JP2003111490A - Inverter control method and device - Google Patents

Abstract

(57) [Summary] In a direct torque control method, by switching a voltage vector during a control operation time of a microcomputer, torque ripple and magnetic flux ripple are minimized, an inverter maximum output voltage is improved, and inverter switching is performed. Achieve higher frequency. SOLUTION: DC voltage V _dc (n), rotational angular velocity ω
_e (n), the phase angle φ (n + 1), φ0 (n + 1), the area number, the judgment output of the torque deviation calculating unit 45 and the judgment output of the magnetic flux comparing unit 48 are input, and the voltage vector output time is calculated. It has a voltage vector output time calculation section 51 that is set in the PWM timer section 42 as a timer value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inverter control method and an apparatus therefor, and more specifically, it detects line current and terminal voltage of an AC motor connected to an inverter, and detects torque and stator from these detected values. The present invention relates to an inverter control method and apparatus for calculating a magnetic flux, selecting a voltage vector based on these calculation results and a corresponding command value, and switching controlling each transistor of the inverter.

[0002]

2. Description of the Related Art Brushless DC motors having permanent magnets mounted on a rotor have higher efficiency than induction motors and are widely used in applications where energy-saving operation is desired first.

In order to control the torque, the brushless DC motor needs voltage and current control synchronized with the rotational position.

However, when a system in which a rotary position sensor is mounted on a brushless DC motor and the position signal is used to control an inverter is used, the cost of the entire system increases due to the mounting of the rotary position sensor. With
To become larger. Further, the reliability is lowered due to the disconnection of the sensor signal line. Further, in an application in which the brushless DC motor and the controller are installed separately, there is a possibility that the position sensor information cannot be accurately transmitted to the controller.

Therefore, in recent years, in order to deal with these problems, research on a position sensorless brushless DC motor in which a rotational position sensor is omitted has been advanced.

The sensorless control methods for brushless DC motors, which have been studied so far, are roughly classified into three methods: a back electromotive voltage zero-cross detection method, a motor model utilization method, and a direct torque control method. (1) The back electromotive voltage zero-cross detection method is a method in which the speed electromotive force induced by the rotor magnet magnetic flux is detected from the motor terminal voltage, and a position signal is obtained to perform synchronous control of voltage, current and rotational position.

More specifically, there are a method of detecting the zero cross of the speed electromotive force of the non-excited phase by controlling the inverter conduction by 120 °, a method of detecting the zero cross of the third harmonic component contained in the speed electromotive force, and the like. It can be illustrated. The position detection resolution is 60 ° in electrical angle in any method. (2) The motor model utilization method detects the motor terminal voltage and current, and configures a position estimator and a detector from these detected amounts and the motor model, obtains a position signal, and outputs the voltage, current and rotational position. This is a method of performing synchronous control of.

Specifically, in the case of a brushless DC motor in which a permanent magnet is embedded inside a rotor, a rotational position is determined by a method of constructing a position estimator based on modern control theory such as an observer or a model reference adaptive system. An example is a method of detecting a change in inductance due to the above-described harmonic ripple amplitude of the line current. The position detection resolution is several degrees or less in electrical angle in any method. (3) In the direct torque control method, the motor terminal voltage and current are detected, the motor stator magnetic flux and the torque are calculated from these detected amounts, and the inverter is directly switching-controlled so that the calculation result follows the command value. It is a method to do. Since this method controls the torque in principle, no position detection process is involved.

Therefore, in order to control the position and speed, it is necessary to separately combine these detectors and estimators.

[0010]

When the back electromotive voltage zero-cross detection method is adopted, there is a disadvantage that the inverter energization waveform and the motor electromagnetic structure are restricted, and the position detection resolution is low, so that high-speed torque is obtained. The disadvantage is that no response can be achieved.

When the motor model utilizing method is adopted, a position detection error occurs when the motor device constants such as the inductance and the electromotive force constant used in the motor model change due to the influence of temperature and magnetic saturation, and in the worst case, In addition to the inconvenience that the brushless DC motor becomes unstable, there is the inconvenience that the torque response speed is restricted by the convergence time of the position estimator.

On the other hand, in the direct torque control method, the torque is calculated and the inverter is directly controlled so as to follow the command value, so that a high-speed torque response can be stably obtained. Further, since the motor parameter required for the calculation is only the winding resistance, the control performance does not deteriorate due to magnetic saturation. In order to reflect the change in the motor device constant and improve the control calculation accuracy, only the temperature detector or estimator of the winding resistance needs to be added, and the controllability can be easily improved.

Therefore, the direct torque control method has attracted attention as the best method for controlling the torque of a brushless DC motor at high speed without using a rotational position sensor.
This system can be used for controlling an induction motor or a reluctance motor in addition to the brushless DC motor.

A switching control method of the inverter of this system, that is, a PWM control method is disclosed in Japanese Patent Laid-Open No. 2001-8.
As shown in Japanese Patent No. 6795, an instantaneous value comparison method is adopted. This instantaneous value comparison method has no problem when the control circuit is composed of analog arithmetic units.

However, the cost and size of the control circuit can be reduced.
In microcomputer control adopted in home appliances and general industrial equipment applications where simplification of adjustment is strongly desired, there are inconveniences such as an increase in torque ripple and a decrease in DC voltage utilization rate due to constraints such as calculation time and control cycle.

Further description will be made.

The conventional direct torque control method will be described below.

For convenience, the three-phase (u-v-w) is converted into the two-phase (α-β) by the equation (1).

[0019]

[Equation 1]

When the coordinate transformation of equation 1 is performed, the phase relationship between the three-phase coordinates and the two-phase coordinates and the motor model captured on the two-phase coordinates are as shown in (a) and (b) in FIG. Becomes

In FIG. 9, i _α and i _β are winding currents converted into two phases, I _f is a constant current source simulating a permanent magnet incorporated in the rotor, and θ _m is a position angle (electricity) of the motor rotor. Angle).

Since the motor torque τ can be detected as a stator reaction force, the stator magnetic fluxes λ _α and λ _β and the winding current i
It can be obtained by the outer product calculation with _{α 1} and i _β, and can be expressed as Formula 2. Note that p is the number of motor pole pairs.

[0023]

[Equation 2]

On the other hand, the stator magnetic fluxes λ _α and λ _β are the winding resistance R
This can be calculated by integrating the voltage obtained by subtracting the voltage drop at the terminal voltage from the terminal voltage, and is given by Equation 3.

[0025]

[Equation 3]

The magnitude | λ | of the combined magnetic flux of the α-β stator magnetic flux vector is given by The α-β stator magnetic flux vector is referred to as a rotating magnetic field, and the magnitude of the combined magnetic flux of the α-β stator magnetic flux vector is referred to as a rotating magnetic field.

[0027]

[Equation 4]

Next, regarding the inverter switching control method for making the motor torque τ calculated from the detected values of the terminal voltage and the winding current and the magnitude | λ | of the rotating magnetic field follow the command value, the voltage type inverter circuit of FIG. Will be described with reference to.

This voltage type inverter circuit is provided with a DC power source V
Three upper arm transistors T _u ⁺ , T _v ⁺ , T _w ⁺ whose collector terminals are connected to the positive electrode side of _dc , and three lower arm transistors T _u ⁻ , T _v ⁻ whose emitter terminals are connected to the negative electrode side,
T _w ⁻ , and its cathode terminal and anode terminal are composed of six freewheeling diodes D _u ^{+ to} D _w ⁻ connected to the collector terminal and emitter terminal of the transistor, respectively. Also,
The emitter terminal of the upper arm transistor and the collector terminal of the lower arm transistor are connected, and the three connection points are connected to the terminals of the brushless DC motor.

In the voltage source inverter, the upper and lower arm transistors are exclusively turned on and off so that the short circuit of the DC power supply V _dc does not occur due to the circuit configuration, so that the switching states that can be taken are eight as shown in Table 1. In Table 1, V0 to V7 are voltage vectors defined to indicate the ON state of each phase transistor.

[0031]

[Table 1]

FIG. 11 shows the voltage vector captured on the two-phase coordinates.

Voltage vectors V0 and V7 (hereinafter, these are referred to as zero vectors) in which the two-phase voltages v _α and v _β are both 0, and six types of voltage vectors V1 to V in which the voltage is not 0
Consider the method of controlling the magnetic flux and the torque by using 6.

Since the magnetic flux is given by the time product of the voltage, for example, when the voltage vectors V1 to V6 are output for a predetermined time, the magnetic flux changes on the α-β coordinate in the direction indicated by the arrow in FIG. Therefore, as shown in FIG. 12, the α-β coordinate is divided into regions I to VI every 60 °, and the non-zero voltage vector output in each region is restricted as shown in FIG. For example, in the region I, only the voltage vector V4 and the voltage vector V6 are restricted.

By alternately selecting the voltage vector V4 and the voltage vector V6 in this region I, the rotating magnetic field is advanced in the clockwise direction, and the magnitude of the rotating magnetic field is increased by the voltage vector V4 output and decreased by the voltage vector V6 output. Can be made.

Therefore, the magnitude of the rotating magnetic field can be controlled by comparing the magnitude of the rotating magnetic field obtained by the equation 4 with the command value and switching to an appropriate voltage vector in each region depending on the magnitude.

Further, the magnitudes of the voltage vectors V1 to V6 are such that the DC voltage _Vdc is constant, and therefore the rotation speed of the rotating magnetic field cannot be controlled only by switching these voltage vectors. However, the rotation speed of the rotating magnetic field can be controlled by using the zero vector that does not change the magnetic flux, that is, can stagnant the trajectory of the magnetic flux. That is, the torque calculated by the equation 2 is compared with the torque command value, and the voltage vector and the zero vector may be switched depending on the magnitude. Thereby, the phase difference between the reference line passing through the center of the rotor fixed in a specific direction of the rotor and the rotating magnetic field can be maintained at a value corresponding to the desired torque, and the rotation speeds of both can be equalized. .

FIG. 13 shows a state in which torque is compared by a hysteresis comparator and the inverter is switching-controlled in response to the output thereof [magnetic flux locus {see (a) in FIG. 13}, and voltage vector and torque of region II. Change {see (b) in FIG. 13}].

FIG. 14 is a block diagram of a direct torque control system for obtaining the control performance of FIG. 13 {see (a) in FIG. 14}, characteristics of a hysteresis comparator for instantaneous magnetic flux comparison {see (b) in FIG. 14}, And characteristics of hysteresis comparator that compares torque instantaneously ((c) in FIG. 14)
Reference}.

In this direct torque control system, the magnetic flux and the torque are instantaneously calculated by an arithmetic unit composed of an integrating circuit using an OP amplifier and an analog multiplier.

Table 2 shows a switching pattern table in which a switching pattern (voltage vector) to be selected in response to the comparator output of the magnetic flux and the torque is set, and Table 3 shows the αβ magnetic flux for judging the regions I to VI. And the magnitude of the magnetic flux. Note that Table 2 is usually stored in ROM and is looked up by the output of the comparator.

[0042]

[Table 2]

[0043]

[Table 3]

Here, FIG. 14 shows a torque comparator.
The ternary method shown in the middle (c) is adopted when Sτ = −
This is because the reverse rotation (that is, the reverse torque) is controlled by selecting the voltage vector in which the rotating magnetic field rotates in the counterclockwise direction at 1. Therefore, in applications where rapid deceleration control is unnecessary in one rotation direction, for example, applications such as compressors and pumps, instantaneous comparison of torque is performed by a binary comparator and Sτ =
The control system can be simplified by omitting the -1 part.

Next, consider the operation when the system shown in FIG. 14A is replaced with microcomputer control.

The frequency of calculation of the torque and the magnetic flux is restricted by the arithmetic processing speed of the microcomputer, and there is a time delay until the calculation is completed before the switching pattern is determined by comparison with the command value. Therefore, the method of reducing the influence of the calculation delay to one sample by the torque prediction calculation is "High-performance torque control of induction motor using digital signal processor", Ichiro Miyashita et al., IEEJ Transactions, Vol. 107, 2
No., 1987.

FIG. 15 shows a result of simulating the torque control state in the region II when the control calculation time is changed by the control system based on this idea. Further, FIG. 15 also shows the voltage vector selected to increase or decrease the torque. Note that (a) in FIG. 15 shows the case where the calculation time is 10 μs, and (b) in FIG.
Shows the case where the calculation time is 50 μs, and FIG. 15C shows the case where the calculation time is 100 μs.

Generally, the operation time is expensive and the DSP is fast.
In the case of adopting the above, about 30 μs to 50 μs is required, and in the case of adopting a high-speed single-chip control microcomputer or an inexpensive DSP, about 100 μs to 200 μs is required.

As can be seen from FIG. 15, when the control operation time is set to 10 μs, the characteristics almost comparable to those of FIG. 13 can be obtained, but the control operation time is 50 μs.
When set to 100 μs, the interval compared with the hysteresis level, that is, the frequency of switching is sparse, and in some cases, the voltage vector V5 that creates the reverse rotating magnetic field is selected and the torque is rapidly attenuated. Which causes an increase in torque ripple.

The switching cycle per phase of the inverter is 10 μs, 50 μs, 10 μs for the control calculation time.
In the case of 0 μs, it is about 30 μs (33 kH)
z), about 150 μs (6.7 kHz), while about 300 μs when the control calculation time is 100 μs
(3.3 kHz), and the performance of high-speed switching transistors such as IGBT cannot be fully utilized.

Furthermore, since the voltage vector cannot be switched during the control calculation time, the zero vector output average time tends to increase with the control calculation time.
As a result, the maximum output voltage of the inverter is lowered.

[0052]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and minimizes torque ripple and magnetic flux ripple and improves the inverter maximum output voltage by switching the voltage vector during the control calculation time. An object of the present invention is to provide an inverter control method and an apparatus therefor capable of achieving a higher switching frequency of the inverter.

[0053]

According to a first aspect of the present invention, there is provided an inverter control method for detecting a line current and a terminal voltage of an AC motor connected to an inverter and fixing the detected values in a predetermined direction of a motor stator. Coordinate (α-
(β coordinate), and the converted α-β current, α
Calculates the motor torque and alpha-beta stator flux vector by -β voltage, and selects the voltage vector by the corresponding command value and these calculation results, when the switching control of the transistors of the inverter, the predetermined interval T _c The above process is performed for each and the angle φ formed with the reference line provided on the α-β coordinate of the α-β stator magnetic flux vector
Based on the above, two or more types of voltage vectors are selected, and the time for outputting each selected voltage vector is set by the deviation of the command value corresponding to the torque calculation result of the AC motor and α-
β Pulse width to PWM timer that performs on / off control of each transistor of the inverter based on the rotational angular velocity of the stator magnetic flux vector or the rotational angular velocity of the motor rotor, and based on the voltage vector selection result and output time calculation result. Is stored and the switching operation is performed twice or more during the interval T _c .

The inverter control method according to claim 2 is α-β
The deviation of the command value corresponding to the magnitude of the stator magnetic flux vector is calculated, and in response to the result of this calculation being within a predetermined range, the reference provided on the α-β coordinate of the α-β stator magnetic flux vector. This is a method of selecting three types of voltage vectors based on the angle φ formed with the line and calculating the output time distribution of each voltage vector.

According to the third aspect of the inverter control method, α-β
The deviation of the command value corresponding to the magnitude of the stator magnetic flux vector is calculated, and in response to the result of the calculation being outside the predetermined range, the α−β of the stator magnetic flux vector α− at each predetermined interval T _c. Angle φ formed with the reference line on the β coordinate
It is a method of selecting two types of voltage vectors based on the above and calculating the output time distribution of each voltage vector.

According to the fourth aspect of the inverter control method, the output time t _{6 of} the first voltage vector, the output time t ₂ of the _second voltage vector, and the output time t ₀ of the third voltage vector are each expressed by α- 0 to 60 calculated from the angle φ formed between the β stator magnetic flux vector and the reference line provided on the α-β coordinate
Using the phase angle φ ₀ that changes within °, the deviation of the command value corresponding to the torque calculation result, the rotational angular velocity of the α-β stator magnetic flux vector, or the number K _s determined based on the rotational angular velocity of the motor rotor, , T ₆ / T _c = K _s · sin (π / 3−φ ₀ ) t ₂ / T _c = K _s · sin φ ₀ t ₀ / T _c = 1-t ₆ / T _c −t ₂ / T _c It is a method of calculating by.

According to the fifth aspect of the inverter control method, α-β
In response to the magnitude of the stator magnetic flux vector being smaller than the corresponding command value, the output time t of the first voltage vector
₆ , the output time t ₀ of the third voltage vector, α
-Β The phase angle φ ₀ that changes within 60 ° from 0 ° calculated from the angle φ formed with the reference line provided on the α-β coordinates of the stator magnetic flux vector, and the command value corresponding to the torque calculation result Using the deviation and the rotational angular velocity of the α-β stator magnetic flux vector or the number K _s determined based on the rotational angular velocity of the motor rotor, t ₆ / T _c = 3 ^1/2 / 2 · K _s / cos φ ₀ t ₀ / T _c = 1-t ₆ / T _c, and in response to the magnitude of the α-β stator magnetic flux vector being larger than the corresponding command value, the output time t ₂ of the second voltage vector , The output time t ₀ of the third voltage vector is t ₂ / T _c = 3 ^1/2 / 2 · K _s / cos (π / 3−φ ₀ ) t ₀ / T _c = 1-t ₂ This is a method of calculating by / T _c .

According to a sixth aspect of the inverter control method of the present invention, the interval is based on the angle φ formed between the α-β stator magnetic flux vector calculated at each predetermined interval T _c and the reference line provided on the α-β coordinate. The average rotational angular velocity ω of the α-β stator magnetic flux vector for a predetermined period of T _c or more is calculated, and the magnitude of the α-β stator magnetic flux vector | λ | and the deviation Δ of the torque calculation result and the corresponding command value Δ. τ and the DC voltage V _dc of the inverter, the number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc .

According to the seventh aspect of the inverter control method, the interval T _c is calculated based on the rotor position angle calculated based on the α-β stator magnetic flux vector and the α-β current calculated at each predetermined interval T _c. The average rotational angular velocity ω _e of the rotor position angle during the above predetermined period is calculated, and α-β
Based on the magnitude | λ | of the stator magnetic flux vector, the deviation Δτ of the command value corresponding to the torque calculation result, the DC voltage V _{dc of the} inverter, and the proportional gain G, the above number K _s is K _s = (2 ^1/2 · This is a method of calculating by | λ | · ω _e + G · Δτ) / V _dc .

According to another aspect of the present invention, an inverter control device detects a line current and a terminal voltage of an AC motor connected to an inverter, and these detected values are orthogonal to two axes fixed in a predetermined direction of a motor stator. While converting to coordinates (α-β coordinates), the motor torque and the α-β stator magnetic flux vector are calculated by the converted α-β current and α-β voltage,
A voltage vector is selected by a command value corresponding to these calculation results and switching control of each transistor of the inverter is performed, and the above process is performed at every predetermined interval T _c to obtain α-β of the α-β stator magnetic flux vector. β
Two or more types of voltage vectors are selected based on the angle φ formed with the reference line provided on the coordinates, and the time for outputting each selected voltage vector is determined by the torque calculation result of the AC motor and the deviation of the corresponding command value. ON-OFF control of each transistor of the inverter based on the voltage vector calculation means for calculating the rotational angular velocity of the α-β stator magnetic flux vector or the rotational angular velocity of the motor rotor, the selection result of the voltage vector, and the output time calculation result. PWM to do
Switching control means for storing the pulse width in the timer and performing the switching operation twice or more during the interval T _c is included.

According to a ninth aspect of the present invention, the inverter control device, as the voltage vector calculation means, calculates the deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector, and the calculation result is within a predetermined range. In response to, the three types of voltage vectors are selected based on the angle φ formed between the α-β stator magnetic flux vector and the reference line provided on the α-β coordinates, and the output time distribution of each voltage vector is calculated. The one that adopts the thing.

According to a tenth aspect of the present invention, the inverter control device calculates, as the voltage vector calculation means, the deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector, and the calculation result is outside the predetermined range. In response to each of the voltage vectors, two types of voltage vectors are selected based on the angle φ formed between the α-β stator flux vector and the reference line provided on the α-β coordinates at predetermined intervals T _c. The calculation of the output time distribution of is adopted.

According to another aspect of the inverter control device of the present invention, as the voltage vector calculating means, a first voltage vector output time t ₆ , a second voltage vector output time t ₂ , and a third voltage vector output time t. Each of ₀ is a phase angle φ ₀ that changes within 0 ° to 60 ° calculated from an angle φ formed with the reference line provided on the α-β coordinates of the α-β stator magnetic flux vector, and the torque calculation result. And the number K determined based on the deviation of the corresponding command value and the rotational angular velocity of the α-β stator magnetic flux vector or the rotational angular velocity of the motor rotor.
with _{s, t 6 / T c =} K s · sin (π / 3-φ 0) t 2 / T c = K s · sin φ 0 t 0 / T c = 1-t 6 / T c -t _The calculation is performed by ₂ / T _c .

According to the twelfth aspect of the present invention, in the inverter controller, as the voltage vector calculating means, in response to the magnitude of the α-β stator magnetic flux vector being smaller than the corresponding command value, the output of the first voltage vector is output. Each of the time t ₆ and the output time t ₀ of the third voltage vector is calculated from the angle φ formed by the reference line provided on the α-β coordinates of the α-β stator magnetic flux vector from 0 ° to 60 °. The phase angle φ ₀ that varies within, the deviation of the command value corresponding to the torque calculation result, and the rotational angular velocity of the α-β stator magnetic flux vector, or
Using the number K _s determined based on the rotational angular velocity of the motor rotor, t ₆ / T _c = 3 ^1/2 / 2 · K _s / cos φ ₀ t ₀ / T _c = 1-t ₆ / T _c Then, in response to the magnitude of the α-β stator magnetic flux vector being larger than the corresponding command value, the output time t ₂ of the _second voltage vector and the output time t ₀ of the third voltage vector are set to is intended to employ those calculated by ^{_{t 2 / T c = 3 1/2}} / 2 · K s / cos (π / 3-φ 0) t 0 / T c = 1-t 2 / T c.

According to a thirteenth aspect of the present invention, in the inverter controller, as the voltage vector calculation means, an angle formed between the α-β stator magnetic flux vector calculated at each predetermined interval T _c and a reference line provided on the α-β coordinates. Based on φ, the average rotational angular velocity ω of the α-β stator magnetic flux vector for a predetermined period of the interval T _c or more is calculated, and the magnitude | λ | of the α-β stator magnetic flux vector and the torque calculation result are calculated. A method is used in which the number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc by the deviation Δτ of the corresponding command value and the DC voltage V _dc of the inverter. It is a thing.

According to a fourteenth aspect of the present invention, in the inverter control device as the voltage vector calculating means, the rotor position angle calculated based on the α-β stator magnetic flux vector calculated at each predetermined interval T _c and the α-β current is calculated. Based on the above, the average rotational angular velocity ω _e of the rotor position angle for a predetermined period of the interval T _c or more is calculated, and the magnitude | λ | of the α-β stator magnetic flux vector and the command value corresponding to the torque calculation result. Deviation Δτ, inverter DC voltage V _dc , and proportional gain G to calculate the above number K _s by K _s = (2 ^1/2 · | λ | · ω _e + G · Δτ) / V _dc To be adopted.

[0067]

According to the inverter control method of the first aspect, the line current and the terminal voltage of the AC motor connected to the inverter are detected, and the detected values are detected by the two axes fixed in a predetermined direction of the motor stator. In addition to converting to orthogonal coordinates (α-β coordinates), the motor torque and α-β stator magnetic flux vector are calculated from the converted α-β current and α-β voltage, and the command values corresponding to these calculation results A voltage vector is selected according to, and when switching control of each transistor of the inverter is performed, a predetermined interval T _c
The above process is performed for each of the
2 based on the angle φ with the reference line provided on the β coordinate
While selecting more than one kind of voltage vector, the time to output each selected voltage vector, the deviation of the command value corresponding to the torque calculation result of the AC motor and the rotation angular velocity of the α-β stator magnetic flux vector or the motor rotor A pulse width is stored in a PWM timer that performs on-off control of each transistor of the inverter based on the result of voltage vector selection and the result of output time calculation, based on the rotational angular velocity, and is stored twice or more during the interval T _c. Since the switching operation is performed, the voltage vector can be switched even during the control calculation time, the zero vector output averaging time can be shortened, and eventually, the reduction of the inverter maximum output voltage can be prevented. You can
In addition, torque ripple can be suppressed and
The switching cycle can be shortened and the performance of the high speed switching transistor can be fully utilized.

According to the inverter control method of claim 2,
The deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector is calculated, and in response to the result of this calculation being within a predetermined range, the deviation is displayed on the α-β coordinate of the α-β stator magnetic flux vector. Three types of voltage vectors are selected based on the angle φ formed with the provided reference line, and the output time distribution of each voltage vector is calculated. Therefore, the magnitude of the α-β stator magnetic flux vector is to some extent a command value. When following, it is possible to achieve the same effect as in claim 1.

According to the inverter control method of claim 3,
The deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector is calculated, and in response to the result of this calculation being outside the predetermined range, the α-β stator magnetic flux vector is generated at predetermined intervals T _c. Since two types of voltage vectors are selected on the basis of the angle φ formed with the reference line provided on the α-β coordinate of, the output time distribution of each voltage vector is calculated.
Even when the magnitude of the β-stator magnetic flux vector does not follow the command value so much, the same operation as in claim 1 can be achieved.

According to the inverter control method of claim 4,
Each of the output time t _{6 of} the first voltage vector, the output time t ₂ of the _second voltage vector, and the output time t ₀ of the third voltage vector on the α-β coordinate of the α-β stator magnetic flux vector. The phase angle φ ₀ varying from 0 ° to 60 ° calculated from the angle φ formed with the provided reference line, the deviation of the command value corresponding to the torque calculation result, and the rotational angular velocity of the α-β stator magnetic flux vector or , T ₆ / T _c = K _s · sin (π / 3−φ ₀ ) t ₂ / T _c = K _s · sin φ ₀ t ₀ using the number K _s determined based on the rotational angular velocity of the motor rotor. because of being calculated by the _{/ T c = 1-t 6} / T c -t 2 / T c, it is possible to achieve effects similar to those of claim 2.

According to the inverter control method of claim 5,
In response to the magnitude of the α-β stator magnetic flux vector being smaller than the corresponding command value, the output time t _{6 of} the first voltage vector and the output time t ₀ of the third voltage vector are changed to α -[Beta] to 60 [deg.] Calculated from the angle [phi] formed with the reference line provided on the [alpha]-[beta] coordinate of the stator flux vector
Using the phase angle φ ₀ that changes within and the deviation of the command value corresponding to the torque calculation result and the rotational angular velocity of the α-β stator magnetic flux vector or the number K _s determined based on the rotational angular velocity of the motor rotor, t ₆ / T _c = 3 ^1/2 / 2 · K _s / cos φ ₀ t ₀ / T _c = 1-t ₆ / T _c , and the magnitude of the α-β stator magnetic flux vector corresponds to the command value. In response to the output time t ₂ of the _second voltage vector and the output time t ₀ of the third voltage vector, t ₂ / T _c = 3 ^1/2 / 2 · K _s / Since the calculation is performed by cos (π / 3−φ ₀ ) t ₀ / T _c = 1−t ₂ / T _c , the same effect as that of claim 3 can be achieved.

According to the inverter control method of claim 6,
Based on the angle φ formed between the α-β stator magnetic flux vector calculated for each predetermined interval T _c and the reference line provided on the α-β coordinates, α for a predetermined period of time equal to or greater than the interval T _c.
The average rotational angular velocity ω of the −β stator magnetic flux vector is calculated, and the magnitude | λ | of the α-β stator magnetic flux vector, the deviation Δτ of the command value corresponding to the torque calculation result, and the DC voltage V _dc of the inverter are used. 6. The number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc , so that the number K _s is calculated according to claim 4 or claim 5.
The same effect as can be achieved.

According to the inverter control method of claim 7,
Based on the rotor position angle calculated based on a predetermined interval T _c computed alpha-beta stator flux vector and alpha-beta currents for each, the average of the rotor position angle of the interval T _c more predetermined time period While calculating the rotational angular velocity ω _e ,
According to the magnitude | λ | of the α-β stator magnetic flux vector, the deviation Δτ of the command value corresponding to the torque calculation result, the DC voltage V _{dc of the} inverter, and the proportional gain G, the above number K _s is K _s = (2 ^{1 / 2} · | λ | · ω _e + G · Δτ) / V _{dc, so} that the calculation is performed according to claim 4 or claim 5.
The same effect as can be achieved.

According to the eighth aspect of the inverter control device,
The line current and the terminal voltage of the AC motor connected to the inverter are detected, and these detected values are converted into coordinates (α-β coordinates) in which the two axes fixed in a predetermined direction of the motor stator are orthogonal, and The motor torque and the α-β stator magnetic flux vector are calculated by the converted α-β current and α-β voltage, the voltage vector is selected by the command value corresponding to these calculation results, and each transistor of the inverter is switched. In controlling, the voltage vector calculation means performs the above-mentioned processing at every predetermined interval T _c , and two types are obtained based on the angle φ formed between the α-β stator magnetic flux vector and the reference line provided on the α-β coordinates. In addition to selecting the above voltage vector, the output time of each selected voltage vector is determined by the deviation of the command value corresponding to the torque calculation result of the AC motor and the α-β stator. A PWM timer that performs an operation based on the rotational angular velocity of the bundle vector or the rotational angular velocity of the motor rotor, and performs on / off control of each transistor of the inverter based on the selection result of the voltage vector and the output time operation result by the switching control means. The pulse width can be stored and the switching operation can be performed twice or more during the interval T _c .

Therefore, the voltage vector can be switched even during the control calculation time, the zero vector output averaging time can be shortened, and eventually the inverter maximum output voltage can be prevented from lowering. Further, torque ripple can be suppressed, the switching cycle can be shortened, and the performance of the high-speed switching transistor can be fully utilized.

According to the inverter control device of claim 9,
As the voltage vector calculation means, the deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector is calculated, and in response to the calculation result being within a predetermined range, α-β
Three types of voltage vectors are selected based on the angle φ formed with the reference line provided on the α-β coordinate of the stator magnetic flux vector,
Since the one that calculates the output time distribution of each voltage vector is adopted, when the magnitude of the α-β stator magnetic flux vector follows the command value to some extent, the same operation as in claim 8 is achieved. be able to.

According to the inverter control device of the tenth aspect, the voltage vector calculation means calculates the deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector,
In response to the result of this operation being outside the predetermined range, 2 is calculated based on the angle φ formed by the reference line provided on the α-β coordinate of the α-β stator magnetic flux vector at each predetermined interval T _c. Since a type of voltage vector is selected and the output time distribution of each voltage vector is calculated, α−
Even when the magnitude of the β-stator magnetic flux vector does not follow the command value so much, the same operation as in claim 8 can be achieved.

According to the eleventh aspect of the present invention, as the voltage vector calculating means, the output time t _{6 of} the first voltage vector, the output time t ₂ of the _second voltage vector, and the output of the third voltage vector are output. Each time t ₀ is 0 ° to 60 ° calculated from the angle φ formed with the reference line provided on the α-β coordinate of the α-β stator magnetic flux vector.
Using the phase angle φ ₀ that changes within and the deviation of the command value corresponding to the torque calculation result and the rotational angular velocity of the α-β stator magnetic flux vector or the number K _s determined based on the rotational angular velocity of the motor rotor, t ₆ / T _c = K _s · sin (π / 3−φ ₀ ) t ₂ / T _c = K _s · sin φ ₀ t ₀ / T _c = 1-t ₆ / T _c −t ₂ / T _c Claim 9 because it employs an arithmetic operation
The same effect as can be achieved.

In the inverter control device according to the twelfth aspect of the present invention, the voltage vector computing means responds to the fact that the magnitude of the α-β stator magnetic flux vector is smaller than the corresponding command value, in response to the first voltage vector. From 0 ° calculated from the angle φ formed by the reference line provided on the α-β coordinate of the α-β stator magnetic flux vector with respect to each of the output time t ₆ and the output time t ₀ of the third voltage vector. A phase angle φ ₀ changing within 60 °, a deviation between the command value corresponding to the torque calculation result, and the rotational angular velocity of the α-β stator magnetic flux vector or the number K _s determined based on the rotational angular velocity of the motor rotor are used. Then, t ₆ / T _c = 3 ^1/2 / 2 · K _s / cos φ ₀ t ₀ / T _c = 1-t ₆ / T _c , and the magnitude of the α-β stator magnetic flux vector corresponds. In response to being larger than the command value, the second voltage Output time t ₂ of the vector, the respective output time t ₀ of the third voltage _{vector, t 2 / T c = 3} 1/2 / 2 · K s / cos (π / 3-φ 0) t 0 / T since it is to adopt those calculated by _{c = 1-t 2 / T} c, claim 1
An effect similar to 0 can be achieved.

According to the inverter control device of the thirteenth aspect, as the voltage vector calculating means, the rotor position calculated based on the α-β stator magnetic flux vector calculated at each predetermined interval T _c and the α-β current. Based on the angle, the average rotational angular velocity ω _e of the rotor position angle for a predetermined period of the interval T _c or more is calculated, and the magnitude | λ | of the α-β stator magnetic flux vector corresponds to the torque calculation result. Using the deviation Δτ of the command value, the DC voltage V _{dc of the} inverter, and the proportional gain G, the above number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω _e + G · Δτ) / V _dc Because it adopts one, claim 1
It is possible to achieve the same effect as the first or the twelfth aspect.

According to the fourteenth aspect of the present invention, in the inverter control device of the fourteenth aspect, the voltage vector calculation means is the α-β of the magnetic flux vector of the α-β stator calculated at the predetermined intervals T _c.
Based on the angle φ formed with the reference line provided on the β-coordinate, the average rotational angular velocity ω of the α-β stator magnetic flux vector during the predetermined period of the interval T _c or more is calculated, and α-β
Size of stator magnetic flux vector | λ |, deviation of command value corresponding to torque calculation result Δτ, DC voltage V _{dc of} inverter
Therefore, the number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc.
It is possible to achieve the same effect as the first or the twelfth aspect.

[0082]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of an inverter control method and apparatus of the present invention will be described in detail below with reference to the accompanying drawings.

FIG. 1 is a schematic diagram showing the configuration of an AC motor drive system for driving an AC motor using an inverter controlled by the inverter control method or inverter control device of the present invention.

In this AC motor drive system, the DC power supply 1 is supplied to the three-phase inverter 2 to convert it into three-phase AC power, and this three-phase AC power is supplied to the three-phase AC motor 3. The motor 3 is driven. A microcomputer 4 for controlling the three-phase inverter 2 is provided. Here, as the three-phase AC motor 3, a brushless DC motor, an induction motor, a reluctance motor, etc. can be exemplified.

FIG. 2 is a block diagram showing the internal processing arrangement of the microcomputer 4.

The microcomputer 4 comprises a CPU core 4a and a built-in peripheral circuit 4b.

The built-in peripheral circuit 4b calculates the motor currents i _u and i _{w for the} two phases (the remaining current i _v can be calculated by using the sum of the three-phase currents being 0) and the DC voltage V _dc . An AD converter 41 for converting into a digital value in synchronization with the control processing interval T _c , and a three-phase P for outputting a transistor switching signal for each phase of the inverter to be supplied to each switching transistor of the three-phase inverter 3 as inputs.
It has a WM timer unit 42.

The CPU core 4a receives the motor currents i _u (n) and i _w (n) for the two phases converted into digital values as an input and performs a three-phase two-phase coordinate conversion process to obtain an αβ-axis current i.
Three-phase / two-phase coordinate conversion unit 4 that outputs _α (n) and i _β (n)
3, a torque calculator 44 that calculates a torque τ (n) by inputting αβ-axis currents i _α (n), i _β (n) and αβ-axis magnetic fluxes λ _α (n + 1) and λ _β (n + 1), and A torque deviation calculator 45 that calculates a deviation Δτ (= τ * −τ (n)) between the torque τ (n) and an externally applied torque command τ *, and αβ axis currents i _α (n), i _β ( n), the DC voltage V _dc (n) converted into a digital value, and the output time of each voltage vector, αβ-axis magnetic flux λ _α (n + 1),
and .alpha..beta flux calculating unit 46 for outputting a _{λ β (n + 1),} αβ
A stator magnetic flux calculation unit 47 that calculates the magnitude | λ | (n + 1) of the stator magnetic flux with the axial magnetic fluxes λ _α (n + 1) and λ _β (n + 1) as input, and the magnitude | λ | (n + 1) of the stator magnetic flux. )
And a magnetic flux comparison unit 48 for determining whether a deviation between a magnetic flux command | λ | * given from the outside is within a predetermined range, and αβ
A phase calculator 49 that outputs phase angles φ (n + 1), φ ₀ (n + 1), and a region number with the axial magnetic fluxes λ _α (n + 1) and λ _β (n + 1) as input, and the αβ axial magnetic flux λ _α (n), λ
A speed calculation unit 5 that inputs _β (n) and α β axis currents i _α (n) and i _β (n) and outputs a rotational angular speed ω _e (n).
0, the DC voltage V _dc (n), the rotational angular velocity ω _e (n), the phase angle φ (n + 1), φ ₀ (n + 1), the region number, the output Δτ of the torque deviation calculation unit 45, and the magnetic flux comparison unit 48. It has a voltage vector output time calculation unit 51 that selects a voltage vector using the determination output as an input, calculates the output time thereof, and sets the output time in the three-phase PWM timer unit 42.

The operation of the microcomputer having the above configuration will be described with reference to FIGS.
The flowchart will be described with reference to the flowchart. The process of the flowchart of FIG. 3 is started at every interval T _c by a trigger signal from the interrupt timer.

First, the control law will be described. (1) Three types of voltage vector output time allocation method: In this method, the voltage vector output time allocation for obtaining the tangential magnetic flux locus of the rotating magnetic field is calculated. This is called method 1.

FIG. 5 is an enlarged view of a magnetic flux locus (voltage vector time product) obtained by three kinds of voltage vectors including the zero vector in the area II in a minute section.

In FIG. 5, voltage vectors V6, V2, V7
Are appropriately arranged. And P in FIG.₀To P₁To
Time to T_cAnd an appropriate variable V₁If is adopted, P₀P₁
= V ₁・ T_cThe approximate expression of holds. Also, in the voltage vector
Configured triangle △ P₀qP₁And apply the sine wave theorem
Then, the relationship of Expression 5 is obtained.

[0093]

[Equation 5]

When T _c = t ₀ + t ₆ + t _{2 is taken} into consideration, the relations of Expressions 5 to 6 can be obtained, and the output time of each voltage vector is determined by the magnitude of K _s .

[0095]

[Equation 6]

Here, there is a relationship of equation 7.

[0097]

[Equation 7]

Therefore, if the time distribution of the three types of voltage vectors is determined by the equation 6, a magnetic flux locus equivalent to the output of the voltage vector in the tangential direction of the rotating magnetic field can be drawn.

In addition to the voltage vector output pattern shown in FIG. 5 (see (a) in FIG. 6), the pattern in which the output order of voltage vectors is changed {(b) (c) in FIG. 6)
(Refer to (d)} can also be adopted.

In other areas, voltage vectors may be selected based on FIG. 12, and the respective voltage vector output time distributions may be calculated from equation 6. Table 4 shows the voltage vector of each region corresponding to the output time of the equation (6). Here, the output order of each voltage vector is the pattern of (a) in FIG.

[0101]

[Table 4]

When the voltage vector is selected according to Table 4, the inverter output voltage captured by the α-β coordinate is t ₆
The one corresponding to the voltage vector output in time is v _{α #} mai
Table 5 shows v _{α #} sub and v _{β #} sub which correspond to the voltage vector output for n, v _{β #} main, t ₂ time.

[0103]

[Table 5]

The phase angle φ ₀ can be calculated as shown in Table 6 by the stator magnetic fluxes λ _α and λ _β . The β axis is adopted as the reference line for determining the phase φ of the α-β stator magnetic flux vector, and the direction of the β axis is φ = 0 °.

[0105]

[Table 6]

In the above method, the direction of the magnetic flux vector obtained by synthesizing the three types of voltage vectors is the tangential direction of the rotating magnetic field. Therefore, the normal direction, in other words, the magnitude of the rotating magnetic field has three types of voltage vectors. There is almost no change between the start and end points of. Therefore, it is applied when the deviation between the actual magnetic flux and the command magnetic flux is within a predetermined range and it is not necessary to change the rotating magnetic field. (2) Two kinds of voltage vector time allocation methods: This method is such that, as shown in FIG. 7 (illustrated for the area II), the magnetic flux vector obtained by combining the voltage vectors moves in the tangential direction, and In the normal direction, in other words, two kinds of voltage vectors that increase and decrease the magnitude of the magnetic flux at the start point and the end point of the voltage vector are selected. Then, the time distribution of the two types of voltage vectors including the zero vector is calculated so that the movement amount in the tangential direction becomes a predetermined value. This is called method 2.

Here, if an appropriate variable V ₁ is used, the length of the vector in the tangential direction is set as V ₁ · T _c, and it is noted that the magnetic flux stagnates at the zero vector, from FIG. 7 (a). , The output time t ₆ of the voltage vector V6 is related by the equation 8, and from FIG. 7B, the output time t ₂ of the voltage vector V2 is represented by the equation 9.
Have a relationship.

[0108]

[Equation 8]

[0109]

[Equation 9]

The voltage vector V6 causes the magnetic flux to move in the direction normal to the direction away from the origin of the α-β coordinate, in other words, the magnitude of the rotating magnetic field is increased, and the voltage vector V2 causes the magnetic flux to move in the normal direction. It moves toward the origin of the α-β coordinate, in other words, reduces the magnitude of the rotating magnetic field. Further, the output time of each voltage vector can be obtained based on the equations 7, 8, and 9.

The relationship between the actual value of the magnitude of the rotating magnetic field calculated in the above embodiment, the deviation of the command value, and the pulse width is summarized in Table 7. The allowable value Δλ of the deviation between the actual magnetic flux magnitude and the command magnetic flux magnitude may be set to 0, and the calculation in the second column of Table 7 may be omitted to perform control using only two types of voltage vectors. This can simplify the calculation.

[0112]

[Table 7]

However, t₂> T_cWhen, t₂= T_c, T₆
> T_cWhen, t₆= T_cAnd (3) K_s(Ie voltage vector output time) calculated
Method: The magnitude of the deviation between the actual motor torque and the command value
When 0, the motor rotation angular velocity ω _e(Electrical angle) and Inva
Data output voltage angular velocity, that is, the angular velocity of the rotating magnetic field
Perform control to keep the same, and make the voltage waveform and rotor equal.
Hold the status.

The inverter output voltage V ₁ is given by Equation _{1 according} to the magnitude of the rotating magnetic field | λ | and the motor rotation angular velocity ω _e (electrical angle).
It can be written as 0. Then, the relationship of Expression 11 is obtained from Expression 7 and Expression 10.

[0115]

[Equation 10]

[0116]

[Equation 11]

Therefore, the motor rotation speed is detected and K
_s can be calculated by Equation 11, and the output time of each voltage vector can be determined according to Table 7.

On the other hand, when there is a deviation between the actual motor torque and the command value, it is necessary to increase or decrease K _s , that is, to accelerate or decelerate the angular velocity of the rotating magnetic field to follow the motor rotational angular velocity.

Therefore, the voltage proportional to the deviation Δτ between the actual motor torque τ and the command value τ * is added to the equation 11 (see the equation 12).

[0120]

[Equation 12]

Here, G is an appropriate proportional gain. (4) Rotational angular velocity detection method: The steady-state motor rotational angular velocity ω _e can be obtained by time-differentiating the phase angle φ of the rotating magnetic field. However, this method has a problem that the rotational angular velocity of the rotating magnetic field and the rotational angular velocity of the rotor are not equal during a transition such as a sudden change in torque.

To solve this problem, "Sensorless position estimation method and stability of salient-pole brushless DC motor", Chen Zhi-Ken et al., 1998 Annual Meeting of the Japan Institute of Electrical Engineers, Industrial Application Division, pp. 179-182. In addition, by calculating the rotor position angle θ _e by a conventionally known rotor position angle estimation method and differentiating this time, it is possible to calculate the rotational angular velocity including the transient state.

The α-β stator magnetic flux can be expressed by Equation 13 by the d-q axis inductances L _d and L _q and the magnetic flux Λ of the permanent magnet.

[0124]

[Equation 13]

The relationships of Expressions 13 to 14 are obtained.

[0126]

[Equation 14]

Therefore, the rotor position angle θ _e can be calculated from the table.

[0128]

[Table 8]

Then, the rotational angular velocity of the rotor can be obtained by the time differentiation of the obtained rotor position angle θ _e .

By adopting the microcomputer having the configuration shown in FIG. 2, the above processing can be performed.

Further description will be made.

The AD converter 41 is a multiplexer.
(Not shown) allows 2-phase motor current out of 3-phase motor current.
Current i_u, I_wAnd DC voltage V_dcAnd the control processing interface
Bar T _cCPU core 4
supply to a.

In the 3-phase / 2-phase coordinate conversion section 43,
Mathematical _Expression 1 using _Expression 1 and the relationship of i _u + i _v + i _w = 0
By performing the calculation of αβ-axis current i _α (n), i
_β (n) can be obtained.

[0134]

[Equation 15]

In the αβ magnetic flux calculation section 46,
In consideration of the fact that the calculation result of the stator magnetic flux calculation is performed at the sampling point n is reflected in the voltage at the next sampling point n + 1, the calculation of Equation 16 is performed and the αβ-axis magnetic flux λ _α at the n + 1 point is calculated.
(N + 1) and λ _β (n + 1) are obtained. The voltage vector is converted into a two-phase voltage according to Table 5.

[0136]

[Equation 16]

With this, it is possible to eliminate the influence of the calculation dead time on the control. In the high rotation speed region where the terminal voltage of the motor increases, the contribution of the voltage drop terms (R · i _α , R · i _β ) in the winding resistance R in the equation 16 to the magnetic flux calculation result is small, This may be treated as zero.

In the stator magnetic flux calculation unit 47, the calculation of equation 17 is performed to calculate the magnitude of the stator magnetic flux | λ | (n +
1) is obtained.

[0139]

[Equation 17]

In the torque calculation section 44,
Is calculated to obtain the torque τ (n).

[0141]

[Equation 18]

In the calculation of the equation (18), the value at the sampling point n + 1 is used only for the magnetic flux amount for which the predictive calculation can be performed easily, and the influence of the calculation delay is reduced.

The phase calculator 49 obtains and outputs the phase angles φ (n + 1), φ ₀ (n + 1) and the area number based on Table 6.

In the speed calculator 50, various quantities necessary for the calculation of the rotor position angle are calculated by the equation 19, and the rotational position angle θ _e (n) is calculated from the calculated quantities according to Table 8. , The rotational angular velocity ω _e (n) of the rotor is calculated from the difference between the sample points of the rotational position angle θ _e (n) (see Formula 20).

[0145]

[Formula 19]

[0146]

[Equation 20]

In the voltage vector output time calculation unit 51, in response to the output of the torque deviation calculation unit 45 and the determination result of the magnetic flux comparison unit 48, K _s is calculated according to Table 8 and based on Table 7. The pulse width of each voltage vector is calculated.

The calculated voltage vector output time t ₆ ,
Based on t ₂ and t ₀ , the on / off time of each phase transistor is set in the three-phase PWM timer unit 42 based on Table 4. Based on this set value, from the next sample n + 1 points to n +
On / off control of the transistor is performed during the period of two points.

These processes in the CPU core 4a are
It is executed at a predetermined interval T _c , with a signal output from the interrupt timer at each set time as a trigger.
That is, the elapsed time from the sampling point n to the sampling point n + 1 is T _c .

Since the time constant of the mechanical system is sufficiently longer than that of the electric system, the influence of the calculation dead time can be ignored even if the information of the sampling point n is adopted as the rotational speed as shown in the equation (20). Further, a value obtained by subjecting the obtained speed information to an appropriate number of processes such as a moving average process can be used as the rotation speed information, and the speed calculation cycle can be set to an interval T _c or more.

Further, due to magnetic saturation, the rotational position angle calculation value θ
Even if an error occurs in _e (n), the error ε near the sampling point
Assuming that (n) is almost the same, the following equation 21 is obtained.

[0152]

[Equation 21]

Therefore, in the above embodiment, in the process of calculating the rotational angular velocity ω _{e of the} rotor, the q-axis inductance of equation (20) changes due to magnetic saturation, and the rotational position angle θ
Despite the adoption of a conventionally known technique that causes a calculation error of _e (n), it is essentially insusceptible to magnetic saturation.

Next, the processing of the flowcharts of FIGS. 3 and 4 will be described.

First, the processing of the flowchart of FIG. 3 will be described.

At step SP1, the magnetic flux command | λ |
* And the torque command τ * are input, the DC voltage V _dc (n) is input in step SP2, and the motor currents i _u (n) and i _w (n) for two phases are input in step SP3. .

Then, at step SP4,
Is converted into a two-phase current (αβ axis current) by step SP
5, the voltage vector is converted into a two-phase voltage (αβ
(Axis voltage), in step SP6, the stator magnetic flux is calculated by the equation 16, and in step 7, the equation 17
The magnitude of the stator magnetic flux is calculated by.

Then, in step SP8, the phase angles φ (n + 1), φ ₀ (n + 1), and the region number are calculated from Table 6, the torque is calculated in step SP9 by the equation 18, and the rotational angular velocity is calculated in step SP10. A subroutine for calculating ω _e (n) is executed.

Then, in step SP11, the number 1
2, K _s is calculated in step 2, the pulse width is calculated in table 7 in step SP12, the pulse width of each phase is scheduled in the PWM timer in step SP13 based on table 4, and the process returns to the original processing.

Next, the flowchart of FIG. 4 will be described.

In step SP1, the calculation of the equation 19 is performed to calculate various amounts required for the calculation of the rotor position angle, and in step SP2, the rotor position angle θ is calculated according to Table 8.
_e (n) is calculated, and in step 3, the rotational angular velocity ω _e (n) is calculated by Expression 20, and the process directly returns to the original process.

However, as the rotational angular velocity calculation unit, it is possible to employ a unit that executes another conventionally known rotational position angle calculation method or angular velocity calculation method. In applications where rapid acceleration / deceleration is not repeated, for example, in applications such as compressors and pumps, the angular velocity is calculated by calculating the difference between sample points of the phase φ of the stator magnetic flux vector, and the calculation result is low-pass filtered. However, it is also possible to adopt it as the rotational angular velocity ω _e (n).

FIG. 8 shows the simulation result of the torque response obtained by adopting the microcomputer of FIG. 2 and the processing of the flow charts of FIGS.

Until time t _R , steady operation is performed near the command value. At this time, the value of K _s calculated by Expression 12 is substantially the same as the value calculated by Expression 11 because the deviation between the actual torque value and the command value is sufficiently small. In response to K _s calculated by Equation 12 and the deviation between the actual magnetic flux magnitude and the command magnetic flux. The voltage vector output time is determined from Table 7, and each phase P from Table 4 is determined based on the region determined by the phase φ.
The number of pulses set in the WM timer is determined, and the pulse width is stored in each PWM timer. As a result, the switching is performed twice or more during the calculation, and the torque ripple can be controlled to be small.

When a stepwise torque command exceeding a predetermined deviation is input at time t _R , the calculation started at point S is completed by point R and the result is stored in the pre-register of the PWM timer. Then, the stored result is transferred from the pre-register to the PWM timer at the point R and is reflected in the switching control of each transistor of the inverter.

Since the deviation between the command and the actual torque has increased,
After the R point, K _s increases based on the equation 12, and the actual torque rises rapidly.

When the deviation between the command and the actual torque becomes small at point Q, Ks calculated by equation (12) becomes substantially equal to the value calculated based on equation (11). Move.

Although the case where the AC motor 4 is normally rotated has been described above, it is also possible to cope with the normal and reverse rotation by adding a voltage vector for reversely rotating the rotating magnetic field.

[0169]

According to the first aspect of the present invention, the voltage vector can be switched even during the control calculation time, the zero vector output averaging time can be shortened, and eventually the inverter maximum output voltage can be reduced. It is possible to prevent it, and further, it is possible to suppress the torque ripple, shorten the switching cycle, and bring out the unique effect that the performance of the high-speed switching transistor can be fully utilized.

According to the invention of claim 2, when the magnitude of the α-β stator magnetic flux vector follows the command value to some extent,
The same effect as that of claim 1 is achieved.

The invention of claim 3 has the same effect as that of claim 1 even when the magnitude of the α-β stator magnetic flux vector does not follow the command value so much.

The invention of claim 4 has the same effect as that of claim 2.

The invention of claim 5 has the same effect as that of claim 3.

The invention of claim 6 has the same effect as that of claim 4 or claim 5.

The invention of claim 7 has the same effect as that of claim 4 or claim 5.

According to the eighth aspect of the present invention, the voltage vector can be switched even during the control calculation time, and the zero vector output average time can be shortened.
It is possible to prevent the inverter maximum output voltage from decreasing.
Moreover, torque ripple can be suppressed, and the switching cycle can be shortened, so that the performance of the high-speed switching transistor can be fully utilized, which is a unique effect.

According to a ninth aspect of the invention, when the magnitude of the α-β stator magnetic flux vector follows the command value to some extent,
An effect similar to that of claim 8 is achieved.

The invention of claim 10 has the same effect as that of claim 8 even when the magnitude of the α-β stator magnetic flux vector does not follow the command value so much.

The invention of claim 11 has the same effect as that of claim 9.

The invention of claim 12 has the same effect as that of claim 10.

The invention of claim 13 has the same effect as that of claim 11 or claim 12.

The invention of claim 14 has the same effect as that of claim 11 or claim 12.

[Brief description of drawings]

FIG. 1 is a schematic diagram showing a configuration of an AC motor drive system that drives an AC motor using an inverter controlled by an inverter control method or an inverter control device of the present invention.

FIG. 2 is a block diagram showing an internal processing configuration of a microcomputer.

FIG. 3 is a flowchart illustrating processing of a microcomputer.

FIG. 4 is a flowchart illustrating in detail the processing of step SP10 of the flowchart of FIG.

FIG. 5 is a diagram showing a magnetic flux locus (voltage vector time product) obtained by three types of voltage vectors in a region II in an enlarged manner in a minute section.

FIG. 6 is a diagram showing a pattern of voltage vector output.

FIG. 7 is an enlarged view showing a magnetic flux locus (voltage vector time product) obtained by two types of voltage vectors in a region II in a minute section.

8 is a diagram showing a simulation result of torque response obtained by adopting the microcomputer of FIG. 2 and the processes of the flowcharts of FIGS.

FIG. 9 is a diagram showing a phase relationship between three-phase coordinates and two-phase coordinates and a motor model captured on the two-phase coordinates.

FIG. 10 is an electric circuit diagram showing the configuration of a voltage source inverter.

FIG. 11 is a diagram showing voltage vectors on two-phase coordinates.

FIG. 12 is a diagram showing a magnetic flux change direction depending on each voltage vector.

FIG. 13 shows the magnitude of an α-β stator magnetic flux vector and
FIG. 7 is a diagram showing a magnetic flux locus, a voltage vector in a region II, and a change in torque when the torque is compared by a hysteresis comparator and the inverter is switching-controlled in response to the output.

FIG. 14 is a diagram showing a block configuration of a conventional direct torque control system, characteristics of a magnetic flux hysteresis comparator, and characteristics of a torque hysteresis comparator.

FIG. 15 is a diagram for explaining changes in torque ripple with changes in control calculation time.

[Explanation of symbols]

2 3 phase inverter 3 AC motor 45 Torque deviation calculation unit 48 Magnetic flux comparison unit 51 Voltage vector output time calculator

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 5H576 BB02 BB04 BB06 BB09 CC01                       DD02 DD04 DD07 DD09 EE01                       EE14 EE15 EE16 GG04 HB02                       JJ03 JJ06 JJ14 JJ17 JJ18                       JJ22 LL13 LL22 LL24 LL34                       LL39

Claims (14)

[Claims] 1. Coordinates in which two axes orthogonal to each other are detected by detecting a line current and a terminal voltage of an AC motor (3) connected to an inverter (2) and fixing these detected values in a predetermined direction of a motor stator. (Α-β coordinate) and at the same time, the converted α-β current and α-β voltage cause the motor torque and α
In the inverter control method for calculating the -β stator magnetic flux vector, selecting the voltage vector according to the command value corresponding to these calculation results, and switching controlling each transistor of the inverter (2), at every predetermined interval T _c . The above process is performed to select two or more types of voltage vectors based on the angle φ formed between the α-β stator magnetic flux vector and the reference line provided on the α-β coordinates, and output each selected voltage vector. The time is the deviation of the command value corresponding to the torque calculation result of the AC motor and the rotation angular velocity of the α-β stator magnetic flux vector, or
A pulse width is stored in a PWM timer that performs an on-off control of each transistor of the inverter (2) based on the calculation result of the voltage vector and the output time calculation result based on the rotation angular velocity of the motor rotor, and the interval T
An inverter control method characterized in that switching operation is performed twice or more during a period _c . 2. The deviation of the command value corresponding to the magnitude of the α-β stator magnetic flux vector is calculated, and in response to the result of the calculation being within a predetermined range, the α-β stator magnetic flux vector α The inverter control method according to claim 1, wherein three types of voltage vectors are selected based on an angle φ formed with a reference line provided on the −β coordinate, and output time distribution of each voltage vector is calculated. 3. A deviation of a command value corresponding to the magnitude of an α-β stator magnetic flux vector is calculated, and in response to the result of this calculation being outside a predetermined range, α- at every predetermined interval T _c. The inverter according to claim 1, wherein two kinds of voltage vectors are selected based on an angle φ formed between a β stator magnetic flux vector and a reference line provided on the α-β coordinate, and an output time distribution of each voltage vector is calculated. Control method. 4. The output time t _{6 of} the first voltage vector, the output time t ₂ of the _second voltage vector, and the output time t ₀ of the third voltage vector are respectively defined as α-β of the α-β stator magnetic flux vector. -Phase angle φ ₀ changing within 0 ° to 60 ° calculated from the angle φ formed with the reference line provided on the β coordinate, and the deviation of the command value corresponding to the torque calculation result and α
Using the number K _s determined based on the rotational angular velocity of the −β stator magnetic flux vector or the rotational angular velocity of the motor rotor, t ₆ / T _c = K _s · sin (π / 3−φ ₀ ) t ₂ / T The inverter control method according to claim 2, wherein the calculation is performed by _c = K _s · sin φ ₀ t ₀ / T _c = 1−t ₆ / T _c −t ₂ / T _c . 5. The magnitudes of the α-β stator magnetic flux vectors are paired.
In response to being smaller than the corresponding command value, the first voltage
Vector output time t₆, When outputting the third voltage vector
Interval t₀Of the α-β stator flux vector α-
Calculated from the angle φ formed with the reference line provided on the β coordinate
Phase angle φ that varies from 0 ° to 60 ° ₀And
Deviation of command value corresponding to Luke calculation result and α-β fixed
Rotational angular velocity of the child magnetic flux vector or of the motor rotor
Number K determined based on the angular velocity of rotation_sUsing, t₆/ T_c= 3^1/2/ 2 · K_s/ Cos φ₀ t₀/ T_c= 1-t₆/ T_c And the magnitude of the α-β stator flux vector is calculated as
In response to a larger value than the corresponding command value, the second voltage
Vector output time t₂, When outputting the third voltage vector
Interval t₀Each of t₂/ T_c= 3^1/2/ 2 · K_s/ Cos (π / 3-φ₀) t₀/ T_c= 1-t₂/ T_c The inverter control method according to claim 3, which is calculated by 6. α calculated at each predetermined interval T _c
The average rotational angular velocity ω of the α-β stator magnetic flux vector during a predetermined period of the interval T _c or more is calculated based on the angle φ formed between the −β stator magnetic flux vector and the reference line provided on the α-β coordinate. In addition, the magnitude of the α-β stator magnetic flux vector | λ | and the deviation of the command value corresponding to the torque calculation result Δ
6. The inverter according to claim 4, wherein the number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc by τ and the DC voltage V _dc of the inverter. Control method. 7. The predetermined interval T_cΑ calculated for each
Calculated based on −β stator flux vector and α-β current
The interval T based on the rotor position angle _cthat's all
Average rotational angular velocity ω of the rotor position angle during a predetermined period of_eCalculate
And the magnitude of the α-β stator magnetic flux vector | λ
│, the deviation of the command value corresponding to the torque calculation result Δτ,
DC voltage V of the burner_dc, The above number K by proportional gain G_s
To K_s= (2^1/2・ ｜ λ ｜ ・ ω_e+ G ・ △ τ) / V_dc Inverter according to claim 4 or claim 5, which is calculated by
Data control method. 8. Coordinates in which two axes orthogonal to each other are detected by detecting a line current and a terminal voltage of an AC motor (3) connected to an inverter (2) and fixing these detected values in a predetermined direction of a motor stator. (Α-β coordinate) and at the same time, the converted α-β current and α-β voltage cause the motor torque and α
In an inverter control device that calculates a -β stator magnetic flux vector, selects a voltage vector according to a command value corresponding to these calculation results, and controls switching of each transistor of the inverter (2) at predetermined intervals T _c . The above process is performed to select two or more types of voltage vectors based on the angle φ formed between the α-β stator magnetic flux vector and the reference line provided on the α-β coordinates, and output each selected voltage vector. The time is the deviation of the command value corresponding to the torque calculation result of the AC motor and the rotation angular velocity of the α-β stator magnetic flux vector, or
The pulse width is stored in the voltage vector calculation means that calculates based on the rotation angular velocity of the motor rotor, and in the PWM timer that performs on / off control of each transistor of the inverter (2) based on the selection result of the voltage vector and the output time calculation result. However, the inverter control device includes switching control means for performing the switching operation twice or more during the interval T _c . 9. The voltage vector calculation means calculates a deviation of a command value corresponding to the magnitude of the α-β stator magnetic flux vector, and in response to the calculation result being within a predetermined range, α- 9. The method according to claim 8, wherein three types of voltage vectors are selected based on the angle φ formed by the β stator magnetic flux vector and the reference line provided on the α-β coordinate, and the output time distribution of each voltage vector is calculated. The described inverter control device. 10. The voltage vector calculation means is α-β
The deviation of the command value corresponding to the magnitude of the stator magnetic flux vector is calculated, and in response to the result of the calculation being outside the predetermined range, the α−β of the stator magnetic flux vector α− at each predetermined interval T _c. Angle φ formed with the reference line on the β coordinate
9. The inverter control device according to claim 8, wherein two types of voltage vectors are selected based on the above, and the output time distribution of each voltage vector is calculated. 11. The voltage vector calculation means sets the output time t _{6 of} the first voltage vector, the output time t ₂ of the _second voltage vector, and the output time t ₀ of the third voltage vector to α− 0 to 60 calculated from the angle φ formed between the β stator magnetic flux vector and the reference line provided on the α-β coordinate
Using the phase angle φ ₀ that changes within °, the deviation of the command value corresponding to the torque calculation result, the rotational angular velocity of the α-β stator magnetic flux vector, or the number K _s determined based on the rotational angular velocity of the motor rotor, , T ₆ / T _c = K _s · sin (π / 3−φ ₀ ) t ₂ / T _c = K _s · sin φ ₀ t ₀ / T _c = 1-t ₆ / T _c −t ₂ / T _c The inverter control device according to claim 9, which is calculated by the following. 12. The voltage vector calculation means is α-β.
In response to the magnitude of the stator magnetic flux vector being smaller than the corresponding command value, the output time t of the first voltage vector
₆ , the output time t ₀ of the third voltage vector, α
-Β The phase angle φ ₀ that changes within 60 ° from 0 ° calculated from the angle φ formed with the reference line provided on the α-β coordinates of the stator magnetic flux vector, and the command value corresponding to the torque calculation result Using the deviation and the rotational angular velocity of the α-β stator magnetic flux vector or the number K _s determined based on the rotational angular velocity of the motor rotor, t ₆ / T _c = 3 ^1/2 / 2 · K _s / cos φ ₀ t ₀ / T _c = 1-t ₆ / T _c, and in response to the magnitude of the α-β stator magnetic flux vector being larger than the corresponding command value, the output time t ₂ of the second voltage vector , The output time t ₀ of the third voltage vector is t ₂ / T _c = 3 ^1/2 / 2 · K _s / cos (π / 3−φ ₀ ) t ₀ / T _c = 1-t ₂ The inverter control device according to claim 10, which is _operated by / T _c . 13. The voltage vector calculation means, based on an angle φ formed by a reference line provided on the α-β coordinates of the α-β stator magnetic flux vector calculated at predetermined intervals T _c , based on the interval φ. The average rotational angular velocity ω of the α-β stator magnetic flux vector for a predetermined period of T _c or more is calculated, and the magnitude of the α-β stator magnetic flux vector | λ | and the deviation Δ of the torque calculation result and the corresponding command value Δ. 13. The number K _s is calculated by K _s = (2 ^1/2 · | λ | · ω + G · Δτ) / V _dc with τ and the DC voltage V _dc of the inverter.
Inverter control device according to. 14. The voltage vector calculation means, based on the α-β stator magnetic flux vector calculated at predetermined intervals T _c and the rotor position angle calculated based on the α-β current, the interval T _c. The average rotational angular velocity ω _e of the rotor position angle during the above predetermined period is calculated, and α-β
The magnitude of the stator flux vector | lambda |, the deviation of the torque calculation result as the corresponding command value △ tau, the DC voltage V _dc of the inverter, the number of K _s by a proportional gain ^{_{G, K s = (2 1/2}} · 13. The calculation according to | λ | · ω _e + G · Δτ) / V _dc.
Inverter control device according to.