JPH0379548B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for controlling knocking by determining the presence or absence of knocking based on an electrical signal from a knock sensor and advancing or retarding the ignition timing of an internal combustion engine.

There is a knocking control method that uses a knock sensor, which is an element provided in the main body of an internal combustion engine and that electrically converts mechanical vibrations. In the conventional method, the output of the knock sensor is compared with a certain determination level, and if the output is greater than that, it is determined that knocking has occurred, and control is performed to delay the ignition timing. in this case,
If the comparison reference value is fixed, it will be affected by variations in the knock sensor and changes over time. Therefore, the comparison reference value is not fixed, but is calculated based on an output signal from the knock sensor that is considered to be unrelated to knocking (this is referred to as a background signal). That is, a peak hold circuit detects and stores the amplitude value from the knock sensor at a crank angle at which no knocking occurs, and the stored amplitude value is multiplied by a predetermined coefficient to determine whether or not knocking occurs. The presence or absence of knocking is determined by determining the magnitude of the peak hold value in the amplitude obtained from the knock sensor at a crank angle at which knocking is considered to occur.

However, in the conventional control method in which knocking is determined at one level, the ignition timing is controlled by the same amount regardless of the magnitude of knocking, and therefore knocking cannot be efficiently prevented. For example, if control is performed using a judgment level based on small knocking, when the engine conditions are such that large knocking occurs, the ignition timing control amount will be too small, and it will take time to reach the ignition timing that does not cause knocking. This may result in several knockings. On the other hand, if control is performed using a determination level based on large knocking, the ignition timing will be delayed more than necessary when small knocking is detected.

Japanese Utility Model Application No. 56-15457 discloses that in order to obtain an appropriate amount of retardation of the ignition timing, there are two determination levels for knocking, and the strength of knocking is determined by comparing the magnitude of knocking with each determination level. A configuration is disclosed. Furthermore, Japanese Patent Application Laid-Open No. 54-87308 discloses that the output signal of the vibration detector is subjected to high-speed AD conversion and smoothed to determine a judgment level, and the number of local maximum values of the output signal of the vibration detector that are greater than this judgment level is disclosed. A configuration is disclosed in which the intensity of knocking is determined based on this, and the amount of retardation of the ignition timing is controlled based on this.

However, with such a configuration in which the amount of retardation of the ignition timing is controlled only according to the intensity of knocking, in a transient operating state of the engine, for example, when small knocking occurs continuously,
There is a problem in that knocking cannot be quickly reduced.

An object of the present invention is to provide a control method that can quickly reduce knocking even in a transient operating state of the engine.

The knocking control method for an internal combustion engine according to the present invention includes forming at least two judgment levels for the output signal from the knock sensor, large and small, and comparing the output from the knock sensor and the knocking judgment level.
The magnitude of knocking and the frequency of knocking are detected, and the amount of retardation of the ignition timing is controlled according to the magnitude and frequency of knocking. As for the frequency of occurrence of knocking, the retard amount is increased as the frequency of knocking increases, even if the knocking itself is small.

The following will be explained with reference to the drawings. FIG. 1 schematically shows the overall configuration of an internal combustion engine according to the present invention.
0 is a cylinder block of the engine, and 12 is a knock sensor attached to the cylinder block 10. The knock sensor 12 is a well-known device that is composed of, for example, a piezoelectric element or an electromagnetic element, and converts mechanical vibration into electrical vibration fluctuation. In FIG. 1, 14 further indicates a distributor, and this distributor 14 is provided with crank angle sensors 16 and 18.
The crank angle sensor 16 is for cylinder discrimination, and if this engine has 6 cylinders, one pulse is generated every time the distributor shaft makes one revolution, that is, every two revolutions of the crankshaft (every 720°C). occurs.
The occurrence position is set, for example, to the top dead center of the first cylinder. The crank angle sensor 18 generates 24 pulses each time the distributor shaft rotates once.
Therefore, a pulse is generated every 30 degrees of crank angle.

Electrical signals from the knock sensor 12 and crank angle sensors 16 and 18 are sent to a control circuit 20. The control circuit 20 is further fed with a signal representing the intake air flow rate from an air flow sensor 24 provided in the intake passage 22 of the engine. On the other hand, the control circuit 20 outputs an ignition signal to the igniter 26, and the spark current generated by the igniter 26 is distributed to the spark plugs 28 of each cylinder via the distributor 14.

The engine is usually provided with various other sensors that detect operating state parameters, and the control circuit 20 also controls the fuel injection valves 30 and the like, but these are not directly related to the present invention. , all of these will be omitted in the following description.

FIG. 2 is a block diagram showing an example of the configuration of the control circuit 20 shown in FIG. Air flow sensor 24
The voltage signal from is sent to the analog multiplexer 32 via the buffer 30, selected according to instructions from the microcomputer, applied to the A/D converter 34, converted to a binary signal, and then sent to the input/output port. 36 into the microcomputer.

Pulses for every 720 degrees of crank angle from the crank angle sensor 16 are applied to the interrupt request signal forming circuit 40 via the buffer 38. On the other hand, the pulses from the crank angle sensor 18 every 30 degrees of crank angle are
Interrupt request signal forming circuit 4 via buffer 42
0 and is applied to the speed signal forming circuit 44. The interrupt request signal forming circuit 40 forms various interrupt request signals from pulses at every 720° and every 30° crank angle. These interrupt request signals are applied to the microcomputer via the input/output port 46. The speed signal forming circuit 44 determines the crank angle.
A binary signal representing the rotational speed Ne of the engine is formed from the pulse period of every 30°. The formed rotational speed signal is sent to the microcomputer via the input/output port 46.

The output signal of the knock sensor 12 is sent to a peak hold circuit 50 through a circuit 48 comprising a buffer for impedance conversion and a bandpass filter whose pass band is a frequency band (7 to 8 kHz) specific to knocking. The peak hold circuit 50 takes in the output signal from the knock sensor 12 and holds the maximum amplitude only when a "1" level signal is applied from the microcomputer via the line 52 and the input/output port 46. conduct. The output of the peak hold circuit 50 is
converted into a binary signal by an A/D converter 54,
It is sent to the microcomputer via the input/output port 46. However, the A/D conversion of the A/D converter 54 starts at the input/output port 46 and the line 56.
This is performed by an A/D conversion activation signal applied from the microcomputer via the microcomputer. Also,
When the A/D conversion is completed, the A/D converter 54
A/D conversion completion notification is sent to the microcomputer via line 58 and input/output port 46.

On the other hand, when an ignition signal is output from the microcomputer to the drive circuit 60 via the input/output port 46, this is converted into a drive signal and the igniter 26 is energized, depending on the duration and duration of the ignition signal. ignition control is performed.

The microcomputer has the aforementioned input/output ports 36 and 46 and a microprocessor (MPU).
62, random access memory (RAM) 64,
It mainly consists of a read-only memory (ROM) 66, a clock generation circuit (not shown), a memory control circuit, a bus 68 connecting these, etc., and performs various processes according to the control program stored in the ROM 66. .

FIGS. 3 to 8 are flowcharts for explaining ignition timing control, which is the premise of the present invention. According to this processing routine, at least two judgment levels are formed from the knock sensor 12, large and small, and the magnitude of knocking is determined by comparing these two judgment levels with the signal from the knock sensor. Ignition timing can be corrected.
That is, this general idea will be explained with reference to FIG. 9. This figure is a diagram showing the output characteristics from the knock sensor. (B) When the knot king is small; (B)
is assumed to be large. At a relatively stable crank angle where knocking does not occur and the electrical signal from the knock sensor is only basic vibration (e.g. crank angle t ₁ to _{t 2} before compression top dead center), the output amplitude b from the knock sensor is set as a background signal. Take it out as Then, a predetermined coefficient k ₁ is added to this b,
k ₁ b multiplied by k ₂ (>k ₁ ; e.g. k ₂ = 2k ₁ ),
Let k ₂ b be the small and large knocking judgment levels, respectively. Read the output peak value a from the knock sensor at the crank angle at which knocking is considered to occur (for example, the crank angle between t ₃ and _{t 4} after compression top dead center), and compare this with reference values k ₁ b and k ₂ b. . As shown in Figure 9 A, a>k ₁ b, but a<
When K ₂ b, it is determined that the knocking is small, and the ignition timing correction range is reduced accordingly. Further, as shown in FIG. 9B, if a>k ₂ b, it is determined that knocking is large, and the ignition timing correction range is increased accordingly. Next, returning to the flowcharts of FIGS. 3 to 8, a more detailed explanation will be given.

From the interrupt request signal forming circuit 40, a position at a predetermined crank angle θ ₀ CA before the compression top dead center of each cylinder or a predetermined specific cylinder, for example, 60
When a predetermined interrupt request signal is applied at .degree. C.A.BTDC, the MPU 62 executes the interrupt processing routine shown in FIG. That is, in step 70, the time _t1 at which the background detection period starts is calculated. This time t ₁ is the time of the timer on the software, and includes the crank angle position at which the background detection period begins, the current crank angle position (θ 0 CA BTDC), the current crank angle position (θ ₀ CA BTDC), and the current time of the timer. It is clear that it can be easily calculated if the time t ₀ and the rotational speed Ne of the engine are known. When the processing in step 70 is completed, the process returns to the main routine.

When the soft timer reaches time _t1 , a time interrupt request is generated, which causes the MPU 62 to execute the interrupt processing routine shown in FIG. First, in step 71, the peak hold circuit 50 starts a peak hold operation of the output of the knock sensor 12.
This is connected to peak hold circuit 50 via line 52.
This is accomplished by sending a "1" level peak hold instruction signal to. Then step
72, the time at which the background detection period ends
Calculate t ₂ . Since the crank angle position at which the background detection period ends (for example, 10° A.BTDC) is determined in advance, the calculation method in step 72 is the same as that in step 70. Next, in step 73, the A/D converter 54 is instructed to start A/D conversion, and then the process returns to the main routine. Note that the maximum amplitude value b during peak hold is stored at a predetermined location in the RAM.

When the soft timer reaches time _t2 , a time interrupt request is generated, which causes the MPU 62 to execute the interrupt processing routine shown in FIG. First step
At 74, the peak hold operation of the peak hold circuit 50 is ended. This is accomplished by inverting the peak hold instruction signal to "0". Next, in step 75, time _t3 at which the knocking detection period starts is calculated. Crank angle position at which the knocking detection period starts (e.g. 10°C)
A・ATDC) is predetermined, so the calculation method in step 75 is almost the same as that in step 70. The program then returns to the main routine.

When the soft timer reaches time _t3 , a time interrupt request is generated, which causes the MPU 62 to execute the interrupt processing routine shown in FIG. First, in step 84, the peak hold instruction signal is set to "1".
to start peak hold operation.
Next, in step 85, time _t4 at which the knocking detection period ends is calculated. Crank angle position at the end of the knocking detection period (for example, 50℃
ATDC) is determined in advance, so the calculation method in step 85 is almost the same as that in step 70. Then in step 86 A/
After starting the D conversion, the process returns to the main routine.

When the soft timer reaches time _t4 , a time interrupt request is generated, which causes the MPU 62 to execute the interrupt processing routine shown in FIG. i.e. step
At 87, the peak hold instruction signal is inverted to "0" and the peak hold operation is ended. The maximum amplitude value a from the knock sensor generated during this knocking period is stored at a predetermined location in the RAM 64. Then return to the main routine.

Next, when a predetermined interrupt request is input from the interrupt request signal forming circuit at a position a predetermined crank angle θ ₁ CA after compression top dead center, the MPU 62 executes the routine shown in FIG. First, in step 90, the intake air flow rate Q detected by the air flow meter 24 and the rotation speed generated by the rotation speed signal forming circuit 44 are used.
The basic ignition advance angle θ is calculated from Ne using a well-known method. In step 91, the values of peak values b and a in the background detection period (t ₁ to t ₂ ) and the knocking detection period (t ₃ to t ₄ ) obtained by the time interrupt processing shown in FIGS. 4 to 7 are calculated. is fetched from RAM64. Next, in step 92, a is calculated from the peak hold value b of the amplitude during the background period, the peak hold value a during the knocking period, and the constant _k1 .
A determination is made as to whether ≧k ₁ ·b. If a≧ _k1 ·b, it is recognized that knocking has occurred, and in step 93, a counter for counting the number of ignitions is cleared to n=0. Next, in step 94, k ₂ is greater than k ₁ .
Based on the value of (for example, k ₂ =2k ₁ ), it is determined whether a≧k ₂ b.

If a≧k ₂ b, it is recognized as a large knocking (Fig. 9 (b)), and the process goes to step 95, where the ignition timing correction value θ _ki is set by a predetermined angle, for example 2, from the previous correction value θ _ki .
The subtracted value is used as a correction value, and the process goes to step 96, where the ignition timing is determined by adding θ _ki =θ _ki −2 to the basic advance angle. The ignition timing during large knotting is θ
= θ+θ _ki −2, and the amount of correction to the retard side is increased.

In step 94, if a<k ₂ b, it is determined that the knocking is small, and the process goes to step 98, where the previous correction value θ _ki of the ignition timing is subtracted by a predetermined angle, for example 1°, which is smaller than that for large knocking, i.e. θ _ki = θ _ki −
Set 1 as the corrected ignition timing and proceed to step 96. Therefore, the ignition timing at the time of small firing is θ=θ+θ _ki
-1. That is, the amount of correction to the retard side of the ignition timing is made smaller than that in the case of large knocking.

At step 92, if a<k ₁ b, it is recognized that knocking has not occurred, and the process goes to step 100, where the count value of the ignition number counter is a predetermined number of times, e.g.
Determine whether it is greater than 10. At first, since n<10, the process branches NO and goes to step 101, where 1 is added to the count value of the counter. And 102
Go to step 96, set the ignition timing correction value θ _ki the same as last time. Therefore, the ignition timing at this time is θ=θ+θ _ki . In other words, since the engine has just struck, the ignition timing is neither advanced nor retarded.

If n≧10 in 100 steps, it means that ignition was carried out for a considerable time without knocking after knocking occurred. At this time, step
After clearing the counter for counting the number of ignitions after knocking in step 104, in step 105, the ignition timing correction value θ _ki is added to the previous correction value θ _ki by a predetermined small angle, for example, 1°. Therefore, the ignition timing calculated in step 96 is θ=θ+θ _ki +1
Therefore, the ignition timing is adjusted to the advanced side.

On the premise of the above-mentioned normal ignition timing control, _FIG .
The processing in the figure has been changed, and the same steps as in FIG. 8 are designated by the same reference numerals and explanations are omitted, and only the different parts will be explained. N is Konotsuking 2
This is a counter that checks whether the ignition has occurred repeatedly, and is incremented by 1 for each ignition, similar to the counter n. If it is determined in step 92 that no knocking has occurred, this counter is cleared in step 108.
If the determination in step 94 is NO, that is, small knocking, the process branches to step 110, where it is determined whether the number of ignitions N after small knocking is N≧1. If N≧1, it means that small knocking occurred continuously. Therefore, in this case, it is considered necessary to retard the ignition timing even if the firing is small. Therefore, after the counter N representing the number of ignitions after the knock is cleared in step 111, the process goes to step 95, where the ignition timing correction amount θ _ki is set to θ _ki −2. step 110
If N≧1, that is, there is a small notching, but no notching occurs continuously, then 113
After incrementing the counter N, the process goes to step 102. At this time, the ignition timing correction amount θ _ki is not changed.
Incidentally, it may be modified as shown in 98 in FIG. In addition, when it is a big knot king, clear the counter N with 115 and prepare for the next time.

According to the present invention, two knocking determination levels (k ₂ ×b and k ₁ ×b) are provided, and when the higher determination level k ₂ ×b is exceeded, the amount of retardation of the ignition timing is increased ( Steps 94-95 in Fig. 10), when the small level k ₁ ×b is exceeded, when the frequency of knocking is high (in the example, when small knocking occurs twice in a row), the knocking level is Even if the amount of retardation of the ignition timing is small, the amount of retardation of the ignition timing is increased (steps 110-111-95 in FIG. 10). As described above, in the present invention, by performing retard control based on both the magnitude of knocking and the frequency of knocking in parallel, it is possible to perform control with good followability in transient operating conditions. In other words, if the knocking level were to be used for control, the control would be delayed when small knocking occurs frequently during transient operation, but as in the present invention, the control is delayed depending on the knocking level and the knocking frequency. By using this together with control, quick and reliable knocking control can be achieved.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram of an internal combustion engine according to the present invention,
FIG. 2 is a block diagram of the control circuit shown in FIG. 1, FIGS. 3 to 8 are flowcharts showing each interrupt process in ignition timing control, and FIG.
10 is a flowchart of an interrupt routine for each crank angle θ ₁ showing ignition timing control of the present invention. 10...Engine body, 12...Knock sensor, 14...Distributor, 16, 18...
...Crank angle sensor, 20...Control circuit.

Claims (1)

[Claims] 1. The presence or absence of knocking is determined by detecting the electrical output from a knock sensor, which is a mechanical vibration-electrical vibration changing element provided in the main body of the internal combustion engine,
In a method for controlling ignition timing, the magnitude of knocking and the frequency of knocking can be determined by forming at least two judgment levels of the output signal from the knock sensor, large and small, and comparing the output from the knock sensor with the judgment level. The amount of retardation of the ignition timing is controlled according to the size and frequency of knocking, and the larger the knocking is, the larger the retardation amount is. A knocking control method for an internal combustion engine, characterized in that the greater the knocking occurrence frequency, the larger the retard amount.