DE10356133A1 - Diesel engine combustion engine combustion initiation time measurement procedure uses acceleration value from differentiated crank shaft angular velocity meaurement - Google Patents

Abstract

A diesel engine combustion initiation time measurement procedure uses a crank shaft angular velocity measurement to derive angular acceleration from differentiation of an interpolated error corrected angle sensor filtered with an low pass FIR (Finite Impulse Response) filter. Independent claims are included for equipment using the procedure.

Description

field of use

The The present invention relates to a method and an apparatus for determining the time of commencement of the combustion process in internal combustion engines, hereinafter referred to as start of combustion, in particular of diesel engines.

A essential parameter for the assessment the combustion process in internal combustion engines is the start of combustion. The knowledge about the exact start of combustion is for the regulation and control of the combustion process as well as the diesel engine and the gasoline engine necessary. When gasoline engine is the ignition of the fuel-air mixture by a spark plug initiated. Depending on the development of the flame kernel, a certain amount passes Time until a predetermined proportion of the amount of heat supplied by the fuel was implemented. When gasoline engine, one strives to this so-called ignition delay low to avoid uncontrolled auto-ignition.

At the Diesel engine passes between injection time of fuel into the combustion chamber and the time of ignition also one certain ignition delay. The ignition delay depends on the type of fuel, the temperature and the pressure in the combustion chamber Therefore, the exact timing of the start of combustion is difficult predictable. Again, the aim is to keep the ignition delay low, to the noise Cheap to design, so-called knocking sounds to avoid and one to achieve high efficiency. On the other hand, a later start combustion a reduction in peak temperature and therefore one Reduction of nitric oxide formation. The start of burning should therefore be as possible can be regulated that he mentioned reasons not too soon and not too late he follows. Especially with newer combustion methods, such as homogeneous or partially homogeneous combustion, one is dependent on To be able to determine the start of burning precisely to the combustion process to control or regulate.

State of technology

Out The prior art discloses direct and indirect methods of determination the beginning of combustion known. Direct procedures are based on the measurement and evaluation of the combustion chamber pressure curve. For this purpose is the pressure curve measured by sensors directly in the combustion chamber thermodynamic evaluated and you get thus a measure of the start of burning. This method is the standard procedure in engine development It has the disadvantage that the effort for the sensors and the evaluation is very high, especially since usually in each Cylinder requires a sensor becomes. The installation of the combustion chamber pressure sensors in the cylinder head is quite expensive. Another problem is that the sensors in the combustion chamber are there, there high pressure and combustion directly are exposed and usually not the required fatigue strength exhibit. Therefore, this direct procedure is usually only used in the laboratory.

indirect Methods, for example, initiate the start of combustion from acceleration signals which are measured on the cylinder head of one or more cylinders. The start of combustion determination on the basis of acceleration signals is in the dissertation "Model-based Regulation of selected Drive components in the motor vehicle "by Ralf Schernewski, Institute of Industrial Information Technology of the University of Karlsruhe, 1999, p.132-143 been described. These signals will be similar to the knock detection processed. Using the envelope analysis (Bandpass, rectification, low pass) become the acceleration signals demodulated. Subsequently the demodulation result is integrated and the angle at which exceeded a certain threshold becomes, as measure for the beginning of burning used.

In addition, there are methods that gain the start of combustion from ion current signals. Such a procedure is over EP1113255A2 known. Here, an ion current measurement is carried out directly in the combustion chamber of a diesel engine in order to make a statement about the combustion process in the combustion chamber based on the thus obtained ion current signal.

The acceleration methods based on indirect methods the disadvantage of occurring in practice low signal to noise ratio compared to the measured signals Non-combustion-related structure-borne sound. opening and closing operations of Gas exchange valves and other impulsive noises are superimposed on the combustion noise. The detection of the actual Burning start is due to the poor signal to noise ratio with great uncertainties connected. This also applies to Methods based on ion current signals.

Task of invention

The The object of the present invention is a method and a device for the safe determination of the start of burning with low sensory Effort with good signal to noise ratio specify.

presentation the invention

The object is achieved by the method of the main claim 1 solved. Claim 18 specifies a device for carrying out the method according to claim 1. Advantageous embodiments of the method are the subject of the dependent claims.

Of the The present invention is based on the finding that from the Measuring the angular velocity of the crankshaft and a digital Further processing of the measured signals with relatively little effort the start of burning accurate and reliable can be determined. In contrast to the one already mentioned above known methods can do without measurements directly in the combustion chamber become. This has the advantage that the sensors used for Measuring the angular velocity of the crankshaft a far exposed to less wear are as the sensors in the combustion chamber in the known method. moreover are the disturbance parts in the invention less than in the methods of the prior Technology.

According to the invention, first the angular velocity of the crankshaft is determined. For this purpose, methods are already known from the prior art. A preferred embodiment is in the patent DE4445684C2 The method is based on the preferably non-contact detection of the angular velocity on a sprocket wheel with sprocket mounted on the crankshaft Such a rotary encoder wheel with a sensor for detecting the rotational movement is today By differentiation, in particular with an FIR differentiation filter, the angular acceleration is calculated In patent application WO03 / 062620A1 "Method for Determining and Compensating Geometric Errors of a Rotary Encoder Wheel" is described, as in this method geometrical errors of the encoder wheel can be determined and compensated, and thus a highly accurate determination of the angular velocity can be realized. With the angular acceleration determined by these known methods, the start of combustion b is then determined by further digital signal processing estimmt. For this purpose, several approaches are proposed according to the invention: On the one hand, with the aid of the angular pressure, the second derivative of the angular velocity, the start of combustion can be determined. For all other methods described, the net torque must first be determined. Three methods are described which determine the start of burning on the basis of the phase angle of the cylinder specific net torque. It is also possible to computationally determine the torque component as a result of compression and expansion and thus calculate the purely combustion-related torque component. On the basis of the purely combustion-related torque component, the start of combustion can be derived, for example, by the position of a threshold value overshoot. Another method is to calculate the cylinder-specific pressure curve on the basis of the net torque with the help of model-based equations and a subsequent pressure curve analysis, with which the start of combustion is determined.

By the relatively easy to implement, accurate determination of the start of burning let yourself the time of the start of combustion even during normal driving precisely record and control, this is particularly the case with new combustion methods, as the homogeneous or partially homogeneous combustion important. A late beginning For example, combustion is often used as a means of reducing Peak temperature and thus used for the reduction of nitrogen oxide formation. this makes possible a fuel-efficient and almost soot-free combustion in diesel engines. Another important application of the present invention is in the reduction of disturbing knocking sounds by controlling the start of combustion.

preferred embodiments The present invention will be described below with reference to FIGS attached drawings explained in detail. Show it:

1 a schematic block diagram of a device according to the invention for determining the angular acceleration;

2 a schematic block diagram of a device according to the invention for determining the start of firing based on the angular pressure;

3 a schematic block diagram of a device according to the invention for determining the Brennbe ginns based on the cylinder specific phase angle of the net torque;

4 Simulation results for the course of the combustion chamber pressure at a mean pressure of about 6 bar and a speed of 3000 rpm; the start of combustion was varied between 0 ° and 10 ° crank angle;

5 Simulation results for the course of the alternating torque at a mean pressure of about 6 bar and a speed of 3000 rpm in a crank angle range of 120 °; the start of combustion was varied between 0 ° and 10 ° crank angle;

6 a simulation result for the course of the alternating torque at a mean pressure of about 6 bar and a speed of 3000 rpm before and after digital low-pass filtering;

7 Simulation results for the course of the alternating torque at a mean pressure of about 6 bar and a speed of 3000 rpm after digital low-pass filtering in a range of 30 ° crank angle after top dead center (TDC); the start of combustion was varied between 0 ° and 10 ° crank angle;

8th a schematic block diagram of a device according to the invention for determining the start of combustion based on the pure combustion-related torque component;

9 a schematic block diagram of an inventive device for determining the start of combustion based on a pressure curve analysis of the cylinder-specific pressure curve.

mit dem als ideal angenommenen, konstanten Winkelabstand Δφ_soll zwischen zwei Inkrementen und der Periodendauer ΔT , die aus der Auswertung des Sensorsignals bestimmt wird. Da das reale Winkelinkrement Δφ_real fehlerbehaftet ist, d.h. Δφ_real = Δφ_soll + Δφ_e, ist auch die gemessene Winkelgeschwindigkeit fehlerbehaftet. Um die Korrektur durchführen zu können, muss vorab der individuelle Drehgeberradfehler durch Messungen bestimmt werden. In Einrichtung 17 wird der Geberradfehler ermittelt, der Fehler wird dann an die Einrichtung 12 weitergegeben, in der die mit dem Geberradfehler behaftete Winkelgeschwindigkeit korrigiert wird. Zur Ermittlung des Drehgeberradfehlers und zur Korrektur des Fehlers sind aus dem Stand der Technik Verfahren bekannt, bevorzugt wird das Verfahren, das aus der Patentanmeldung WO03/062620A1 bekannt ist. Hier wird bei einer zeitlich veränderlichen Wellendrehzahl eine Messung des Winkelgeschwindigkeitsverlaufs der Kurbelwelle, sowie eine Mittelung über die bei der Messung gewonnenen Wellendrehzahlsignale durchgeführt. Die Mittelung wird innerhalb eines Wellendrehzahlbereichs durchgeführt, in dem sich die Auswirkungen der im Verbrennungsmotor auf die Kurbelwelle einwirkenden Gas- und Massenmomente auf die Kurbelwellenwinkelgeschwindigkeit gegenseitig zumindest statistisch aufheben. Auf Grundlage des Winkelgeschwindigkeitsverlaufs werden die Geometriefehler des Drehgeberrads ermittelt. Die zugehörigen Verfahrensschritte zur Ermittlung und Korrektur des Geberradfehlers sind der genannten Veröffentlichung zu entnehmen. 1 shows a schematic block diagram of a device according to the invention for determining the angular acceleration. In facility 11 First, the angular velocity of the crankshaft, in particular based on a non-contact detection of the rotational movement of a rotary encoder, measured. The rotary encoder wheel preferably has a sprocket, so that the time sequence of the teeth can be detected in a conventional manner by a sensor. The detection of the teeth as increments is typically optical or inductive. The encoder wheel has manufacturing or assembly-related reasons unavoidable geometrical Drehgeberradfehler, eg not exact distances of the teeth of the ring gear, which lead to measurement errors of the angular velocity. The measured angular velocity ω _mess results from

with the assumed as ideal, constant angular distance Δφ _soll between two increments and the period .DELTA.T, which is determined from the evaluation of the sensor signal. Since the real angle increment Δφ _{real is} error-prone, ie Δφ _real = Δφ _soll + Δφ _e , the measured angular velocity is also error-prone. In order to carry out the correction, the individual encoder error must first be determined by measurements. In facility 17 the encoder wheel error is determined, the error is then sent to the device 12 passed, in which the angular error associated with the encoder wheel error is corrected. In order to determine the encoder wheel error and to correct the error, methods are known from the prior art; the method known from the patent application WO03 / 062620A1 is preferred. Here, at a time-variable shaft speed, a measurement of the angular velocity curve of the crankshaft, as well as an averaging over the wave speed signals obtained during the measurement is performed. The averaging is carried out within a shaft speed range in which the effects of the gas and mass torques acting on the crankshaft in the internal combustion engine on the crankshaft angular velocity cancel each other out at least statistically. Based on the angular velocity curve, the geometrical errors of the encoder wheel are determined. The associated method steps for determining and correcting the encoder wheel error can be found in the cited publication.

Correcting the encoder wheel error gives a very accurate angular velocity signal. This signal is in setup 13 interpolated, preferably to a resolution of 1 ° to 2 °, and then done in facility 14 a filtering with a low pass, preferably with a FIR low pass. The interpolation can be performed with any known interpolation method, such. As linear interpolation or splines interpolation, are performed. The cutoff frequency of the low pass depends on the further procedure. If the start of combustion is determined by means of angular pressure, the ideal cut-off frequency must be determined experimentally. If the start of combustion is determined with the aid of the cylinder-specific phase angle of the net torque, it is advantageous to select the ignition frequency of the engine as the cutoff frequency. To determine the angular acceleration is in device 15 the signal differentiates, preferably the differentiation is performed with an FIR differentiation filter. The use of FIR filters has the advantage that Runtime distortions are avoided and FIR filters are stable. The running time of the filter arrangement is constant and only dependent on the filter length. The FIR differentiation filter has the further advantage that the differentiation bandwidth can be selected. Limiting the bandwidth is usually required in differentiation.

2 shows a schematic block diagram of a device according to the invention for determining the start of combustion based on the angular pressure. The angular pressure is the second derivative of the angular velocity and thus describes the change in the angular acceleration. After either the ignition in the Otto engine or the injection in the diesel engine, the gradient of the combustion chamber pressure is higher after the onset of combustion than in the expansion without combustion. As a result, the gradient of the net torque of the crankshaft is also higher than in the expansion without combustion. The higher gradient at the onset of combustion manifests itself in increased angular pressure. The position of the maximum angular pressure represents a measure of the start of combustion. The angular pressure is in the device 21 from the determined course of the angular acceleration 16 by differentiation, again preferably with an FIR differentiator. Depending on the signal-to-noise ratio (SNR), the bandwidth of the FIR differentiator must be adapted to the bandwidth that is also used when calculating the angular acceleration in the device 15 has been used. In facility 22 In the course of the angular pressure, the time of maximum angular pressure in cylinder-specific crank angle windows is searched for. The left-hand boundaries of the search windows are dynamically adjusted by the ignition or injection timing. The values can be taken from the motor control. The width of the search window is in the range 10 ° to 40 ° crank angle. The time of the maximum angle pressure in a cylinder-specific crank angle window indicates the cylinder-specific start of combustion.

φ . : Winkelgeschwindigkeit der Kurbelwelle
Θ : Trägheitsmoment des Motors
Θ': Ableitung des Trägheitsmomentes nach dem Kurbelwinkel M _g: Gleichanteil (Mittelwert) des Gasdrehmomentes
M ∼_g: Wechselanteil des Gasdrehmoments M _r: Reibungsdrehmoment M _l: Lastdrehmoment 3 shows a schematic block diagram of a device according to the invention for determining the start of combustion based on the cylinder-specific phase angle of the net torque. In addition to a higher gradient in the course of the moment, the change in the start of combustion causes a change in the phase angle of the cylinder-specific torque component. The cylinder-specific torque component reflects the expansion process of the respective cylinder. From the specific angular acceleration 16 will be in setup 31 first determines the net torque applied to the crankshaft. The net torque can be determined on the basis of the angular velocity as described in the publication "Torque Determination in Internal Combustion Engines by Evaluation of Crankshaft Angular Velocity" by H. Fehrenbach, W. Held and F. Zuther in Motortechnische Zeitschrift 5/1998 The calculation is based on the following Model presentation of the torque balance:

φ. : Angular velocity of the crankshaft
Θ: Moment of inertia of the motor
Θ ': Derivation of the moment of inertia after the crank angle M _g : DC component (mean value) of the gas torque
M ~ _g : alternating part of the gas torque M _r : friction torque M _l : load torque

In steady state, ie load and speed are constant, are the drive torque M _g and the sum of friction torque M _r and load torque M _l in balance, leaving GI. (2) results

The left side of the GI. (3) is called net torque and settles determine as follows: Θ and Θ 'are from the motor data for geometry and to calculate masses.

sind, wie oben bereits beschrieben, auf Grundlage der Messungen der Winkelgeschwindigkeit der Kurbelwelle zu bestimmen. Im stationären Fall ergibt demnach die Bestimmung des Nettdrehmoments das mittelwertfreie Wechseldrehmoment. Im instationären Fall gilt die Beziehung aus GI. (3) nicht, das Nettodrehmoment enthält einen Gleichanteil, um den der Wechselanteil schwingt, d.h. das Nettodrehmoment ist nicht mehr mittelwertfrei. Der Gleichanteil kann bei Beschleunigungs- oder Bremsvorgängen auch innerhalb eines Arbeitsspiels, das von 0° bis 720° Kurbelwinkel reicht, zeitlich veränderlich sein. Näheres ist der oben genannten Veröffentlichung zu entnehmen.The angular velocity φ. and angular acceleration

are to be determined on the basis of the measurements of the angular velocity of the crankshaft, as already described above. In the stationary case, therefore, the determination of the net torque results in the mean-value-free alternating torque. In the transient case, the relationship is GI. (3) not, the net torque contains a DC component to that of the Alternating component oscillates, ie the net torque is no longer averaging. The DC component can also be variable in time during acceleration or braking processes within a working cycle that ranges from 0 ° to 720 ° crank angle. For details, see the above publication.

From that in the institution 31 certain net torque is in facility 32 determines the cylinder-specific phase angle. The cylinder-specific phase angle can be determined in various ways: One possibility is to determine the positions of the zero crossings in the course of the alternating torque at steady-state operating points where speeds and load are constant. For this purpose, first the alternating torque of a low-pass filtering, preferably with a low-pass FIR, should be subjected. To illustrate this, simulations were performed with a simple, thermodynamic model. The model assumes polytropic compression and expansion. The resulting pressure curve is superimposed on the combustion of the fuel-air mixture, which is described with a simple Vibebrenngesetz. According to the laws of the crank mechanism, the course of the torque is calculated from this. After elimination of the mean torque results in the following change torque. The following simulations were carried out using the example of a six-cylinder diesel engine at a steady-state operating point with a mean pressure of approx. 6 bar and a speed of 3000 rpm. The start of combustion was varied between 0 ° and 10 ° crank angle.

4 shows simulation results for the course of the combustion chamber pressure for the differently varied times of the start of combustion. As expected, the peak caused by the combustion shifts towards later crank angles in the course of the pressure.

5 shows simulation results for the course of the alternating torque according to the pressure curves 4 , The curves show a section of the crank angle range of 120 ° in the expansion range of a cylinder and the subsequent cylinder following it, which is in the compression phase. The shift of the combustion-related pressure peak is reflected in a shift of the corresponding torque curve. However, the position of the zero crossing in the alternating torque is not yet a clear feature for the displacement of the alternating torque. The alternating torque curves are therefore subjected to digital low-pass filtering in a further processing step.

6 shows a simulation result for the course of the alternating torque before and after the digital low-pass filtering. The low pass was an FIR filter. It is advantageous for the cut-off frequency of the FIR filter to select the ignition frequency of the motor, in the present case therefore in the 6th order.

7 shows simulation results for the history of the alternating torque after the digital low-pass filtering in a range of 30 ° crank angle after top dead center (TDC). After FIR filtering, the zero crossing of the alternating torque can be used as a suitable measure of the shift of the start of combustion. It can be seen that the shift of the start of combustion from 0 ° to 10 ° crank angle results in a shift of the zero crossing in the alternating torque of 5 ° to 12 ° crank angle.

A another possibility to determine the cylinder-specific phase angles is to during transient operations the Position of the points of intersection of the net torque with a respective one Balancing polynomial of net torque within each Determine working cycle. In the transient case contains the net torque from GI. (2) a DC share, which spreads over several work times change can. The equalization polynomial of the net torque corresponds to one estimate of the temporally variable DC component the net torque. It is preferred to use a balancing polynomial a trend line, often referred to in literature as the regression line, calculated. As an estimate of the DC component is a trend line as a sufficiently accurate estimate of the Considered equal share and has the advantage that the calculation less expensive is. Unlike the stationary one Case therefore give in the transient Case not the intersections of the alternating torques with the zero line, but the intersections of the net torque with the equalization polynomial or the trend line is a measure of the start of combustion at. Again, the net torque should firstly be low pass filtered, preferably with an FIR filter.

Another approach for determining the cylinder-specific phase angle is the analysis of the phase spectrum. For the analysis, within the crank angle range of a respective working cycle, which ranges from 0 ° crank angle to 720 ° crank angle, corresponding to the number of cylinders N crank angle intervals defined. Each interval thus assigned to a specific cylinder is to be chosen such that it includes the relevant expansion process of the respective cylinder. Within each of these cylinder-specific intervals, the phase spectrum is determined by means of the Fourier transformation, the net torque curve. The Fourier transformation is preferably carried out with a Fast Fourier Transformation (FFT), which can be implemented particularly effectively. From the orders to be determined as dominant in the phase spectrum, a resulting phase value is determined for each phase spectrum, which is used as a measure of the start of combustion. The relationship between the phase value and the start of combustion must be taught here in the form of a characteristic map in which the dependency of the phase value from the start of combustion as well as the speed and load is defined, ie for various stationary operating points, the points in the map are determined. A load value can be obtained either directly from the motor control or indirectly from the relationship between net torque and load torque to be determined. For the latter method for determining the load value, reference is made to the already mentioned publication "Torque determination in internal combustion engines by evaluation of the crankshaft angular velocity" by H. Fehrenbach, W. Held and F. Zuther in Motortechnische Zeitschrift 5/1998 the chassis dynamometer is uniquely related to steady-state conditions between the net rms value of the net torque and the approximately constant load torque M _l, and a load-torque value can be assigned to each working cycle by means of a corresponding map in which the measured values of speed and net effective torque are assigned during operation, the instantaneous load torque can be read off.

mit dem Anfangsdruck p₁, dem Druck des Endzustands p₂, dem Anfangsvolumen V₁, dem Volumen des Endzustands V₂ und dem Polytropenexponenten n, der allein von den technischen Bedingungen der Zustandsänderung abhängt. Als Anfangsdruck der Verdichtung wird der gemessene Ladedruck verwendet. Der entsprechende Sensor zur Messung des Ladedrucks ist in der Regel ein Standardsensor, der hinter dem Lader angebracht ist. Aus dem berechneten Druckverlauf basierend auf Kompression und Expansion wird entsprechend den Gesetzen der Kurbelmechanik der Verlauf des Drehmomentenanteils berechnet. Der rein verbrennungsbedingte Drehmomentenanteil wird durch Subtraktion 82 des Drehmomentenanteils resultierend aus Kompression und Expansion von dem in Einrichtung 31 bestimmten Nettodrehmomentenanteil berechnet. In Einrichtung 83 wird aus dem Verlauf des verbrennungsbedingten Drehmomentenanteils ein Maß für den Brennbeginn abgeleitet. Dies kann beispielsweise die Lage einer Schwellwertüberschreitung oder die Lage des größten Gradienten sein. 8th shows a schematic block diagram of a device according to the invention for determining the start of combustion based on the pure combustion-related torque component. First, in facility 31 the net torque from GI. (2) determined as described above, this can be done at steady state operating points as well as during transient operations. In facility 81 Then, under the model assumption of a polytropic compression, the torque component as a result of compression and expansion is calculated. According to the first law of thermodynamics, the model for the pressure curve is polytrope change of state

with the initial pressure p ₁ , the pressure of the final state p ₂ , the initial volume V ₁ , the volume of the final state V ₂ and the polytropic exponent n, which depends solely on the technical conditions of the state change. As the initial pressure of the compression, the measured boost pressure is used. The corresponding sensor for measuring the boost pressure is usually a standard sensor, which is mounted behind the loader. From the calculated pressure curve based on compression and expansion, the curve of the torque component is calculated in accordance with the laws of the crank mechanism. The pure combustion-related torque component is subtracted 82 the torque component resulting from compression and expansion of the device 31 calculated net torque proportion. In facility 83 is derived from the course of the combustion-related torque component, a measure of the start of combustion. This can be, for example, the location of a threshold value violation or the position of the largest gradient.

• – Ist innerhalb des Kurbelwinkelbereichs nur das Einlassventil im jeweiligen Zylinder geöffnet, so wird für den Zylinderdruck der gemessene Ladedruck angenommen.
• – Ist innerhalb des Kurbelwinkelbereichs nur das Auslassventil im jeweiligen Zylinder geöffnet, so wird für den Zylinderdruck der gemessene Abgasgegendruck angenommen.
• – Sind innerhalb des Kurbelwinkelbereichs sowohl das Einlass- als auch das Auslassventil im jeweiligen Zylinder geöffnet, so wird für den Zylinderdruck der Mittelwert aus gemessenem Ladedruck und gemessenem Abgasgegendruck angenommen.
• – Befindet sich der Zylinder innerhalb des Kurbelwinkelbereichs bei geschlossenem Einlass- und Auslassventil in der Kompressionsphase, so wird der Zylinderdruck gemäß einer polytropen Verdichtung mit dem vor der Kompression vorgelegenem Ladedruck als Anfangsdruck ermittelt.
• – Befindet sich der Zylinder innerhalb des Kurbelwinkelbereichs bei geschlossenem Einlass- und Auslassventil in der Expansionsphase ohne Verbrennung, so wird der Zylinderdruck gemäß einer polytropen Expansion ermittelt, die als spiegelbildlich zur vorher erfolgten polytropen Kompression mit gemessenem Ladedruck als Anfangsdruck angenommen wird.

9 shows a schematic block diagram of a device according to the invention for determining the start of combustion based on a pressure curve analysis of the cylinder-specific pressure curve. First, in facility 31 the net torque from GI. (2) determined as described above, this can be done at steady state operating points as well as during transient operations. In facility 91 The cylinder-specific pressure curve is calculated on the basis of model-based equations. The mathematical superposition of the moments obtained from the combustion chamber pressures of the individual cylinders and weighting with the respective lever arm of the sliding crank mechanism and the piston surface and equating the superposition result with the net torque results in an equation for the combustion chamber pressures for each considered crank angle. In order to obtain the combustion chamber pressure profile of the cylinder in question in the region of the crank angle range in which the combustion takes place, the following assumptions are made with regard to the pressures in the other cylinders which do not burn within the considered crank angle range:

• - If only the inlet valve in the respective cylinder is opened within the crank angle range, then the measured boost pressure is assumed for the cylinder pressure.
• - If only the exhaust valve in the respective cylinder is opened within the crank angle range, then the measured exhaust back pressure is assumed for the cylinder pressure.
• - If within the crank angle range both the inlet and the exhaust valve in the respective cylinder are open, the cylinder pressure is the mean value of the measured boost pressure and measured assumed exhaust gas back pressure.
• - Is the cylinder within the crank angle range with the intake and exhaust valves closed in the compression phase, the cylinder pressure is determined according to a polytropic compression with the pre-compression pressure before the compression as the initial pressure.
• If the cylinder is within the crank angle range with the intake and exhaust valves closed in the expansion phase without combustion, the cylinder pressure is determined according to a polytropic expansion, which is assumed to be a mirror image of the previous polytropic compression with measured charge pressure as initial pressure.

As a solution of the equations thus obtained for each individual angular position within the cylinder-specific crank angle range, the pressure profile in the combustion cylinder is obtained. Since no pressure values can be determined due to the vanishing lever arm in the region of the top dead center in question here, the pressure profile at these points must be interpolated. In facility 92 the calculated cylinder-specific pressure curve is subjected to a classical pressure analysis based on the evaluation of the heating law and from this, estimated values for the start of combustion are determined.

The also described methods for determining the start of firing be advantageously combined with each other. It is possible, for example, two or perform several procedures in parallel. Due to the different Calculations usually give slightly different results for the Time of burning. A particularly exact determination of the Burning starts by averaging the different calculation results from the different methods described. Another possibility the advantageous combination of the methods is the use the different procedures depending on the operating point, i.e. dependent from the momentary load and speed. It is assumed that the different procedures for different working points are particularly suitable and provide particularly accurate results. Accordingly, the method should then be chosen depending on the operating point.

Claims (19)

Method for determining the time of commencement of the combustion process in internal combustion engines (start of combustion), in particular of diesel engines, characterized in that the determination of the start of combustion on the measurement ( 11 ) the angular velocity of the crankshaft is based. A method according to claim 1, characterized in that the determination of the start of combustion comprises the following steps: - Determination of the angular acceleration ( 16 ) on the basis of measurements ( 11 ) the angular velocity of the crankshaft; and determination of the start of combustion on the basis of the determined angular acceleration ( 16 ). A method according to claim 2, characterized in that the determination of the angular acceleration comprises the following steps: - Measurement ( 11 ) the angular velocity of the crankshaft by means of a sender wheel; and - correction ( 12 ) of the encoder wheel specific encoder wheel error; and - interpolation ( 13 ) of the angular velocity signal; and filtering ( 14 ) of the interpolated angular velocity signal with a low pass; and - differentiation ( 15 ) of the lowpass filtered angular velocity signal. Method according to Claim 3, characterized in that the interpolation ( 13 ) of the angular velocity signal to a resolution of 1 ° to 2 ° crank angle. Method according to one of claims 3 or 4, characterized in that as low pass ( 14 ) a FIR filter, in particular a steep edge FIR filter is used. Method according to one of claims 3 to 5, characterized in that for differentiation ( 15 ) of the angular velocity signal, an FIR differentiation filter is used. Method according to one of claims 2 to 6, characterized in that the determination of the start of combustion based on the angular acceleration ( 16 ) comprises the following steps: - differentiation ( 21 ) the angular acceleration for calculating the angular pressure; and - determination ( 22 ) the position of the maxima of the angular pressure in crank angle windows as a measure of the start of combustion. Method according to claim 7, characterized in that for differentiation ( 21 ) of the angular acceleration signal, an FIR differentiation filter is used. Method according to one of claims 2 to 6, characterized in that the determination of the start of combustion based on the angular acceleration ( 16 ) comprises the following steps: - determination ( 31 ) a net torque; and - determination ( 32 ) of the cylinder-specific phase angle of the net torque as a measure of the start of combustion. Method according to claim 9, characterized in that for the determination ( 32 ) of the cylinder-specific phase angle, the position of the zero crossing in the course of the net torque at stationary operating points is determined. Method according to claim 9, characterized in that for the determination ( 32 ) of the cylinder-specific phase angle, the position of the point of intersection of the net torque with a respective compensation polynomial, in particular a trend line, of the net torque within the respective working cycle is determined during transient operations. Method according to claim 9, characterized in that that the cylinder-specific phase angle from the cylinder-specific Phase spectrum is determined. Method according to one of claims 2 to 6, characterized in that the determination of the start of combustion based on the angular acceleration ( 16 ) comprises the following steps: - determination ( 31 ) a net torque; and - determination ( 81 ) the torque component resulting from compression and expansion, using a known boost pressure as initial compression pressure; and determination of the combustion-related component of torque by subtraction ( 82 ) the torque component resulting from compression and expansion from the net torque; and - analysis ( 83 ) of the course of the combustion-related torque component for determining the start of combustion, in particular by determining the position of a threshold value exceeding or the largest gradient. Method according to one of claims 2 to 6, characterized in that the determination of the start of combustion based on the angular acceleration ( 16 ) comprises the following steps: - determination ( 31 ) a net torque; and - determination ( 91 ) of the cylinder-specific pressure curve based on the net torque based on model-based equations; and - pressure history analysis ( 92 ) of the cylinder-specific pressure curve for determining a cylinder-specific value for the start of combustion. Method according to at least one of claims 7 to 14, characterized in that the methods for determining the Combustion can be combined. Method according to claim 15, characterized in that that for the combination of the methods for determining the start of firing two or more procedures are performed in parallel and the respective ones Values for the time of the start of burning are averaged. Method according to claim 15, characterized in that that for the combination of the methods for determining the start of firing the procedures in dependence of the respective operating point. Device for determining the point in time of the beginning of the combustion process in internal combustion engines (start of combustion), in particular of diesel engines, characterized in that the determination of the start of combustion on the measurement ( 11 ) the angular velocity of the crankshaft is based. Apparatus according to claim 18 for determining the start of combustion with - a device for determining the angular acceleration ( 16 ) on the basis of measurements ( 11 ) the determined angular velocity of the crankshaft; and - means for determining the start of burning on the basis of the angular acceleration ( 16 ).