JPS6255465A - Controlling device for internal combustion engine - Google Patents

Abstract

PURPOSE:To restrain the occurrence of knocking by controlling an ignition timing by either of the maximum value of detected combustion pressure oscillations or its corresponding crank angle together with representative output values of a cylinder and knocking, and also judgements being made whether abnormalities are occurred or not therein. CONSTITUTION:Combustion pressure oscillations are detected by a detecting means A. The maximum value of a combustion pressure or its corresponding crank angle is detected by a detecting means B based on the results, furthermore a representative output value of a cylinder and knocking are detected by a detecting means C. And, the representative output values of a cylinder and knocking are judged whether abnormalities are occurred or not by a first and second judging means D, E respectively. Then, at the occurrence of abnormalities in detecting said values of a cylinder and knocking, respective ignition timings are controlled by an ignition timing control means F according to only the representative output value of knocking and the maximum value of a combustion pressure or its corresponding crank angle. Thus it is possible to avoid such troubles as a reduction of a generated torque, a reduction of fuel consumption, an increase in exhaust gas temperature and the like while restraining the occurrence of knocking.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a control device for an internal combustion engine, and more particularly to a control device for controlling ignition timing.

[Conventional technology]

Conventionally, as an ignition timing control device for controlling the ignition timing of an internal combustion engine, for example, a cylinder pressure sensor (combustion pressure vibration sensor) that detects the cylinder pressure of the engine is provided for each cylinder, and the knocking vibration frequency is detected from the detection signal of the cylinder pressure sensor. The signal components in the band are separated and extracted, and the value obtained by integrating this extracted signal over the compression stroke of the internal combustion engine is taken as the cylinder's representative output value (B), and the value integrated over the explosion stroke is taken as the knock representative output value (K). There is a system that corrects and controls the ignition timing for each cylinder so that the ratio (K/B) becomes a predetermined target value (Japanese Patent Laid-Open No. 59-1
(See Publication No. 16030).

In other words, in an engine, the vibration component in the knocking vibration frequency band during the compression stroke does not change even if knocking occurs, whereas the vibration component in the knocking vibration frequency band during the explosion stroke increases due to the occurrence of knocking, and the knocking intensity increases. There is strong vibration.

Therefore, the components in the knocking vibration frequency band are separated and extracted from the engine cylinder pressure, integrated in the compression stroke and the explosion stroke, respectively, and the representative output value of the cylinder in the section unrelated to knocking (
Detects the representative knock output value (K) corresponding to B) and the knocking intensity, and controls the ignition timing for each cylinder so that the ratio (K/B) exceeds a predetermined target value at a constant frequency. This is to avoid knocking without reducing the output of the internal combustion engine.

[Problem that the invention is trying to solve] However,
In such an internal combustion engine control device, a representative output value (B) of a cylinder, which is an integral value of a compression stroke, is used to control the ignition timing.

Therefore, when the integrator malfunctions or abnormal vibration occurs in the compression stroke part of the detection result of the cylinder pressure sensor, the cylinder representative output value (B) becomes an abnormal value, and ignition timing control is performed based on this. The ignition timing of the cylinder that is being used may become inappropriate.

In addition, when the cylinder pressure sensor malfunctions, the integrator malfunctions, or abnormal vibration occurs in the explosion stroke portion of the cylinder pressure sensor detection result, the representative knock output value (K) becomes an abnormal value.
The ignition timing of the cylinder for which ignition timing control is performed based on this may become inappropriate.

For example, if the ignition timing is retarded than the normal ignition timing, the generated torque may decrease, fuel efficiency may decrease, exhaust temperature may rise excessively, and if the ignition timing is advanced from the normal ignition timing, knocking may occur. arise. In particular, heavy knock caused by advancing the ignition timing too much can lead to engine destruction in the worst case scenario.

[Means for solving problems]

Therefore, the representative output detection means B detects the maximum value of combustion pressure or the crank angle of the maximum value based on the detection result of the control device movement detection means A of the internal combustion engine according to the present invention, and the combustion pressure vibration detection means A detects A representative output detection means C detects the representative output value of the cylinder and the representative output value of knock based on the results, and an abnormality has occurred in the detection result of the representative output value of the cylinder based on the detection result of the representative output detection means C. a first determining means for determining whether or not the representative output value is detected; a second determining means E for determining whether an abnormality has occurred in the detection of the representative output value of knocking based on the detection result of the representative output detecting means C; It is provided with an ignition timing control means F that controls the ignition timing based on the detection results of the output detection means B, the detection results of the representative output detection means C, the first judgment means, and the judgment results of the second judgment means E. be.

[For production]

When an abnormality occurs in the detection of the representative output value of a cylinder, the ignition timing is controlled only by the representative output value of knock, and when an abnormality occurs in the detection of the representative output value of knock, the maximum value Pmas of the combustion pressure (in-cylinder pressure) is controlled. Alternatively, it becomes possible to control the ignition timing using the crank angle θpmax at that time.

〔Example〕

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

FIG. 2 is a block diagram showing the configuration of a control device for a four-cylinder WI engine embodying the present invention.

The cylinder pressure sensor 1A of the first cylinder is a piezoelectric conversion type pressure sensor, and is installed as a washer for the spark plug 22A installed in the cylinder head 21A as shown in FIGS. A charge signal S11 corresponding to the in-cylinder pressure is output.

Note that the cylinder pressure sensors 1B to 1D for the second to fourth cylinders are also installed in the same way as the cylinder pressure sensor 1A, and
Charge signals 812 to s corresponding to the in-cylinder pressures of the cylinders to the fourth cylinder
Output e4.

The charge amplifier 2A converts the charge signal Sl+ from the first cylinder in-cylinder pressure sensor 1A into a voltage signal, and then amplifies it.
It is output as a detection signal S21.

The same applies to charge amplifiers 2B to 2D.

Similarly to the charge amplifier 2A, each cylinder pressure sensor 1B to 1
The charge signals S12 to S+4 from D are converted into voltage signals, amplified, and output as detection signals 322 to S24.

In other words, these cylinder pressure sensors 1A to 1) and charge amplifiers 2A to 2D constitute the combustion pressure vibration detection means A of FIG.

In addition, the crank angle sensor 3 detects the position before the compression top dead center of each cylinder (
BTDC) At the same time as outputting the reference signal S2 at 70 degrees, the position signal S3 is output every 1 or 2 degrees of the crank angle.

Note that, of the reference signal S2, the reference signal corresponding to the first cylinder has a wider pulse width than the reference signals corresponding to the other cylinders.

The idle switch 4 detects whether the engine is in an idle state and outputs an idle state detection signal SI.

On the other hand, the control unit 5 is a circuit that also serves as representative output detection means B2 representative output means C9 first determination means, second determination means E, and ignition timing control means F in FIG. 6, a signal processing circuit 7, a main control circuit 8, etc.

The multiplexer 6 selects the detection signal S21''-824 inputted from the charge amplifiers 2A to 2D according to the selection signal from the main control circuit 8, and outputs the detection signal S21''-824.
Output as 2n.

The signal processing circuit 7 receives the detection signal S from the multiplexer 6.
2n is subjected to predetermined signal processing and converted into a correlation value that correlates with combustion pressure vibration energy during non-knocking, a correlation value that correlates with combustion pressure vibration energy during knocking, and the like.

For example, as shown in FIG.
A predetermined frequency, that is, a frequency band related to knocking (
Only signal components of approximately 6 to 17 KHz are extracted.

Then, the detection signal (extraction signal) S4 output from the charge amplifier 7A is amplified by the amplifier circuit 7B, and the amplified detection signal S5 is half-wave rectified by the rectifier circuit 7C.

Thereafter, the half-wave rectified detection signal S output from the rectifier circuit 7C is integrated by the integrator 7D whose integration operation is controlled by the set/reset signal SSR from the main control circuit 8, and the combustion pressure vibration energy is calculated. The integral signal S7 is output as an integral signal S7 indicating an integral value as a correlation value correlated to the .

Returning to FIG. 2, the main control circuit 8 is CPUIQ.

It is constituted by a microcomputer such as 11013 with built-in ROM 11, RAM 12, A/D converter, etc.

The main control circuit 8 receives the reference signal S2 and position signal S3 from the crank angle sensor 3, the idle state detection signal SI from the idle switch 4, the integral signal 87 from the signal processing circuit 7, and the like.

Then, based on the reference signal S2 and position signal S3 from the crank angle sensor 3, a set/reset signal SSR is output to the integrator 7D of the signal processing circuit 7 to control its integration operation.

In addition, here, the main control circuit 8 is 40 degrees before compression top dead center (
BTDC, 40''), the integrator 7D starts the integration operation, and stops the integration operation at the compression top dead center (TDC).
The integral operation is started again at 5 degrees after compression top dead center (ATDC), and is stopped at 45 degrees ATDC.

In addition, the main control circuit 8 detects an abnormality in the detection of the detection 1 representative output value of the crank angle θpmax at which the combustion pressure (in-cylinder pressure) is maximum, based on the above-mentioned signals and various input signals such as an intake air amount signal (not shown). judgement.

Judgment regarding knocking 2. Determining the amount of ignition timing correction.

Performs processing related to ignition timing control such as determining ignition timing.

Then, the power transistor 16 of the ignition device 15 is controlled on/off based on the determined ignition timing to control the ignition timing.

This ignition timing control (on/off control of the power transistor 16) is controlled by an advance angle value (ADV) register (not shown) provided inside the 1013, and a dwell angle (DWELL) register (not shown).
) Set the values (advance angle value, dwell angle) corresponding to the determined ignition timing in the registers, compare the values of these registers with the value of the counter that counts the position signal S3,
At the time of coincidence, the power transistor 16 is turned on or off.

In addition, in the ignition device 15, the power transistor 16 is controlled on and off, so that the primary current of the ignition coil 18, which is supplied with power from the battery 17, is interrupted and a high voltage is generated on its secondary side. High voltage is selectively distributed and applied to the spark plugs 20A to 20D of the first to fourth cylinders by a distributor to generate spark discharge and ignite.

Note that this main control circuit 8 also performs control other than control related to ignition timing, but a description thereof will be omitted.

Next, the operation of this embodiment configured as described above will be explained with reference to FIG. 5 and subsequent figures.

First, the knocking detection principle in this embodiment will be explained.

It is generally believed that the knock level is determined by human auditory sense based on the relative intensity difference between the sound pressure level due to constantly occurring background noise and the sound pressure level due to knocking vibration.

Therefore, by directly comparing the vibration energy of cylinder pressure (combustion pressure vibration energy) during non-knocking with the vibration energy of cylinder pressure during knock, it is possible to detect a knocking level that closely matches the sensory evaluation of one person. .

By the way, it can be said from experience that knocking does not occur before top dead center, so the cylinder pressure vibration before top dead center (
The integral value obtained by (rectifying) the detected value of (combustion pressure vibration) is:
Regardless of whether or not knocking occurs after top dead center, it can be said that this is a predicted value of the vibration energy of the cylinder pressure during the expansion stroke after top dead center in the non-knocking state.

Therefore, the (rectified) integral value of cylinder pressure vibration within a specified crank angle range before top dead center, and the cylinder pressure within a specified crank angle range after top dead center or within a specified range including the range before top dead center. By comparing the (rectified) integral value of the vibration, the vibration energy of the cylinder pressure during non-knocking is directly compared with the vibration energy of the cylinder pressure during the combustion stroke.

In this case, the integration interval must be selected so as not to be affected by the vibration of the spark plug caused by the vibration of the seating and unseating of the intake and exhaust valves. ~TDC and ATDC5°~AT
DC45' is selected.

Next, the principle of detecting knocking using a representative output value of knocking, that is, a rectified integral value of in-cylinder pressure vibration during the expansion stroke will be explained.

As mentioned above, the (rectification) integral value of cylinder pressure vibration before top dead center is a predicted value of the (!1 flow) integral value of cylinder pressure vibration in the expansion stroke after top dead center in the non-knocking state.

Here, as a substitute for the (rectified) integral value before top dead center, we use the (II flow) of cylinder pressure vibration in the expansion stroke after top dead center.
Let us consider the average value of the integral values, that is, the average value of the representative knock output values.

The average value of the representative output value of this knock is 1'! , , corresponds to the (rectified) integral value of the cylinder pressure vibration during the expansion stroke after top dead center in the non-knocking state.

As long as the ignition timing is controlled normally, the frequency of occurrence of knocking will not increase, and the average value of the representative output values of knocking substantially corresponds to the representative output value during non-knocking.

Therefore, by comparing the average value of the (rectified) integral value of cylinder pressure vibration in the expansion stroke after top dead center with the (rectified) integral value of cylinder pressure vibration in the expansion stroke after top dead center, it is possible to By directly comparing the cylinder pressure vibration energy during the combustion stroke with the cylinder pressure vibration energy during the combustion stroke, it is possible to detect a knocking level that closely matches human sensory evaluation.

When strong knocking is detected, calculate the average value of the (rectified) integral value of the cylinder pressure vibration in the expansion stroke after top dead center, and calculate the (rectified) integral of the cylinder pressure vibration in the expansion stroke after top dead center at that time. By excluding the value, the knocking level can be detected more accurately.

Next, the control of the integration operation of the integrator 7D of the signal processing circuit 7 by the main control circuit 8 for performing such processing will be explained with reference to FIG.

First, in a four-cylinder internal combustion engine, the first cylinder #1 to the fourth cylinder
The ignition of cylinder #4 is controlled in the order of #1→#3→#4→#2→#1.

At this time, the crank angle sensor 3 outputs a reference signal S2 at 70 degrees before the top dead center (TDC) of each cylinder as shown in FIG. A position signal S3 is output every 1° (or 2@) of crank angle. Note that, as described above, the pulse width of the reference signal S2 for the first cylinder is wider than the pulse width of the reference signal S2 for the other cylinders.

On the other hand, cylinder pressure sensors 1A to ID and charge amplifier 2A
~When 2D is normal, from charge amplifier 2A,
A detection signal S21 as shown in FIG.
Since the signals S2 to S24 are outputted, the multiplexer 6 outputs a detection signal S2n as shown in FIG.

Thereby, the detection signal S from this multiplexer 6
When only a signal of a predetermined frequency is extracted from 2n by the bandpass filter 7A of the signal processing circuit 7 and amplified by the amplifier circuit 7B, the detection signal s5 as shown in FIG. By rectifying this signal in a rectifier circuit 7C, a detection signal S6 as shown in FIG.

Therefore, the main control circuit 8 activates an internal counter and starts counting the position signal S3 from the time when the reference signal S2 from the crank angle sensor 3 is input.

Then, as shown in FIG.
The integration operation is started, and at time t2 when TDC is reached, the set/reset signal SSR is set to L'' to stop the integration operation.

Thereafter, at time t3 when ATDC 5' is reached, the integrator 7D starts the integration operation in the same way, and is stopped at time t4 when ATDC 45[deg.].

Thereby, the integral signal S7 output from the integrator 7D
between time points t1 and t4, for example, as shown in FIG. The integral operation between 13 and 14 provides a correlation value (integral value) that correlates with the vibration energy at the time of knocking.

Since the main control circuit 8 controls the integration operation of the integrator 7D at the same timing for the second to fourth cylinders, the integral signal S7 outputted from the integrator 7D as a whole is shown in FIG. It becomes as shown in .

Therefore, in a process not shown, the main control circuit 8 A/D converts the integral signal S7 at each TDC, and converts the integrated signal S7 at each TDC into a
The /D conversion value is stored at a predetermined address in the RAM 12 as an amount B correlated to the vibration energy during non-knocking, which is the cylinder's representative output value.

In addition, the integral signal S7 at each ATD, C45'' is
D conversion is performed, and this A/D conversion value is converted into an amount correlated to the vibration energy at the time of knocking, which is the representative output value of knocking.
Store it at a predetermined address in M12.

Regarding knocking control, the ratio between this ftB and the amount (K/B) or the ratio between the average value π of the amount and the amount (K
/K) is calculated and used.

Next, the relationship among the ignition timing, the knocking intensity, and the crank angle θρmax at which the in-cylinder pressure becomes maximum will be explained.

Generally, the relationship between ignition timing, knocking intensity, and crank angle θpmax at which the cylinder pressure is maximum is as shown in FIG. 6. The characteristics shown in FIG. 6 are the results of an experiment using a 4-cylinder engine at 2400 RPM and full load.

That is, the ignition timing and the knocking intensity have a good correlation as long as the operating conditions are determined, and the ignition timing and the crank angle θpmax at which the cylinder pressure is maximum also have a good correlation.

From these factors, the knocking strength and crank angle θp
o+ax has a good correlation as long as the operating conditions are determined. For example, in the example shown in FIG. 6, in order to control to trace knock, the ignition timing may be controlled so that 'θpmax becomes 13° ATDC.

Therefore, a target Opmax value that causes trace knock is stored in advance for each operating condition, and the crank angle θpmax at which the cylinder pressure is maximum is the target θpmax.
The ignition timing may be controlled so as to approach x.

Next, a detection signal S2n from the multiplexer 6 as shown at the crank angle (d) at which the cylinder pressure is maximum is input.

Therefore, the main control circuit 8, for example,
The detection signal S2n from the multiplexer 6 is A/D converted every 2 degrees up to 40 degrees DC.

After that, these A/D conversion values are compared, and the crank angle corresponding to the largest value is determined as the crank angle θp.
max at a predetermined address in RAM 12,
Used for control.

Next, a description will be given of the principle of determining an abnormality in the detection of an amount that is a representative knock output value (K).

Generally, iK, which corresponds to the vibration energy of the cylinder pressure during the combustion stroke, has a minimum value when the internal combustion engine is idling, and
Under no other operating conditions will it ever be less than the idle amount.

FIG. 7 shows the results of measurements of ftK for each operating condition of a four-cylinder internal combustion engine under conditions where knock does not occur by the present applicant.According to various experiments, the relationship shown in the figure is It can be assumed that this holds true for most engines.

Therefore, the value of the amount during idling operation, that is, the minimum value of the amount, is stored in advance at a predetermined address in the RAM 12 as the minimum representative knock value KMIN.

By comparing the amount under operating conditions other than idling with the minimum knock representative value KFi711N,
<KMIN, it can be determined that an abnormality has occurred in the detection of the knock representative output value (K) for that cylinder.

Here, to explain how to obtain the minimum knock representative value KMIN, the amount K during idling operation may be used as the minimum knock representative value KMIN, or in order to obtain a more stable knock minimum representative value KMrN. , the moving average value of ff1K during idling operation is KMIN = (1-α) KMIN + α K However, the value calculated by calculating (0<α<1) may be used as the minimum knock representative value KMIN. .

Further, instead of the value B, which is the vibration energy correlation value during knocking, the amount B, which is the representative output value of the cylinder described above, can be used as the vibration energy correlation value during non-knocking.

In other words, the value of ftB during idling, that is, the amount B
Assume the minimum value of cylinder minimum representative value 8KN.

By storing the amount B in advance at a predetermined address in the RAM 12 and comparing the cylinder minimum representative value BMI N under operating conditions other than idling, when B<BMN, the amount is determined. Typical output value of the cylinder (
It can be determined that an abnormality has occurred in the detection of B).

Also, to explain how to obtain the cylinder minimum representative value BMI N, fitB during idling operation may be used as is as the cylinder minimum representative value 8KN, or to obtain a more stable cylinder minimum representative value BMI N. is the moving average value of B during idling operation. The value may be BMIN.

Next, details of abnormality determination and ignition timing control executed by the main control circuit 8 will be explained with reference to FIG. 8 and subsequent figures.

Referring to FIG. 8, cylinder discrimination processing is performed at 5TEP i. This is done by starting the internal force counter to count the position signal S3 from the crank angle sensor 3 when the reference signal S2 from the crank angle sensor 3 is input (when it rises), and when the reference signal S2 rises. It is determined whether it is the first cylinder or not from the count value when the cylinder falls.

Other cylinders are discriminated based on this discrimination result.

In other words, as mentioned above, the pulse width of the reference signal S2 outputted at 70 degrees before the top dead center of the first cylinder is wider than the pulse width of the reference signal S2 outputted at 70 degrees before the top dead center of the other cylinders. . For example, the pulse width of the reference signal S2 for the first cylinder is about 14 degrees, while the pulse width of the reference signal S2 for the second to fourth cylinders is 4 degrees to 5 degrees.

Therefore, by measuring the pulse width of the input reference signal S2, for example, if the count value is 10 degrees or more, it can be determined that the cylinder is the first cylinder. 4 cylinders.

Since the order is the second cylinder, each cylinder can be identified.

Then, drive the multiplexer 6 with 5TEP2, and
Detection signals 821-S from charge amplifiers 2A-2D corresponding to nine cylinders according to the cylinder discrimination result in TIEP 1
24 is selected and output as the detection signal S2n.

After that, as described above in 5TEP3, the integration timing of the integrator 7D of the signal processing circuit 7 is set.

Between BTDC40' and TDC and ATDC5' and 45°
The integrator 7D is caused to perform an integration operation between.

Then, in 5TEP4 to 10, a first cylinder correction amount determination process to a fourth cylinder correction amount determination process are performed to determine an abnormality in the ignition timing control system and determine an ignition timing correction amount according to the cylinder discrimination result in 5TEP1. .

FIG. S is a flowchart showing in detail the first (t=1 to 4) cylinder correction amount determination process in FIG.

First, the meanings of the abbreviations used in the figure (excluding those already explained) will be explained.

dci: Ignition timing correction amount, 1st cylinder is dcl
, the second cylinder is dc2. The third cylinder is represented by dC3°, and the fourth cylinder is represented by de4.

Note that if dat is positive, it represents an advance angle correction amount, and if it is negative, it represents a retard angle correction amount. In other words, it is assumed that the larger the correction amount is, the more the ignition timing advances.

KM engineering Ni: Quantity K during idling operation, that is, the minimum representative value of knock, the 1st cylinder is KM engineering N1 + the 2nd cylinder is K
Fili engineering N2 + No. N2 + 3rd cylinder N3゜ 4th cylinder is K
Represented by M engineering N4.

α: Means the correction coefficient of the minimum knock representative value KM engineering N, 0
This is the value where α is 1.

BM engineering Ni: Quantity B during idling operation, that is, the cylinder minimum representative value, BM engineering N1 for the 1st cylinder + B for the 2nd cylinder
Fi'l I N 21 The third cylinder is expressed as BP/llN3°, and the fourth cylinder is expressed as BM engineering N4.

α′: means the correction coefficient of the cylinder minimum representative value BM engineering N,
This is the value of 0<α' minus 1.

Next, the correction amount determination process for the i-th cylinder (i=1 to 4) will be explained with reference to FIG.

Detection signal S2n from multiplexer 6 at 5TEPII
A/D every 26 from ATDC2' to ATDC40''
Convert each A/D conversion value and crank angle to RAM.
It is stored in 12 data areas.

Further, the rectified integral signal S7 in a predetermined angle range of the detection signal S2n is A/D converted by TDC to obtain the amount B, and the ATD
A/D conversion is performed at C45° to obtain the quantity K.

Then, at 5TEP12, once at 5TEPII, RAM12
The crank angle that is the maximum value among the A/D conversion values every 2 degrees stored in is determined as θρmax.

Then, at 5TEP13, the idle state detection signal S from the idle switch 4 is checked to determine whether or not the vehicle is in the idle state.

At this time, if the engine is in an idling state, in 5TEP 14, the minimum representative knock value KM-N+ corresponding to the amount K of the i-th cylinder during idling is rewritten.

In other words, the moving average value of the idle amount is (1-α
)・K M 'I N i + α is calculated,
This calculated moving average value is used as the new minimum knock representative value KM.
It is stored at a predetermined address in RAM 12 as N+.

Then, in 5TEP15, the cylinder minimum representative value B MIN i corresponding to 5B during idling of the i-th cylinder is rewritten.

In other words, the moving average value of the amount B during idling is expressed as (1-α
')・BM engineering Ni+α'B is calculated, and the calculated moving average value is stored at a predetermined address in the RAM 12 as a new cylinder minimum representative value BMI~i.

Then, at 5TEP16, the ignition timing correction amount dc of the i-th cylinder
Set i to "0" (dci←0). That is, in this embodiment, control is performed using the basic ignition timing during idling.

On the other hand, if it is not idling at 5TEP l 3, then at 5TEP 17 [K is compared with the minimum representative knock value KM engineering Ni corresponding to the amount K of the i-th cylinder at idling, and it is determined whether K≧KMINi. Determine whether or not.

At this time, if K≧KMIN i, that is, the i-th
If the cylinder knock representative output value (K) is detected normally,
Quantity B at 5TEP 18 and quantity B at idle of the i-th cylinder
By comparing the cylinder minimum representative value BMINi corresponding to B
Determine whether ≧BM Ni.

Then, if B≧B M I N i, that is, if the detection of the representative output value (B) of the i-th cylinder is normal, the amount B, which is the representative output value (B) of the cylinder, and the knock are determined in 5TEP19. The ratio (K/B) to iK, which is the representative output value (K) of
is calculated, and the ignition timing correction i d c i of the i-th cylinder is calculated according to the calculation result.

Also, if B≧BMINi in 5TEP18.

In other words, if the detection of the representative output value (B) of the i-th cylinder is abnormal, the knock representative output value (K) is detected at 5TEP20.
The ratio (K/K) between a certain amount and the average value of this amount is calculated, and the ignition timing correction amount dci of the i-th cylinder is calculated according to the calculation result.

In other words, when the detection of the representative output value (B) of the i-th cylinder becomes abnormal, control using the representative value (B) of that cylinder is stopped and only the representative output value of knock (K) is detected. Switch to the ignition timing control used.

On the other hand, if 5TEP 17 is not ≧KMINi, that is, the representative knock output value (K) of the i-th cylinder
When an error occurs in the detection of 5TEP21,
A process is performed to calculate the i-th cylinder ignition timing correction amount dci based on θpmax determined in TEP12.

Next, details of the ignition timing correction amount calculation process (STEP 19) using the ratio (K/B) in FIG. 9 will be explained with reference to FIG. 10.

First, the meaning of each abbreviation in the figure (excluding those already explained) will be explained.

SL: A reference value for determining the presence or absence of knocking.

KFLG: A flag used to determine the presence or absence of knocking.

BCNT: Value indicating the number of ignitions since the flag KFLG was reset (hereinafter referred to as r count value BCNTJ) KCNT: Value indicating the number of ignitions since the flag KFLG was set (hereinafter referred to as ``count value KCNTJ'') Note that KFLG, KCNT, and BCNT are individually provided for each cylinder correction amount determination process.

First, calculate the ratio (K/B value) between the amount B that is the vibration energy correlation value during non-knocking and the amount that is the vibration energy correlation value during knocking described in 5TEP31, and calculate iK based on iB. To normalize, /B calculation processing is performed.

Then, in 5TEP32, the K/B value calculated by the above-described process is compared with the reference value SL to determine whether the K/B value>SL, thereby determining whether knocking has occurred.

At this time, if the K/B value>SL, that is, if knocking has occurred, the process moves to 5TEP43, which will be described later.

On the other hand, if the K/B value is not > SL, that is, /B value ≦ SL, and knocking has not occurred, 5TEP33 and 5TEP44, which will be described later, set the value when knocking occurs ("1"). ) Flag KFLG
It is determined whether or not is "0".

At this time, if the flag KFLG is rOJ, that is, if knocking has not occurred, then STEP 34 to 37
When the state of /B value≦SL continues for 20 cycles or more after knocking occurs, the ignition timing is advanced by 1 degree.

That is, after the count value BCNT is incremented (+1) in 5TEP34, it is determined in 5TEP35 whether or not the count value BCNT exceeds r20J (BCNT>20).

At this time, if BCNT>20, the process is terminated, and if BCNT>20, the ignition timing is advanced by 1 degree by incrementing the ignition timing correction amount dci (+1)L at 5TEP36, and then 5TEP37 clears the count value BCNT (BCNT=0) and ends the process.

On the other hand, if the flag KFLG is not "0", that is, if knocking has occurred in the past, 5T
When the state of /B value≦SL continues for 20 cycles or more from the time when /B value>SL in EP38 to EP42, processing is performed to make it non-knock.

In other words, at 5TEP38, the count value KCNT is incremented (+1)L, and at 5TEP39, the count value K
It is determined whether CNT exceeds "20"(KCNT>20).

At this time, if KCNT>20, the process ends, and if KCNT>20, the flag KFLG is reset at 5TEP40, and then the count value KCNT is cleared at 5TEP41 (KCNT=O) L,, 5
At TEP42, the count value BCNT is cleared and the process ends.

On the other hand, when /B value > SL at 5TEP32, that is, when knock occurs, 5TEP4
At step 3, it is checked whether the flag KFLG is "0" or not to determine whether or not knocking has occurred for the first time.

At this time, if the flag KFLG is "0", that is, if it is the first knocking, the flag KFLG is set at 5TEP44.
1.5TEP45 after setting LG (KFLG=1)
The count value KCNT is cleared and the process ends.

On the other hand, if the flag KFLG is not rOJ, that is, if knocking occurs for the second time or later, 5TE
P46, the number of past ignitions is within 10 (KCNT≦10)
), i.e. within 10 cycles /B value > SL
Determine whether it has become.

At this time, if KCNT≦10, the above-mentioned 5TE
Execute P45 to end the process, and if KCNT≦10, 5TEP47 decrements (-1) the ignition timing correction amount dci to retard the ignition timing by 1 degree, and then the STY! Execute P45 and end the process.

In this manner, when knocking occurs, the frequency of knocking is determined, and the amount of correction of the ignition timing is determined based on the frequency determination result.

In addition, the correction amount dc in each of the above 5TEPs 36 and 47
Regarding i, the ignition timing is advanced or retarded by more than a predetermined value by determining whether the correction amount dci after correction does not exceed a predetermined value and limiting the value of the correction amount dci. You can also choose not to have any corners.

Also, the amount of retardation to be decremented by 5TEP47 is as follows.

The angle is not limited to 1 degree, such as 1/2 degree or 1/4 degree, but can also be set to a value depending on the magnitude of the /B value, that is, the intensity or degree of the knock.

Next, the ignition timing correction amount calculation process (STEP 20 ) using the ratio (K/K) in FIG. S will be explained with reference to FIG. 11.

This process is almost the same as the ignition timing correction amount calculation process using the ratio (K/B) shown in FIG. 10, so the abbreviation SL has the same meaning as in FIG.

Add r' J to KFLG, BCNT, and KCNT, respectively, and write SL', KFLG', BCNT' KCN
T' and 5TEPs 31' to 47' in contrast to 5TEPs 31 to 47 in FIG.

First, calculate the ratio (K/value) to the average value of the quantity, which is the vibration energy correlation value during knocking mentioned above in 5TEP31', and calculate the quantity K based on the average value k of the quantity. To normalize / perform calculation processing. In addition, this 5TEP31
' is also used to calculate the amount at the same time.

As a method for calculating the average value of ff1K, for example,
The moving average value of the background is obtained by exactly calculating 0<β<1 on (1−β)K+β, and this value is used as the average value.

Then, set the value calculated in 5TEP32' to the reference value S
It is determined whether knocking has occurred by comparing it with /L' and determining whether the value of /L>SL.

Since the subsequent process is the same as the ignition timing correction amount calculation process using the ratio (K/B) explained in FIG. 10, the explanation will be omitted.

Next, details of the ignition timing correction amount calculation process (STEP 21 ) based on θρIIIax in FIG. 9 will be described with reference to FIG. 12.

During this process, the angle θρ+*ax is the crank angle at which the target cylinder pressure, which is determined in advance for each operating condition and stored in the ROM 11, is at its maximum. ing.

First, compare the crank angle θpmax at which the cylinder pressure is maximum at 5TEP51 with the target value I (3pmax,
It is determined whether pmax=Iθpmax.

At this time, if QpmaX=1θ concave ax, it is assumed that control is being performed at the target ignition timing, and the ignition timing is maintained as it is.

On the other hand, if θpmax=Iθpmax,
5TEP52 determines whether θpmax)Iθpmax or not.

At this time, if θpmas〉■θpn+ax, it means that 1 ignition timing is too retarded, so 5TEP53
Add "0,1°" to the ignition timing correction amount dci (dci
+0.1°) to advance the angle.

Also, if θpmax>Iθpmax, it means that the ignition timing is too advanced, so subtract "0°pi" from the ignition timing correction amount dci in 5TEP54 (dci-o
, i”) and retard the angle.

Next, the ignition timing control process will be explained with reference to FIG. 13.

This process is entered and executed when, for example, the reference signal S2 from the crank angle sensor 3 is input.

First, in 5TEP61, the basic ignition timing is determined according to the intake air amount, engine speed, etc. In addition, Koreba ROM
This is done by looking up a table of characteristic values stored in 1l, as shown in FIG. 14, for example.

Then, in 5TEP62 to 68, the determined basic ignition timing and the correction i d determined by the above-mentioned processing are
Based on c i, calculate (70 (D + dc i)), convert BTDC (D + d c i) into an angle from the input timing of reference signal S2, and use this calculation result in ST[EP69 as described above in l101B. Set the advanced angle value (ADV) register.

In this way, in this internal combustion engine control device, the crank angle Opmax at which the in-cylinder pressure (combustion pressure) detected based on the detection result of combustion pressure vibration is the maximum value, the representative output value of the cylinder, and the representative knock The ignition timing is controlled based on the output value and each determination result as to whether or not there is an abnormality detected for each of these representative output values.

Therefore, when the detection of the cylinder's representative output value (B) becomes abnormal, the ignition timing is controlled using the ratio (K/B) between the cylinder's representative output value (B) and the knock representative output value (K). Instead, the ignition timing is controlled using only the knock representative output value (K), and if the detection of the knock representative output value (K) becomes abnormal, the crank angle Oprnax at which the cylinder pressure is maximized is set. Ignition timing can be controlled using

As a result, even if an abnormality occurs in the detection of the representative output value of a cylinder or the representative output value of knock, the generated torque will decrease, fuel efficiency will decrease, and exhaust temperature will increase.

Inconveniences such as occurrence of knocking can be avoided.

In the above embodiment, an example was described in which the crank angle θρmax at which the cylinder pressure (combustion pressure) reaches the maximum value is detected and used for ignition timing control, but the maximum value Pmax of the cylinder pressure (combustion pressure) It can also be detected and used for ignition timing control.

Further, in each of the above embodiments, the cylinder pressure sensor is provided in the spark plug, but a so-called vibration sensor may be provided in the cylinder block.

Further, in the above embodiment, an example was described in which the ignition timing is controlled for each cylinder, but the present invention can be similarly implemented in a case where a plurality of cylinders are set as one group and the ignition timing is controlled for each cylinder group.

〔Effect of the invention〕

As explained above, according to the present invention, θpmax or Pwa detected based on the detection result of combustion pressure vibration
x, the representative output value of the cylinder and the representative output value of knock, which are also detected based on the detection result of combustion vibration pressure vibration, and the determination result of whether or not each representative output value is detected abnormally. Since the timing is controlled, when the detection of the representative output value of a cylinder becomes abnormal, the ignition timing control is performed using only the representative output value of knock instead of the ignition timing control using the representative output value of that cylinder. or when the detection of the representative knock output value becomes abnormal, the ignition timing control using the representative knock output value is replaced by θ.
It is now possible to control ignition timing using pHaX or P wax, and prevent knocking due to abnormality in detection of each representative output value due to malfunction of the integrator or abnormal vibrations on the detection signal of in-cylinder pressure (combustion pressure vibration). It is possible to avoid disadvantages such as a decrease in generated torque, a decrease in fuel efficiency, and an increase in exhaust temperature while suppressing the occurrence of the problem.

[Brief explanation of the drawing]

FIG. 1 is a functional block diagram showing the configuration of the present invention, FIG. 2 is a block diagram showing the configuration of a control device for an internal combustion engine implementing the present invention, and FIG. 3 is a sectional view showing an example of the cylinder pressure sensor. 4 is a block diagram showing an example of the signal processing circuit; FIG. 5 is a timing chart for explaining the integral operation control process of the integrator executed by the main control circuit; FIG. 6 is the ignition timing; A characteristic diagram for explaining the relationship between the knock intensity and Opmax. FIG. 7 is a diagram showing the engine speed/intake air amount-iK characteristic for explaining the principle of detection abnormality determination of the representative knock output value. FIG. 8 is a flowchart showing abnormality determination and ignition executed by the main control circuit; FIG. 10 is a flowchart showing an example of the ignition timing correction amount calculation process using the ratio (K/B) in FIG. 9; Fig. 11 is a flowchart showing an example of the ignition timing correction amount calculation process using the ratio (K/K) in Fig. S, and Fig. 12 is an example of the ignition timing correction amount calculation process based on O prnax in Fig. S. Flow diagram shown. Fig. 13 is a flowchart showing an example of the ignition control process executed by the main control circuit, and Fig. 14 shows the engine speed/intake air flow rate - advance angle value characteristic to explain the basic ignition timing calculation process shown in Fig. 13. FIG. 1A to 1D... Cylinder pressure sensor 2A to 2D... Charge amplifier 3... Crank angle sensor 4... Idle switch 5... Control unit 7... Signal processing circuit 7C... Rectifier 7D. ...Integrator 8...Main control circuit 15...Ignition device Fig. 3 Fig. 4 Fig. 5 Fig. 1 C1 1+ 11 Oni sja Fig. 6 Ignition timing (BTDσ) Fig. 7 Fig. 8 Fig. 12 Figure 13 Figure 14

Claims (1)

[Claims] 1. A control device for an internal combustion engine that controls ignition timing based on a detection result of combustion pressure vibration of the internal combustion engine, comprising a combustion pressure vibration detection means for detecting combustion pressure vibration of the engine, and a detection result of the combustion pressure vibration detection means. a maximum cylinder pressure detection means for detecting a maximum value of combustion pressure or a crank angle of the maximum value based on the combustion pressure, and a representative output value of the cylinder and a representative output value of knock based on the detection result of the combustion pressure vibration detection means. representative output detection means for detecting a representative output value; first determination means for determining whether or not an abnormality has occurred in the detection result of the representative output value of the cylinder based on the detection result of the representative output detection means; and detection of the representative output detection means. a second determination means for determining whether an abnormality has occurred in the detection of the representative knock output value based on the result;
It is characterized by providing an ignition timing control means for controlling the ignition timing based on the detection result of the maximum cylinder pressure detection means, the detection result of the representative output detection means, and each determination result of the first determination means and the second determination means. A control device for an internal combustion engine.