JP2543379B2 - Knocking control device for internal combustion engine - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial application] The present invention determines knocking of an internal combustion engine based on a detection signal from a knock sensor, and determines knocking factors such as ignition timing, supercharging pressure, and exhaust gas recirculation amount. The present invention relates to a knocking control device for an internal combustion engine that controls the engine.

[Prior Art] Conventionally, in a knocking control device for an internal combustion engine, a detection signal from a knock sensor for detecting vibration of the internal combustion engine exceeds a knock determination level, as described in, for example, Japanese Patent Laid-Open No. 56-115861. In this case, the ignition timing is retarded on the assumption that knock occurs in the internal combustion engine, and conversely, the ignition timing is advanced by advancing the ignition timing when the detection signal from the knock sensor falls below the knock determination level. It is configured to always control near the knocking occurrence limit to maximize fuel efficiency and output characteristics of the internal combustion engine.

Further, in this type of device, in order to set the knock determination level to an appropriate value according to the characteristics of the knock sensor and the internal combustion engine, a value obtained by integrating the detection signal from the knock sensor is set in advance for each rotational speed of the internal combustion engine. The knock determination level is created by multiplying the determined constant K.

However, if the above constant K is closely matched for each revolution of the internal combustion engine through experiments and the like, and the knock determination level is created using the set constant K in the internal combustion engine of the same model, the internal combustion engine of the same model is However, since there is an error in the characteristics in manufacturing, the knock determination level may not match the engine characteristics, and accurate knock detection may not be possible. Even if the constant K is set for each internal combustion engine, the constant K may not match the engine characteristic due to the change in the engine characteristic over time, and accurate knock detection may not be performed.

Therefore, in recent years, as described in Japanese Patent Application Laid-Open No. 60-243369, knocking is performed so that the distribution shape of the logarithmic conversion value logV of the maximum value V of the detection signal output from the knock sensor within a predetermined section has a predetermined shape. It is considered that the accuracy of knock detection is improved by correcting the determination level.

In other words, this kind of control device, when the rotational speed of the internal combustion engine is constant and no knock occurs, as shown in FIG. 9, the cumulative probability of the distribution of the logarithmic conversion value logV of the maximum value V. Change uniformly, and when knock occurs, as shown in FIG. 9, there is a sudden change at a break point P at which there is a change rate of the cumulative probability of the distribution of the logarithmic transformed value logV, and the break point P also occurs. It is constructed by paying attention to the fact that it moves in the direction of lower cumulative probability as the frequency increases, so that the knocking sound becomes constant regardless of the rotation speed of the internal combustion engine, that is, the break point P is the rotation of the internal combustion engine. The knock determination level is corrected so that it becomes as shown in FIG. 10 according to the speed.

As shown in FIG. 10, the target position of the break point P is changed according to the rotational speed of the internal combustion engine because the knocking sound felt by a human changes depending on the number of knocking occurrences within a certain time, This is because the number of knock occurrences per hit (that is, the knock occurrence frequency) does not match.

[Problems to be Solved by the Invention] In the knocking control device configured as described above, the knock determination level can be sequentially updated according to the knocking occurrence state during operation of the internal combustion engine. Even if the determination level is not carefully set for each rotation speed of the internal combustion engine, the frequency of knocking per hour (that is, knocking sound) can be constantly controlled at a constant level.
If the air-fuel ratio becomes lean and heavy knock occurs in the internal combustion engine during low speed operation of the internal combustion engine, the knock determination level is erroneously corrected to a large value as if the internal combustion engine is not knocked. There was such a problem.

That is, when a heavy knock occurs while the internal combustion engine is operating at a high rotation speed, as shown in (a) of FIG.
The break point P is located at the center of the cumulative probability, and its distribution shape can be determined without any problem, but when the rotation speed of the internal combustion engine decreases, as shown in (b) of FIG. 11, in the same heavy knock state. Even if there is a break point P, the cumulative probability of the logarithmic transformed value logV of the maximum value V may change uniformly at the change rate at the time of knock occurrence, depending on the case. There is a possibility that the heavy knock will be erroneously determined to be in a state where no knock has occurred and the knock determination level will be corrected.

In FIGS. 11 (a) and 11 (b), represents the distribution shape of the logarithmic conversion value logV when no knock has occurred, and represents the distribution shape of the logarithmic conversion value logV in the light knock state. There is.

Therefore, the present invention has been made for the purpose of providing a knocking control device that does not erroneously correct the knock determination level when a heavy knock occurs during low-speed operation of an internal combustion engine as described above.

[Means for Solving the Problems] That is, as shown in FIG. 1, the configuration of the present invention made to achieve the above object includes a knock sensor M1 for detecting knocking generated in an internal combustion engine, and a knock sensor M1. Knocking determination means based on the detection signal of the knocking control means M2 for controlling a predetermined knocking control factor according to the determination result, and the maximum value of the detection signal output from the knocking sensor M1 within a predetermined section. Maximum value detection means M3 and determination level correction means for correcting the knock determination level used in the knock determination by the knocking control means M2 so that the distribution shape of the logarithmic conversion value of the detected maximum value becomes a predetermined distribution shape. In a knocking control device for an internal combustion engine including M4 and M4, a rotation speed detecting means M5 for detecting the rotation speed of the internal combustion engine is provided.
And a correction prohibiting means M6 for prohibiting correction of the knock determination level by the determination level correcting means M4 when the detected rotation speed is equal to or lower than a predetermined speed, and a knocking control of an internal combustion engine, The device is the main point.

[Operation] In the knocking control device of the present invention configured as described above, when the rotation speed of the internal combustion engine detected by the rotation speed detection means M5 is equal to or lower than the predetermined speed, the correction prohibiting means M6 determines the determination level correction means M4. Correction of knock determination level by is prohibited. Therefore, the determination level is not corrected when the internal combustion engine is running at low speed, and even if a heavy knock occurs at this time, the determination level will not be erroneously corrected.

Here, the knock control factor is one that can control the occurrence frequency of knocking, such as ignition timing, supercharging pressure by the supercharger, or the amount of exhaust gas recirculation by the exhaust gas recirculation device. The control amount in the ignition timing control device, the supercharging pressure control device, the exhaust gas recirculation amount control device, or the like is corrected according to the knocking determination result.

Next, as the predetermined section in which the maximum value detection means M3 detects the maximum value of the output signal from the knock sensor M1, a predetermined section during the internal combustion engine combustion stroke that is conventionally set as a knock determination section (for example, 10 ° C) 90 ° C from A-ATDC A-
You can set it to ATDC).

Next, the determination level correction means M4 is for correcting the knock determination level so that the above distribution shape becomes a preset desired distribution shape. Specifically, for example, the maximum value detection means M3 detects Then, the median of the maximum values is calculated, the predetermined values are arithmetically operated on the median, and the upper and lower limits of the maximum value are set, and the maximum value becomes equal to or larger than the calculated upper limit value at every predetermined time. The knock determination level is corrected in accordance with the magnitude relation between the number of times of occurrence and the number of times when the number is equal to or less than the lower limit value.
Various construction methods described in JP-A-60-243369 can be applied.

Embodiments Embodiments of the present invention will be described below with reference to the drawings.

First, FIG. 2 shows a 6-cylinder internal combustion engine 2 to which the present invention is applied.
FIG. 3 is a schematic configuration diagram showing a peripheral device and its peripheral devices.

As shown in the figure, the amount of air flowing into the intake pipe 4 of the internal combustion engine 2 via the throttle valve 6 (the amount of intake air)
Air flow meter 8 for detecting the temperature, an intake air temperature sensor 10 for detecting its temperature (intake air temperature), and a throttle opening sensor for detecting the opening of the throttle valve 6 (throttle opening)
Twelve are equipped. Further, the intake pipe 4 is provided with a fuel injection valve 14 for injecting the fuel pressure-fed from a fuel pump (not shown) for each cylinder of the internal combustion engine, and the fuel injected from the fuel injection valve 14 and the throttle valve 6 are provided. It can be mixed with the inflowing air and supplied to the internal combustion engine 2.

Further, in the internal combustion engine 2, an air-fuel ratio sensor 18 for detecting the air-fuel ratio of the fuel mixture supplied to the internal combustion engine 2 from the oxygen concentration in the exhaust gas flowing through the exhaust pipe 16, a water temperature sensor 19 for detecting the cooling water temperature, the internal combustion engine Knock sensor 20 for detecting knocking occurring in 2 and rotation speed sensor 24 for generating a pulse signal for detecting rotation speed NE of internal combustion engine 2 at every predetermined rotation angle (for example, 30 ° C. A) of distributor 22. , And a cylinder discrimination sensor 26 that outputs a pulse signal for determining the fuel injection timing and the ignition timing once every one rotation of the distributor 22 (that is, once every two rotations of the internal combustion engine 2).
Is provided, the air flow meter 8, the intake air temperature sensor 10,
The throttle opening sensor 12 as well as its operating state can be detected.

Then, the detection signals from the knock sensor 20 are output to the knock determination circuit 30, and the detection signals from the other sensors are output to the control circuit 40.

The knock determination circuit 30 is for determining the knock intensity generated in the internal combustion engine 2 based on the detection signal output from the knock sensor 20, and is a central processing unit that executes a knock determination process described later according to a preset control program. The unit (CPU) 30a, the read-only memory (ROM) 30b in which the control programs and various data necessary for executing the knock determination processing in the CPU 30a are recorded in advance, and the various data necessary for executing the arithmetic processing in the CPU 30a are stored. Random access memory (RAM) 30c that is temporarily read and written, and inputs detection signals from knock sensor 20 after A / D conversion, and also exchanges various control data with control circuit 40 via data bus 41. 1 by the input / output unit 30d for performing
It is configured as a chip microcomputer.

Further, the control circuit 40 uses the igniter based on the knock determination result by the knock determining circuit 30 and the detection signals from the respective sensors.
42 and the fuel injection valve 14 are driven to execute ignition timing control and fuel injection control including knocking control. Various calculation processes for fuel injection control and ignition timing control are performed according to a preset control program. CPU40a, CPU4
A ROM 40b in which control programs and various data necessary for executing arithmetic processing at 0a are recorded in advance, a RAM 40c at which various data necessary for executing arithmetic processing at the CPU 40a are temporarily read and written, and each of the above sensors I / O unit 40 for inputting a detection signal from the fuel injection valve 14 and the igniter 42 and outputting a driving signal to the knock determining circuit 30 via the data bus 41.
It is configured as a one-chip microcomputer by d and the like.

Next, the ignition timing control executed by the control circuit 40 and the knock determination processing executed by the knock determination circuit 30 will be described with reference to the flowcharts shown in FIGS. 3 to 6.

First, FIG. 3 is a flowchart showing an ignition timing calculation process executed as one of the main routines in order to perform the ignition timing control including the knocking control in the control circuit 40.

As shown in the figure, when this process starts, the first step
100 is executed, the air flow meter 8 and the rotation speed sensor 2
Based on the intake air amount Q and the rotational speed NE detected by using 4, the basic ignition timing corresponding to the engine load Q / NE is calculated.

Next, at step 110, it is judged if the knocking control execution condition is currently satisfied. This step 110
Then, when the cooling water temperature THW detected using the water temperature sensor 19 is equal to or higher than a predetermined temperature and the load Q / NE of the internal combustion engine 2 is equal to or higher than a predetermined value, it is determined that the knocking control execution condition is satisfied. If the execution condition of knocking control is satisfied, the routine proceeds to the next step 120, and the knocking strength determination result output from the knocking determination circuit 30 by the knocking determination process described later is read. And the following steps
In 130, the retard amount of the ignition timing is calculated according to the read knock intensity, and the process proceeds to the next step 140, in which a knock determination permission signal is transmitted to the knock determination circuit 30.

Note that this knock determination permission signal indicates that the knock determination circuit 30 is currently under the condition for executing knock control.
This is for notifying that knock determination needs to be performed, and the knock determination circuit 30 performs knock determination based on this knock determination permission signal.

When the ignition retard amount for knocking control is calculated in this way, in the subsequent step 150, it is determined whether or not the internal combustion engine 2 is currently in steady operation based on changes in the rotational speed NE of the internal combustion engine 2. In step 150, the internal combustion engine 2
If it is determined that is a steady operation state, the following step 155
Then, it is determined whether the rotational speed NE of the internal combustion engine 2 is within a predetermined rotational speed range (for example, 2400r.pm to 5200r.pm). When it is determined in step 155 that the rotation speed NE of the internal combustion engine 2 is within the predetermined rotation speed range, it is determined that the knock determination level correction condition is satisfied, and the routine proceeds to the next step 160, where the knock determination circuit A knock permission level correction permission signal is transmitted to 30.

The correction permission signal is for prompting the knock determination circuit 30 to correct the knock determination level (specifically, update the correction value ΔK used to set the knock determination level). The knock determination circuit 30 also executes the correction control of the knock determination level at the time of executing the knock determination process by the correction permission signal.

Next, when the knock determination level correction permission signal is transmitted in step 160, step 170 is executed, and the ignition timing correction amount for controls other than knocking control is calculated. That is, as is well known, the ignition timing control is the cooling water temperature TH
Warm-up correction that advances the ignition timing when W is low, cooling water temperature T
High temperature correction for advancing / retarding the ignition timing according to the idle state of the internal combustion engine 2 when HW is high, depending on the deviation of the rotational speed NE when the rotational speed NE is lower than the target rotational speed during idle operation of the internal combustion engine 2. Since various correction controls such as the idle rotation correction for advancing the ignition timing are executed by this, in step 170, the correction amount for the ignition timing control other than the knocking control is set according to the operating state of the internal combustion engine 2. It is calculated by

When the various correction amounts of the ignition timing are calculated in this manner, the basic ignition timing calculated in step 110 is advanced or retarded based on the calculated correction amount in step 180, and the internal combustion engine as the control target is corrected. The ignition timing of 2 is calculated.

On the other hand, if it is determined in step 110 that the knocking control condition is not satisfied, the transmission of the knock determination permission signal is stopped in step 190, and the transmission of the knock determination level correction permission signal is stopped in step 200. Step 17
Move to 0.

If it is determined in step 150 that the internal combustion engine 2 is not in steady operation, or if it is determined in step 155 that the rotational speed NE of the internal combustion engine 2 is within the predetermined rotational speed range, the knock determination level is set. When it is determined that the correction condition is not satisfied, the process proceeds to step 200, and after the transmission of the knock determination level correction permission signal is stopped, the process proceeds to step 170.
The series of processes executed in step 155 and step 200 corresponds to the above-described correction prohibiting means M6.

As described above, the control circuit 40 calculates the target ignition timing for the ignition timing control by the same method as the conventional method, executes the ignition timing control, and when the knocking control condition is satisfied, the knock determination is performed. A knock determination permission signal is sent to the circuit 30 to prompt execution of knock determination processing, and when the knock determination level correction condition is satisfied, the knock determination circuit 30 is enabled to correct the knock determination level. A signal is transmitted to prompt correction control of the knock determination level.

Further, although not shown, in the control circuit 40, the rotation speed sensor 24
Further, based on the output signal from the cylinder discrimination sensor 26, every time each cylinder of the internal combustion engine 2 enters the explosive stroke (in this embodiment, since the internal combustion engine 2 has 6 cylinders, it becomes 120 ° C every 120 ° C). At the dead point, a knock determination command signal transmission process for transmitting a command signal for activating the knock determination process to the knock determination circuit 30 is executed, and the knock determination circuit 30 synchronizes the knock determination process with the rotation of the internal combustion engine 120 It is designed to be executed every ° C.

Here, the processing of step 150 and step 155 is processing for determining whether or not the correction condition for correcting the knock determination level is satisfied in the knock determination processing executed by the knock determination circuit 30, and is performed in the present embodiment. In the example, when the internal combustion engine 2 is in the transient operating state or the rotation speed thereof is not within the predetermined rotation speed range, it is determined that the knock determination level correction condition is not satisfied, and the knock determination level correction permission is given in step 200. The signal transmission is stopped, and the knock determination level correction control (that is, the update processing of the knock determination level calculation correction value ΔK executed in the knock determination processing described later) is prohibited.

This is because the normal stable knocking control cannot be executed during transient operation such as rapid acceleration of the internal combustion engine, and knocking is likely to occur. Therefore, if the knock determination level is corrected, the knocking control during steady operation of the internal combustion engine 2 is good. When the internal combustion engine 2 is operating at a low speed, the distribution shape of the maximum value VPEAK is erroneously determined and the knock determination level is erroneously determined as described in the section of the problem to be solved by the invention. This is because there is a risk of correction. Further, in this embodiment, the rotation speed of the internal combustion engine 2 is a predetermined rotation speed (for example,
(5200r.pm) or more, the knock determination level correction is prohibited, because knocking is less likely to occur at high engine speeds and there is no need to correct the knock determination level in this region. .

Further, by prohibiting the execution of the knock determination level correction control in the region where the knock determination level correction is not substantially required, the load on the CPU 30a can be reduced. That is, when the internal combustion engine is rotating at high speed, the number of times per time this knock determination process is executed is increased, and the time during which the knock determination process can be executed is shortened. By prohibiting the correction control of, the execution time of the knock determination process when the internal combustion engine is at high rotation speed can be shortened, and the knock determination can be satisfactorily executed without using a CPU that performs high-speed processing. Of.

Next, FIG. 4 is a flow chart showing knock determination processing executed by knock determination circuit 30.

This process is started by a command signal output from the control circuit 40 each time each cylinder of the internal combustion engine 2 enters the explosion stroke, and the knock intensity generated in the internal combustion engine 2 is classified into four types: none, small, medium and large. Is a process for making a decision and transmitting the decision result to the control circuit 40.

When this process starts as shown in the figure, the first step
210 is executed, and it is determined whether or not the knock determination permission signal is output from the control circuit 40, that is, whether or not the knocking control execution condition is currently satisfied.

If it is determined in step 210 that the knocking control execution condition is satisfied, the routine proceeds to the following step 220, and the knock determination level VLEVL ^m is calculated. Knock judgment level V
LEVL ^m serves as a reference for determining the knock strength of the cylinder currently in the explosion stroke, and the average value VMEAN ^m of the knock detection signal obtained when the knock determination process was executed for this cylinder last time. (K-1), a reference value K for calculating a knock determination level that is set in advance according to the rotational speed NE of the internal combustion engine 2, and a correction value updating process described later so that the knock sound becomes constant at predetermined time intervals. The correction value ΔK ^m that is updated as follows is calculated by the following equation (1) using as parameters.

VLEVL ^m = VMEAN ^m (k-1) · (K + ΔK ^m ) ... (1) The subscript ^m attached to the above-mentioned smoothed value VMEAN and knock determination level VLEVEL is the number of the cylinder currently subject to knock determination. (I.e., the number of the cylinder currently in the explosion stroke), and the value calculated for each cylinder will be described with a subscript ^m in the description below. Also, the correction value ΔK ^m
3 types are set for low speed, medium speed, and high speed according to the rotational speed NE of the internal combustion engine 2. When calculating the knock determination level VLEVL ^m , among the three correction values, A correction value ΔK ^m corresponding to the rotational speed NE of the internal combustion engine 2 is selected and used.

When the knock determination level VLEVL ^m is calculated in this manner, the routine proceeds to the subsequent step 230, this time the knock sensor 20
The knock intensity is determined based on the knock detection signal output from the above and the knock determination level VLEVL ^m calculated above, and the maximum value VPEAK ^m of the knock detection signal in this determination section is calculated, and the routine proceeds to step 240. Step 2
At 40, the determined knock intensity is transmitted to the control circuit 40 as control data used for calculating the ignition retard amount for performing knock control in the ignition timing calculation process.
Knock strength and maximum value VPEA executed in step 230
The calculation of K ^m will be described in detail later with reference to the flowchart of FIG.

In this way, the knock strength judgment and the judgment result control circuit
When it is transmitted to 40, the process proceeds to the following step 250,
It is determined whether or not the knock determination level correction permission signal is input from the control circuit 40, that is, whether or not the knock determination level correction condition is satisfied. If the condition for correcting the knock determination level is satisfied, the following step 26
Shifting to 0, a correction value updating process for updating the correction value ΔK ^m for correcting the knock determination level VLEVL ^{m so} that the knock sound becomes constant is executed, and the process of this routine is once ended. The correction value updating process will be described later in detail with reference to the flowchart of FIG.

On the other hand, if it is determined in step 210 that the knock determination permission signal is not output from the control circuit 40, or
If it is determined in 250 that the correction permission signal for the knock determination level is not output from the control circuit 40, step 27
At step 260, the counter and the like used for updating the correction value ΔK ^m are initialized, and the processing of this routine is once terminated.

Next, FIG. 5 is a flow chart showing a process for calculating the knock intensity and the knock detection signal maximum value VPEAK ^m executed in step 230.

When this process is started as shown in the figure, first, step 30
0 is executed, and 10 ℃ A-ATDC of the cylinder currently in the explosion stroke
The start time and end time of the determination section from the present time are obtained based on the rotational speed NE of the internal combustion engine 2 with the section of 90 ° C A-ATDC as the knock determination section, and the time is set in a timer (not shown).

Then, in the following step 310, it is determined whether or not the time has reached the start time of the knock determination by the set timer, and when the determination section is entered, the process proceeds to the following step 320 and is input via the input / output unit 40d. Knock detection signal A
The / D conversion value VAD ^m (hereinafter referred to as the detection signal level) is read, and the process proceeds to step 330.

In step 330, the knock determination level VLEVL ^m calculated in step 220 of FIG. 4 and the detection signal level VAD ^m are compared in magnitude. And in this step 330 VAD ^m > VL
If it is determined that EVL ^m, it is determined that knock has occurred in the internal combustion engine 2 and the knock determination counter C is determined in step 340.
After incrementing x, the process proceeds to step 350, and if not, the process directly proceeds to step 350.

In step 350, the knock signal is knocked by the following equation (2) using the read detection signal level VAD ^m and the average value VMAD (i-1) of the knock detection signals calculated when the step 350 was executed last time as parameters. The average value VMAD (i) of the detection signals is calculated, and the process proceeds to step 360.

VMAD (i) = (15 · VMAD (i−1) + VAD ^m ) / 16 (2) Next, in step 360, it is determined whether or not the detection signal level VAD ^m read in step 320 exceeds the maximum value VPEAK ^m. to decide. When it is determined in this step 360 that VAD ^m > VPEAK ^m , VAD ^m is updated to the maximum value VPEAK ^{m in the} next step 370 and then the process proceeds to step 380. Otherwise, the process directly proceeds to step 380.

The processes of steps 360 and 370 are processes for obtaining the maximum value VPEAK ^m of the knock detection signal in the knock determination section, and correspond to the above-mentioned maximum value detecting means M3.

In step 380, the timer set in step 300 determines whether or not the time has reached the end time of knock determination. When it is determined that the end time of the knock determination section has not yet elapsed, the process proceeds to step 320 again and the processes of steps 320 to 370 are executed.

On the other hand, if it is determined in step 380 that the knock determination section has ended, the routine proceeds to step 390, this time, the average value VMAD of the knock detection signals updated in the knock determination section.
(I) and the knock determination level VLEV in step 220 described above.
The smoothed value VM of the knock detection signal used to calculate L ^m
The average value VMEAN ^m (i) of the knock detection signal used to calculate the knock determination level VLEVL ^m in the next process is calculated by the following equation (3) using EAN ^m (i-1) as a parameter.

VMEAN ^m (i) = (3 · VMEAN ^m (i-1) + VMAD (i)) / 4 ...
(3) Then, in the following step 400, the counter C that is incremented by the processing in step 340 in the knock determination section
The knock intensity is determined based on the value of x. This step 400
Then, for example, when the value of the counter C is 1 or less, it is determined that there is no knock, when the value of the counter C is 2 to 5, the knock intensity is small, and when the value of the counter C is 6 to 9, the knock intensity is medium. , If the value of the counter C is 10 or more, the knock intensity is high,
In this way, the knock strength is judged in four stages. Then, when this determination process is executed, the process proceeds to the subsequent step 410, the value of the counter Cx is initialized for the next knock determination, and the process proceeds to step 420.

Next, at step 420, it is determined whether or not the knock strength is high, medium, or low at step 410, that is, the knocking occurs in the internal combustion engine 2 and the ignition timing retard control by the knocking control is performed. Is executed. When it is determined in step 420 that the ignition timing retard control is executed, the process proceeds to the following step 430, the counter CKNK ^m is incremented, and this process is ended. If it is determined in step 420 that knocking has not occurred and ignition timing retard control by knocking control is not performed, the process ends.

As described above, in step 230 of FIG. 4, the knock intensity is determined, and the smoothed value VMEAN ^m of the knock detection signal for calculating the knock determination level used for the determination.
A well-known process for calculating is calculated, and the maximum value VPEAK ^m required to update the correction value ΔK ^m for knock determination level correction is calculated.

Next, FIG. 6 is a flowchart showing the correction value updating process executed in step 260 of FIG.

When this process is started as shown in the figure, first, step 50
0 is executed, and the counter CPEAK ^m for counting the number of executions of this updating process is incremented for each cylinder. Then, in subsequent step 510, the maximum value VPEAK ^m of the calculated knock detection signal determines whether or not greater than the median V50 ^m being calculated currently, the process proceeds to step 520 if VPEAK ^m> V50 ^m update the median V50 ^m by adding the set value Derutabui50 ^m for the central value update to the median value V50 ^m Te, the process proceeds to step 530.

On the other hand, when it is determined in step 510 that VPEAK ^m > V50 ^m is not established, the process proceeds to step 540, and this time the maximum value VPEAK ^m
Is smaller than the currently calculated median V50 ^m . And in the case of VPEAK ^{m <V50 m,} the steps after updating the median V50 ^m from median V50 ^m shifts by subtracting the set value Derutabui50 ^m for the central value update step 550 530
If not, go to step 530 as it is.

In step 530, the set value ΔV50 ^m for updating the median value used when updating the median value V50 ^m in the next process is calculated by the following equation (4).
Is calculated, and the process proceeds to step 560.

ΔV50 ^m = | VPEAK ^m −V50 ^m | / 16 (4) In the processing of steps 510 to 550, when the maximum value VPEAK ^m is larger than the median value V50 ^m , the median value V50 ^m is increased by ΔV50 ^m , by reducing the median V50 ^m only Derutabui50 ^m when the maximum value VPEAK ^m is smaller than the median V50 ^m, in the process for making such a median V50 ^m is always located at the cumulative 50% point of the maximum value VPEAK ^m is there.

Next, in step 570, the maximum value VPEAK ^m
Maximum value VPEAK for discriminating the distribution shape of logarithmic transformation value of
lower limit VH ^m on the ^m, husband VL ^m s equation (5), is calculated using (6).

VH ^m = (A ^m + D) · V50 ^m … (5) VL ^m = V50 ^m / A ^m … (6) (where A ^m is the upper and lower limit value setting variable, D is the target knock sound adjustment coefficient) When the upper and lower limit values VH ^m and VL ^m are calculated, the process proceeds to step 580 and it is determined whether or not the maximum value VPEAK ^m is below the lower limit value VL ^m . If VPEAK ^m <VL ^m , the routine proceeds to step 590, where the distribution shape determination counter CHL ^m is decremented, and at step 600 VPEAK ^m
<CL for counting the number of times when it becomes VL ^m
Increment ^m and move to step 610.

On the other hand, if it is determined in step 580 that VPEAK ^m <VL ^m , the process proceeds to step 620, and this time the maximum value VPEAK ^m.
It is equal to or exceeds the upper limit value VH ^m. And VPEA
If K ^m > VH ^m , the process proceeds to step 630, the distribution shape determination counter VHL ^m is incremented, and the step
Move to 610. If it is determined in step 620 that VPEAK ^m > VH ^m is not established, the process directly proceeds to step 610.

In step 610, the maximum value VPEAK ^m calculated in the knock determination section and used for calculating the median value V50 ^m and determining the distribution shape as described above is initialized, and a new maximum value can be calculated in the next knock determination section. Then, the process proceeds to step 615.

In step 615, it is determined whether or not a predetermined time (for example, 0.7 sec.) Has elapsed since the correction value ΔK ^m was updated in the processing described later last time. Then, if the predetermined time has not elapsed, this processing is ended as it is.

On the other hand, if it is determined in step 615 that the predetermined time has elapsed after updating the previous correction value ΔK ^m , the process proceeds to step 640,
This time, it is determined whether or not the value of the counter CPEAK ^m counted in step 500 is equal to or larger than a predetermined value (for example, 16) within the elapsed time. In this processing, if the number of times the processing of steps 500 to 630 is executed during a predetermined time has elapsed, the distribution shape of the logarithmic conversion value of the maximum value VPEAK ^m is accurately determined based on the value of the counter CHL ^m. Since it may not be possible to update the correction value ΔK ^m to an erroneous value in the processing of step 650 and subsequent steps, in such a case, the update of the correction value ΔK ^m is prohibited. Processing. Therefore, in this step 640, the counter CPEAK
^When it is determined that the value of ^m is less than the predetermined value, step 650
The subsequent correction value update processing is not executed, and the process proceeds to step 740.

On the other hand, if it is determined in step 640 that the value of the counter CPEAK ^m is equal to or larger than the predetermined value, the process proceeds to step 650, and the counter CHL updated in step 590 or step 630 is updated.
It is determined whether the value of ^m is a positive value. And CHL ^m
> 0 (that is, the maximum value VPEAK ^m is the upper limit value V
If the number of times of exceeding H ^m is larger than the number of times of falling below the lower limit value VL ^m ), the break point P position of the distribution shape of the current maximum value VPEAK ^m is lower in cumulative probability than the target position, and knocking occurs. When it is determined that the frequency is higher than the target, the process proceeds to step 660.

Then, in the following step 660, it is determined whether or not the value of the counter CKNK ^m is 1 or less, and if CKNK ^m ≦ 1, the correction value ΔK ^m is subtracted from the correction value ΔK ^{m in the} next step 670. After updating, shifts to step 680, and if not, shifts to step 680 as it is.

In other words, if the value of CKNK ^m is 1 or less and the ignition timing retard control by the knocking control is 1 time or less while the knocking frequency is higher than the target, the fourth The correction value ΔK ^m is reduced because the knock determination level VLEVL ^m calculated in step 220 in the figure is too large.

Note that the processing of steps 650 to 710 is performed by successively updating the correction value ΔK ^m used for calculating the knock determination level VLEVL ^{m so} that the distribution shape of the maximum value VPEAK ^m becomes a desired shape. The level VLEVL ^m is for correcting the knocking sound to be constant, and therefore corresponds to the above-mentioned determination level correcting means M4.

Next, when it is determined in step 650 that CHL ^m > 0 is not established, the process proceeds to step 690 and it is determined whether or not the value of the counter CHL ^m updated in step 590 or step 630 is a negative value. . When CHL ^m <0 (that is, when the maximum value VPEAK ^m falls below the lower limit value VL ^m is larger than the upper limit value VH ^m ), the distribution shape of the current maximum value VPEAK ^m The break point P position is higher in cumulative probability than the target position, and it is determined that the knock occurrence frequency is lower than the target, and the process proceeds to step 700.

In step 700, the counter CKNK value of ^m is determined whether exceeds 1, CKNK ^m> update the correction value [Delta] K ^m by adding a predetermined value β to the correction value [Delta] K ^m if 1 in the next step 710 After that, the process proceeds to step 680, and if not, the process proceeds to step 680 as it is.

In other words, if the value of CKNK ^m exceeds 1 and the ignition timing retard control by the knocking control is executed a plurality of times while the knocking occurrence frequency is smaller than the target value, the knocking control is executed multiple times. The correction value ΔK ^m is reduced because the knock determination level VLEVL ^m calculated in step 220 of FIG. 4 is too small.

As described above, in the present embodiment, the correction value ΔK ^m is set to 3 for low speed, medium speed, and high speed according to the rotational speed NE of the internal combustion engine 2.
Since the kinds of things have been set, when updating the correction value [Delta] K ^m in step 670 or step 710, the correction value [Delta] K ^m corresponding to the rotational speed NE of the internal combustion engine 2 from the three types of correction values
Is to select and update its value.

Next, at step 680, at step 600 the maximum value VPEAK ^m
Is below the lower limit value VL ^m , it is determined whether or not the value of the counter CL ^m that is incremented is 0. And CL ^m =
If it is 0, the process proceeds to step 720 and the upper / lower limit value setting variable A is set.
Update the upper and lower limit values set for the variable A ^m from ^m by subtracting a predetermined value gamma, adds a predetermined value gamma in upper and lower limit setting variable A ^m goes to CL ^{m =} 0 Otherwise step 730 in the opposite Then, the upper / lower limit value setting variable ^Am is updated, and the routine proceeds to step 740.

Then, in the following step 740, after a lapse of a predetermined time, a counter or the like used for updating the correction value is initialized so that the correction value ΔK ^m can be updated in the processing of steps 650 to 730, and this processing is performed. finish.

As described above, in this embodiment, the correction value updating process
Knock determination level for knock control execution is corrected by the correction value [Delta] K ^m, the correction value [Delta] K ^m is, as shown in FIG. 7, the number of NH and the lower limit value VL of the maximum value VPEAK ^m exceeds the upper limit value VH ^m It is updated every predetermined time based on the deviation from the number of times NH that is less than ^m (that is, the value of the counter CHL ^m ). The upper limit VH ^m is calculated by dividing the median V50 ^m obtained at step 510 to step 550 at the upper or lower limit setting variable A ^m, the lower limit value VL ^m upper and lower limits set to the median value V50 ^m It is calculated by multiplying a value obtained by adding the target variable A ^m and the target knock sound adjustment coefficient D, and the upper and lower limit value setting variable A ^m is the number NL of times when the maximum value VPEAK ^m is below the lower limit value VL ^m (that is, the counter CL Depending on the value of ^m ), CL ^m is set to a value of 1 or less.

This means that when the maximum value VPEAK exceeds the upper limit value VH and the lower limit value VL falls below the lower limit value VL in the predetermined time, the break point P is set to the target position as shown by the solid line in FIG. Yes If it can be determined that the distribution shape is the target shape, and if NH> NL, the break point P is below the target position as shown by the alternate long and short dash line in FIG. This is because it can be determined that it is large, and conversely, if NL> NH, it can be determined that the break point P is above the target position as shown by the chain double-dashed line in FIG. .

The next use the target knock adjustment factor D to calculate the upper limit value VH ^m as described above, for example, upper and lower limit VH ^m only at the upper or lower limit setting variable A ^m, as shown in FIG. 8, VL ^m If the knock determination level is corrected so that NL = NH as described above, the break point should be above the target position Po (that is, the frequency of knock occurrence should be less than the target). Although it is possible to fix it, it is not possible to fix it at the target position Po. In other words, by using this knock sound adjustment coefficient D, it becomes possible to make NL = NH when the break point P reaches the target position. Therefore, if the value of the target knock sound adjustment coefficient D is changed, the distribution shape, that is, the position of the break point P is changed, and the knock occurrence frequency can be changed to a desired value.

The rotational speed of the upper to update the lower limit value setting variable A ^m as described above, the internal combustion engine M1 knock occurrence frequency per time
This is to keep it constant regardless of NE. That is, as shown in FIG. 10 described above, in order to make the knock occurrence frequency constant, it is necessary to change the target break point P position in accordance with the rotational speed NE of the internal combustion engine M1. the upper and lower limit values set for the variable a ^m, by the number NL of the maximum value VPEAK is below the lower limit value VL within a predetermined time is set to a predetermined number of times, and the target in accordance with the rotational speed NE of the internal combustion engine M1 That is, the position of the break point P (that is, the distribution shape) is changed.

That is, for example, when the cumulative value of 20% of the points where NL2 (= NL2 / sec.) Per hour when the rotational speed NE of the internal combustion engine 2 is 2000 rpm is 20%, the doubled rotational speed (4000 rpm.
pm), the point where NL4 (-NL4 / sec.) per hour becomes a predetermined value is cumulative 10%, so the upper and lower limit value setting variables set at each rotation speed are from the median value V50, respectively. Cumulative 20
Distance to the% point becomes a distance of up to 10% points accumulated probability thus calculated on basis of the lower limit setting for the variable A ^m set the upper and lower limit values VH, knock determination level correction control using the VL By doing so, the position of the break point P can be changed according to the rotation speed.

As described above, in the present embodiment, the transmission of the knock determination level correction permission signal from the control circuit 40 is prohibited when the rotation speed of the internal combustion engine 2 is equal to or lower than the predetermined rotation speed.
The updating process of the correction value ΔK is not executed on the knock determination circuit 30 side. Therefore, when the air-fuel ratio becomes lean and heavy knock occurs in the internal combustion engine due to component failure or the like during low-speed operation of the internal combustion engine 2, the distribution shape of the logarithmic conversion value of the maximum value VPEAK is erroneously determined and the correction value ΔK That is, the knock determination level is not erroneously corrected to a large value.

Further, in the present embodiment, the transmission of the knock determination level correction permission signal from the control circuit 40 is prohibited even when the internal combustion engine 2 is not in steady operation or when the internal combustion engine 2 is operating at a high rotational speed. To be done. Therefore, in a region where the knocking control cannot be stably executed during the transient operation of the internal combustion engine 2 or the correction value ΔK does not substantially need to be updated, the correction value updating process (that is, the knock determination level correction) is performed.
Can be prohibited and the correction value ΔK due to frequent knocks
It is possible to prevent erroneous correction of and reduce the load on the CPU 30a.

As described above in detail, according to the knocking control device of the present invention, when the rotation speed of the internal combustion engine is equal to or lower than the predetermined speed,
Since correction of the knock determination level is prohibited, when the air-fuel ratio becomes lean and heavy knock occurs in the internal combustion engine due to component failure etc. at low engine speed, the distribution shape of the logarithmic conversion value of the maximum value is incorrect. It is possible to prevent the determination and prevent erroneous correction of the knock determination level.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of the present invention, FIG. 2 is a schematic configuration diagram showing an internal combustion engine and its peripheral devices of an embodiment, and FIG. 3 is a flowchart showing an ignition timing calculation process executed by a control circuit. FIG. 4 is a flow chart showing knock determination processing executed by the knock determination circuit, FIG. 5 is a flow chart showing processing for knock intensity and maximum value calculation executed in step 230 of knock determination processing, and FIG. FIG. 7 is a flow chart showing the correction value update processing executed in step 260 of the judgment processing, FIG. 7 is a diagram for explaining the method of judging the distribution shape of the logarithmic conversion value of the maximum value of the embodiment, and FIG. 8 is a knock sound adjustment coefficient. FIG. 9 is a diagram for explaining the problems when determining the distribution shape without using D, and FIG. 9 shows the maximum value of the knock detection signal when the knock is not generated and when the knock is generated. Fig. 10 is a diagram showing the difference in the distribution shape of the number conversion value, Fig. 10 is a diagram showing the break point P position of the distribution shape set according to the rotation speed, and Fig. 11 is the maximum due to the difference in the rotation speed of the internal combustion engine. It is a diagram showing the change of distribution shape of the logarithmic conversion value of the value. M1, 20 …… Knock sensor M2 …… Knocking control means M3 …… Maximum value detection means M4 …… Judgment level correction means M5 …… Rotation speed detection means (24 …… rotation speed sensor) M6 …… Correction inhibition means 30… … Knock judgment circuit, 40 …… Control circuit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Koji Sakakibara, 1-1, Showa-cho, Kariya city, Aichi Prefecture, Nippon Denso Co., Ltd. (72) Inventor, Hiroshi Harada, 1-1, Showa-cho, Kariya city, Aichi prefecture Nippon Denso Co., Ltd. (56) References JP-A-59-43974 (JP, A) JP-A-60-243369 (JP, A)

Claims (1)

(57) [Claims] 1. A knock sensor for detecting knocking generated in an internal combustion engine, and knocking control means for determining knocking based on a detection signal from the knock sensor and controlling a predetermined knock control factor according to the determination result. A maximum value detecting means for detecting a maximum value of a detection signal output from the knock sensor within a predetermined section, and the knocking so that the distribution shape of the logarithmic conversion value of the detected maximum value has a predetermined distribution shape. In a knocking control device for an internal combustion engine, the control means includes a determination level correction means for correcting a knock determination level used for knock determination, and a rotation speed detection means for detecting a rotation speed of the internal combustion engine, When the rotational speed is equal to or lower than a predetermined speed, a correction prohibiting unit that prohibits the correction of the knock determination level by the determination level correcting unit is provided. And a knocking control device for an internal combustion engine.