculated based on the updated time integral value. Thereby, it is possible to calculate the downstream feedback correction amount during the overlap period more appropriately than when calculating based on the time integral value in which the output deviation in the overlap period is not reflected. As a result, in the overlap period, the downstream air-fuel ratio can be quickly brought close to the downstream target air-fuel ratio.

  On the other hand, during the learning condition failure period, the downstream feedback learning correction amount is maintained constant regardless of whether or not it is an overlapping period. Further, at the end of the learning condition failure period (second time point), the time integral value is a predetermined value that is not related to the output deviation in the learning condition failure time period (period from the first time point to the second time point) (that is, It is set to a value based on the time integrated value of the output deviation calculated by time integrating the output deviation only in the learning condition satisfaction period (for example, the time integrated value at the first time point).

  As a result, it is possible to avoid the output deviation in the learning condition non-satisfied period (particularly, the overlap period) being reflected in the downstream feedback learning correction amount, thereby preventing the downstream feedback learning correction amount from becoming an inappropriate value. be able to. As a result, the output deviation of the upstream air-fuel ratio sensor can be compensated appropriately, and the actual air-fuel ratio of the internal combustion engine can be made to match the target air-fuel ratio with certainty.

On the other hand, another object of the present invention is to provide a fuel injection amount control device for an internal combustion engine that can quickly match the actual air-fuel ratio with the target air-fuel ratio.
In order to achieve the above object, a fuel injection amount control device for an internal combustion engine according to the present invention is applied to the internal combustion engine described above,
Fuel injection means for injecting fuel supplied to the internal combustion engine;
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
And controlling the fuel injection amount which is the amount of fuel injected by the fuel injection means.

  The fuel injection amount control device further includes fuel cut means, fuel increase means, downstream feedback correction amount calculation means, downstream feedback learning correction amount calculation means, and fuel injection amount feedback control means.

The fuel cut means is means for stopping fuel injection by the fuel injection means during a fuel cut condition establishment period, which is a period in which a predetermined fuel cut execution condition is established.
The fuel increasing means is configured so that the air-fuel ratio of the mixed gas supplied to the internal combustion engine is higher than the stoichiometric air-fuel ratio during the fuel increasing period from when the fuel cut condition is satisfied until a predetermined fuel increasing time elapses. It is means for controlling the fuel injection amount so as to achieve a predetermined fuel increase air-fuel ratio on the rich side.

  The downstream feedback correction amount calculation means is a downstream feedback condition establishment period which is a period in which a predetermined downstream feedback execution condition is established among the period not including the fuel cut condition establishment period and the fuel increase period. The means for calculating the downstream feedback correction amount based on the output deviation which is the difference between the output value of the downstream air-fuel ratio sensor and the predetermined downstream target value corresponding to the theoretical air-fuel ratio.

  The downstream-side feedback learning correction amount calculating means establishes a learning condition that is a period that does not include a transient period from the end of the fuel increase period of the downstream-side feedback condition establishment period until a predetermined transient time elapses. During the period, the downstream feedback learning correction amount is calculated based on the output deviation, and the downstream feedback learning correction amount is maintained constant during the learning condition non-establishment period other than the learning condition establishment period.

  The fuel injection amount feedback control means performs upstream feedback correction based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor during the downstream feedback condition establishment period. The fuel injection amount is controlled based on the upstream feedback correction amount calculated and the downstream feedback execution condition is not satisfied, and is at least a period within the downstream feedback condition non-satisfaction period During the period not including the fuel cut condition establishment period, the upstream feedback correction amount is calculated based on the downstream feedback learning correction amount and the output value of the upstream air-fuel ratio sensor. The fuel injection amount is controlled on the basis of

  According to this, the transition period is a period within the downstream feedback condition establishment period. Furthermore, the transition period is not included in the learning condition establishment period. That is, the transition period is a period (overlapping period) in which the downstream feedback condition establishment period and the learning condition non-establishment period overlap. In the overlap period, the downstream feedback correction amount is calculated based on the output deviation in the overlap period. Thereby, the downstream feedback correction amount during the overlapping period can be calculated more appropriately than when the output deviation in the overlapping period is not reflected in the downstream feedback correction amount. As a result, in the overlap period, the downstream air-fuel ratio can be quickly brought close to the downstream target air-fuel ratio.

  Further, during the learning condition failure period, the downstream feedback learning correction amount is maintained constant (update is stopped) regardless of whether or not it is an overlapping period. As a result, it is possible to avoid updating the downstream feedback learning correction amount based on the output deviation in the transition period in which the feedback control of the air-fuel ratio using the downstream feedback correction amount is not performed for a sufficiently long time. Accordingly, it is possible to prevent the downstream feedback learning correction amount from becoming an inappropriate value. As a result, the output deviation of the upstream air-fuel ratio sensor can be compensated appropriately, and the actual air-fuel ratio of the internal combustion engine can be made to match the target air-fuel ratio with certainty.

<First Embodiment>
Hereinafter, embodiments of an air-fuel ratio control device (fuel injection amount control device) for an internal combustion engine according to the present invention will be described with reference to the drawings. FIG. 2 shows a schematic configuration of a system in which the air-fuel ratio control apparatus for an internal combustion engine according to the first embodiment of the present invention is applied to a multi-cylinder (four-cylinder) internal combustion engine 10 operated by a four-cycle spark ignition system. FIG. 2 shows only a cross section of the specific cylinder, but the other cylinders have the same configuration.

  The internal combustion engine 10 includes a cylinder block unit 20 including a cylinder block, a cylinder block lower case, an oil pan, and the like, a cylinder head unit 30 fixed on the cylinder block unit 20, and a mixed gas (main In the example, an intake system 40 for supplying gasoline mixture) and an exhaust system 50 for releasing exhaust gas from the cylinder block unit 20 to the outside are included.

  The cylinder block unit 20 includes a cylinder 21, a piston 22, a connecting rod 23, and a crankshaft 24. The piston 22 reciprocates in the cylinder 21, and the reciprocating motion of the piston 22 is transmitted to the crankshaft 24 through the connecting rod 23, whereby the crankshaft 24 rotates. The cylinder 21, the head of the piston 22, and the cylinder head portion 30 form a combustion chamber (cylinder) 25.

  The cylinder head portion 30 includes an intake port 31 that communicates with the combustion chamber 25, an intake valve 32 that opens and closes the intake port 31, an intake cam shaft that drives the intake valve 32, and continuously changes the phase angle of the intake cam shaft. A variable intake timing device 33, an actuator 33a of the variable intake timing device 33, an exhaust port 34 communicating with the combustion chamber 25, an exhaust valve 35 for opening and closing the exhaust port 34, an exhaust camshaft 36 for driving the exhaust valve 35, an ignition plug 37, An igniter 38 including an ignition coil that generates a high voltage to be applied to the spark plug 37 and an injector (fuel injection means) 39 for supplying fuel into the combustion chamber 25 by injecting fuel into the intake port 31 are provided.

  The intake system 40 includes an intake pipe 41 including an intake manifold that communicates with the intake port 31 of each cylinder and forms an intake passage together with the intake port 31, an air filter 42 provided at an end of the intake pipe 41, and an intake pipe 41. A throttle valve 43 for changing the opening cross-sectional area of the intake passage and a throttle valve actuator 43a including a DC motor are provided. The intake system 40 takes in air from the outside and supplies the taken-in air to the combustion chamber 25 of each cylinder.

  The exhaust system 50 includes an exhaust manifold 51 composed of a plurality of independent passages communicating with the exhaust ports 34 of the respective cylinders, and a collecting portion for collecting these passages downstream, and an exhaust pipe connected to the collecting portion of the exhaust manifold 51. (Exhaust pipe) 52, a three-way catalyst 53 (also referred to as an upstream catalytic converter or a start catalytic converter) as an exhaust purification catalyst disposed (interposed) in the exhaust pipe 52, hereinafter referred to as "first catalyst 53" ) And the downstream three-way catalyst 54 (disposed below the floor of the vehicle) disposed (interposed) in the exhaust pipe 52 downstream of the first catalyst 53. This is also referred to as a “second catalyst 54”. The exhaust port 34, the exhaust manifold 51, and the exhaust pipe 52 constitute an exhaust passage. In the exhaust passage, exhaust gas, which is a combustion gas generated by combustion of a mixed gas containing fuel and air in the combustion chamber 25, passes therethrough.

  Each of the first catalyst 53 and the second catalyst 54 occludes oxygen in the exhaust gas. Further, each of the first catalyst 53 and the second catalyst 54 promotes the reaction between the unburned components of the fuel in the exhaust gas and the oxygen in the exhaust gas or the stored oxygen to remove harmful substances in the exhaust gas. It purifies (purifies exhaust gas).

  On the other hand, this system includes a hot-wire air flow meter 61, a throttle position sensor 62, a cam position sensor 63, a crank position sensor 64, a water temperature sensor 65, an exhaust passage upstream of the first catalyst 53 (in this example, the exhaust manifold 51). An air-fuel ratio sensor 66 (hereinafter referred to as an “upstream air-fuel ratio sensor 66”) disposed in the exhaust gas passage of the first catalyst 53 and upstream of the second catalyst 54. An air-fuel ratio sensor 67 (hereinafter referred to as “downstream air-fuel ratio sensor 67”), an accelerator opening sensor 68, and an electric control device 70 are provided.

The air flow meter 61 detects the mass flow rate per unit time of the intake air passing through the intake pipe 41 and outputs a signal representing the intake air flow rate Ga.
The throttle position sensor 62 detects the opening of the throttle valve 43 and outputs a signal representing the throttle valve opening TA.
The cam position sensor 63 generates a signal (G2 signal) having one pulse that is generated every time the intake camshaft rotates 90 ° (that is, every time the crankshaft 24 rotates 180 °).

The crank position sensor 64 outputs a signal having a narrow pulse generated every time the crankshaft 24 rotates 10 ° and a wide pulse generated every time the crankshaft 24 rotates 360 °. This signal represents the engine speed NE.
The water temperature sensor 65 detects the temperature of the cooling water for cooling the internal combustion engine 10 and outputs a signal representing the cooling water temperature THW.

  The upstream air-fuel ratio sensor 66 is a limit current type air-fuel ratio sensor. The upstream air-fuel ratio sensor 66 is based on the oxygen concentration in the detection target gas (in this example, the exhaust gas upstream of the first catalyst 53) and the unburned component (for example, hydrocarbon) concentration of the fuel. A voltage Vabyfs corresponding to the output current is output as an output value.

  As shown in FIG. 3, the output value Vabyfs of the upstream air-fuel ratio sensor 66 increases monotonously as the air-fuel ratio A / F becomes a leaner air-fuel ratio and is approximately proportional to the air-fuel ratio A / F. Change. Further, the output value Vabyfs of the upstream air-fuel ratio sensor 66 coincides with the theoretical air-fuel ratio corresponding output value Vst when the air-fuel ratio A / F is the stoichiometric air-fuel ratio. With such a configuration, the upstream air-fuel ratio sensor 66 accurately detects the air-fuel ratio A / F over a wide range.

  The downstream air-fuel ratio sensor 67 is an electromotive force type (concentration cell type) air-fuel ratio sensor. The downstream air-fuel ratio sensor 67 outputs the voltage Voxs corresponding to the air-fuel ratio A / F as an output value based on the oxygen concentration in the detection target gas (in this example, the exhaust gas downstream of the first catalyst 53). It has become.

  As shown in FIG. 4, the output value Voxs of the downstream air-fuel ratio sensor 67 monotonously decreases as the air-fuel ratio A / F becomes a leaner air-fuel ratio. The output value Voxs changes suddenly in a region D where the air-fuel ratio A / F is a value near the stoichiometric air-fuel ratio. That is, the output value Voxs changes suddenly when the air-fuel ratio A / F changes from the lean side to the rich side with respect to the stoichiometric air-fuel ratio, or when it changes in the opposite direction. The output value Voxs increases the air-fuel ratio while taking a value of approximately 0.1 (V) when the air-fuel ratio A / F is an air-fuel ratio other than the region D and is an air-fuel ratio leaner than the stoichiometric air-fuel ratio. Accordingly, when the air-fuel ratio A / F is an air-fuel ratio other than the region D and is richer than the stoichiometric air-fuel ratio, the air-fuel ratio becomes approximately 0.9 (V) while taking a value of approximately 0.9 (V). It gradually increases with a decrease. The output value Voxs is 0.5 (V) when the air-fuel ratio A / F is the stoichiometric air-fuel ratio.

  Referring again to FIG. 2, the accelerator opening sensor 68 detects the operation amount of the accelerator pedal 81 operated by the driver, and represents the operation amount (accelerator pedal operation amount) Accp of the accelerator pedal 81. A signal is output.

  The electric control device 70 includes a CPU 71 connected by a bus, a routine (program) executed by the CPU 71, a ROM 72 pre-stored with tables (look-up tables, maps), constants, and the like, and the CPU 71 temporarily stores data as necessary. The microcomputer 73 includes a RAM 73 that stores data, a backup RAM 74 that stores data while the power is on, and retains the stored data even when the power is shut off, and an interface 75 that includes an AD converter. The interface 75 is connected to the sensors 61 to 68, supplies signals from the sensors 61 to 68 to the CPU 71, and in response to instructions from the CPU 71, the actuator 33a, the igniter 38, the injector 39, and the throttle valve of the variable intake timing device 33. A drive signal is sent to the actuator 43a.

(Outline of air-fuel ratio feedback control)
Next, air-fuel ratio feedback (FB) control (air-fuel ratio FB control) performed by the air-fuel ratio control device (control device) of the internal combustion engine configured as described above will be described.

The air-fuel ratio FB control is control for determining the fuel injection amount Fi so that the air-fuel ratio of the mixed gas supplied to the internal combustion engine 10 matches the target air-fuel ratio. In the air-fuel ratio FB control, the fuel injection amount Fi is determined by correcting the basic fuel injection amount Fbase with the upstream FB correction amount DFi according to the following equation (1).
Fi = Fbase + DFi (1)

The basic fuel injection amount Fbase is an injection amount considered necessary for obtaining the target air-fuel ratio, and is determined by dividing the separately obtained in-cylinder intake air amount by the target air-fuel ratio. The upstream FB correction amount DFi is an amount for correcting “the excess / deficiency with respect to the injection amount for obtaining the target air-fuel ratio” of the basic fuel injection amount Fbase. The upstream FB correction amount DFi is set to the output value Vabyfs of the upstream air-fuel ratio sensor 66, the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb (actually, the sum thereof), as will be described in detail later. Based on. That is, when the predetermined function is represented by f, the upstream FB correction amount DFi is represented by the following equation (2).
DFi = f (Vabyfs + Vafsfb + Gsfb) (2)

  In the air-fuel ratio FB control, the upstream FB correction amount DFi in the above equation (1) is set to a value according to the success or failure of the upstream FB execution condition. Further, in the air-fuel ratio FB control, the downstream FB correction amount Vafsfb in the above equation (2) is set to a value according to the success or failure of the downstream FB execution condition. In addition, the air-fuel ratio FB control updates the downstream FB learning correction amount Gsfb in the above equation (2) according to the success or failure of the learning execution condition. Hereinafter, each condition will be described.

(1) Upstream FB execution condition (main feedback execution condition) and upstream FB control The upstream FB execution condition is that the upstream air-fuel ratio sensor 66 is normal (including that it is activated), and will be described later. When the fuel cut execution condition to be satisfied is not satisfied, it is satisfied regardless of the target air-fuel ratio. When the upstream FB execution condition is satisfied, the air-fuel ratio FB control sets the upstream FB correction amount DFi based on the above equation (2). In the air-fuel ratio FB control, the fuel injection amount Fi is determined by correcting the basic fuel injection amount Fbase with the upstream FB correction amount DFi. That is, upstream FB (main FB) control is executed. On the other hand, when the upstream FB execution condition is not satisfied, the air-fuel ratio FB control sets the upstream FB correction amount DFi to “0”. As a result, the basic fuel injection amount Fbase is not corrected by the upstream FB correction amount DFi. That is, the upstream FB control is not executed.

(2) Downstream FB execution condition (sub-feedback execution condition) and downstream FB control As for the downstream FB execution condition, the upstream FB execution condition is satisfied and the downstream air-fuel ratio sensor 67 is normal (activated). This is true when the target air-fuel ratio is the stoichiometric air-fuel ratio. As will be described later, in the fuel increase period after the fuel cut is performed (when the fuel increase condition after the fuel cut is satisfied), the target air-fuel ratio is on the richer side than the stoichiometric air-fuel ratio because the fuel increase condition is satisfied. The fuel increase control is executed after the fuel cut with the air-fuel ratio set.

Therefore, the downstream FB execution condition is
Both the upstream air-fuel ratio sensor 66 and the downstream air-fuel ratio sensor 67 are normal, and
・ The fuel cut execution condition is not satisfied, and
-The fuel increase condition after fuel cut is not satisfied (that is, the target air-fuel ratio is the stoichiometric air-fuel ratio),
Sometimes true.

When the downstream side FB execution condition is satisfied, the air-fuel ratio FB control is performed with the output value Voxs of the downstream side air-fuel ratio sensor 67 (that is, the air-fuel ratio downstream of the first catalyst 53) and the downstream each time a predetermined time elapses. The output deviation DVoxs, which is the difference from the side target value (the output value of the downstream side air-fuel ratio sensor 67 corresponding to the theoretical air-fuel ratio), is obtained. In the air-fuel ratio FB control, the time integrated value (downstream FB time integrated value, first time integrated value) SDVoxs1 of the output deviation DVoxs is updated according to the equation (3). Δt is a fixed minute time.
SDVoxs1 (current value) = SDVoxs1 (previous value) + DVoxs · Δt (3)

In addition, the air-fuel ratio FB control calculates the integral term (Ki · SDVoxs1) by multiplying the updated latest downstream FB time integral value SDVoxs1 by a predetermined gain Ki, and the integral term (Ki · SDVoxs1) The downstream FB correction amount Vafsfb including is calculated according to the following equation (4). The other term “Other” is a proportional term and a differential term described later.
Vafsfb ＝ Ki ・ SDVoxs1 ＋ Other… (4)

  As described above, the air-fuel ratio FB control uses the downstream FB correction amount Vafsfb when obtaining the upstream FB correction amount DFi. That is, downstream FB (sub FB) control is executed.

  On the other hand, when the downstream FB execution condition is not satisfied, the air-fuel ratio FB control sets the downstream FB correction amount Vafsfb to “0”. That is, the downstream FB control is not executed. Further, the air-fuel ratio FB control maintains the value of the downstream side FB time integration value SDVoxs1 to the value of the downstream side FB time integration value SDVoxs1 when the downstream side FB execution condition is not satisfied.

(3) Learning execution condition and learning control The learning execution condition is the output of the downstream air-fuel ratio sensor 67 from the time when the downstream FB execution condition is satisfied and the fuel increase period after the fuel cut is completed. This holds true when the value Voxs is not a transitional period up to the point of outputting a predetermined threshold output value.

Therefore, the learning execution condition is
Both the upstream air-fuel ratio sensor 66 and the downstream air-fuel ratio sensor 67 are normal, and
・ The fuel cut execution condition is not satisfied, and
-The fuel increase condition after the fuel cut is not satisfied (that is, when the target air-fuel ratio is the stoichiometric air-fuel ratio), and
・ The above transition period has elapsed since the fuel increase control was executed after the fuel cut.
Sometimes true.

When the learning execution condition is satisfied, the air-fuel ratio FB control calculates the time integration value (learning time integration value, second time integration value) SDVoxs2 of the output deviation DVoxs (5 Update according to formula. Δt is a fixed minute time.
SDVoxs2 (current value) = SDVoxs2 (previous value) + DVoxs · Δt (5)

Further, as shown in the following equation (6), the air-fuel ratio FB control is a downstream FB learning correction by a value (Kg · SDVoxs2) obtained by multiplying the updated latest learning integration value SDVoxs2 by a predetermined coefficient Kg. By adding to the amount Gsfb, the downstream FB learning correction amount Gsfb is calculated (updated). In the air-fuel ratio FB control, when the downstream FB learning correction amount Gsfb is updated, the value Kg · SDVoxs2 is subtracted from each of the downstream FB time integrated value SDVoxs1 and the learning time integrated value SDVoxs2.
Gsfb (current value) = Gsfb (previous value) + Kg · SDVoxs2 (6)

When the learning execution condition is not satisfied, the air-fuel ratio FB control maintains the value of the downstream FB learning correction amount Gsfb at the value of the downstream FB learning correction amount Gsfb when the learning execution condition is no longer satisfied, The value of the learning time integration value SDVoxs2 is maintained at the value of the learning time integration value SDVoxs2 when the learning execution condition is no longer satisfied.
The air-fuel ratio FB control as described above is executed by the control device.

(Outline of actual operation before and after fuel cut execution)
Further, the control device executes the fuel cut when the fuel cut execution condition is satisfied, and then executes the fuel increase control. Specifically, the control device operates as follows before and after the fuel cut execution condition is satisfied. It is assumed that both the upstream air-fuel ratio sensor 66 and the downstream air-fuel ratio sensor 67 are normal.

(1) Before executing fuel cut Before executing fuel cut, the control device sets the target air-fuel ratio to the stoichiometric air-fuel ratio. Therefore, all conditions of the upstream FB execution condition, the downstream FB execution condition, and the learning execution condition are satisfied. Accordingly, the downstream FB time integration value SDVoxs1, the downstream FB correction amount Vafsfb, the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb are updated. Then, the control device calculates the upstream FB correction amount DFi according to the above equation (2), and corrects the basic fuel injection amount Fbase based on the calculated upstream FB correction amount DFi and the above equation (1).

(2) During fuel cut execution When the fuel cut execution condition is satisfied, the control device executes a fuel cut (fuel cut) to stop fuel injection. At this time, all of the upstream FB execution condition, the downstream FB execution condition, and the learning execution condition are not satisfied. Therefore, the upstream FB correction amount DFi is set to “0”. Further, the downstream FB correction amount Vafsfb is set to “0”, and the downstream side FB time integration value SDVoxs1 is maintained at the value at the time when the fuel cut is started. In addition, the downstream FB learning correction amount Gsfb and the learning time integration value SDVoxs2 are maintained at respective values at the time when the fuel cut starts to be executed.

(3) While fuel increase control is being executed immediately after the end of the fuel cut The fuel increase condition after the fuel cut is satisfied when the fuel cut execution condition is not satisfied and the fuel increase condition after the fuel cut is satisfied. Is set to a value on the rich side. At this time, the upstream FB execution condition is satisfied. On the other hand, neither the downstream FB execution condition nor the learning execution condition is satisfied. Accordingly, the downstream side FB correction amount Vafsfb continues to be set to “0”, and the downstream side FB time integration value SDVoxs1 continues to be maintained at the value at the time when the fuel cut starts to be executed. In addition, the downstream FB learning correction amount Gsfb and the learning time integration value SDVoxs2 continue to be maintained at respective values at the time when the fuel cut is started. As a result, “0” is substituted for the downstream FB correction amount Vafsfb in the above equation (2), and the upstream FB correction amount DFi is output with the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the fuel cut is started. It is obtained based only on the downstream FB learning correction amount Gsfb maintained at the value at the time. Then, the control device corrects the basic fuel injection amount Fbase according to the above equation (1).

(4) Transition period after completion of execution of fuel increase control In the transient period after execution of fuel increase control after fuel cut is terminated because the fuel increase condition is not satisfied, the target air-fuel ratio is set to the theoretical air-fuel ratio, The upstream FB execution condition and the downstream FB execution condition are satisfied. On the other hand, the learning execution condition is not satisfied. Accordingly, the downstream side FB time integration value SDVoxs1 and the downstream side FB correction amount Vafsfb are updated. The downstream FB learning correction amount Gsfb and the learning time integration value SDVoxs2 are maintained at the respective values at the time when the fuel cut starts to be executed. Then, the control device calculates the upstream FB correction amount DFi according to the above equation (2), and corrects the basic fuel injection amount Fbase according to the above equation (1).

(5) After the lapse of the transition period after the completion of the fuel increase control execution When the transition period has elapsed, all of the upstream FB execution condition, the downstream FB execution condition, and the learning execution condition are the same as before the fuel cut execution. The condition is met. Accordingly, the downstream FB time integration value SDVoxs1, the downstream FB correction amount Vafsfb, the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb are updated. Then, the control device calculates the upstream FB correction amount DFi according to the above equation (2), and corrects the basic fuel injection amount Fbase according to the above equation (1).

  Thus, the present control device starts updating the downstream side FB time integration value SDVoxs1 and the downstream side FB correction amount Vafsfb from the start of the transition period after the completion of the fuel increase control. Thereby, the air-fuel ratio downstream of the first catalyst 53 can be brought close to the stoichiometric air-fuel ratio (downstream target air-fuel ratio) quickly during the transition period.

  Furthermore, the control device continues to maintain the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb at the respective values at the time when the fuel cut starts to be executed during the transition period, and when the transition period elapses. Each of the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb is updated. As a result, the output deviation DVoxs during the transition period can be avoided from being reflected in the downstream FB learning correction amount Gsfb, so that the downstream FB learning correction amount Gsfb can be prevented from becoming an inappropriate value. .

During the fuel cut, the amount of oxygen stored in the first catalyst 53 becomes excessive. Therefore, by performing the fuel increase control after the fuel cut is executed, the amount of oxygen stored in the first catalyst 53 is quickly reduced to an appropriate value.
The control device operates as described above before and after the fuel cut execution condition is established.

(Flag used for control)
Next, a flag used when the control device performs the above-described air-fuel ratio FB control will be described. The control device uses six flags including a fuel cut flag XFC, a fuel increase flag XRC, an upstream FB control execution flag XMFB, a downstream FB control execution flag XSFB, a transition period display flag XKT, and a learning execution flag XG (FIG. See 5). These flags are set to “0” in an initial routine executed when an ignition key (not shown) is changed from OFF to ON.

  The fuel cut flag XFC is a flag indicating whether or not to stop fuel injection (perform fuel cut). That is, when the value of the fuel cut flag XFC is set to “1”, fuel injection is stopped (fuel injection is not performed). On the other hand, when the value of the fuel cut flag XFC is set to “0”, fuel injection is executed (fuel injection is performed).

  The fuel cut flag XFC is such that the engine rotational speed NE is higher than the threshold rotational speed α, and the throttle valve opening TA is a threshold opening (in this example, the fully closed opening at which the throttle valve 43 is fully closed) β. Set to “1” when: The fuel cut flag XFC is equal to or less than a value obtained by subtracting a predetermined positive value α1 from the threshold rotational speed α (α−α1), or the throttle valve opening TA is greater than the threshold opening β. When it becomes larger, it is set to “0”.

  The fuel increase flag XRC is a flag indicating whether or not the target air-fuel ratio of the internal combustion engine 10 is set to the fuel increase air-fuel ratio (fuel increase control after fuel cut is performed). That is, when the value of the fuel increase flag XRC is set to “1”, the target air-fuel ratio of the internal combustion engine 10 is set to the fuel increase air-fuel ratio AFrc which is relatively larger than the stoichiometric air-fuel ratio and is rich. On the other hand, when the value of the fuel increase flag XRC is set to “0”, the target air-fuel ratio of the internal combustion engine 10 is set to the stoichiometric air-fuel ratio AFst.

  The value of the fuel increase flag XRC is changed (set) from “0” to “1” when the fuel cut flag XFC changes from “1” to “0”. The value of the fuel increase flag XRC is set to “0” when the fuel increase time τrc has elapsed since the value was changed from “0” to “1”.

  The upstream FB control execution flag XMFB is a flag indicating whether or not the upstream FB control is performed. That is, when the value of the upstream FB control execution flag XMFB is set to “1”, the upstream FB control is performed. On the other hand, when the value of the upstream FB control execution flag XMFB is set to “0”, the upstream FB control is not performed.

  The value of the upstream FB control execution flag XMFB is set to “1” when the upstream air-fuel ratio sensor 66 is normal. However, while the fuel cut flag XFC is set to “1”, the value of the upstream FB control execution flag XMFB is set to “0”. In the present embodiment, it is assumed that the upstream air-fuel ratio sensor 66 and the downstream air-fuel ratio sensor 67 are always normal during the operation of the internal combustion engine 10.

  The downstream FB control execution flag XSFB is a flag indicating whether or not the downstream FB control is performed. That is, when the value of the downstream FB control execution flag XSFB is set to “1”, the downstream FB control is performed. On the other hand, when the value of the downstream FB control execution flag XSFB is set to “0”, the downstream FB control is not performed.

  The value of the downstream FB control execution flag XSFB is set to “1” when both the upstream air-fuel ratio sensor 66 and the downstream air-fuel ratio sensor 67 are normal. However, the value of the downstream FB control execution flag XSFB is set to “0” while either the fuel cut flag XFC or the fuel increase flag XRC is set to “1”.

  The transition period display flag XKT is a flag indicating that the current time is the transition period. The value of the transition period display flag XKT is set to “1” when the value of the fuel increase flag XRC changes from “1” to “0”, and the value of the transition period display flag XKT is set to “1”. When the output value Voxs of the downstream air-fuel ratio sensor 67 is greater than a preset threshold output value (in this example, the downstream target value Voxsref), the value is set to “0”.

  The learning execution flag XG is a flag indicating whether or not to update the downstream FB learning correction amount Gsfb. That is, when the value of the learning execution flag XG is set to “1”, the downstream FB learning correction amount Gsfb is updated. On the other hand, when the value of the learning execution flag XG is set to “0”, the downstream FB learning correction amount Gsfb is not updated.

  The value of the learning execution flag XG is set to “1” when the downstream FB control execution flag XSFB is “1” and the value of the transition period display flag XKT is “0”. On the other hand, the value of the learning execution flag XG is the value when the value of the downstream FB control execution flag XSFB is “0” or the value of the downstream FB control execution flag XSFB is “1” and the transient period display flag XKT. When the value of “1” is “1”, it is set to “0”.

(Actual operation)
Next, the actual operation of the control device will be described. The CPU 71 of the control device repeatedly executes a fuel cut condition determination routine for determining whether or not to execute the fuel cut shown in the flowchart of FIG. 6 every time a predetermined calculation cycle elapses. Note that the execution of the routine of FIG. 6 corresponds to the achievement of part of the function of the fuel cut means.

  Accordingly, when the predetermined timing is reached, the CPU 71 starts processing from step 600 and proceeds to step 605 to determine whether or not the value of the fuel cut flag XFC is “0”.

  Now, when the driver of the internal combustion engine 10 operates the accelerator pedal 81, the accelerator pedal operation amount Accp is maintained constant at the predetermined operation amount Accp1, and thereafter is maintained at "0" for a predetermined period. The description will be continued with reference to FIG. 5 assuming that the predetermined operation amount Accp1 is maintained again. In this case, it is assumed that the engine rotational speed NE is always higher than the threshold rotational speed α. In addition, when the accelerator pedal operation amount Accp is the predetermined operation amount Accp1, the CPU 71 sets the throttle valve opening TA to a predetermined opening TA1 that is larger than the threshold opening (full opening in this example) β. In addition, the throttle valve actuator 43a is controlled so that the throttle valve opening degree TA becomes the fully closed opening degree when the accelerator pedal operation amount Accp is "0".

  First, the accelerator pedal operation amount Accp is a predetermined operation before the accelerator pedal operation amount Accp is changed to “0” when the driver of the internal combustion engine 10 releases the accelerator pedal 81 from the accelerator pedal 81. From the time point when the amount Accp1 is maintained constant (the time point before the time point t0 in FIG. 5), the values of the fuel cut flag XFC, the fuel increase flag XRC, and the transient period display flag XKT are “ 0 ", and the values of the upstream FB control execution flag XMFB, the downstream FB control execution flag XSFB, and the learning execution flag XG are set to" 1 ".

  Accordingly, the CPU 71 makes a “Yes” determination at step 605 to proceed to step 610, where the current engine speed NE is higher than the preset threshold speed α, and the throttle valve opening degree TA is It is determined whether or not it is equal to or less than the preset threshold opening β.

  According to the above assumption, the engine rotational speed NE is higher than the threshold rotational speed α, and the throttle valve opening TA is a predetermined opening TA1 larger than the threshold opening β. Accordingly, the CPU 71 makes a “No” determination at step 610 to directly proceed to step 699 to end the present routine tentatively. That is, the value of the fuel cut flag XFC is maintained at “0”.

  On the other hand, the CPU 71 is configured to execute a fuel increase condition determination routine for determining whether or not to execute the fuel increase control shown by the flowchart in FIG. 7 following the fuel cut condition determination routine. The execution of the routine of FIG. 7 corresponds to the achievement of part of the function of the fuel increasing means.

  Accordingly, when the execution of the fuel cut condition determination routine is completed, the CPU 71 starts processing from step 700 and proceeds to step 705, where the current value immediately after the value of the fuel cut flag XFC has changed from “1” to “0”. It is determined whether it is a time.

  At this time, the value of the fuel cut flag XFC is maintained at “0”. Accordingly, the CPU 71 makes a “No” determination at step 705 to proceed to step 710, where the fuel increase time τrc has elapsed since the current value of the fuel increase flag XRC changed from “0” to “1”. It is determined whether or not it is the time point. At this time, since the value of the fuel increase flag XRC is maintained at “0”, the CPU 71 makes a “No” determination at step 710 to directly proceed to step 799 to end the present routine tentatively.

  Further, the CPU 71 is configured to execute an upstream FB condition determination routine for determining whether or not to execute the upstream FB control shown in the flowchart of FIG. 8 following the fuel increase condition determination routine. . The execution of the routine of FIG. 8 corresponds to the achievement of part of the function of the air-fuel ratio feedback control means.

  Accordingly, when the execution of the fuel increase condition determination routine is completed, the CPU 71 starts processing from step 800 and proceeds to step 805 to determine whether or not the upstream air-fuel ratio sensor 66 is normal. According to the above assumption, the upstream air-fuel ratio sensor 66 is normal. Accordingly, the CPU 71 determines “Yes” in step 805 and proceeds to step 810 to determine whether or not the value of the fuel cut flag XFC is “0”.

At this time, the value of the fuel cut flag XFC is maintained at “0”. Accordingly, the CPU 71 determines “Yes” at step 810 and proceeds to step 815 to set the value of the upstream FB control execution flag XMFB to “1”. That is, the value of the upstream FB control execution flag XMFB is maintained at “1”.
Then, the CPU 71 proceeds to step 899 to end the present routine tentatively.

  Further, the CPU 71 executes a downstream FB condition determining routine for determining whether or not to execute the downstream FB control shown by the flowchart in FIG. 9 following the upstream FB condition determining routine. Yes. The execution of the routine of FIG. 9 corresponds to the achievement of part of the function of the air-fuel ratio feedback control means.

  Therefore, when the execution of the upstream FB condition determination routine ends, the CPU 71 starts the process from step 900 and proceeds to step 905 to determine whether or not the upstream air-fuel ratio sensor 66 is normal. According to the above assumption, the upstream air-fuel ratio sensor 66 is normal.

  Therefore, the CPU 71 determines “Yes” in step 905 and proceeds to step 910 to determine whether or not the downstream air-fuel ratio sensor 67 is normal. According to the above assumption, the downstream air-fuel ratio sensor 67 is normal. Therefore, the CPU 71 determines “Yes” in step 910 and proceeds to step 915 to determine whether or not the value of the fuel cut flag XFC is “0”.

  At this time, the value of the fuel cut flag XFC is maintained at “0”. Accordingly, the CPU 71 determines “Yes” in step 915 and proceeds to step 920 to determine whether or not the value of the fuel increase flag XRC is “0”.

At this time, the value of the fuel increase flag XRC is maintained at “0”. Accordingly, the CPU 71 determines “Yes” in step 920 and proceeds to step 925 to set the value of the downstream FB control execution flag XSFB to “1”. That is, the value of the downstream FB control execution flag XSFB is maintained at “1”.
Then, the CPU 71 proceeds to step 999 to end this routine once.

  Further, the CPU 71 executes a learning condition determination routine for determining whether or not to update the downstream FB learning correction amount shown in the flowchart of FIG. 10 following the downstream FB condition determination routine. ing.

  Therefore, when the execution of the downstream FB condition determination routine is completed, the CPU 71 starts processing from step 1000 and proceeds to step 1005 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”. .

  At this time, the value of the downstream FB control execution flag XSFB is maintained at “1”. Therefore, the CPU 71 determines “Yes” in step 1005 and proceeds to step 1010 to determine whether or not the value of the transition period display flag XKT is “0”.

At this time, the value of the transition period display flag XKT is maintained at “0”. Accordingly, the CPU 71 determines “Yes” in step 1010, proceeds to step 1015, and sets the value of the learning execution flag XG to “1”. That is, the value of the learning execution flag XG is maintained at “1”.
Then, the CPU 71 proceeds to step 1099 to end the present routine tentatively.

(Fuel injection amount control)
On the other hand, the CPU 71 performs the fuel injection amount control routine for instructing the calculation of the fuel injection amount Fi and the fuel injection shown in the flowchart of FIG. Each time the crank angle reaches a predetermined crank angle (in this example, BTDC 90 ° CA, that is, an angle advanced by 90 ° from the exhaust top dead center) before completion and top dead center at which intake starts substantially) It is supposed to run. The execution of the routine of FIG. 11 corresponds to the achievement of part of the function of the air-fuel ratio feedback control means (fuel injection amount feedback control means).

  Accordingly, when the crank angle of an arbitrary cylinder (hereinafter also referred to as “fuel injection cylinder”) reaches the predetermined crank angle, the CPU 71 starts processing from step 1100 and proceeds to step 1105 to proceed to the fuel cut flag XFC. It is determined whether or not the value of “0” is “0”.

  At this time, the value of the fuel cut flag XFC is “0”. Accordingly, the CPU 71 makes a “Yes” determination at step 1105 to proceed to step 1110, where the intake air flow rate Ga detected by the air flow meter 61, the engine speed NE detected by the crank position sensor 64, and the intake Based on the table MapMc that defines the relationship between the air flow rate Ga and the engine rotational speed NE and the in-cylinder intake air amount Mc that is sucked (introduced) into the cylinder for one combustion cycle, An in-cylinder intake air amount Mc (k) (= MapMc (Ga, NE)) to be taken into the fuel injection cylinder is obtained.

Here, k is the number of intake strokes (ie, the number of strokes) generated in the internal combustion engine 10 regardless of which cylinder is the intake stroke. Further, the subscript (k) indicates that the in-cylinder intake air amount Mc (k) is a value for the k-th (current) intake stroke (stroke) (hereinafter, the same applies to other physical quantities). ).
The CPU 71 stores (stores) the obtained cylinder intake air amount Mc in the RAM 73 in association with the stroke number k.

  In the following description, a table represented as MapX (a) means a table that defines the relationship between the variable a and the value X. Further, obtaining the value X based on the table MapX (a) means obtaining (determining) the value X based on the current variable a and the table MapX (a). There may be two or more variables.

  Next, the CPU 71 proceeds to step 1115 to determine whether or not the value of the fuel increase flag XRC is “0”. At this time, the value of the fuel increase flag XRC is “0”. Accordingly, the CPU 71 determines “Yes” in step 1115 and proceeds to step 1120 to set the current upstream target air-fuel ratio abyfr (k) to the theoretical air-fuel ratio AFst. The CPU 71 stores (stores) the determined upstream target air-fuel ratio abyfr in the RAM 73 in association with the stroke number k.

  Next, the CPU 71 proceeds to step 1125 to divide the in-cylinder intake air amount Mc (k) determined in step 1110 by the upstream target air-fuel ratio abyfr (k) set in step 1120. The fuel injection amount Fbase (the basic fuel injection amount Fis corresponding to the theoretical air-fuel ratio AFst at this time) is determined. Note that the upstream target air-fuel ratio abyfr (k) used in step 1125 is a value also referred to as the target air-fuel ratio of the internal combustion engine 10 in this specification.

  Then, the CPU 71 proceeds to step 1130 to add the latest upstream FB correction amount DFi obtained in the upstream FB correction amount calculation routine described later to the basic fuel injection amount Fbase determined in step 1125. The fuel injection amount Fi is determined.

Then, the CPU 71 proceeds to step 1135 and injects the fuel with the fuel injection amount Fi determined at step 1130 into the injector 39 so that the fuel is injected by the injector 39 of the fuel injection cylinder for the current combustion cycle. Send an instruction signal.
Thereafter, the CPU 71 proceeds to step 1199 to end the present routine tentatively.

(Upstream FB correction amount calculation)
On the other hand, the CPU 71 repeatedly executes the upstream FB correction amount calculation routine for calculating the upstream FB correction amount DFi shown in the flowchart of FIG. 12 every time a predetermined calculation cycle elapses. The execution of the routine of FIG. 12 corresponds to the achievement of part of the function of the air-fuel ratio feedback control means and part of the function of the upstream feedback correction amount calculation means.

  Therefore, when the predetermined timing comes, the CPU 71 starts processing from step 1200 and proceeds to step 1205 to determine whether or not the value of the upstream FB control execution flag XMFB is “1”. At this time, the value of the upstream FB control execution flag XMFB is set to “1”. Accordingly, the CPU 71 determines “Yes” in step 1205 and proceeds to step 1210 to determine the delay stroke number N based on the table MapN (Mc (k), NE). Here, the number of delay strokes N is the output value Vabyfs of the upstream air-fuel ratio sensor 66 when the fuel injection is instructed, and the air-fuel ratio of the mixed gas when the fuel injected by this instruction is used for combustion. It is the number of strokes corresponding to the delay time until it appears as. The table MapN (Mc (k), NE) is preset based on experimentally measured values.

  Next, the CPU 71 proceeds to step 1215, the table Mapabyfs defining the relationship between the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the air-fuel ratio A / F shown in FIG. 3 and the current upstream air-fuel ratio sensor 66. Output value Vabyfs, the downstream FB correction amount Vafsfb determined in the downstream FB correction amount calculation routine described later, and the downstream FB learning correction amount Gsfb determined in the downstream FB correction amount calculation routine. The control air-fuel ratio abyfs (= Mapabyfs (Vabyfs + Vafsfb + Gsfb)) at the present time is obtained based on the control air-fuel ratio equivalent output value (Vabyfs + Vafsfb + Gsfb).

  Next, the CPU 71 proceeds to step 1220 and of the cylinder intake air amount Mc stored in the RAM 73, the cylinder that is the intake air amount of the cylinder that has reached the intake stroke before N strokes (N intake strokes) from the present time. By dividing the internal intake air amount Mc (kN) by the control air-fuel ratio abyfs obtained in step 1215, the actual in-cylinder supplied fuel amount Fc (kN) N strokes before the present time is obtained.

  Subsequently, the CPU 71 proceeds to step 1225 to set the in-cylinder intake air amount Mc (kN) before the N stroke from the current time to the upstream of the upstream target air-fuel ratio abyfr stored in the RAM 73 before the N stroke from the current time. By dividing by the target air-fuel ratio abyfr (kN), the target in-cylinder supply fuel amount Fcr (kN) N strokes before the present time is obtained.

  Next, the CPU 71 proceeds to step 1230, and from the current in-cylinder fuel supply amount Fcr (kN) N strokes before the current time obtained in step 1225, the actual value before the N strokes from the current time obtained in step 1220. The in-cylinder supplied fuel amount deviation DFc (= Fcr (kN) −Fc (kN)) is obtained by subtracting the in-cylinder supplied fuel amount Fc (kN). That is, the in-cylinder supplied fuel amount deviation DFc is an amount representing an excess or deficiency of the fuel supplied into the cylinder at a time point N strokes before the current time point.

Next, the CPU 71 proceeds to step 1235, in which the following equation (7), the in-cylinder supply fuel amount deviation DFc calculated in the above step 1230, and the in-cylinder supply fuel amount deviation DFc calculated in step 1240 described later. The upstream FB correction amount DFi is obtained based on the time integral value SDFc. That is, the upstream FB correction amount DFi is obtained by performing proportional / integral processing (PI processing) on the cylinder fuel supply amount deviation DFc.
DFi ＝ (Gp ・ DFc + Gi ・ SDFc) ・ KFB (7)

Here, Gp is a preset proportional gain (proportional constant), and Gi is a preset integral gain (integral constant). Each gain Gp and Gi is a positive value. Further, in this example, the coefficient KFB is unchanged and is set to “1” in advance. The coefficient KFB may be variable. In this case, it is preferable that the coefficient KFB is set based on the engine rotational speed NE and the cylinder intake air amount Mc.
The upstream FB correction amount DFi is used when the fuel injection amount Fi is determined in step 1130 of FIG. 11 described above.

Then, the CPU 71 proceeds to step 1240 and calculates the product of the in-cylinder supplied fuel amount deviation DFc and the time step Δt obtained in step 1230 as the time integrated value SDFc of the in-cylinder supplied fuel amount deviation DFc, Is added to the latest time integration value SDFc calculated in this way to calculate the latest value of the time integration value SDFc (updates the time integration value SDFc). Here, the time step Δt is the time from the time when this routine was executed last time to the time when this routine was executed this time.
Thereafter, the CPU 71 proceeds to step 1299 to end the present routine tentatively.

  As described above, the present air-fuel ratio control apparatus matches the target in-cylinder fuel supply amount Fcr (kN) N strokes before the current stroke with the actual in-cylinder fuel supply amount Fc (kN) N strokes before the current time point. As shown, the output value Vabyfs of the upstream air-fuel ratio sensor 66, the downstream FB correction amount Vafsfb (obtained based on the output value Vaxs of the downstream air-fuel ratio sensor 67), and the output value Vaxs of the downstream air-fuel ratio sensor 67 The air-fuel ratio is FB controlled based on the downstream FB learning correction amount Gsfb (obtained based on this). In other words, the air-fuel ratio control apparatus FB-controls the air-fuel ratio so that the current control air-fuel ratio abyfs matches the upstream target air-fuel ratio abyfr (N stroke before the current time).

(Downstream FB correction amount calculation)
Further, the CPU 71 repeatedly executes the downstream FB correction amount calculation routine for calculating the downstream FB correction amount Vafsfb shown in the flowchart of FIG. 13 every time a predetermined calculation cycle elapses. Note that the processing of the routine in FIG. 13 is executed because part of the function of the air-fuel ratio feedback control unit, part of the function of the upstream feedback correction amount calculation unit, function of the downstream feedback correction amount calculation unit, and downstream This corresponds to the achievement of the function of the side feedback learning correction amount calculating means and the function of the time integral value setting means.

  Therefore, when the predetermined timing comes, the CPU 71 starts processing from step 1300 and proceeds to step 1305 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”. At this time, the value of the downstream FB control execution flag XSFB is maintained at “1”. Therefore, the CPU 71 determines “Yes” in step 1305 and proceeds to step 1310 to obtain the output deviation DVoxs by subtracting the current output value Voxs of the downstream air-fuel ratio sensor 67 from the downstream target value Voxsref. .

  In this example, the downstream target value Voxsref is preset to a value (0.5 [V]) corresponding to the theoretical air-fuel ratio. The downstream target value Voxsref is a time average value of the air-fuel ratio of the gas flowing into the first catalyst 53 when the state of the first catalyst 53 becomes a predetermined deterioration state (when the catalyst is deteriorated) and during the warm-up operation. May be set to be slightly richer than the stoichiometric air-fuel ratio. Further, the downstream target value Voxsref may be changed based on the engine speed NE, the throttle valve opening TA, and the like.

  Next, the CPU 71 proceeds to step 1315 and, as shown in the above equation (3), calculates the product of the output deviation DVoxs obtained in step 1310 and the time step Δt as the downstream FB of the output deviation DVoxs. The time integration value (first time integration value) SDVoxs1 is calculated and added to the latest downstream FB time integration value SDVoxs1, which is currently calculated, to calculate the latest value of the downstream FB time integration value SDVoxs1. (Update the time integration value SDVoxs1 for the downstream FB).

  Then, the CPU 71 proceeds to step 1320 and determines whether or not the value of the learning execution flag XG is “1”. At this time, the value of the learning execution flag XG is set to “1”. Therefore, the CPU 71 determines “Yes” in step 1320 and proceeds to step 1325 to obtain the difference between the output deviation DVoxs obtained in step 1310 and the time step Δt as shown in the above equation (5). By adding the product to the learning time integration value SDVoxs2 of the output deviation DVoxs (second time integration value) SDVoxs2, the learning time integration value SDVoxs2 is calculated. The latest value is calculated (the learning time integration value SDVoxs2 is updated).

  Next, the CPU 71 proceeds to step 1330 to calculate the product of the learning time integration value SDVoxs2 obtained in step 1325 and the predetermined learning coefficient Kg at the present time as shown in the above equation (6). The latest value of the downstream FB learning correction amount Gsfb is calculated by adding to the latest downstream FB learning correction amount Gsfb (the downstream FB learning correction amount Gsfb is updated). The downstream FB learning correction amount Gsfb is used when the control air-fuel ratio abyfs is obtained in step 1215 of the upstream FB correction amount calculation routine.

  Then, the CPU 71 proceeds to step 1335, and obtains the product of the learning time integration value SDVoxs2 obtained in step 1325 and the learning coefficient Kg from the downstream side FB time integration value SDVoxs1 obtained in step 1315. By subtracting, the time integration value SDVoxs1 for downstream FB is corrected (updated). That is, the downstream side FB time integration value SDVoxs1 is reduced by the amount newly added to the downstream side FB learning correction amount Gsfb in step 1330.

  Next, the CPU 71 proceeds to step 1340 and subtracts the product of the learning time integral value SDVoxs2 obtained in step 1325 and the learning coefficient Kg from the learning time integral value SDVoxs2 obtained in step 1325. To correct (update) the learning time integration value SDVoxs2. That is, the learning time integration value SDVoxs2 is reduced by the amount newly added to the downstream FB learning correction amount Gsfb in step 1330.

  Next, the CPU 71 proceeds to step 1345 and sets the value of the downstream FB control execution flag XSFB to “1” at the past time point when this routine was executed from the output deviation DVoxs obtained in step 1310. The value obtained by subtracting the past output deviation DVoxs1 set (updated) in step 1355 described later at the latest time point (final time point when the downstream FB control is executed) when this routine is executed in the time step Δt To obtain a time differential value (approximate value) DDVoxs of the output deviation DVoxs. Here, the time step Δt is a time from the time when this routine is executed last time to the time when this routine is executed (that is, the calculation cycle). If the current time is the time immediately after the value of the downstream FB control execution flag XSFB is changed from “0” to “1”, the time differential value DDVoxs of the output deviation may be set to “0”. .

Next, the CPU 71 proceeds to step 1350, and formula (8) below, the output deviation DVoxs calculated in step 1310, the downstream side FB time integration value SDVoxs1 calculated in step 1335, and the above step. A downstream FB correction amount Vafsfb is obtained based on the time differential value DDVoxs of the output deviation calculated in 1345. That is, the downstream FB correction amount Vafsfb is calculated by performing proportional / integral / differential processing (PID processing) on the output deviation DVoxs.
Vafsfb = Kp · DVoxs + Ki · SDVoxs1 + Kd · DDVoxs… (8)

  Here, Kp is a preset proportional gain (proportional constant), Ki is a preset integral gain (integral constant), and Kd is a preset differential gain (differential constant). Each of the gains Kp, Ki, and Kd is a positive value.

  In this way, the present air-fuel ratio control apparatus, based on the output deviation DVoxs that is the difference between the downstream target value Voxsref and the output value Voxs of the downstream air-fuel ratio sensor 67, the downstream FB correction amount Vafsfb and the downstream FB A learning correction amount Gsfb is obtained. The downstream FB correction amount Vafsfb and the downstream FB learning correction amount Gsfb are used when the control air-fuel ratio abyfs is obtained in step 1215 of the upstream FB correction amount calculation routine.

Then, the CPU 71 proceeds to step 1355 to set the past output deviation DVoxs1 to the current output deviation DVoxs obtained in step 1310.
Thereafter, the CPU 71 proceeds to step 1399 to end the present routine tentatively.

  Thus, at the time point before time point t0 shown in FIG. 5 (time point in period A), the control device updates the downstream side FB time integration value SDVoxs1 at step 1315 and at step 1325. Update the learning time integration value SDVoxs2. Further, the control device calculates the downstream FB correction amount Vafsfb based on the downstream FB time integral value SDVoxs1 in step 1350, and in step 1330, the downstream FB correction value Vafsfb based on the learning time integral value SDVoxs2. A learning correction amount Gsfb is calculated. In addition, the control device outputs the downstream side FB learning correction amount Gsfb calculated from the output value Vabyfs of the upstream side air-fuel ratio sensor 66 (that is, the air-fuel ratio upstream of the first catalyst 53) and the upstream target air-fuel ratio abyfr (kN). The upstream FB correction amount DFi is determined based on the calculated downstream FB correction amount Vafsfb, and the fuel injection amount Fi is determined based on the determined upstream FB correction amount DFi, so that the air-fuel ratio is FB controlled. To do. Thereby, the air fuel ratio (downstream air fuel ratio) downstream of the first catalyst 53 substantially matches the air fuel ratio (downstream target air fuel ratio) corresponding to the downstream target value Voxsref. As a result, emission can be improved.

  This period A is a period in which the downstream feedback execution condition is established, and is also referred to as a downstream feedback condition establishment period in this specification. In the period A, the downstream feedback is performed under the condition that the state of the internal combustion engine 10 is in a state (steady state) in which the air-fuel ratio FB control using the downstream FB correction amount Vafsfb is continued for a sufficiently long time. This is a period in which a learning execution condition added to the execution condition is satisfied, and is also referred to as a learning condition establishment period in this specification.

At time t0, when the driver depresses the accelerator pedal 81 and releases the accelerator pedal 81, the accelerator pedal operation amount Accp becomes “0”, so that the throttle valve opening TA is fully closed and opened. Degree (that is, the threshold opening β). Accordingly, when the CPU 71 starts the process of the fuel cut condition determination routine of FIG. 6 at this time t0 and proceeds to step 610, the CPU 71 determines “Yes” at step 610 and proceeds to step 615. The value of the fuel cut flag XFC is set to “1”. That is, the value of the fuel cut flag XFC is changed from “0” to “1”. Note that the determination of “Yes” in step 610 corresponds to the fuel cut execution condition being satisfied.
Then, the CPU 71 proceeds to step 699 to end the fuel cut condition determination routine once.

  Next, when the CPU 71 starts executing the fuel increase condition determination routine of FIG. 7, the CPU 71 determines “No” in both steps 705 and 710 and proceeds to step 799 to temporarily execute the fuel increase condition determination routine. finish.

Further, when the CPU 71 starts executing the upstream FB condition determination routine of FIG. 8, the CPU 71 proceeds to step 810 to determine whether or not the value of the fuel cut flag XFC is “0”. In step 820, the value of the upstream FB control execution flag XMFB is set to “0”. That is, the upstream FB execution condition is not satisfied, and the value of the upstream FB control execution flag XMFB is changed from “1” to “0”.
Then, the CPU 71 proceeds to step 899 to end the upstream FB condition determination routine once.

When the CPU 71 starts executing the downstream FB condition determination routine of FIG. 9, the CPU 71 proceeds to step 915 to determine whether or not the value of the fuel cut flag XFC is “0”. In step 930, the value of the downstream FB control execution flag XSFB is set to “0”. That is, the downstream FB execution condition is not satisfied, and the value of the downstream FB control execution flag XSFB is changed from “1” to “0”.
Then, the CPU 71 proceeds to step 999 to end the downstream FB condition determination routine once.

Furthermore, when the CPU 71 starts executing the learning condition determination routine of FIG. 10, the CPU 71 proceeds to step 1005 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1005, it is determined as “No”, and the process proceeds to step 1020, where the value of the learning execution flag XG is set to “0”. That is, the learning execution condition is not satisfied, and the value of the learning execution flag XG is changed from “1” to “0”.
Then, the CPU 71 proceeds to step 1099 to end the learning condition determination routine once.

  Further, when the CPU 71 starts executing the fuel injection amount control routine of FIG. 11, the CPU 71 proceeds to step 1105 to determine whether or not the value of the fuel cut flag XFC is “0”. "No" is determined, the process proceeds directly to step 1199, and this routine is temporarily terminated. Therefore, at this time t0, fuel is not injected by the injector 39 of the fuel injection cylinder. That is, fuel injection is stopped (fuel cut is executed). Note that the execution of the processing of step 1105 corresponds to the achievement of part of the function of the fuel cut means.

On the other hand, when the CPU 71 starts executing the upstream FB correction amount calculation routine of FIG. 12, the CPU 71 proceeds to step 1205 for determining whether or not the value of the upstream FB control execution flag XMFB is “1”. In step 1205, “No” is determined, the process proceeds to step 1250, and the upstream FB correction amount DFi is set to “0”. Thereby, the upstream FB control is not substantially performed (that is, the downstream FB control is not substantially performed).
Then, the CPU 71 proceeds to step 1299 to end the upstream FB correction amount calculation routine once.

When the CPU 71 starts executing the downstream FB correction amount calculation routine of FIG. 13, the CPU 71 proceeds to step 1305 for determining whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1305, “No” is determined, the process proceeds to step 1360, and the downstream FB correction amount Vafsfb is set to “0”. Thereby, the downstream FB control is not substantially performed.
Then, the CPU 71 proceeds to step 1399 to end the downstream FB correction amount calculation routine once.

  When the calculation cycle elapses, the CPU 71 again starts the process of the fuel cut condition determination routine of FIG. 6 and proceeds to step 605 to determine whether or not the value of the fuel cut flag XFC is “0”. . At this time, the value of the fuel cut flag XFC is set to “1”.

  Therefore, the CPU 71 makes a “No” determination at step 605 to proceed to step 620, where the engine rotational speed NE is equal to or less than a value obtained by subtracting a predetermined positive value α1 from the threshold rotational speed α (α−α1). And at least one of the conditions that the throttle valve opening TA is larger than the threshold opening β is determined.

  According to the above assumption, the engine rotational speed NE is higher than the threshold rotational speed α. At this time, the throttle valve opening TA is equal to the threshold opening β. Therefore, the CPU 71 makes a “No” determination at step 620 to directly proceed to step 699 to end the fuel cut condition determination routine once. That is, the value of the fuel cut flag XFC is maintained at “1”.

  Since the value of the fuel cut flag XFC is maintained at “1”, the same processing as that at the time point t0 described above is executed in the other routines described above. Such a process is repeatedly executed until time t1 when a fuel cut execution condition described later is not satisfied. Therefore, in the fuel cut period (time t0 to time t1) Tfc shown in FIG. 5, the fuel injection is stopped, and as a result, the amount of oxygen stored in the first catalyst 53 (oxygen storage amount) increases. . At time t0a, the oxygen storage amount reaches the maximum value of the amount of oxygen that can be stored by the first catalyst 53 (maximum value of the oxygen storage amount). Accordingly, at a time point after time point t0a, the downstream air-fuel ratio sensor 67 outputs an output value corresponding to a lean air-fuel ratio that is leaner than the stoichiometric air-fuel ratio.

  Further, since the control device does not execute any of Step 1315, Step 1325, and Step 1330, any update of the downstream FB time integration value SDVoxs1, the learning time integration value SDVoxs2, and the downstream FB learning correction amount Gsfb. Can also be stopped. Thereby, the downstream side FB time integration value SDVoxs1 (curve C2 in FIG. 5), the learning time integration value SDVoxs2 (curve C1 in FIG. 5), and the downstream FB learning correction amount Gsfb are maintained constant. Further, since the control device executes step 1360 instead of step 1350, the downstream FB correction amount Vafsfb is maintained at “0”.

  The time point t0 is a time point at which the downstream feedback execution condition is not satisfied and the learning execution condition is not satisfied, and is also referred to as a first time point in the present specification. Further, the fuel cut period Tfc is a period also referred to as a fuel cut condition establishment period in this specification. Further, the fuel cut period Tfc is a period in which the downstream feedback execution condition is not satisfied, and is a part of the downstream feedback condition not satisfied period. Furthermore, the fuel cut period Tfc is a period in which the learning execution condition is not satisfied, and is a part of the learning condition non-satisfied period.

At time t1, when the driver depresses the accelerator pedal 81, the throttle valve opening TA coincides with a predetermined opening TA1 equal to or greater than the threshold opening β. Therefore, when the CPU 71 starts the process of the fuel cut condition determination routine of FIG. 6 and proceeds to step 620 at this time t1, the CPU 71 determines “Yes” in step 620 and proceeds to step 625. The value of the fuel cut flag XFC is set to “0”. That is, the value of the fuel cut flag XFC is changed from “1” to “0”. Note that changing the value of the fuel cut flag XFC from “1” to “0” corresponds to satisfying the fuel increase condition.
Then, the CPU 71 proceeds to step 699 to end the fuel cut condition determination routine once.

Further, when the CPU 71 starts executing the fuel increase condition determination routine of FIG. 7, the CPU 71 determines whether or not the current time is immediately after the value of the fuel cut flag XFC changes from “1” to “0”. When the process proceeds to step 705, it is determined “Yes” in step 705, and the process proceeds to step 715, where the value of the fuel increase flag XRC is set to “1”. That is, the value of the fuel increase flag XRC is changed from “0” to “1”.
Then, the CPU 71 proceeds to step 799 to end the fuel increase condition determination routine once.

When the CPU 71 starts executing the upstream FB condition determination routine of FIG. 8, the CPU 71 proceeds to step 810 to determine whether or not the value of the fuel cut flag XFC is “0”. In step 815, the value of the upstream FB control execution flag XMFB is set to “1”. That is, the value of the upstream FB control execution flag XMFB is changed from “0” to “1”.
Then, the CPU 71 proceeds to step 899 to end the upstream FB condition determination routine once.

  Furthermore, when the CPU 71 starts executing the downstream FB condition determination routine of FIG. 9, the CPU 71 proceeds to step 915 to determine whether or not the value of the fuel cut flag XFC is “0”. The process proceeds to step 920 in which it is determined as “Yes” and it is determined whether or not the value of the fuel increase flag XRC is “0”.

At this time, the value of the fuel increase flag XRC is set to “1”. Accordingly, the CPU 71 makes a “No” determination at step 920 to proceed to step 930 to set the value of the downstream FB control execution flag XSFB to “0”. That is, the value of the downstream FB control execution flag XSFB is maintained at “0”.
Then, the CPU 71 proceeds to step 999 to end the downstream FB condition determination routine once.

Furthermore, when the CPU 71 starts executing the learning condition determination routine of FIG. 10, the CPU 71 proceeds to step 1005 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1005, it is determined as “No”, and the process proceeds to step 1020, where the value of the learning execution flag XG is set to “0”. That is, the value of the learning execution flag XG is maintained at “0”.
Then, the CPU 71 proceeds to step 1099 to end the learning condition determination routine once.

  On the other hand, when the CPU 71 starts executing the fuel injection amount control routine of FIG. 11, the CPU 71 proceeds to step 1105 to determine whether or not the value of the fuel cut flag XFC is “0”. If “Yes” is determined, the process proceeds to step 1110 to determine the in-cylinder intake air amount Mc (k) as described above.

  Next, the CPU 71 proceeds to step 1115 to determine whether or not the value of the fuel increase flag XRC is “0”. At this time, the value of the fuel increase flag XRC is “1”. Accordingly, the CPU 71 makes a “No” determination at step 1115 to proceed to step 1150, where the current upstream target air-fuel ratio abyfr (k) is set to be relatively larger than the stoichiometric air-fuel ratio and to the rich side fuel increasing air-fuel ratio AFrc. Set to. The CPU 71 stores the determined upstream target air-fuel ratio abyfr in the RAM 73 in association with the stroke number k. Note that the execution of the processing of step 1115 and step 1150 corresponds to the achievement of part of the function of the fuel increasing means.

  Then, the CPU 71 proceeds to step 1125 to divide the in-cylinder intake air amount Mc (k) determined in step 1110 by the upstream target air-fuel ratio abyfr (k) set in step 1150. The fuel injection amount Fbase (the basic fuel injection amount Fir corresponding to the fuel increase air-fuel ratio AFrc at this time) is determined.

  Next, the CPU 71 proceeds to the steps after step 1130, and determines the fuel injection amount Fi based on the basic fuel injection amount Fbase and the upstream FB correction amount DFi determined in step 1125. After sending an injection instruction signal to the injector 39 so that fuel of the fuel injection amount Fi is injected, the routine of FIG. 11 is temporarily ended.

When the CPU 71 starts executing the upstream FB correction amount calculation routine of FIG. 12, the CPU 71 proceeds to step 1205 in which it is determined whether or not the value of the upstream FB control execution flag XMFB is “1”. In step 1205, “Yes” is determined, and the process proceeds to step 1210 and the subsequent steps. As described above, the upstream FB correction amount DFi is calculated.
Then, the CPU 71 proceeds to step 1299 to end the upstream FB correction amount calculation routine once.

Further, when the CPU 71 starts executing the downstream FB correction amount calculation routine of FIG. 13, the CPU 71 proceeds to step 1305 for determining whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1305, “No” is determined, the process proceeds to step 1360, and the downstream FB correction amount Vafsfb is set to “0”.
Then, the CPU 71 proceeds to step 1399 to end the downstream FB correction amount calculation routine once.

  When the calculation cycle elapses, the CPU 71 again starts the process of the fuel cut condition determination routine of FIG. 6 and proceeds to step 605 to determine whether or not the value of the fuel cut flag XFC is “0”. . At this time, the value of the fuel cut flag XFC is set to “0”.

  Accordingly, the CPU 71 makes a “Yes” determination at step 605 to proceed to step 610, where the current engine speed NE is higher than the threshold rotational speed α, and the throttle valve opening TA is the threshold opening. It is determined whether or not β or less.

  According to the above assumption, the engine rotational speed NE is higher than the threshold rotational speed α. At this time, the throttle valve opening TA coincides with a predetermined opening TA1 larger than the threshold opening β. Accordingly, the CPU 71 makes a “No” determination at step 610 to directly proceed to step 699 to end the fuel cut condition determination routine once. That is, the value of the fuel cut flag XFC is maintained at “0”.

  Further, when the CPU 71 starts executing the fuel increase condition determination routine of FIG. 7, the CPU 71 determines whether or not the current time is immediately after the value of the fuel cut flag XFC changes from “1” to “0”. When the process proceeds to step 705, it is determined as “No” in step 705 and the process proceeds to step 710, where the current time is the fuel increase time after the value of the fuel increase flag XRC changes from “0” to “1”. It is determined whether or not τrc has elapsed.

  This time is immediately after the value of the fuel increase flag XRC is changed from “0” to “1”. Accordingly, the CPU 71 makes a “No” determination at step 710 to directly proceed to step 799 to end the present routine tentatively.

  Since the value of the fuel cut flag XFC is maintained at “0” and the value of the fuel increase flag XRC is maintained at “1”, the same processing as that at the time point t1 described above is performed in the other routines described above. Is executed. Such a process is repeatedly executed until a later-described time point t2 when a predetermined fuel increase time τrc elapses after the value of the fuel increase flag XRC changes from “0” to “1”. Accordingly, in the fuel increase period (time t1 to time t2) Trc shown in FIG. 5, the downstream FB correction amount Vafsfb is set to “0” in step 1360 of the downstream FB correction amount calculation routine. The upstream FB based on the output value Vabyfs of the upstream air-fuel ratio sensor 66, the upstream target air-fuel ratio abyfr (kN), and the downstream FB learning correction amount Gsfb without being based on the downstream FB correction amount Vafsfb. The correction amount DFi is determined (that is, the downstream FB control is stopped).

  As a result, the air-fuel ratio of the internal combustion engine 10 is controlled to the fuel increase air-fuel ratio AFrc (fuel increase control is performed), and as a result, the oxygen storage amount decreases. Here, the fuel increase time τrc is controlled by controlling the air-fuel ratio of the internal combustion engine 10 to the fuel increase air-fuel ratio AFrc for the same fuel increase time τrc from the time when the oxygen storage amount is maximum, so that the oxygen storage amount is an appropriate amount ( For example, it is set so as to coincide with the half of the maximum oxygen storage amount.

  Further, in the fuel increase period Trc, the control device does not execute any of Step 1315, Step 1325, and Step 1330. Therefore, the downstream side FB time integration value SDVoxs1, the learning time integration value SDVoxs2, and the downstream side FB Any update of the learning correction amount Gsfb is stopped. Thereby, the downstream side FB time integration value SDVoxs1 (curve C2 in FIG. 5), the learning time integration value SDVoxs2 (curve C1 in FIG. 5), and the downstream FB learning correction amount Gsfb are maintained constant.

  The fuel increase period Trc is a part of the downstream feedback condition non-satisfied period or a part of the learning condition non-satisfied period.

Then, at time t2, the CPU 71 starts executing the fuel increase condition determination routine of FIG. 7, and at this time, since the value of the fuel increase flag XRC changes from “0” to “1”, the fuel increase time τrc is When the process proceeds to step 710 where it is determined whether or not the time has elapsed, the CPU 71 determines “Yes” at step 710 and proceeds to step 720 to set the value of the fuel increase flag XRC to “0”. To do. Next, the CPU 71 proceeds to step 725 to set the value of the transition period display flag XKT to “1”. That is, the value of the fuel increase flag XRC is changed from “1” to “0”, and the value of the transition period display flag XKT is changed from “0” to “1”. Note that the elapse of the fuel increase time τrc after the value of the fuel increase flag XRC changes from “0” to “1” corresponds to the fact that the fuel increase condition is not satisfied.
Then, the CPU 71 proceeds to step 799 to end the fuel increase condition determination routine once.

Further, when the CPU 71 starts executing the downstream FB condition determination routine of FIG. 9, the CPU 71 proceeds to step 920 to determine whether or not the value of the fuel increase flag XRC is “0”. The determination is “Yes” and the process proceeds to step 925. Then, the CPU 71 sets the value of the downstream FB control execution flag XSFB to “1” in step 925. That is, the value of the downstream FB control execution flag XSFB is changed from “0” to “1”.
Next, the CPU 71 proceeds to step 999 to end the downstream FB condition determination routine once.

  Furthermore, when the CPU 71 starts executing the learning condition determination routine of FIG. 10, the CPU 71 proceeds to step 1005 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1005, it is determined as “Yes”, and the process proceeds to step 1010 to determine whether or not the value of the transition period display flag XKT is “0”.

  At this time, the value of the transition period display flag XKT is set to “1”. Accordingly, the CPU 71 makes a “No” determination at step 1010 to proceed to step 1025 to determine whether or not the output value Voxs of the downstream air-fuel ratio sensor 67 is larger than the threshold output value Voxsref.

  By the way, when the fuel corresponding to the air-fuel ratio richer than the stoichiometric air-fuel ratio is injected, the oxygen storage amount of the first catalyst 53 decreases with a relatively large time delay from the time when the fuel is injected. . Therefore, when the fuel increase control is executed for the fuel increase time τrc from the time when the oxygen storage amount in the first catalyst 53 is the maximum, the actual oxygen storage amount of the first catalyst 53 is completed. After that, it becomes an appropriate amount with a relatively large time delay.

Accordingly, since the actual oxygen storage amount of the first catalyst 53 is relatively large at this time t2, the output value Voxs of the downstream side air-fuel ratio sensor 67 is equal to the output value corresponding to the air-fuel ratio leaner than the stoichiometric air-fuel ratio. Become. That is, the output value Voxs of the downstream air-fuel ratio sensor 67 is smaller than the threshold output value Voxsref. Accordingly, the CPU 71 makes a “No” determination at step 1025 to proceed to step 1020 to set the value of the learning execution flag XG to “0”. That is, the value of the learning execution flag XG is maintained at “0”.
Then, the CPU 71 proceeds to step 1099 to end the learning condition determination routine once.

  When the CPU 71 starts executing the fuel injection amount control routine of FIG. 11, the CPU 71 proceeds to step 1115 to determine whether or not the value of the fuel increase flag XRC is “0”. Then, the process proceeds to step 1120, and the current upstream target air-fuel ratio abyfr (k) is set to the theoretical air-fuel ratio AFst. The CPU 71 stores the determined upstream target air-fuel ratio abyfr in the RAM 73 in association with the stroke number k.

  Then, the CPU 71 proceeds to Step 1125 and basically divides the in-cylinder intake air amount Mc (k) determined in Step 1110 by the upstream target air-fuel ratio abyfr (k) set in Step 1120. The fuel injection amount Fbase (the basic fuel injection amount Fis corresponding to the theoretical air-fuel ratio AFst at this time) is determined.

  Next, the CPU 71 proceeds to the steps after step 1130, and determines the fuel injection amount Fi based on the basic fuel injection amount Fbase and the upstream FB correction amount DFi determined in step 1125. After sending an injection instruction signal to the injector 39 so that fuel of the fuel injection amount Fi is injected, the routine of FIG. 11 is temporarily ended.

On the other hand, when the CPU 71 starts executing the upstream FB correction amount calculation routine of FIG. 12, the CPU 71 proceeds to step 1205 for determining whether or not the value of the upstream FB control execution flag XMFB is “1”. In step 1205, “Yes” is determined, and the process proceeds to step 1210 and the subsequent steps. As described above, the upstream FB correction amount DFi is calculated.
Then, the CPU 71 proceeds to step 1299 to end the upstream FB correction amount calculation routine once.

  When the CPU 71 starts executing the downstream FB correction amount calculation routine of FIG. 13, the CPU 71 proceeds to step 1305 for determining whether or not the value of the downstream FB control execution flag XSFB is “1”. In step 1305, “Yes” is determined, and the process proceeds to step 1310. As described above, the output deviation DVoxs is calculated.

  Next, the CPU 71 proceeds to step 1315 and, as described above, based on the latest downstream FB time integration value SDVoxs1 calculated at the present time and the output deviation DVoxs calculated in step 1310. The latest value of the downstream side FB time integration value SDVoxs1 is calculated (the downstream side FB time integration value SDVoxs1 is updated).

Then, the CPU 71 proceeds to step 1320 in which it is determined whether or not the value of the learning execution flag XG is “1”. In step 1320, the CPU 71 determines “No” and proceeds to step 1345 and the subsequent steps. Thus, the downstream FB correction amount Vafsfb is calculated.
Then, the CPU 71 proceeds to step 1399 to end the downstream FB correction amount calculation routine once.

  The above processing is repeatedly executed until a later-described time point t3 when the output value Voxs of the downstream air-fuel ratio sensor 67 becomes larger than the threshold output value Voxsref. Therefore, in the transition period (time t2 to time t3) Tk shown in FIG. 5, the downstream side FB time integration value SDVoxs1 is updated based on the output deviation DVoxs in the transition period Tk (curve C2 in FIG. 5). A downstream FB correction amount Vafsfb is calculated based on the updated downstream FB time integration value SDVoxs1. As a result, the downstream FB correction amount Vafsfb during the transition period Tk can be calculated more appropriately than when the output deviation DVoxs during the transition period Tk is not reflected.

  Based on the output value Vabyfs of the upstream air-fuel ratio sensor 66, the upstream target air-fuel ratio abyfr (kN), the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb, the upstream FB correction amount. DFi is determined (ie, both upstream FB control and downstream FB control are performed). As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio in the transition period Tk.

  On the other hand, in the transition period Tk, the control device does not execute Step 1325 and Step 1330. Therefore, the update of the learning time integration value SDVoxs2 and the update of the downstream FB learning correction amount Gsfb are not resumed (they remain stopped). is there). Thereby, the respective values of the learning time integration value SDVoxs2 (curve C1 in FIG. 5) and the downstream FB learning correction amount Gsfb are kept constant.

  The transition period Tk is a part of the downstream feedback condition establishment period or a part of the learning condition establishment period. Furthermore, the transition period (time t2 to time t3) Tk is a period called an overlapping period in which the downstream feedback condition establishment period and the learning condition non-establishment period overlap. Further, the time elapsed from the time point t2 to the time point t3 is a time referred to as a transient time in this specification.

  At time t3, the output value Voxs of the downstream air-fuel ratio sensor 67 exceeds the threshold output value Voxsref. Therefore, when the CPU 71 starts the processing of the learning condition determination routine of FIG. 10 at this time t3 and proceeds to step 1025, the CPU 71 determines “Yes” at step 1025 and proceeds to step 1030. The value of the period display flag XKT is set to “0”. That is, the value of the transition period display flag XKT is changed from “1” to “0”.

Next, the CPU 71 proceeds to step 1015 to set the value of the learning execution flag XG to “1”. That is, the learning execution condition is satisfied, and the value of the learning execution flag XG is changed from “0” to “1”.
Then, the CPU 71 proceeds to step 1099 to end the learning condition determination routine once.

In this state, when the CPU 71 starts executing the downstream FB correction amount calculation routine of FIG. 13, the CPU 71 proceeds to step 1320 to determine whether or not the value of the learning execution flag XG is “1”. In step 1320, “Yes” is determined, and the process proceeds to step 1325 and subsequent steps. As described above, the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb are updated and the downstream FB correction amount Vafsfb is set. calculate.
Then, the CPU 71 proceeds to step 1399 to end the downstream FB correction amount calculation routine once.

When the calculation cycle has elapsed, the CPU 71 starts the process of the learning condition determination routine of FIG. 10 again and proceeds to step 1010 to determine whether or not the value of the transient period display flag XKT is “0”. . At this time, the value of the transition period display flag XKT is set to “0”. Accordingly, the CPU 71 determines “Yes” in step 1010, proceeds to step 1015, and sets the value of the learning execution flag XG to “1”. That is, the value of the learning execution flag XG is maintained at “1”.
Then, the CPU 71 proceeds to step 1099 to end the learning condition determination routine once.

  Since the value of the learning execution flag XG is maintained at “1”, the same processing as that at the time point t3 described above is executed in the other routines described above. Such processing is repeatedly executed thereafter. Therefore, in the period B after the time point t3 shown in FIG. 5, the value of the downstream FB control execution flag XSFB is “1”, and therefore, based on the output deviation DVoxs in the period B as in the transient period Tk. Thus, the downstream FB time integration value SDVoxs1 is updated, and the downstream FB correction amount Vafsfb is calculated based on the updated downstream FB time integration value SDVoxs1. Further, since the value of the learning execution flag XG is “1” in the period B, the learning time integration value SDVoxs2 is updated based on the output deviation DVoxs in the period B, and is updated to the updated learning time integration value SDVoxs2. Based on this, the downstream FB learning correction amount Gsfb is calculated. That is, the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb are updated.

As in the transition period Tk, the output value Vabyfs of the upstream air-fuel ratio sensor 66, the upstream target air-fuel ratio abyfr (kN), the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb Based on this, the upstream FB correction amount DFi is determined (that is, both the upstream FB control and the downstream FB control are performed). As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio.
Further, since the output deviation DVoxs in the fuel cut period Tfc, the fuel increase period Trc and the transient period Tk is not reflected in the learning time integration value SDVoxs2 and the downstream FB learning correction amount Gsfb, the air-fuel ratio is controlled appropriately. be able to.

  The time point t3 is a time point at which the learning execution condition that is no longer satisfied at the time point t0 is satisfied again, and is a time point also referred to as a second time point in this specification. The period B is a part of the downstream feedback condition establishment period or a part of the learning condition establishment period.

  As described above, according to the first embodiment of the air-fuel ratio control apparatus for an internal combustion engine according to the present invention, in the transient period Tk, based on the output deviation DVoxs in the transient period Tk, the time integration value SDVoxs1 ( The first time integration value) is updated, and the downstream FB correction amount Vafsfb is calculated based on the updated downstream FB time integration value SDVoxs1. As a result, the downstream FB correction amount Vafsfb during the transition period Tk can be calculated more appropriately than when the output deviation DVoxs during the transition period Tk is not reflected. As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio in the transition period Tk.

  Furthermore, according to the first embodiment, the downstream FB learning correction amount Gsfb is kept constant during the learning condition non-fulfilling period (fuel cut period Tfc, fuel increase period Trc, and transition period Tk). In addition, at the end time t3 of the learning condition failure period, the learning time integration value SDVoxs2 is a predetermined value unrelated to the output deviation DVoxs in the learning condition failure period (that is, for learning at the start time t0 of the learning condition failure period). Time integration value SDVoxs2).

  As a result, it is possible to avoid the output deviation in the learning condition failure period (particularly, the transition period Tk) being reflected in the downstream FB learning correction amount Gsfb, and therefore the downstream FB learning correction amount Gsfb becomes an inappropriate value. Can be prevented. As a result, a deviation (upstream air-fuel ratio) of the correspondence relationship between the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the actual air-fuel ratio upstream of the first catalyst 53 (upstream air-fuel ratio) is expected. Therefore, the actual air-fuel ratio of the internal combustion engine 10 can be reliably matched with the target air-fuel ratio.

  In the first embodiment, the first value is set as the learning time integral value SDVoxs2 at the second time point (time point t3) and is not related to the output deviation DVoxs during the learning condition failure period. Although the learning time integration value SDVoxs2 at the time point (time point t0) is adopted, a value obtained by multiplying the learning time integration value SDVoxs2 at the first time point (time point t0) by a predetermined coefficient may be adopted.

<Second Embodiment>
Next, an air-fuel ratio control apparatus for an internal combustion engine according to a second embodiment of the present invention will be described. The air-fuel ratio control apparatus according to the second embodiment includes the downstream side FB time integral value SDVoxs1 updated in the transition period Tk, and the learning time integration value SDVoxs2 maintained constant in the learning condition non-satisfied period including the transition period Tk. For the air-fuel ratio control apparatus according to the first embodiment that calculates two time integral values independently of each other, only one time integral value is calculated, the time integral value is updated in a transient period, and a learning condition The only difference is that the time integration value is set to the time integration value at the start of the learning condition non-establishment period at the end of the non-establishment period (in this example, the end of the transition period).

  This control apparatus executes the routine of FIG. 14 instead of the routine of FIG. 13 as a downstream FB correction amount calculation routine for calculating the downstream FB correction amount Vafsfb. Note that the processing of the routine of FIG. 14 is executed because part of the function of the air-fuel ratio feedback control means, part of the function of the upstream feedback correction amount calculation means, function of the downstream feedback correction amount calculation means, and downstream This corresponds to the achievement of the function of the side feedback learning correction amount calculating means, the function of the time integral value updating means, and the function of the time integral value setting means.

  Therefore, when the predetermined timing is reached, the CPU 71 starts processing from step 1400, proceeds to step 1405, and the current time is the time immediately after the value of the learning execution flag XG changes from “0” to “1”. It is determined whether or not.

  As in the first embodiment, when the driver of the internal combustion engine 10 operates the accelerator pedal 81, the accelerator pedal operation amount Accp is maintained constant at the predetermined operation amount Accp1. The description will be continued with reference to FIG. 15 assuming that it is maintained at “0” and then maintained at the predetermined operation amount Accp1 again.

  First, the accelerator pedal operation amount Accp is a predetermined operation before the accelerator pedal operation amount Accp is changed to “0” when the driver of the internal combustion engine 10 releases the accelerator pedal 81 from the accelerator pedal 81. A description will be given from the time point at which the amount Accp1 is maintained constant (time point before time point t0 in FIG. 15). At this time, the values of the upstream FB control execution flag XMFB, the downstream FB control execution flag XSFB, and the learning execution flag XG are continuously set to “1”.

  Therefore, the CPU 71 determines “No” in step 1405 and proceeds to step 1410 to determine whether or not the value of the downstream FB control execution flag XSFB is “1”.

  According to the above assumption, the CPU 71 determines “Yes” in step 1410 and proceeds to step 1415, and similarly to step 1310, the current output value Voxs of the downstream air-fuel ratio sensor 67 from the downstream target value Voxsref. The output deviation DVoxs is obtained by subtracting.

  Next, the CPU 71 proceeds to step 1420 to calculate the product of the output deviation DVoxs obtained in step 1415 and the time step Δt as the time integral value SDVoxs of the output deviation DVoxs at the present time. The latest value of the time integration value SDVoxs is calculated by adding to the latest time integration value SDVoxs (the time integration value SDVoxs is updated). Here, the time step Δt is a time from the time when this routine is executed last time to the time when this routine is executed (that is, the calculation cycle).

  Then, the CPU 71 proceeds to step 1425 to determine whether or not the value of the learning execution flag XG is “1”. At this time, the value of the learning execution flag XG is “1”. Accordingly, the CPU 71 determines “Yes” in step 1425 and proceeds to step 1430 to calculate the product of the time integration value SDVoxs obtained in step 1420 and the predetermined learning coefficient Kg at the present time. The latest value of the downstream FB learning correction amount Gsfb is calculated by adding to the latest downstream FB learning correction amount Gsfb (the downstream FB learning correction amount Gsfb is updated).

  Then, the CPU 71 proceeds to step 1435 to correct (update) the time integration value SDVoxs by subtracting the product of the time integration value SDVoxs and the learning coefficient Kg from the time integration value SDVoxs obtained in step 1420. That is, the time integration value SDVoxs is reduced by the amount newly added to the downstream FB learning correction amount Gsfb in step 1430.

  Next, the CPU 71 proceeds to step 1440 and sets the value of the downstream FB control execution flag XSFB to “1” at the past time point when this routine was executed from the output deviation DVoxs obtained in step 1415. The value obtained by subtracting the past output deviation DVoxs1 set (updated) in step 1450, which will be described later, at the latest time point when the routine is executed in the state where the downstream side FB control is executed (time step Δt). To obtain a time differential value (approximate value) DDVoxs of the output deviation DVoxs. If the current time is the time immediately after the value of the downstream FB control execution flag XSFB is changed from “0” to “1”, the time differential value DDVoxs of the output deviation may be set to “0”. .

Next, the CPU 71 proceeds to step 1445 to calculate the following equation (9), the output deviation DVoxs calculated in step 1415, the time integration value SDVoxs calculated in step 1435, and the step 1440. A downstream FB correction amount Vafsfb is obtained based on the time differential value DDVoxs of the output deviation. That is, the downstream FB correction amount Vafsfb is calculated by performing proportional / integral / differential processing (PID processing) on the output deviation DVoxs.
Vafsfb ＝ Kp ・ DVoxs ＋ Ki ・ SDVoxs ＋ Kd ・ DDVoxs… (9)

  Here, Kp is a preset proportional gain (proportional constant), Ki is a preset integral gain (integral constant), and Kd is a preset differential gain (differential constant). Each of the gains Kp, Ki, and Kd is a positive value.

  In this way, the present air-fuel ratio control apparatus, based on the output deviation DVoxs that is the difference between the downstream target value Voxsref and the output value Voxs of the downstream air-fuel ratio sensor 67, the downstream FB correction amount Vafsfb and the downstream FB A learning correction amount Gsfb is obtained. The downstream FB correction amount Vafsfb and the downstream FB learning correction amount Gsfb are used when the control air-fuel ratio abyfs is obtained in step 1215 of the routine of FIG.

  Then, the CPU 71 proceeds to step 1450 to set the past output deviation DVoxs1 to the current output deviation DVoxs obtained in step 1410.

  Next, the CPU 71 proceeds to step 1455 to determine whether or not the current time is the time immediately after the value of the learning execution flag XG is changed from “1” to “0”. As described above, at this time, the value of the learning execution flag XG is set to “1”. Accordingly, the CPU 71 makes a “No” determination at step 1455 to proceed to step 1499 to end the present routine tentatively.

Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS.
At this time, the value of the upstream FB control execution flag XMFB is set to “1”. Accordingly, the upstream FB correction amount DFi is calculated based on the downstream FB correction amount Vafsfb and the downstream FB learning correction amount Gsfb calculated in the routine of FIG. Further, since both the value of the fuel cut flag XFC and the value of the fuel increase flag XRC are set to “0”, the basic fuel injection amount Fis corresponding to the theoretical air-fuel ratio AFst and the calculated upstream FB correction amount DFi An injection instruction signal is sent to the injector 39 so that the fuel of the fuel injection amount Fi determined based on is injected. That is, both the upstream FB control and the downstream FB control are performed.

  The above processing is repeatedly executed until time t0 when the fuel cut execution condition is satisfied. Therefore, in the period A before the time point t0 shown in FIG. 15, the processing of step 1420, step 1430, and step 1445 is executed, so the time integration value SDVoxs, the downstream FB correction amount Vafsfb, and the downstream FB learning Each value of the correction amount Gsfb is continuously updated.

  At time t0, when the driver depresses the accelerator pedal 81 and releases the accelerator pedal 81, the accelerator pedal operation amount Accp becomes “0”, so that the throttle valve opening TA is fully closed and opened. Degree (that is, the threshold opening β). Therefore, as in the first embodiment, the value of the fuel cut flag XFC is changed to “1” when the CPU 71 executes the routine of FIG.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. As a result, the value of the upstream FB control execution flag XMFB, the value of the downstream FB control execution flag XSFB, and the value of the learning execution flag XG are each changed to “0”. On the other hand, the value of the fuel increase flag XRC is maintained at “0”.

  Therefore, when the CPU 71 starts the processing of the downstream FB correction amount calculation routine of FIG. 14 and determines “No” in step 1405 and proceeds to step 1410, the CPU 71 determines “No” in step 1410. The determination proceeds to step 1460, where the downstream FB correction amount Vafsfb is set to “0”. Thereby, the downstream FB control is not substantially performed.

Next, the CPU 71 proceeds to step 1455 in which it is determined whether or not the current time is immediately after the value of the learning execution flag XG has changed from “1” to “0”. This time point t0 is a time point immediately after the value of the learning execution flag XG is changed from “1” to “0”. Therefore, the CPU 71 determines “Yes” in step 1455 and proceeds to step 1465. In step 1465, the CPU 71 sets the time integral value storage value SDVoxs0 to the current time integration value SDVoxs (stores the current time integration value SDVoxs in the time integration value storage value SDVoxs0). That is, the CPU 71 stores the time integration value SDVoxs at time t0 as the stored value SDVoxs0 when the learning condition is not satisfied (when the value of the learning execution flag XG is changed from “1” to “0”).
Thereafter, the CPU 71 proceeds to step 1499 to end the present routine tentatively.

  In addition, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. At this time point t0, the value of the fuel cut flag XFC is set to “1”. Therefore, as in the first embodiment, even if the CPU 71 executes the routine of FIG. 39 does not inject fuel. That is, fuel injection is stopped (fuel cut is executed).

  The above processing is repeatedly executed until time t1 when the fuel cut execution condition is not satisfied. Therefore, during the fuel cut period (time t0 to time t1) Tfc shown in FIG. 15, the fuel injection is stopped, and as a result, the oxygen storage amount in the first catalyst 53 increases.

  Further, in the fuel cut period Tfc, the value of the downstream FB control execution flag XSFB is “0”, so the processing of step 1420, step 1430, and step 1445 is not executed. Accordingly, the values of the time integration value SDVoxs, the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb are kept constant.

  The time point t0 is a time point at which the downstream feedback execution condition is not satisfied and the learning execution condition is not satisfied, and is also referred to as a first time point in the present specification.

  At time t1, when the driver depresses the accelerator pedal 81, the throttle valve opening TA coincides with a predetermined opening TA1 equal to or greater than the threshold opening β. Accordingly, as in the first embodiment, the CPU 71 executes the routine of FIG. 6 at this time t1, whereby the value of the fuel cut flag XFC is changed to “0”. Thereby, the fuel increase condition is established.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. As a result, the value of the fuel increase flag XRC and the value of the upstream FB control execution flag XMFB are changed to “1”. On the other hand, the value of the downstream FB control execution flag XSFB and the value of the learning execution flag XG are each maintained at “0”.

  In addition, the CPU 71 executes the routine of FIG. At this time t1, since the value of the downstream FB control execution flag XSFB is maintained at “0”, the downstream FB correction amount Vafsfb is set to “0” as in the case of the fuel cut period Tfc described above. The Further, the values of the time integration value SDVoxs, the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb are kept constant.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. At this time t1, since the value of the upstream FB control execution flag XMFB is set to “1”, the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the downstream FB learning correction amount Gsfb (maintained constant from time t0). The upstream FB correction amount DFi is calculated based on the calculated value). As described above, since the downstream FB correction amount Vafsfb is set to “0”, it is not substantially used for calculating the upstream FB correction amount DFi.

  In addition, since the value of the fuel increase flag XRC is set to “1”, it is determined based on the basic fuel injection amount Fir corresponding to the fuel increase air-fuel ratio AFrc and the calculated upstream FB correction amount DFi. An injection instruction signal is sent to the injector 39 so that the fuel injection amount Fi is injected. That is, fuel increase control is performed and only upstream FB control is performed (downstream FB control is not performed).

  The above processing is repeated until time t2 when the fuel increase condition is not satisfied. Accordingly, in the fuel increase period (time t1 to time t2) Trc shown in FIG. 15, the processing of step 1420, step 1430, and step 1445 is not executed as in the case of the fuel cut period Tfc, so the time integration value SDVoxs The values of the downstream FB correction amount Vafsfb and the downstream FB learning correction amount Gsfb are kept constant.

  Next, at time t2, the fuel increase condition is not satisfied when the fuel increase time τrc elapses after the value of the fuel increase flag XRC changes from “0” to “1”. Therefore, as in the first embodiment, the CPU 71 executes the routine of FIG. 7 at this time t2, whereby the value of the fuel increase flag XRC is changed to “0”.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. As a result, the value of the downstream FB control execution flag XSFB is changed to “1”. On the other hand, the value of the upstream FB control execution flag XMFB is maintained at “1”. Further, at the time point t2, the value of the learning execution flag XG is maintained at “0” because the output value Voxs of the downstream air-fuel ratio sensor 67 is equal to or less than the threshold output value Voxsref, as in the first embodiment. The

  In addition, the CPU 71 executes the routine of FIG. At this time t2, when the CPU 71 starts the processing of the downstream FB correction amount calculation routine of FIG. 14 and proceeds to step 1410, the CPU 71 determines “Yes” in step 1410, and performs steps 1415 and 1420. Then, the output deviation DVoxs is calculated and the time integration value SDVoxs is updated based on the calculated output deviation DVoxs.

  Next, the CPU 71 proceeds to step 1425 in which it is determined whether or not the value of the learning execution flag XG is “1”. In step 1425, the CPU 71 determines “No” and proceeds to step 1440 and the subsequent steps. After calculating (updating) the FB correction amount Vafsfb, the routine of FIG. 14 is temporarily ended.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. At this time t2, since the value of the upstream FB control execution flag XMFB is set to “1”, the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the downstream FB learning correction amount Gsfb (maintained constant from time t0). 14) and the downstream FB correction amount Vafsfb calculated in the routine of FIG. 14 above, the upstream FB correction amount DFi is calculated.

  In addition, since the value of the fuel cut flag XFC and the value of the fuel increase flag XRC are both set to “0”, the basic fuel injection amount Fis corresponding to the theoretical air-fuel ratio AFst and the calculated upstream FB correction amount DFi An injection instruction signal is sent to the injector 39 so that fuel of the fuel injection amount Fi determined based on the above is injected.

  The above process is repeated until time t3. Accordingly, in the transition period (time t2 to time t3) Tk shown in FIG. 15, the processing of step 1420 and step 1445 is executed, so that the time integration value SDVoxs is updated based on the output deviation DVoxs in the transition period Tk. Then, the downstream FB correction amount Vafsfb is calculated based on the updated time integration value SDVoxs (see curve C3 in FIG. 15). As a result, the downstream FB correction amount Vafsfb during the transition period Tk can be calculated more appropriately than when the output deviation DVoxs during the transition period Tk is not reflected.

  On the other hand, in the transition period Tk, the processing of step 1430 is not executed, so the value of the downstream FB learning correction amount Gsfb is kept constant.

  As in the period A described above, the output value Vabyfs of the upstream air-fuel ratio sensor 66, the upstream target air-fuel ratio abyfr (kN), the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb , The upstream FB correction amount DFi is determined (that is, both the upstream FB control and the downstream FB control are performed). As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio in the transition period Tk.

  At time t3, the output value Voxs of the downstream air-fuel ratio sensor 67 exceeds the threshold output value Voxsref as in the first embodiment. Therefore, as in the first embodiment, the CPU 71 executes the routine of FIG. 10 at this time t3, whereby the value of the learning execution flag XG is changed to “1”.

  In addition, the CPU 71 executes the routine of FIG. At this time t3, when the CPU 71 starts processing of the downstream FB correction amount calculation routine of FIG. 14 and proceeds to step 1405, the CPU 71 determines “Yes” in step 1405 and proceeds to step 1470.

  Then, the CPU 71 sets the time integration value SDVoxs to the storage value SDVoxs0 of the time integration value set in step 1465 at time t0. Thus, the time integration value SDVoxs is set to the time integration value SDVoxs (= SDVoxs0) at the start time t0 of the learning condition non-satisfied period, which is a value unrelated to the output deviation DVoxs in the learning condition non-satisfied period (time t0 to time t3). Is done.

  Next, the CPU 71 proceeds to step 1410, determines “Yes” in step 1410, proceeds to steps 1415 and 1420, calculates the output deviation DVoxs, and calculates the time integrated value SDVoxs. Update based on DVoxs.

Next, the CPU 71 proceeds to step 1425 in which it is determined whether or not the value of the learning execution flag XG is “1”. In step 1425, the CPU 71 determines “Yes” and proceeds to steps 1430 and 1435. Then, the downstream side FB learning correction amount Gsfb is updated and the time integration value SDVoxs is corrected. That is, the update of the downstream FB learning correction amount Gsfb is resumed.
Thereafter, the CPU 71 proceeds to the steps after step 1440, calculates the downstream FB correction amount Vafsfb, and then ends the routine of FIG.

  Further, as in the first embodiment, the CPU 71 executes the processes of the routines shown in FIGS. At this time t3, since the value of the upstream FB control execution flag XMFB is set to “1”, the output value Vabyfs of the upstream air-fuel ratio sensor 66 and the downstream FB learning updated in the routine of FIG. An upstream FB correction amount DFi is calculated based on the correction amount Gsfb and the downstream FB correction amount Vafsfb.

  In addition, since the value of the fuel cut flag XFC and the value of the fuel increase flag XRC are both set to “0”, the basic fuel injection amount Fis corresponding to the theoretical air-fuel ratio AFst and the calculated upstream FB correction amount DFi An injection instruction signal is sent to the injector 39 so that fuel of the fuel injection amount Fi determined based on the above is injected. That is, both the upstream FB control and the downstream FB control are performed.

  The time point t3 is a time point at which the learning execution condition that is no longer satisfied at the time point t0 is satisfied again, and is a time point that is also referred to as a second time point in this specification.

  In this way, in the period B after the time point t3 shown in FIG. 15, the processing in step 1420 and step 1445 is executed as in the transient period Tk, and therefore the time based on the output deviation DVoxs in the period B The integral value SDVoxs is updated, and the downstream FB correction amount Vafsfb is calculated based on the updated time integral value SDVoxs. Further, in the period B, unlike the case in the transition period Tk, the processing of step 1430 is also executed, so the downstream FB learning correction amount Gsfb is updated (calculated) based on the updated time integration value SDVoxs.

As in the transition period Tk, the output value Vabyfs of the upstream air-fuel ratio sensor 66, the upstream target air-fuel ratio abyfr (kN), the downstream FB correction amount Vafsfb, and the downstream FB learning correction amount Gsfb Based on this, the upstream FB correction amount DFi is determined. As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio.
Furthermore, since the time integration value SDVoxs and the downstream FB learning correction amount Gsfb do not reflect the output deviation DVoxs in the learning condition non-satisfactory period (fuel cut period Tfc, fuel increase period Trc and transient period Tk), the downstream FB The learning correction amount Gsfb is updated appropriately. As a result, the air-fuel ratio can be controlled appropriately.

  As described above, according to the second embodiment of the air-fuel ratio control apparatus for an internal combustion engine according to the present invention, in the transition period Tk, based on the output deviation DVoxs in the transition period Tk, as in the first embodiment. The time integration value SDVoxs is updated, and the downstream FB correction amount Vafsfb is calculated based on the updated time integration value SDVoxs. As a result, the downstream FB correction amount Vafsfb during the transition period Tk can be calculated more appropriately than when the output deviation DVoxs during the transition period Tk is not reflected. As a result, the downstream air-fuel ratio can be quickly matched with the downstream target air-fuel ratio in the transition period Tk.

  Furthermore, according to the second embodiment, the downstream FB learning correction amount Gsfb is kept constant during the learning condition non-fulfilling period (the fuel cut period Tfc, the fuel increase period Trc, and the transition period Tk). In addition, at the end time t3 of the learning condition failure period, the time integration value SDVoxs is a predetermined value unrelated to the output deviation DVoxs in the learning condition failure period (that is, the time integration value at the start time t0 of the learning condition failure period) SDVoxs). As a result, it is possible to avoid the output deviation in the learning condition failure period (particularly, the transition period Tk) being reflected in the downstream FB learning correction amount Gsfb, and therefore the downstream FB learning correction amount Gsfb becomes an inappropriate value. Can be prevented. As a result, the output deviation of the upstream air-fuel ratio sensor 66 can be properly compensated, and the actual air-fuel ratio of the internal combustion engine 10 can be made to coincide with the target air-fuel ratio with certainty.

  Note that, in the second embodiment, the first time point (a value that is set to the time integral value SDVoxs at the second time point (time point t3) and is irrelevant to the output deviation DVoxs in the learning condition non-satisfied period is used. Although the time integration value SDVoxs at the time point t0) is employed, a value obtained by multiplying the time integration value SDVoxs at the first time point (time point t0) by a predetermined coefficient may be employed. In the second embodiment, the time integration value SDVoxs at the first time point (time point t0) is stored as the storage value SDVoxs0 of the time integration value set to the time integration value SDVoxs at the second time point (time point t3). However, since the time integration value SDVoxs is not updated during the period from the time point t0 to the time point t2, it is configured to store the time integration value SDVoxs at a predetermined time point during the period from the time point t0 to the time point t2. May be.

The present invention is not limited to the above embodiments, and various modifications can be employed within the scope of the present invention. For example, the timing for executing the routine for calculating the upstream FB correction amount (FIG. 12) and the routine for calculating the downstream FB correction amount (FIGS. 13 and 14) is the timing at which fuel injection starts for the fuel injection cylinder. It may be the timing when the time (fuel injection start time) arrives.
In each of the above embodiments, the downstream FB correction amount Vafsfb is composed of a proportional term, an integral term, and a derivative term, but may be composed of only a proportional term and an integral term.

  In addition, in each of the above embodiments, the value of the upstream FB control execution flag XMFB is changed to “1” at the start time t1 of the fuel increase period Trc (that is, the end time of the fuel cut period Tfc) (FIG. 8). In step 815), the upstream side FB control is resumed, but the upstream side air-fuel ratio sensor is supplied by supplying the internal combustion engine 10 with the mixed gas having the fuel-increasing air-fuel ratio AFrc in the fuel-increasing period Trc. When the output value Vabyfs of 66 becomes smaller than a predetermined threshold (for example, the theoretical air-fuel ratio corresponding output value Vst corresponding to the theoretical air-fuel ratio), the value of the upstream FB control execution flag XMFB is changed to “1”. May be configured to resume the upstream FB control. Further, the transition period Tk may be ended when a predetermined time has elapsed since the end of the fuel increase period Trc.

  Further, the transition period in which each of the above embodiments is applied is not limited to the transition period Tk after the fuel increase period Trc following the fuel cut period Tfc, and the target air-fuel ratio of the internal combustion engine 10 is compared with the theoretical air-fuel ratio. It may be a transient period from when the stoichiometric air-fuel ratio is set to when a predetermined transient time elapses after the air-fuel ratio is set to be significantly different.

Presence / absence of execution of upstream FB control and downstream FB control, presence / absence of update of downstream FB learning correction amount, fuel injection amount, output value of downstream air-fuel ratio sensor, time integral value of output deviation, and downstream FB learning correction It is the time chart which showed the change with respect to time of quantity. 1 is a schematic configuration diagram of a system in which an air-fuel ratio control apparatus for an internal combustion engine according to a first embodiment of the present invention is applied to an internal combustion engine. 3 is a graph showing the relationship between the output value of the upstream air-fuel ratio sensor shown in FIG. 2 and the air-fuel ratio. 3 is a graph showing the relationship between the output value of the downstream air-fuel ratio sensor shown in FIG. 2 and the air-fuel ratio. It is a time chart which shows the change with respect to time of various flags, fuel injection amount, the output value of an upstream air-fuel ratio sensor, the output value of a downstream air-fuel ratio sensor, and the output deviation. FIG. 3 is a flowchart showing a routine for determining whether or not to execute a fuel cut executed by a CPU shown in FIG. 2. FIG. FIG. 3 is a flowchart showing a routine for determining whether or not to execute fuel increase control executed by a CPU shown in FIG. 2. FIG. 3 is a flowchart showing a routine for determining whether or not to execute upstream FB control executed by a CPU shown in FIG. 2. 3 is a flowchart showing a routine for determining whether or not to execute downstream FB control executed by a CPU shown in FIG. 2. FIG. 3 is a flowchart showing a routine for determining whether or not to update a downstream FB learning correction amount executed by the CPU shown in FIG. 2. FIG. 3 is a flowchart showing a routine for calculating a fuel injection amount and instructing fuel injection executed by a CPU shown in FIG. 2. 3 is a flowchart showing a routine for calculating an upstream FB correction amount executed by a CPU shown in FIG. 2. 3 is a flowchart showing a routine for calculating a downstream FB correction amount executed by a CPU shown in FIG. 2. 14 is a flowchart showing a routine that is executed in place of the routine shown in FIG. 13 by the CPU of the air-fuel ratio control apparatus for an internal combustion engine according to the second embodiment of the present invention in order to calculate the downstream FB correction amount. It is a time chart which shows the change with respect to time of various flags, fuel injection amount, the output value of an upstream air-fuel ratio sensor, the output value of a downstream air-fuel ratio sensor, and the output deviation.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Internal combustion engine, 25 ... Combustion chamber, 39 ... Injector, 43 ... Throttle valve, 52 ... Exhaust pipe, 53 ... Three way catalyst (1st catalyst), 66 ... Upstream air-fuel ratio sensor, 67 ... Downstream air-fuel ratio sensor , 68 ... accelerator opening sensor, 70 ... electric control device, 71 ... CPU, 73 ... RAM, 81 ... accelerator pedal.

Claims (3)

Applied to an internal combustion engine having an exhaust gas purification catalyst disposed in an exhaust passage;
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
An upstream feedback correction amount is calculated based on the output value of the upstream air-fuel ratio sensor and the output value of the downstream air-fuel ratio sensor, and is supplied to the internal combustion engine based on the calculated upstream feedback correction amount. Air-fuel ratio feedback control means for controlling the air-fuel ratio of the mixed gas,
In an air-fuel ratio control apparatus for an internal combustion engine comprising:
The air-fuel ratio feedback control means includes
In a downstream feedback condition satisfaction period in which a predetermined downstream feedback execution condition is satisfied, a first time-integrated output deviation which is a difference between the output value of the downstream air-fuel ratio sensor and a predetermined downstream target value The time integral value is updated based on the first time integral value obtained at the current time and the current output deviation, and the downstream feedback correction amount is calculated based on the updated first time integral value. Downstream feedback correction amount calculating means for calculating,
The second time integration value obtained by time-integrating the output deviation in the learning condition satisfaction period in which a predetermined learning execution condition is further established by further adding a condition to the downstream feedback execution condition. Updating based on the second time integral value and the current output deviation, calculating the downstream feedback learning correction amount based on the updated second time integral value, and satisfying the learning execution condition A downstream feedback learning correction amount calculating means for maintaining the downstream feedback learning correction amount constant during the learning condition not established period,
When the learning execution condition is not satisfied at the first time point and is satisfied at the subsequent second time point, the second time integral value at the second time point is calculated as the output deviation from the first time point to the second time point. Is a time integral value setting means for setting to an unrelated predetermined value;
In the downstream feedback condition establishment period, the upstream feedback correction amount is calculated based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor, and the downstream In the downstream feedback condition non-fulfilling period in which the side feedback execution condition is not satisfied, the upstream side is based on the downstream feedback learning correction amount and the output value of the upstream air-fuel ratio sensor without being based on the downstream feedback correction amount. Upstream feedback correction amount calculating means for calculating the feedback correction amount;
An air-fuel ratio control apparatus for an internal combustion engine comprising: Applied to an internal combustion engine having an exhaust gas purification catalyst disposed in an exhaust passage;
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
An upstream feedback correction amount is calculated based on the output value of the upstream air-fuel ratio sensor and the output value of the downstream air-fuel ratio sensor, and is supplied to the internal combustion engine based on the calculated upstream feedback correction amount. Air-fuel ratio feedback control means for controlling the air-fuel ratio of the mixed gas,
In an air-fuel ratio control apparatus for an internal combustion engine comprising:
The air-fuel ratio feedback control means includes
The time integral value obtained by time-integrating the output deviation, which is the difference between the output value of the downstream air-fuel ratio sensor and the predetermined downstream target value, is set to the same time integral value currently obtained and the current same output deviation. A time integral value updating means for updating based on;
A downstream feedback correction amount calculating means for calculating a downstream feedback correction amount based on the time integral value in a downstream feedback condition satisfaction period in which a predetermined downstream feedback execution condition is satisfied;
A downstream feedback learning correction amount is calculated on the basis of the time integral value in a learning condition establishment period in which a predetermined learning execution condition in which a condition is further added to the downstream feedback execution condition is satisfied, and the learning execution is performed A downstream feedback learning correction amount calculating means for maintaining the downstream feedback learning correction amount constant during a learning condition non-satisfied period in which the condition is not satisfied;
When the learning execution condition is not satisfied at the first time point and is satisfied at the subsequent second time point, the time integration value at the second time point is irrelevant to the output deviation from the first time point to the second time point. Time integral value setting means for setting to a predetermined value;
In the downstream feedback condition establishment period, the upstream feedback correction amount is calculated based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor, and the downstream In the downstream feedback condition non-fulfilling period in which the side feedback execution condition is not satisfied, the upstream side is based on the downstream feedback learning correction amount and the output value of the upstream air-fuel ratio sensor without being based on the downstream feedback correction amount. Upstream feedback correction amount calculating means for calculating the feedback correction amount;
An air-fuel ratio control apparatus for an internal combustion engine comprising: Applied to an internal combustion engine having an exhaust gas purification catalyst disposed in an exhaust passage;
Fuel injection means for injecting fuel supplied to the internal combustion engine;
An upstream air-fuel ratio sensor that is disposed in the exhaust passage upstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A downstream air-fuel ratio sensor that is disposed in the exhaust passage downstream of the exhaust purification catalyst and outputs an output value corresponding to the air-fuel ratio of the gas that actually flows through the exhaust passage;
A fuel injection amount control device for an internal combustion engine that controls a fuel injection amount that is an amount of fuel injected by the fuel injection means,
Fuel cut means for stopping fuel injection by the fuel injection means during a fuel cut condition establishment period, which is a period in which a predetermined fuel cut execution condition is established;
A predetermined fuel whose air-fuel ratio of the mixed gas supplied to the internal combustion engine is richer than the stoichiometric air-fuel ratio during a fuel increase period from when the fuel cut condition is satisfied until a predetermined fuel increase time elapses. Fuel increasing means for controlling the fuel injection amount so as to achieve an increasing air-fuel ratio;
The output value of the downstream air-fuel ratio sensor during a downstream feedback condition satisfaction period, which is a period during which a predetermined downstream feedback execution condition is satisfied among the fuel cut condition satisfaction period and the period not including the fuel increase period. And a downstream feedback correction amount calculating means for calculating a downstream feedback correction amount based on an output deviation that is a difference between a predetermined downstream target value corresponding to the stoichiometric air-fuel ratio,
Based on the output deviation during the learning condition satisfaction period, which is a period that does not include a transient period from the end of the fuel increase period in the downstream feedback condition satisfaction period until a predetermined transient time elapses. A downstream feedback learning correction amount calculating means for calculating a feedback learning correction amount and maintaining the downstream feedback learning correction amount constant during a learning condition non-establishment period other than the learning condition establishment period;
During the downstream feedback condition establishment period, the upstream feedback correction amount is calculated based on the downstream feedback correction amount, the downstream feedback learning correction amount, and the output value of the upstream air-fuel ratio sensor. The fuel injection amount is controlled based on the side feedback correction amount, and is a period within the downstream feedback condition non-satisfied period in which the downstream feedback execution condition is not satisfied, and does not include at least the fuel cut condition satisfaction period During the period, an upstream feedback correction amount is calculated based on the downstream feedback learning correction amount and the output value of the upstream air-fuel ratio sensor, and the fuel injection amount is controlled based on the calculated upstream feedback correction amount. Fuel injection amount feedback control means,
A fuel injection amount control device for an internal combustion engine.

Images