JP2006336604A - Knocking controller for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a knocking controller for an internal combustion engine capable of suppressing influence of a filter and determining the presence or absence of occurrence of knocking accurately irrespective of high or low rotational speed of the engine to control an operation condition of the internal combustion engine properly. <P>SOLUTION: This knocking controller has two pairs of digital filters or more for extracting only a specific frequency component of a vibration wave-shaped output signal of a knock sensor 10 to switch to reduce degree of the digital filter when rotational speed of the engine is higher than a predetermined value. Consequently, a knock wave shape can be accurately extracted even on a high rotational speed side. The extracted knock wave shape may be corrected in accordance with response property of the filter without changing degree of the filter to suppress influence of the filter. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a knocking control device for an internal combustion engine that controls an operating state of the internal combustion engine in accordance with a state of occurrence of knocking that occurs in the internal combustion engine.

  In general, an internal combustion engine knock (hereinafter referred to as “knock”) control device generates a knock when an electrical signal from a knock sensor that detects engine vibration exceeds a predetermined level (knock determination level). The ignition timing is retarded, and when the knock is not detected for a predetermined period, the ignition timing is advanced so that the ignition timing is always controlled near the knock limit. It is what draws out to the maximum.

  In such a knock control device, the knock determination level is extremely important. If the knock determination level is too large, it is not detected despite the occurrence of knocking, so the ignition timing is advanced, knocking occurs frequently, and eventually the engine is damaged. On the other hand, if the knock determination level is too small, the ignition timing is retarded even if no knock has occurred, and the engine output cannot be sufficiently extracted.

  In general, knock detection is performed by a method that captures engine-specific vibration associated with the occurrence of knock, and this knock vibration is a resonance phenomenon generated by a shock wave in the cylinder. At the same time as the occurrence, the output of the knock sensor suddenly increases, and then attenuates in synchronization with the crank angle. However, since the actual time for a predetermined crank angle becomes shorter as the engine speed increases, when the same bandpass filter (BPF) is used, the knock waveform collapses as the engine speed increases, and its characteristics are smoothed. Discrimination from noise waveform becomes difficult.

  In order to reduce this influence, it is conventionally known from Patent Document 1 that the passband width of a bandpass filter is changed according to the engine speed.

JP 59-65728 A

  However, if the passband is changed as in the prior art of Patent Document 1, a frequency band unrelated to knocking is also captured with a gain equivalent to the knocking frequency, which may deteriorate the discrimination from noise. There is a problem.

  The present invention has been made in view of the above problems, and its object is to suppress the influence of a filter, and accurately determine whether knock has occurred regardless of whether the engine speed is high or low. The present invention provides a knocking control device for an internal combustion engine that can appropriately control the engine.

The present invention provides a knocking control device for an internal combustion engine according to each of claims, as means for solving the above-mentioned problems.
The knock control device for an internal combustion engine according to claim 1 has two or more pairs of digital filters that extract only a specific frequency component of a vibration waveform output signal of the knock sensor, and the rotational speed of the internal combustion engine is a predetermined value. When higher, the digital filter is switched to reduce the order, so that the filter responsiveness can be changed with the filter passband almost unchanged, and the higher the rotation, the higher the filter performance. By reducing the order, the knock vibration waveform can be accurately extracted even on the high rotation side without impairing the noise removal performance on the low rotation side. Therefore, it is possible to accurately determine whether knock has occurred.

  A knock control apparatus for an internal combustion engine according to claim 2, wherein a digital filter that extracts only a specific frequency component of an output signal of the knock sensor, a waveform extraction unit that extracts a knock waveform shape after passing through the digital filter, and an extraction Waveform correction means for correcting the knock waveform shape based on the response characteristic of the filter, and performing knocking determination based on the corrected knock waveform shape. By correcting the knock waveform shape extracted based on the response characteristics of the digital filter, the influence of the filter can be suppressed, and the presence or absence of knock can be accurately determined regardless of the engine speed.

The knocking control device according to claim 3, wherein the waveform correcting means obtains a peak position of the knock waveform in a predetermined crank angle period based on the knock waveform shape extracted by the waveform extracting means, and the peak is determined according to the response characteristic of the filter. The position and the waveform before and after the position are corrected, and it is clarified that the correction of the knock waveform is performed based on the peak position.
The knocking control device according to claim 4, wherein the waveform correction means obtains the peak position of the knock waveform during the predetermined crank angle period, and then, after the peak position, the position attenuated by a predetermined amount from the peak value is determined as the vibration start position and the end position. The peak position is corrected according to the response characteristics of the filter, and the corrected peak position, vibration start position, and end position are complemented with a straight line or curve to obtain a knock waveform. The correction of the waveform is clarified more specifically.

Hereinafter, a knocking control device for an internal combustion engine according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of an internal combustion engine knocking control apparatus according to an embodiment of the present invention.
In FIG. 1, reference numeral 10 denotes a knock sensor. The knock sensor 10 is installed on an engine block wall surface of an internal combustion engine (not shown) and detects a vibration, an in-cylinder pressure sensor installed in a cylinder, or the like. It is. The sensor signal from the knock sensor 10 is input to the input circuit 11 and is taken into the knock detection circuit 20 at the subsequent stage. Therefore, the sensor signal is adjusted by the input circuit 11 to be within a predetermined voltage range. Further, the sensor signal is passed through the filter 12 before being taken into the knock detection circuit 20. As the filter 12, an LPF (Low Pass Filter) that cuts a high frequency component for anti-aliasing or a BPF (Band Pass Filter) that extracts only a specific frequency component is set.

Next, the sensor signal that has passed through the filter 12 is taken into the knock detection circuit 20 via an A / D (analog-digital) converter 21. Note that the sampling period for the sensor signal through the A / D converter 21 of the knock detection circuit 20 is fast enough to sufficiently extract the waveform of the knock frequency component peculiar to knock. In this embodiment, the frequency components A / D converted by the A / D converter 21 are input to five digital BPFs 22a, 22b, 22c, 22d, and 22e having different pass bands, and these digital BPFs 22a, 22b, and 22c are used. The three knock frequency components and the digital BPFs 22d and 22e are separated into frequency components (hereinafter simply referred to as “noise frequency components”) in the signal region that do not include the two knock frequency components.
The digital BPFs 22a to 22e are one of the features of the present invention, and will be described in detail later as a first embodiment.

  Each frequency component after separation by the digital BPFs 22a to 22e is input to the interval integrating units 23a to 23e, and the intensity of each frequency component is integrated for each predetermined interval. This integration interval is defined by a crank angle [° CA (Crank Angle)] period. That is, by defining the crank angle instead of the real time, it is possible to extract a characteristic waveform at the time of occurrence of knock regardless of the engine speed. Further, the sum of the knock frequency components integrated by the section calculation units 23a, 23b, and 23c is obtained by the adding unit 23, so that a knock waveform can be extracted even if a knock occurs in which a plurality of vibration modes appear simultaneously. It will be possible.

The knock waveforms extracted by the respective interval integrating units 23a to 23e are subjected to the correction of the knock waveforms by the respective waveform correcting units 24a to 24d.
This waveform correction is also a feature of the present invention, which will be described in detail later as a second embodiment.
In the comparison calculation units 25a to 25d, the knock waveforms of the knock frequency components by the section calculation units 23a to 23c and the addition unit 24 are corrected by the waveform correction units 24a to 24d, and the corrected knock waveforms are ideal. It is compared with an ideal knock waveform having an ideal knock waveform shape stored in advance in the knock waveform storage unit 26, and the knock intensity is corrected so that the knock intensity increases as the ideal knock waveform is closer. As this correction method, for example, a coefficient (error coefficient) E corresponding to the degree of deviation from the ideal knock waveform is calculated, and the knock intensity is multiplied by (1-E) as a correction coefficient. Here, the ideal knock waveform may be determined by theoretical calculation such as a combustion model equation or the knock shape may be set in advance for each operating condition for each internal combustion engine. In addition, when using an in-cylinder pressure sensor, there is a high correlation between the attenuation shape of the resonance vibration peculiar to knock and the attenuation shape of the average pressure in the cylinder caused by combustion and piston movement. An attenuation shape of an ideal knock waveform may be defined.

  Next, in knock intensity selection unit 27, each knock waveform for each frequency component is corrected by comparison operation unit 25a-25d by comparison with an ideal knock waveform for each knock waveform corrected by each waveform correction unit 24a-24d. Among the knock intensities, a modified knock waveform used for knock determination is selected. Here, since the selected knock waveform has a vibration mode that appears remarkably for each knock, the one having the maximum correction knock intensity or the sum of the frequency components is used.

  On the other hand, a BGL (Back Ground Level) creation unit 28 creates BGL based on the noise frequency components from the interval integration units 23d and 23e in order to perform knock determination regardless of the operating state of the internal combustion engine. As this BGL creation method, similarly to the knock frequency component, for each noise frequency component, a corrected noise intensity that is compared and corrected with the ideal knock waveform is obtained, or the sum or average value of each noise frequency component is calculated. .

  Then, the comparison calculation / knock determination unit 29 performs a comparison operation using the knock intensity selected by the knock intensity selection unit 27 and the BGL generated by the BGL generation unit 28, and if it is equal to or greater than a predetermined value, it is determined as knock. As this comparison method, the determination may be made from the ratio between the knock strength and the BGL, or may be made from the difference between the knock strength and the BGL.

Further, a reference position signal and a cylinder discrimination signal of a crankshaft (not shown) of the internal combustion engine are inputted from an internal combustion engine control ECU (Electronic Control Unit) 30 to an A / D converter 21 of the knock detection circuit 20. A knock determination signal corresponding to a knock determination result from a comparison operation / knock determination unit 29 of the knock detection circuit 20 described later is input to the internal combustion engine control ECU 30. In addition, various sensor signals from, for example, a crank angle sensor, an intake air amount sensor, a water temperature sensor, and the like of the internal combustion engine are input to the internal combustion engine control ECU 30. Based on these input signals, the ECU 30 for controlling the internal combustion engine calculates the ignition timing, the fuel injection amount, etc., executes correction according to the knock determination result, and appropriately controls the igniter 40, the injector (fuel injection valve) 50, etc. Will be.
Here, the internal combustion engine control ECU 30 is a central processing unit that executes various known arithmetic processes, a CPU that stores a control program, a control map, etc., a RAM that stores various data, and a B / U (backup). The logic operation circuit includes a RAM, an input / output circuit, a bus line connecting them, and the like.

Next, a schematic operation of knock detection / determination by the knock detection circuit 20 for the vibration waveform signal from the knock sensor 10 will be described.
A waveform sample section signal is generated on the basis of the fall of the reference position signal from the internal combustion engine control ECU 30, and in the integration section where the waveform sample section signal is “ON (on)” and every predetermined section, the section calculation units 23a to 23a. The frequency components Fa to Fe via the BPFs 22a to 22e are respectively integrated by 23e. The knock frequency components Fa, Fb, Fc are applied to the knock frequency components Fa, Fb, Fc during the period from the end of the integration period in which the current waveform sample section signal is “OFF” to the start of the integration section in which the next waveform sample section signal is “ON”. Each knock waveform is obtained as a detected waveform based on the extracted knock waveform, and each extracted knock waveform is corrected by a waveform correcting means to be described later, and is stored in the intensity and ideal knock waveform storage unit 26 for each corrected knock waveform. The comparison calculation units 25a to 25d compare and calculate the intensity of the ideal knock waveform, and the corrected knock intensity is calculated.

  Of these correction knock intensities, the maximum correction knock intensity is selected by the knock intensity selection unit 27. Then, the comparison calculation / knock determination unit 29 uses the maximum correction knock strength from the knock strength selection unit 27 and the BGL from the BGL creation unit 28 based on the noise frequency components Fd and Fe as described later. A knock determination parameter is calculated. This knock determination parameter is compared with a predetermined determination threshold to determine whether knock has occurred or not. Then, the current integration value is reset at the fall of the next reference position signal, that is, at the start of the next integration interval when the waveform sample interval signal is “ON”. At this time, a knock determination signal corresponding to the knock determination result is output to the internal combustion engine control ECU 30, and the same operation is repeatedly executed thereafter.

  Next, a description will be given based on the flowchart of FIG. 2 showing a knock determination processing procedure in the knock detection circuit 20 used in the knock control device for an internal combustion engine according to an example of the embodiment of the present invention. This knock determination routine is repeatedly executed by the knock detection circuit 20 for each combustion cycle of the internal combustion engine based on the reference position signal and the cylinder determination signal from the internal combustion engine control ECU 30.

  In FIG. 2, first, in step S101, the process waits until the integration period is started, and then the process proceeds to step S102, where the vibration waveform signal is separated into a plurality of frequency components, and digital filter processing is performed on each frequency component. The process proceeds to step S103, and intensity integration is performed on the frequency components subjected to the digital filter processing. The process proceeds to step S104 to determine whether a predetermined crank angle has elapsed. When the determination condition of step S104 is not satisfied, that is, when the predetermined crank angle has not elapsed, the process returns to step S102 and the same processing is repeatedly executed.

  On the other hand, when the determination condition in step S104 is satisfied, that is, when the predetermined crank angle has elapsed, the process proceeds to step S105, and the integrated intensity value in step S103 is stored in the RAM (not shown) of the knock detection circuit 20. The process proceeds to step S106, and the integrated intensity value in step S103 is reset. It transfers to step S107 and it is determined whether the integration area is complete | finished. When the determination condition of step S107 is not satisfied, that is, when the integration interval has not ended, the process returns to step S102 and the same processing is repeatedly executed.

  On the other hand, when the determination condition of step S107 is satisfied, that is, when the integration period is ended, the process proceeds to step S108, and the knock waveform is corrected in detail later. Next, the routine proceeds to step 109, where the corrected knock intensity is calculated by comparison processing with the ideal knock waveform described later for the corrected knock waveform. Proceeding to step S110, the maximum correction knock strength of the correction knock strengths calculated at step S109 is selected. The process proceeds to step S111, and a BGL (background level) creation process described later for the noise frequency component is executed.

  Proceeding to step S112, the maximum correction knock intensity selected in step S110 is divided by the BGL (background level) created in step S111 to calculate the knock determination parameter. The process proceeds to step S113, where it is determined whether the knock determination parameter calculated in step S112 exceeds a predetermined determination threshold. When the determination condition of step S113 is satisfied, that is, when the knock determination parameter exceeds the determination threshold, the process proceeds to step S114, where it is determined that the knock is detected, and this routine is terminated. On the other hand, when the determination condition of step S113 is not satisfied, that is, when the knock determination parameter is smaller than the determination threshold value, the process proceeds to step S115, it is determined that there is no knock, and this routine is finished.

  Next, the filtering process in step S102, which is one of the features of the present invention, will be described with reference to FIGS. FIG. 3 is a flowchart showing a processing procedure according to the first embodiment of the present invention for switching the filter order based on the engine speed. In step S201, it is determined whether or not the engine speed is greater than or equal to a predetermined value. If it is greater than or equal to the predetermined value, the process proceeds to step S202, and a low-order digital filter is selected. If it is equal to or smaller than the predetermined value, the process proceeds to step S203, and a high-order digital filter is selected. In this way, various processes are performed based on the knock waveform extracted by the low-order or high-order digital filter, and finally knock determination is performed in step S204.

  FIGS. 4 and 5 are diagrams for explaining the effects of switching the filter order based on the rotational speed in the first embodiment. FIG. 4 shows the shapes of knock waveforms before filter processing, after passing through a high-order filter, and after passing through a low-order filter when the engine is running at a low speed and a high speed. The horizontal axis indicates the crank angle. As can be seen from FIG. 4, in the case of high rotation, the actual time with respect to the crank angle is shortened, so that the knock waveform shape collapses after passing through the high-order filter, and the discrimination from the noise waveform is poor. On the other hand, in the case of low rotation, the actual time with respect to the crank angle is long, so that the knock waveform shape is less distorted even after passing through the high-order filter.

  FIG. 5 is a graph comparing the responsiveness of three examples of (a) a high-order filter, (b) a wideband filter, and (c) a low-order filter. The horizontal axis represents frequency (kHz), and the vertical axis represents Indicates the vibration level (dB). When the filter order is increased from FIG. 5A, attenuation outside the pass band can be made steep. In general, the frequency of knock vibration is almost determined by the cylinder bore diameter. Therefore, on the low rotation side where it is desired to eliminate the influence of direct injection injector noise and the like as much as possible, a high-order filter that can pass frequencies other than the necessary frequency is desirable.

When the broadband filter is used from FIG. 5B, the response is improved, but the frequency band unrelated to the knock is allowed to pass through without being attenuated.
Further, by reducing the filter order from FIG. 5C, it is possible to both attenuate the frequency band unrelated to knocking and improve responsiveness.

From the above, when the engine is running at low speed, the mechanical and electrical noise output such as the operation noise of the direct injection injector is an output level that cannot be ignored with respect to knock vibration, so the filter eliminates the noise output, The filter should have as steep frequency characteristics as possible so that only the knock frequency component can be extracted. On the other hand, when the engine is running at high speed, the actual time with respect to the crank angle is shortened. Therefore, in order to extract an accurate knock waveform, it is required that the responsiveness of the filter is good.
In Embodiment 1 of the present invention, by changing the filter order on the high rotation side and the low rotation side using a digital filter, the response of the filter can be changed with the frequency passband substantially unchanged. Therefore, the knock vibration waveform can be accurately extracted even on the high rotation side without decreasing the noise removal performance on the low rotation side by decreasing the filter order on the high rotation side.

  Next, the waveform correction processing in step 108, which is another feature of the present invention, will be described with reference to FIGS. FIG. 6 is a flowchart illustrating a processing procedure for correcting a knock waveform extracted by a bandpass filter according to the second embodiment of the present invention. First, in step S301, the vibration waveform signal is separated into a plurality of frequency components as described above, and digital filter processing is performed on each frequency component. In step S302 shown in FIG. 7A, the intensity integration for the frequency component subjected to the digital filter processing is repeatedly performed within the integration interval. The knock waveform F extracted by this is shown in FIG. Steps S301 and S302 basically correspond to steps S101 to S107 described above.

In the second embodiment, the extracted knock waveform F shown in FIG. 7B is corrected. First, the peak position P of the knock waveform F extracted in step S303 is detected. Next, the process proceeds to step S304, and the vibration start position S and the vibration end position E are detected as shown in FIG. 7B with a value that is a fraction of the peak value. In step S305, since the peak position Po of the knock waveform is substantially determined by the crank angle, the extracted peak position P of the knock waveform is moved and corrected to the normal peak position Po as shown in FIG. 7C. The corrected knock waveform is obtained by complementing the corrected peak position Po, vibration start position S, and vibration end position E with straight lines. In addition, in FIG.7 (c), although the waveform supplemented with a straight line is shown, you may use a curve instead of a straight line. In step S305, knock determination is performed based on the knock waveform thus corrected. This step S305 is performed by comparison with the stored ideal knock waveform in the same manner as steps S109 to S115 described above.
As in the second embodiment, the influence of the filter is also suppressed as in the first embodiment by correcting the knock waveform extracted based on the response characteristic of the digital filter (a map determined by the time constant and the rotation speed of the filter). it can.

Next, description will be made based on the flowchart of FIG. 8 showing the processing procedure of the comparison operation with the ideal knock waveform in step S109 in the knock determination of FIG.
In FIG. 8, in step S401, the peak value p of each knock frequency component is searched. The process proceeds to step S402, and an ideal knock waveform a (θ) calculation process described later is executed. In step S403, the ideal knock waveform a (θ) calculated in step S402 is multiplied by the peak value p searched in step S401, and scale adjustment is executed. The process proceeds to step S404, and correlation processing for the waveform before the peak value, which will be described later, is executed. The process proceeds to step S405, and the correlation calculation process of the waveform after the peak value described later is executed. The process proceeds to step S406, where the correction intensity before peak value and the correction intensity after peak value are added to calculate a correction knock intensity, and this routine ends.

Next, a description will be given based on the flowchart of FIG. 9 showing the processing procedure of the ideal knock waveform calculation of step S402 in the comparison calculation with the ideal knock waveform of FIG.
In FIG. 9, in-cylinder pressure is detected in step S501. The process proceeds to step S502, and low-pass filter processing is executed. In step S503, the logarithmic decay rate per predetermined crank angle is calculated. The process proceeds to step S504, where the attenuation factor calculation process is executed, and this routine ends. In this way, as shown in FIG. 10, an ideal knock waveform a (θ) based on the in-cylinder pressure attenuation shape is created.

Next, a description will be given based on the flowchart of FIG. 11 showing the processing procedure of the correlation calculation of the waveform before peak value in step S404 in the comparison calculation with the ideal knock waveform of FIG.
In FIG. 11, in step S601, it is determined whether the ideal knock waveform a (θ) is smaller than the corrected waveform s (θ). When the determination condition in step S601 is satisfied, that is, the ideal knock waveform a (θ) is smaller than the corrected waveform s (θ), the process proceeds to step S602, and the pre-peak value corrected knock intensity is calculated by the following equation (1). This routine ends.
Pre-peak value corrected knock intensity ← a (θ) × [1- {s (θ) −a (θ)} / a (θ)]
... (1)
On the other hand, if the determination condition in step S601 is not satisfied, that is, if the ideal knock waveform a (θ) is larger than the corrected waveform s (θ), the process proceeds to step S603, and the pre-peak value corrected knock intensity is the ideal knock waveform a ( θ), and this routine ends.

Next, description will be made based on the flowchart of FIG. 12 showing the processing procedure of the correlation calculation of the waveform after the peak value in step S405 in the comparison calculation with the ideal knock waveform of FIG.
In FIG. 12, in step S701, the difference d (θ) between the ideal knock waveform a (θ) and the corrected waveform s (θ) is calculated by the following equation (2).
d (θ) ← | s (θ) −a (θ) | (2)
Next, the process proceeds to step S702, the corrected knock intensity after peak value is calculated by the following equation (3), and this routine is finished.
Corrected knock intensity after peak value ← a (θ) × {1-d (θ) / a (θ)} (3)

Next, a description will be given based on the flowchart of FIG. 13 showing the processing procedure of BGL (background level) creation in step S111 in the knock determination of FIG.
In FIG. 13, in step S801, the average intensity of the noise frequency component is calculated, and this routine ends.

In the above description, the knocking control apparatus for an internal combustion engine according to the embodiment of the present invention has the means for changing the order of the digital filter based on the engine speed shown in the first embodiment and the filter shown in the second embodiment. Although it has been described that the means for correcting the knock waveform shape extracted based on each processed frequency component is incorporated together, it is necessary that both of these means be incorporated together. However, even if one of the means is provided, the influence of the filter that is the object of the present invention can be suppressed.
In the above description, the knock determination system itself is also described, but other known knock determination systems can also be used.

  As described above, the knocking control device for the internal combustion engine of the present embodiment includes a knock sensor 20 that detects a vibration waveform generated in the engine, and digital filters 22a to 22b that separate the detected vibration waveform signal into a plurality of frequency components. Further, the interval integration for extracting the knock waveform shape by the frequency component consisting of the value obtained by integrating the intensity of the plurality of separated frequency components Fa to Fc for each predetermined section having a predetermined crank angle and the value added for each same section. Waveform extraction means comprising the units 23a to 23c and the addition unit 23, and a waveform comprising waveform correction units 24a to 24d for correcting the peak position P and the waveforms before and after the extracted knock waveform in accordance with the response characteristics of the digital filter. A knock detection circuit for determining whether or not knock has occurred in the engine based on the correcting means and the corrected knock waveform shape Knock determination means comprising an ideal knock waveform storage unit 26, comparison calculation units 25a to 25d, knock intensity selection unit 27, BGL creation unit 28, and comparison calculation / knock determination unit 29, and operation of the internal combustion engine according to the knock determination result And a knock control means comprising an internal combustion engine control ECU 30 for controlling the state.

  Here, in the case of a knock waveform in which one vibration mode is dominant, knock determination can be performed by extracting the vibration component waveform. However, when multiple vibration modes are included in one knock waveform, the sensor's raw waveform has an ideal knock waveform due to factors such as the change of the dominant vibration mode during the knock period. If only the waveform shape of each knock frequency component is viewed, it may be far from the ideal knock waveform and cannot be determined as a knock.

  In contrast, a vibration waveform signal generated in an internal combustion engine is separated into a plurality of frequency components, and a value obtained by integrating the separated frequency components for each predetermined section having a predetermined crank angle, and this value is the same section. According to the knock waveform shape set by at least two frequency components consisting of the values added for each, the knock waveform shape as the sum of each vibration mode can be represented, and by using this knock waveform shape, The presence or absence of knocking in the internal combustion engine can be accurately determined, and the operating state of the internal combustion engine can be controlled appropriately.

  In the knocking control apparatus for an internal combustion engine according to the first embodiment, the digital filters 22a to 22e have a secondary pair or more, and the pass band is almost unchanged by changing the filter order according to the rotational speed of the internal combustion engine. Can change the responsiveness of the filter. Therefore, the knock vibration waveform can be accurately extracted even on the high rotation side by reducing the filter order as the rotation speed increases.

  Further, in the knocking control device for the internal combustion engine of the second embodiment, a specific frequency component is extracted by a digital filter from the vibration waveform output signal detected by the knock sensor 20, and is extracted by the section calculation units 23a to 23e which are extraction waveform means. Since the knock waveform is corrected by the waveform correction units 24a to 24d, which are waveform correction means, for the extracted knock waveform, the influence of the filter can be suppressed, and the presence or absence of the occurrence of the knock can be accurately determined.

It is a block diagram which shows the whole structure in the knocking control apparatus of the internal combustion engine of embodiment of this invention. It is a flowchart which shows the processing means of the knock determination used for the knock control apparatus of the internal combustion engine of embodiment of this invention. FIG. 3 is a flowchart showing a processing procedure in a filter that is Embodiment 1 of the present invention in the filter processing of FIG. 2. FIG. It is a figure explaining the effect of Example 1. FIG. Similarly, it is a figure explaining the effect of Example 1. FIG. In the waveform correction process of FIG. 2, it is a flowchart which shows the waveform correction process procedure of the extracted knock waveform which is Example 2 of this invention. It is a figure explaining the waveform correction process of Example 2. FIG. FIG. 3 is a flowchart showing a processing procedure of a comparison operation with an ideal knock waveform in the knock determination of FIG. 2. FIG. FIG. 9 is a flowchart showing a processing procedure of ideal knock waveform calculation in comparison calculation with the ideal knock waveform of FIG. 8. FIG. 10 is an explanatory diagram showing creation of an ideal knock waveform based on the in-cylinder pressure attenuation shape of FIG. 9. It is a flowchart which shows the process sequence of the correlation calculation of the waveform before a peak value in the comparison calculation with the ideal knock waveform of FIG. FIG. 9 is a flowchart illustrating a processing procedure for correlation calculation of a waveform after a peak value in comparison calculation with the ideal knock waveform of FIG. 8. It is a flowchart which shows the process sequence of BGL preparation in the knock determination of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Knock sensor 20 Knock detection circuit 22a-22e Digital band pass filter (digital BPF)
24a-24d Waveform correction section

Claims (4)

A knocking control device for an internal combustion engine that detects knocking occurring in the internal combustion engine with a knock sensor, determines knocking by an output signal of the knock sensor, and controls the operating state of the internal combustion engine according to the determination result,
A switch having two or more pairs of digital filters for extracting only a specific frequency component of the vibration waveform output signal of the knock sensor, and reducing the order of the digital filter when the rotational speed of the internal combustion engine becomes higher than a predetermined value. , A knocking control apparatus for an internal combustion engine, wherein an accurate knock waveform is extracted. A knocking control device for an internal combustion engine that detects knocking occurring in the internal combustion engine with a knock sensor, determines knocking by an output signal of the knock sensor, and controls the operating state of the internal combustion engine according to the determination result,
A digital filter that extracts only a specific frequency component from the vibration waveform output signal of the knock sensor;
Waveform extracting means for extracting a knock waveform shape after passing through the digital filter;
Waveform correcting means for correcting the extracted knock waveform shape based on the response characteristic of the digital filter;
And a knocking control device for an internal combustion engine, wherein knocking is determined based on the corrected knock waveform shape.   The waveform correcting means obtains the peak position of the knock waveform in a predetermined crank angle period based on the knock waveform shape extracted by the waveform extracting means, and calculates the peak position and the waveforms before and after the peak position according to the response characteristic of the digital filter. The knocking control device for an internal combustion engine according to claim 2, wherein the knocking control device is modified.   The waveform correction means obtains the peak position of the knock waveform in a predetermined crank angle period, and then sets the positions attenuated by a predetermined amount from the peak value before and after the peak position as the vibration start position and end position, and the response characteristics of the digital filter The knock position of the internal combustion engine according to claim 3, wherein the knock position is determined by correcting the peak position in accordance with the correction and complementing the corrected peak position, vibration start position and end position with a straight line or a curve. Control device.

Images