DE102011103427A1 - Method for knock control of internal combustion engine, involves assigning knocking limit, and determining operating point and cylinder-selective default value for gradient of knock level curve in response to firing angle - Google Patents

Abstract

The method involves assigning knocking limit, and determining an operating point and cylinder-selective default value (13) for a gradient of a knock level curve in response to a firing angle (2). The operating point- and cylinder-selective default value for the knock level curve gradient is stored in a control unit. A regulator is provided with an actual value, an adaptable reference value and a manipulated variable. An operating point- and cylinder-selective adaptation is carried out during the operation of an internal combustion engine. An independent claim is provided for a device for performing the knock control method.

Description

Technical area

The present invention relates to a method and a device for knock control of a spark-ignited internal combustion engine, wherein the internal combustion engine is operated at optimum efficiency at the knock limit.

State of the art

In an Otto engine combustion under normal conditions, the mixture is specifically ignited by a spark ignition. During combustion, the flame front moves from the ignition location three-dimensionally through the combustion chamber until the mixture is burnt. In knocking combustion, part of the mixture at hot spots of the combustion chamber spontaneously ignites due to the increased combustion pressure and the increased combustion chamber temperature during combustion, even before the flame front reaches these areas. It comes to an explosive combustion, which increases the combustion speed and results in a correspondingly high combustion pressure. The resulting pressure waves and pressure oscillations propagate through the components of the internal combustion engine as structure-borne noise and are perceived as knocking.

To increase the efficiency of spark-ignition internal combustion engines high compression ratios are sought. In contrast, however, stands the associated increased tendency to knocking burns. Due to the knocking burns, the thermal and mechanical load on the internal combustion engine and thus the risk of destruction of the internal combustion engine is increased. However, low knock levels do not necessarily cause damage even with ongoing knocking burns, so the knock level must always be considered in conjunction with the frequency of knocking burns. For rapid suppression of dangerous beating burns, the ignition timing is adjusted as a priority. Furthermore, other methods are possible to counteract knocking burns.

To detect knocking burns, knock sensors are generally used, which are arranged, for example, on the crankcase of the internal combustion engine and determine the residual body sound of the knocking combustion present there. The knock sensors or structure-borne noise sensors convert the structure-borne sound vibrations resulting from the knocking combustion into alternating-current signals. The thus-detected structure-borne sound signals of the knock sensor are fed to an evaluation unit, for example an engine control unit, and processed accordingly by filters, so that dangerously knocking burns can be detected.

In order to operate the engine as efficiently as possible at the knock limit for optimizing efficiency, knock controls are provided. In these knock control knocking burns are allowed up to a predetermined knock strength or knock intensity, which does not lead to damage to the internal combustion engine. Conventional knock control methods form a signal of the structure-borne sound oscillation which is averaged over several combustion cycles and represents a reference signal. To assess the combustion, a currently determined structure-borne noise signal is then compared with the previously determined reference signal. For this purpose, a distance or a ratio between the currently determined structure-borne noise signal and the reference signal is evaluated. If the reference signal is designated as noise and the currently determined structure-borne noise signal as a signal, it can be assumed that a signal-to-noise ratio or a signal-to-noise ratio to be evaluated is to be evaluated. Assuming that the signal of knocking burns always stands out clearly from the non-knocking burns, so that the signal is noticeably larger than the noise, the detection value of the signal-to-noise ratio, above which knocking combustion occurs, is above cylinder speed and load stored in a map and thus fixed.

Since the ignition point, which is specified as the ignition angle at a distance from the top dead center of the internal combustion engine between the compression and power stroke, the combustion and the tendency to knock significantly affected, this is set later in time with detected knocking combustion. A knocking combustion is detected when a current signal-to-noise ratio exceeds a corresponding detection value. The signal-to-noise ratio at which a change occurs between non-hazardous knocking combustion and dangerously knocking combustion is also referred to as a knock limit.

From these conventional knock control methods, as described, for example, in US Pat DE 195 49 175 A1 or EP 0 166 033 B1 and a variety of other writings are derived, among other things, some disadvantages. For example, it takes a lot of effort to optimize the signal-to-noise ratio. The knock sensor is to be arranged on the internal combustion engine so that mainly combustion noise can be detected in comparison to noise. The noise directly affects the signal-to-noise ratio. The structure-borne sound signal detected by the knock sensor, which is a Knocking level must be subjected to a complex signal processing to minimize the noise component. The knock control procedures must be adapted so that also scatters that are to be expected in mass production of internal combustion engines are included.

In the conventional knock control methods, measures are taken to prevent knocking combustion after the dangerously beating burns have taken place. Since the reference level is formed from the combustion noise of the internal combustion engine itself, any increase in the noise of the internal combustion engine leads to an increase in the reference level. At a fixed signal-to-noise ratio thus increases the minimum detectable actual level, so that knocking burns are not recognized as such and the destructive burns can not be counteracted. These knocking sounds in turn lead to a further increase of the reference level, whereby a process of automatic deactivation of the knock control can be set in motion. In contrast, a reduction in the noise level results in a reduction in the minimum detectable actual level, so that dangerously knocking burns are detected although no destructive combustion has occurred. The ignition angle would be unnecessarily late, which would result in an increased loss of efficiency. In internal combustion engines with low combustion noise, poor transmission of structure-borne noise or unfavorable knock sensor positioning, knock control can only be carried out inadequately.

Due to the relative evaluation of structure-borne sound vibrations, changes in the noise level of the internal combustion engine due to aging or transit time are insufficiently taken into account. A trend evaluation is not possible. Stochastic disturbances due to worn or defective engine components cause increased noise levels, which are falsely recognized as knocking combustion.

Object of the invention

The object of the invention is to provide an improved method for knock control and an apparatus for carrying out the method.

Solution of the task

The object is achieved by the method according to the features of claim 1 and the device according to the features of claim 7.

Description of the invention

The invention provides an improved over the prior art method and an improved device for knock control of internal combustion engines, wherein an improved adaptation of a knock limit is carried out. The advantageous invention provides an adaptable knock control system that independently adapts itself from an internal combustion engine to further model lines of the engine family as well as to individual differences of the respective internal combustion engine from a series production without additional effort. The process steps necessary for the method are divided into a development phase and a use phase of the internal combustion engine. During the development phase engine control unit specifications are performed based on an exemplary internal combustion engine, which are valid for a complete series production of the corresponding internal combustion engines. The adaptation of the knock control is then carried out in the use phase of the respective internal combustion engine from this series production.

For this purpose, an exemplary internal combustion engine is placed on an engine test bench during the development phase or application phase and an ignition angle variation is carried out for each operating point and each cylinder. The operating point is determined by an engine speed and an engine load. The engine load can be described for example by an engine torque or by a combustion medium pressure. The interpolation points of the operating points for the engine speed and engine load are predefined according to a suitable resolution of a characteristic field. The Zündwinkelvariation is performed for each operating point and each cylinder by the ignition angle is adjusted starting from a current ignition angle in the late direction and then in predefinable steps back towards early until the knock limit has been briefly and significantly exceeded. Subsequently, the ignition angle is adjusted again in the direction of late. As ignition angle is the angle during operation of the internal combustion engine to understand in which an ignition is initiated at a certain distance to the temporally following top dead center of the piston between the compression and power stroke. The ignition angle defines the ignition point at which the spark and thus the ignition of the air-fuel mixture is initiated. In the case of an internal combustion engine operating according to the four-stroke principle, this top dead center is referred to as the top dead center or briefly as the ignition TDC. For optimum combustion efficiency, the ignition timing is before top dead center. Accordingly, the ignition angle is retarded when the ignition angle is reduced and the ignition timing is timed closer to the following top dead center. In contrast, an adjustment is made early by increasing the ignition angle. As a termination criterion for the adjustment of the ignition angle, a predefined knock intensity is used, for example, by means of a cylinder pressure measurement, which describes a knock limit. The exceeding of the knock limit during the ignition angle adjustment serves to be able to determine the knock limit itself exactly. During operation of the internal combustion engine on the engine test bench, the combustion pressure in the interior of the cylinder is detected by a pressure sensor and evaluated accordingly as knock strength. The knock intensity is a measure of the extent to which a pressure pattern of a dangerous knocking combustion differs from a normal combustion. For knock strength, a threshold is defined which describes the knock limit. The knock strength is given as pressure. During the adjustment of the ignition angle, the structure-borne sound waves of the internal combustion engine are detected by at least one knock sensor and output as voltage. For the knock limit is then in addition to knock strength and a knock level available.

In an advantageous manner according to the invention, the adjustment of the ignition angle in the direction of late and early in a predefined ignition angle sequence, which can also be specified depending on operating point, performed.

The resulting measured values for knock intensity and knock level as a function of the ignition angle can be described or approximated by compensation functions. From the knocking level ignition angle course represented, for example, by a polynomial of at least a second degree, a gradient curve, that is to say the change in the knocking level as a function of the ignition angle, can be determined. This is done in a simple manner by deriving the compensation function according to the ignition angle. The knock limit can thus be described by the knock strength, the knock level and additionally by the knock level gradient. For the description of the knock limit over the knock level gradient, the non-linear relationship between knock level, ie knock intensity and ignition angle early adjustment, is utilized. The knocking level increases disproportionately with an increasingly earlier ignition angle. Thus, the Klopfpegelgradient increases with increasingly earlier ignition angle. Thus, a threshold value of a knock level gradient can be used as a knock limit.

This threshold value of a knock level gradient is stored as an operating-point-dependent and cylinder-selective default value in an engine control unit for the internal combustion engine. The inventive advantage of the method is that a default value is used, which is valid regardless of the background noise of the internal combustion engine.

In a further step, a mean value of an operating-point-dependent knocking level representative of a series deviation is determined in the development phase from an adequate random sample of internal combustion engines and stored in each individual engine control unit of an internal combustion engine. At the end of the development work on the test stand, a default value for the knock level gradient and a knock level are stored in the engine control unit of the internal combustion engine for each operating point and each cylinder as the mean value of a series spread for representing the knock limit.

In the use phase of the individual internal combustion engine installed, for example, in a vehicle, a knock level adaptation is performed on the basis of the default values already stored in the engine control unit in order to determine the ignition angle for optimal operation at the knock limit. The knock level adaptation is necessary because each individual internal combustion engine has an individual signal transmission of the knocking combustion via the engine structure and the sensors to the evaluation unit and thus the individual internal combustion engine is subject to certain tolerances and changes which influence the function and ultimately the combustion. These tolerances result from the series production, which occurs in the manufacture of internal combustion engines. Series deviations can result, for example, from manufacturing tolerances of mechanical elements or also from differences in signal acquisition and transmission. In addition to the series production of a model series of an internal combustion engine, the internal combustion engines can also be equipped with different units and modules. Also aging and wear phenomena involve a change in the operation of the internal combustion engine. These influences can be reacted by the knock level adaptation. A knock level adaptation is performed during the operation of the internal combustion engine over the entire service life under predetermined conditions and usage intervals.

Advantageously according to the invention, the knock level adaptation takes place during a quasi-stable operating state of the internal combustion engine in which only slight changes in rotational speed and load occur. If such an operating point is detected, the adaptation for the respective operating point is performed individually for each cylinder. For the detection of a quasi-stable operating point, among other things, a predefined threshold of a load and speed change speed can be used. Furthermore, for the decision to adapt Make predictions about imminent signal or operational changes of the internal combustion engine.

The operating-point and cylinder-selective knock level adaptation is effected by an ignition angle variation. In this case, the ignition angle is adjusted by means of a firing angle sequence which corresponds to the firing angle sequence during the application. During the adjustment of the ignition angle of the self-adjusting knock level is detected by the built-in internal combustion engine knock sensor. As a result, a plurality of measured values of the knocking level as a function of the firing angle are available for the respective operating point, so that a profile of the knocking level over the firing angle results. The course is again described or approximated by a compensation function. By the first derivation of the compensation function according to the ignition angle, an operating point and cylinder-selective course of the knocking level gradient is generated for the individual internal combustion engine. For the ignition angle, in which the course of the knocking gradient coincides with the default value stored in the engine control unit during the application, the corresponding knocking level is determined on the basis of the compensation function. In other words, a recalculation to a knocking level takes place on the basis of the operating-point- and cylinder-selective course of the knocking level gradient, for the firing angle of which the default value coincides with the value of the course. The thus determined operating point and cylinder-selective knock level of the individual internal combustion engine represents the knock limit at which the internal combustion engine is to be operated. The adaptive component results from the difference between the mean value of a knocking level of a series deviation already stored in the engine control unit and the individually determined knocking level. According to the invention, the setpoint value of the knock level required for the knock control results according to the invention as the sum of the mean value of a series deviation and the adaptive component. In principle, it would also be possible to use the currently determined knocking level as setpoint. In order to keep the adaptive effort but low and to allow the engine even with not yet done or unsafe adaptation safe operation, the adaptation is based on this mean of a series dispersion and thus the setpoint of a fixed proportion, the average knock level of the knock limit of the Series dispersion, and a variable, adaptable share. By means of a freely selectable weighting, first adaptations to the other operating points without successful adaptation can already be interpolated or extrapolated. Thus, the first successful adaptation already corrects the setpoint in the entire control range.

After each successful adaptation, an operating point and cylinder-dependent status counter is increased by a predefined count value and stored accordingly in a status map. The interpolation points are identical to those of the adaptation parameter field, so that each adaptation value is assigned exactly its status value. The current adaptation result is averaged with the stored value. If the difference between the two values lies outside a tolerance to be defined, the status counter is reduced in order to allow a new adaptation under corresponding conditions. The reliability of the adaptation value is guaranteed if the status counter has reached or exceeded an upper threshold. Occurs during the first part of the adaptation sequence, ie at the ignition angle early adjustment, a significant signal interference, the adaptation is terminated and restarted. If this is the case in the second part of the adaptation sequence, the result is stored in the adaptation map, as described above. The status counter is increased by a count which is reduced compared to the count value of a successful adaptation. If, during the advanced adjustment, a threshold value of a knock level stored in relative dependence on the fixed portion is reached before the desired target gradient, then it can be assumed that the noise level is outside a permissible range and the adaptation is aborted. The start values are kept unchanged. If the behavior is repeated in further adaptation experiments, the relevant cylinder-selective adaptation point is marked as not being adaptable in order to avoid further unnecessary adaptation attempts. A corresponding entry can be made in the diagnostic system.

The necessary for the knock control also actual level, which represents a current and stable knock level during operation of the internal combustion engine as a voltage signal is determined as a moving average over a defined number of cyclically consecutive knock sensor signals. The number of signals used for the averaging is predetermined by signal-influencing parameters. Such parameters may be, for example, speed, load and combustion center position as well as their rate of change in the dynamic operation of the internal combustion engine. Furthermore, changes in the signal conditioning, such as when switching the signal evaluation, or the connection or disconnection of engine components, such as charging units or cargo movement flaps, can be included in the consideration.

Knock control uses a controller that compares the actual knock level with the target knock level. The control variable used is the ignition angle. In an advantageous manner according to the invention, a PID controller is used for the controller, in which the ignition angle change from the I component in addition to the described knock level adaptation, Thus, the adaptation of the target knocking level, determined and stored in cylinder-specific maps operating point-dependent.

The described method provides a knock control with which the individual internal combustion engine can be operated significantly closer to the knock limit and thus significantly closer to the thermodynamic optimum than conventional knock control methods. As a result, the efficiency of the internal combustion engine compared to conventional internal combustion engines increases significantly, which is made clear, for example, by a fuel economy.

The method according to the invention for knock control further enables diagnostic functions. Thus, during operation of the internal combustion engine, a stochastic noise can be detected by a standard deviation of the knock level operating point and cylinder selectively stored in a map. If the mean value calculated from the characteristic field exceeds a predefined threshold value, then a presumed stochastic noise is assumed. A trend detection of the noise level of the internal combustion engine through the analysis of the D component of the controller is possible. During the adaptation, the standard deviation of the knocking level via ignition angle can also be determined. If the gradient of the thus determined curve of the standard deviation of the knock level via ignition angle in a defined range exceeds a predetermined limit value, a stochastic noise is detected. Depending on the result of this diagnosis, the parameters for the adaptation and control can be adapted. A predefined gradient of the standard deviation can also be used as a termination criterion for the ignition angle adjustment of the adaptation.

For optimum knock control, the knock sensor signal is detected, filtered and evaluated in the representative time and frequency domain. For a time window or a crankshaft angle window is defined as a function of load and speed. The frequency filtering is performed by a filter having a passband width from the first to the third combustion chamber natural frequency. The passband bandwidth is defined by a high and lowpass frequency. In contrast to conventional knock control, the window is already opened before top dead center, which is assigned to the ignition timing, in order to be able to reliably detect the actual value during the adaptation. Furthermore, it is also possible to detect pre-ignition by comparing the actual value with a corresponding threshold value and, if it is exceeded, to infer an extremely strong knocking combustion. All information resulting from the diagnostic functions can be further processed in the engine control unit in order to further optimize the operation of the internal combustion engine.

The knock control method according to the invention reduces the outlay for optimum sensor placement. The self-adaptation of the setpoint allows a stable knock control with high control quality, which adapts to series variations between different internal combustion engines, to deviations in the measuring chain, to the aging of the internal combustion engine, to different fuel qualities and to different loud noise. As a result, error detection of dangerous beating burns can be significantly reduced and the efficiency of the internal combustion engine can be increased. Furthermore, even small deviations between the setpoint and actual values can be evaluated and then reacted with small ignition angle changes, whereby the internal combustion engine can be operated correspondingly more stable and closer to the knock limit.

The device for carrying out the method includes a control device for the electronic control of the internal combustion engine. The device therefore contains several functional units with which the different method steps can be carried out. A first functional unit evaluates the structure-borne sound waves recorded by a knock sensor belonging to the device and further operating variables of the internal combustion engine. For this purpose, filter devices are used, which capture and process the signal for further processing. In a second functional unit, the setpoint and actual values required for the knock control are calculated. For this purpose, several memory devices are used, in which various parameters and maps are stored. Furthermore, a PID controller is integrated, which compares the actual values and setpoints with each other and calculates a manipulated variable intervention in a closed loop. About a third functional unit of the targeted intervention is made to the manipulated variable.

embodiment

By way of example, a possible embodiment of the method according to the invention is shown here. In the accompanying figures show:

1 : A schematic representation of the course of a knock level as a function of the ignition angle and

2 : A schematic representation of a recalculation of a specification to a setpoint for the knock level.

The knock control method according to the invention comprises a plurality of method steps, which are subdivided into an application phase or development phase and to a use phase of the internal combustion engine. During the development phase engine control unit specifications are performed based on an exemplary internal combustion engine, which are valid for a complete series production of the corresponding internal combustion engines. The adaptation of the knock control is then carried out in the use phase of the respective internal combustion engine from this series production. Within this embodiment, the procedure of the method is described at a certain operating point and a specific cylinder of an internal combustion engine.

During the development phase, for the selected operating point of the internal combustion engine, a voltage profile of a knocking level ( 1 ) by means of a knock sensor as a function of the ignition angle ( 2 ) certainly. In 1 For this purpose, a diagram for an application is shown which, inter alia, displays a knock level profile (FIG. 3 ) in dependence of the ignition angle ( 2 ) shows. The ignition angle ( 2 ) is on an abscissa axis ( 4 ) and the knock level ( 1 ) on an ordinate axis ( 5 ) applied. For the determination of the voltage curve, an adjustment of the ignition angle ( 2 ) and the corresponding knock level ( 1 ) measured as a voltage signal of the knock sensor. The resulting knock level profile ( 3 ) is then replaced by a knock level compensation function ( 6 ), which is exemplified here as a fourth-degree polynomial. For the exact determination of the knock level compensation function ( 6 ) at least as many measured values ( 6a . 6b . 6c . 6d . 6e ) are available, as are necessary for the determination of the parameters of the compensation function. At the same time to determine the knock level profile ( 3 ), a knocking strength ( 7 ), which is recorded by means of a cylinder pressure sensor, depending on the ignition angle ( 2 ) to a knock strength course ( 8th ) detected. This makes it possible to check whether the knock level course measured by the knock sensor ( 3 ) representative of the measured knock intensity profile ( 8th ). Based on an associated knocking strength compensation function ( 9 ), which analogously to the knock level compensation function ( 6 ), a predetermined knock limit may be determined by a limit knock strength ( 10 ), which correspond to a limit knocking level ( 11 ) corresponds. Through a mathematical derivation of the knock level compensation function ( 6 ), a gradient curve ( 12 ) of the knock level course ( 3 ). This means that the knock limit is next to the limit knock level ( 11 ) also has a knock level gradient of the knock limit, which is available as the default value ( 13 ) is stored in an engine control unit of each internal combustion engine of series production.

The adaptable portion of a setpoint for the knock control is determined during the use phase of an individual internal combustion engine in the vehicle, as soon as the conditions are met. For this purpose, for the individual internal combustion engine, a engine-specific knock level profile ( 14 ) via ignition angle ( 2 ). 2 shows a diagram for an adaptation analogous to the diagram in 1 , The engine-specific knock level profile ( 14 ) is controlled by a motor-specific knock level compensation function ( 15 ) whose derivation is a motor-specific gradient curve ( 16 ) of the engine-specific knock level profile ( 14 ). The engine-specific gradient course ( 16 ) deviates from series deviations and signs of wear as well as differences in the signal processing from the gradient curve ( 12 ), which was determined during the development of the internal combustion engine from. For a limit ignition angle ( 17 ) for which the gradient of the engine-specific knock level profile ( 14 ) equal to the default value ( 13 ), or is within a certain tolerance, is determined on the basis of the engine-specific knock level profile ( 14 ) a motor-specific limit knocking level ( 18 ), which determines the engine-specific limit knock strength ( 19 ) and thus represents the engine-specific knock limit determined. Despite different knock level course ( 3 ) and engine-specific knock level course ( 14 ), the engine-specific limit knock strength ( 19 ) equal to the limit knock strength ( 10 ) be. Thus, the functionality of the method according to the invention is shown. The difference between a fixed proportion of the setpoint, which was determined from a sufficient sample of internal combustion engines of series production, and the motor-specific limit knocking level valid for this internal combustion engine ( 18 ) represents the adaptive component of the setpoint and is stored in the control unit. On the basis of the adaptation thus determined, a knock control is carried out by means of a PID controller using the device provided for this purpose.

The method described in this exemplary embodiment for a specific operating point and a specific cylinder is repeatedly performed for application and adaptation of the knock control in order to obtain at least sufficiently many setpoint values within the operating range and for all cylinders of the internal combustion engine in order to enable improved operation of the internal combustion engine ,

LIST OF REFERENCE NUMBERS

1

knocking level

2

firing angle

3

Knocking level course

4

abscissa

5

axis of ordinates

6

Knocking level compensation function

6a, 6b, 6c, 6d, 6e

readings

7

knock intensity

8th

Knock intensity History

9

Knock intensity compensation function

10

Cross knock intensity

11

Border knocking level

12

gradient profile

13

default

14

Engine-specific knock level profile

15

Engine-specific knock level compensation function

16

engine-specific gradient course

17

Grenzzündwinkel

18

engine-specific limit knock level

19

engine-specific limit knock strength

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 19549175 A1 [0007]
• EP 0166033 B1 [0007]

Claims (7)

Method for knock control of an internal combustion engine, wherein a controller is used with an actual value, an adaptable setpoint and a manipulated variable, wherein the actual value is compared with the desired value and is controlled by a firing angle as a manipulated variable, the actual value of a moving average of a knocking level and the Reference value from a fixed, predetermined proportion and a variable, adaptable component is formed, characterized in that a knock limit associated, operating point and cylinder-selective default value for a gradient of a Klopfpegelverlaufs determined in dependence on a firing angle and stored in a control unit. Method for knock control according to claim 1, characterized in that an operating point and cylinder-selective adaptation during operation of the internal combustion engine is performed, in which a Zündwinkelsequenz is traversed to a target knock level, assuming equality between the default value and a current value for the gradient of Determine knock level. Method for knock control according to one of the preceding claims, characterized in that the operating point and cylinder-selective course of the knocking level is represented by a compensation function, wherein the required for the determination of the compensation function value pairs of the support points, consisting of knock level and ignition angle, determined by a defined ignition angle become. Method for knock control according to one of the preceding claims, characterized in that the gradient is determined from a profile of the knock level as a function of the ignition angle by a first derivative of the compensation function is formed. Method for knock control according to one of the preceding claims, characterized in that the controller is a PID controller. Method for knock control according to one of the preceding claims, characterized in that the knocking level is detected by a knock sensor. Device for carrying out the method for knock control, characterized in that a plurality of functional units are provided, wherein a first functional unit evaluates operating variables of the internal combustion engine, a second functional unit calculates the necessary for the knock control target and actual values and a manipulated variable intervention and a third functional unit for the targeted intervention is provided on the manipulated variable.

Images