JP3236217B2 - Knock sensor signal processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing device for a knock sensor used to control knocking of a vehicle, and more particularly to a knock processing device which can remove noise mixed in at the time of sampling a signal. The present invention relates to a knock sensor signal processing device capable of detecting a sensor abnormality.

[0002]

2. Description of the Related Art A knock sensor is used to detect knocking of a vehicle. The knock sensor is mounted at a specific location on a cylinder block of a vehicle engine. The high-frequency vibration caused by knocking transmitted to the cylinder block is converted into an electric signal by a pickup (knock sensor) utilizing the piezoelectric effect. This high frequency vibration is 6 to 10 kHz.
z. The knock control performs ignition timing control, knock sensor abnormality determination, and the like based on a signal from the knock sensor.

FIG. 25 is a diagram showing an analog peak detection circuit and an analog integration circuit provided in a stage preceding the conventional knock control. As shown in the figure, the output of the band-pass filter 2 connected to the knock sensor 1 and passing only the signal of 6 to 10 kHz is input to the analog peak detection circuit 3 and the analog integration circuit 7 . A Narogupiku detecting circuit 3 is composed of an operational amplifier 4 and the transistor 5 and the capacitor 6. The peak hold voltage (peak value) of knock sensor 1 is obtained by accumulating electric charge in capacitor 6. This peak value indicates the knock intensity. The ignition timing is controlled so as to retard the ignition timing according to the knock intensity. The analog integration circuit 7 includes an operational amplifier 8, resistors 9, 10, 11 and a capacitor 12. The integrated voltage (integrated value) of the signal of knock sensor 1 is obtained by accumulating the electric charge of capacitor 12. When the integrated voltage is equal to or less than a predetermined value, the knock sensor is determined to be abnormal.

[0004]

However, in the analog peak detection circuit 3, it is necessary to accumulate electric charges for several milliseconds in order to secure the accuracy of the peak value. Required. For this reason, there is a problem that it is difficult to increase the density of the circuit and make the circuit into an IC. Therefore, instead of the analog peak detection circuit 3, high-speed sampling is performed to change the signal waveform of the knock sensor to A
It is conceivable to detect the maximum peak value digitized by the / D converter. However, if the sampling cycle and the noise cycle coincide with each other and noise is superimposed instantaneously, the peak value is greatly affected. Such superimposition of instantaneous noise was not a problem in the analog peak detection circuit 3 due to the large capacity of the capacitor 5 and was not a problem. However, it is a new problem for digitization.

Further, in the analog integration circuit 7, it is necessary to use a large-capacity capacitor 12 as described above in order to integrate the signal of the knock sensor. For this reason, there is a problem that it is difficult to increase the density of the circuit and make the circuit into an IC.
Therefore, instead of the analog integration circuit 7, it is conceivable to perform high-speed sampling to integrate the signal waveform of the knock sensor. However, similarly, when the sampling cycle and the noise cycle coincide with each other and the noise is instantaneously superimposed, the integration value is greatly affected, and the knock sensor cannot be detected abnormally. Such superimposition of instantaneous noise is not a problem in the analog integration circuit 7 because the effect of the capacitor 12 is large because the capacitor 12 has a large capacity. However, it becomes a new problem in digitization.

Accordingly, an object of the present invention is to provide a high-speed knock sensor signal sampling apparatus capable of removing the influence of noise on the peak value and integrated value of the knock sensor.

[0007]

Means for Solving the Problems The present invention, in order to solve the above problems, a knock signal for detecting a knock click engine
In the signal processing apparatus of JP, sampling the knock signal
A / D converter for performing analog-to-digital conversion by performing
Certain crank angle in the crank TDC of the engine
At regular time intervals in, start the A / D converter, and write no interrupt controller takes the output value of the A / D converter, the output value of the sampling of the A / D converter <br /> set upper and lower limit determination value against the said output value before
Serial and normal value when in the upper threshold value within the lower limit determination value, the normal value is compared with normal values at the previous sampling, obtains a maximum value in each period of the knock signal, each peripheral
And a determining operation unit knocking the maximum value of the period by statistically processing. By this means, even if noise is superimposed during the loading of the A / D converter, it can be removed.
Since a capacitor is not required unlike the related art, it is possible to reduce the size of the device.

[0008] In the signal processing apparatus of the knock signal for detecting a knock click engine, sampling the knock signal
Contact the A / D converter for digitally converting, from the crank TDC within a certain crank angle of the <br/> engine - to analog
The A / D converter is activated at regular time intervals and the
And write no interrupt controller takes the output value of the A / D converter, against the output value of the sampling of the A / D converter
The upper and lower judgment values are set, and the output value is set to the upper limit.
A normal value when in the judgment value and the lower limit determination value, positive
The normal value was compared with the normal value at the time of the previous sampling, the notch
Obtains a minimum value in each period of the click signal, and a determining operation unit knocking the minimum value of the each cycle <br/> statistical processing to. By this means, the size of the device can be reduced as described above.

[0009] In the arithmetic unit, San before SL A / D converter
If the time from when the output value at the time of pulling is above the center value in the amplitude of the knock signal to below the center value is not within a certain range , an abnormality in the knock sensor Is determined. By this means, even if the noise is superimposed and the level fluctuates, the abnormality is determined from the cycle, so that even if there is noise, it is possible to easily detect the abnormality of the knock sensor.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram for explaining a high-speed sampling processing apparatus for a knock sensor signal according to the present invention. As shown in this figure, knock sensor 1 and air flow sensor 20 are attached to engine 100. The output signal of the knock sensor 1 passes through the band-pass filter 2, and the signal of the air flow sensor is input to the high-speed sampling processor 30 for the knock sensor signal. The knock sensor signal high-speed sampling processing device 30 is provided with a high-speed A / D converter 31 for converting the signals of the knock sensor 1 and the air flow sensor 20 from analog to digital, and further includes a calculation unit (CPU).
32, time interrupt controller 33, ROM34, R
AM 35, an interface (I / F) 36 for inputting an E / G rotation speed signal NE from the engine 100,
An interface (I / F) 37 for outputting an ignition signal to engine 100 is provided.

Here, the signal of the knock sensor is in the range of 0 V to 5 V around 2.5 V, and its signal waveform is close to a sine wave, and has a constant frequency in the range of 6 to 10 kHz as described above. I have. Also, the high-speed A / D converter 31
Performs quantization on, for example, 10-bit data. FIG. 2 is a diagram for explaining the operation of the time interrupt controller 33 of FIG. As shown in the figure, the time interrupt controller 33 detects a top dead center pulse (TDC) every 180 ° CA (crank angle) from the engine speed NE.

The time interrupt controller 33
When detecting the position of 10 ° CA from the top dead center, (b)
Start the ATDC 10 ° CA program. When a position at 90 ° CA from the top dead center is detected, (c) ATDC90
° Start the CA program. Top dead center 10 ° CA and 9
During 0 ° CA, (a) a 4 μs program (routine) is started. These programs are stored in the ROM 34, and when read out by the time interrupt controller 33, are temporarily stored in the RAM 35, and are processed by the arithmetic unit 32 as described later.

The reason why the signal processing of the knock sensor is performed in the range between 10 ° CA and 90 ° CA is that knocking occurs in this range and not in other ranges. First, for ease of understanding, examples of programs for detecting knocking when noise is not considered will be described.

FIG. 3 is a flow chart for explaining a program of ATDC 10 ° CA. In step S1, initialization MAX = 0 is performed. In step S2, the start of the 4 μs routine is permitted, and a return process is performed. FIG. 4 is a flowchart illustrating a 4 μs program.

The 4 μs program generates an interrupt every 4 μs and performs the following processing. In step S11, the high-speed A / D converter 31 for performing signal conversion of the knock sensor 1 is started. In step S12, A /
An A / D value which is a conversion output value of the D converter 31 is taken in.

In step S13, it is determined whether MAX <A / D value. If the determination is "NO", a return process is performed. In step S14, the determination is "YES".
If so, MAX = A / D value is stored in the RAM 35 and return processing is performed. Here, the A / D value is 0 to 5 V for 10 bits (10 ¹⁰ = 102
4v). For convenience, 0v = 1 (minimum value), 2.5v = 512 (center value of vibration), 5v = 10
24 (maximum value) is displayed. Note that a value obtained by averaging MAX as a whole or a value obtained by averaging the maximum value of the peak value of one cycle of the vibrating knock signal in the entire cycle is obtained as a final maximum value (MAXF).

FIG. 5 is a flow chart for explaining a program for ATDC 90 ° CA. In step S21, the start of the 4 μs routine is prohibited. In step S22, it is determined that MAX-512> k · Av. Here, k · Av is a knocking judgment value, k is an arbitrary constant, and Av is M to the previous ignition point.
This is the average value of AX-512.

In step S23, the above judgment is made as "Y
In the case of "ES", a knock determination flag is set. In step S24, the determination in step S22 is “NO”
In this case, the knock determination flag is reset. In step S25, the average value Av is updated by setting Av = (7 · Av + MAX−512) / 8.

Next, a 4 μs program will be described with respect to an example in which the signal of the knock sensor 1 is sampled at a high speed and A / D converted to remove noise when a peak value is detected. FIG. 6 is a flowchart illustrating a 4 μs program according to the first embodiment of the present invention.

In steps S31 and S32, the A / D converter 31 for knocking is started, and the A / D value becomes 2.5v (5
12) Above, when the A / D value increases, the A / D value is captured. In step S33, the upper limit determination value = the previous A / D value × 2−the A / D value of the last two times +
Find 10. Here, “10” is a design margin.

In step S34, the lower limit judgment value = the previous A / D value is obtained. In step S35, for the next 4 μs routine, the value A / D value before last time = the previous value A / D value The previous value A / D value = the current A / D value

In step S36, it is determined whether the lower limit judgment value ≦ A / D value ≦ the upper limit judgment value. If the determination is "NO", the return process is performed, and if there is noise, the MAX determination is not performed. In step S37, if the above determination is "YES", it is determined whether or not MAX <A / D value. That is, when the A / D value rises, MAX
, But not when descending. This judgment is "N
If "O", return processing is performed.

In step S38, the above determination is made as "Y
If "ES", MAX = A / D value, and after MAX is updated, return processing is performed. As described above, if the current A / D value is between the lower limit determination value and the upper limit determination value, the A / D value is regarded as a normal value, and it is determined that there is no influence of noise. Judge that there is an influence of noise.

FIG. 7 is a diagram schematically illustrating the formation of the upper limit judgment value in step S33 of FIG. As shown in the figure, since the output of the knock sensor has a substantially sinusoidal waveform, the gradient becomes smaller with the sampling time when the center value is 2.5 V (512) or more. If the current A / D value is normal, it does not exceed the upper limit determination value determined as in step S33, that is, by performing a linear extrapolation and adding a design margin.

In the above description, the A / D value is set to the central value of 2.5 v (51
Although the case of rising above 2) has been described, A /
The same processing may be performed when the D value falls below the center value 2.5v (512). In this case, in steps S <b> 33 and S <b> 34, −10 is added as a design margin of the upper limit determination value and the lower limit determination value instead of the lower limit determination value. Further, in step S38, the value is changed as MAX> A / D value, and MAX is updated when the A / D value decreases, but not when the A / D value increases. This is the same in the following.

FIG. 8 shows a fourth embodiment according to the second embodiment of the present invention.
It is a flowchart explaining a micros program. Second
The embodiment is a modification of the first embodiment relating to the setting of the upper limit judgment value. In steps S41 and S42, the knocking A / D converter 31 is activated, and the A / D value is fetched.

In step S43, it is determined whether A / D value ≦ upper limit determination value. If the determination is “NO”, the return processing is performed, and if there is noise, the MAX is not updated. In step S44, the latest normal value = A / D value is set.

In step S45, it is determined whether or not MAX <A / D value is satisfied. If the determination is "NO", a return process is performed. In step S46, MAX = A / D value is set, MAX is updated, and then return processing is performed.

In step S47, the upper limit judgment value = latest normal value + 100 is set and a return process is performed. As described above, the A / D value of the maximum value that changes at the sampling time interval (100: equivalent to about 0.5 v) may be predicted, and the upper limit determination value may be set in addition to the latest normal value.

FIG. 9 shows a fourth embodiment according to the third embodiment of the present invention.
It is a flowchart explaining a micros program. Third
The embodiment is a modification of the second embodiment relating to the setting of the upper limit judgment value. In this figure, steps S41 to S4
8 is the same as that in FIG. 8, and in a different step S47A, the upper limit determination value = MAX + 100 is set. The same operation and effect as those of the second embodiment can be obtained.

FIG. 10 is a flowchart illustrating a 4 μs program according to the fourth embodiment of the present invention. The fourth embodiment is a modification of the second embodiment relating to the setting of the upper limit judgment value. In this figure, steps S41 to S41
46 is the same as that of FIG.
Is set as upper limit determination value = upper limit determination value + 100. The same operation and effect as those of the second embodiment can be obtained.

FIG. 11 is a flowchart for explaining a 4 μs program according to the fifth embodiment of the present invention. In steps S51 and S52, the A / D converter 31 for knocking is activated, and the A / D value is fetched. In step S5 3, to determine the success or failure of the A / D value ≦ upper threshold value. If this determination is "NO", the flow proceeds to step S55.

In step S54, the above determination is made as "Y
If “ES”, the upper limit judgment value = A / D value + 100 is set. In step S55, upper limit determination value = upper limit determination value + 100 is set, and a return process is performed.

In step S56, it is determined whether MAX <A / D value is satisfied. If the determination is "NO", a return process is performed. In step S57, MAX = A / D value is set, MAX is updated, and return processing is performed.

As described above, in step S53, A
Even if the / D value does not satisfy the upper limit judgment value, the upper limit judgment value is updated. Until now, when the A / D value did not satisfy the upper limit determination value, the upper limit determination value was not updated, and the original upper limit determination value was used. For this reason, if it is not updated, the next determination in step S53 will be inaccurate, and this embodiment avoids inaccuracy in this determination.

FIG. 12 is a flow chart for explaining a program for ATDC 10 ° CA according to the sixth embodiment of the present invention. In step S61, MAX = 0 is set.

In step S62, the engine (E /
G) Read the number of rotations. In step S63, it is determined whether the engine speed is “small”. If the engine speed is "small" in step S64, step S47 in FIG. 8, step S47A in FIG. 9, step S47 in FIG.
47B, constant A = 10 in steps S54 and S55 in FIG.
Set 0 to constant A = 25.

In step S65, if the engine speed is not "small", it is determined whether the engine speed is "medium". In step S66, if the engine speed is "medium", a constant A = 50 is set. Step S67
, If the engine speed is “large”, the constant A = 10
Set to 0.

In step S68, activation of the 4μ routine is permitted, and return processing is performed. In this way,
By increasing the constant A as the engine speed increases, noise is removed in consideration of the fact that the peak value of knocking increases in proportion to the engine speed. Note that, instead of the engine speed, the intake air amount may be used based on the signal of the air flow sensor 20.

FIG. 13 shows an example of a 4 μs program in consideration of the ATDC 10 ° CA program of FIG. 12, which is a modification of FIG. In this figure, step S51,
52, 53, 56, 57 are the same as in FIG.
In a different step 54A, the upper limit judgment value = A / D value + A is set.

In a different step S55A, upper limit judgment value = upper limit judgment value + A is set. FIG. 14 is a flowchart illustrating a 4 μs program according to the seventh embodiment of the present invention.

In steps S71 and S72, the knocking A / D converter 31 is activated, and the A / D value is fetched. In step S73, it is determined whether A / D value ≧ 512. That is, it is determined whether the A / D value is above the center value. If this determination is "NO", the flow proceeds to step S75.

In step S74, the above determination is made as "Y
In the case of “ES”, A = A−25 is set, and the constant A is gradually reduced with each interruption, that is, with the elapsed time after the upper side. This is because when the knocking waveform is substantially a sine wave and becomes larger than the central value of the A / D value, the rate of increase of the peak value decreases in each sampling cycle.

In step S75, A = 100 is set. This means an initial value. Step S7
At 6, a determination is made as to whether A ≧ 25. If this determination is "NO", the flow proceeds to step S77.

In step S77, A = 25 is set. This means the lowest value. In step S78, it is determined whether A / D value ≦ upper limit determination value. If this determination is "NO", the flow proceeds to step S80.

In step S79, the above determination is made as "Y
If “ES”, the upper limit judgment value = A / D value + A is set. In step S80, the upper limit determination value = upper limit determination value + A is set, and a return process is performed.

In step S81, it is determined whether or not MAX <A / D value. If the determination is "NO", a return process is performed. In step S82, MAX = A / D value is set, MAX is updated, and return processing is performed.

FIG. 15 is a flowchart for explaining a 4 μs program according to the eighth embodiment of the present invention. In steps S91 and S92, the knocking A / D converter 31 is activated, and the A / D value is fetched. In step S93, it is determined whether A / D value ≧ 512. If this determination is "NO", the flow proceeds to step S94.

In step S95, it is determined whether A / D value ≦ upper limit determination value. If this determination is "NO", the flow proceeds to step S96. In step S96, the upper limit determination value = upper limit determination value + 100 is set, and a return process is performed.

In step S97, the above determination is made as "Y
If “ES”, the upper limit judgment value = A / D value + 100 is set. In step S98, it is determined whether MAX <A / D value. If the determination is "NO", a return process is performed.

In step S99, MAX = A / D value is set, MAX is updated, and return processing is performed.
In this way, the upper limit determination value may be initialized for each peak of the knocking waveform.

FIG. 16 is a flowchart illustrating a 4 μs program according to the ninth embodiment of the present invention. In this figure, steps S31 to S37 are the same as those in FIG. In step S101, step S101
If the determination at 37 is "YES", a count-up operation such as C = C + 1 is performed.

In step S102, step S3
If the judgment of 6 or 37 is “NO”, C = 0 is set and the return processing is performed. In step S103, it is determined whether or not C ≧ 3 is satisfied. If the determination is "NO", a return process is performed.

In step S104, MAX = A / D value is set. In this way, the normal value determination is, for example, 3
MAX is updated only when it is repeated a plurality of times. That is, the noise is generated in a wide range in terms of time, and even if the noise is normal at the boundary of the determination, it is removed as noise.

FIG. 17 is a flowchart illustrating a 4 μs program according to the tenth embodiment of the present invention.
In this figure, steps S33 to S38 are the same as those in FIG. In steps S101 and S102, the A / D converter 31 for knock is activated, and the A / D value is fetched.

In step S103, it is determined whether A / D value ≧ 512. If this determination is "NO", the flow proceeds to step S104. In step S104, C = 0 is set, and the process proceeds to step S33.

In step S105, a count-up operation such as C = C + 1 is performed. Step S10
At 6, a determination is made as to whether C ≧ 2. If this determination is "NO", the flow proceeds to step S33.

In step S107, if the determination is "YES", the activation of the 4 μs routine is prohibited for only 40 μs. That is, the conversion operation of the knocking A / D converter 31 is stopped. In step S108, C = -10 is set, a prohibition release condition is formed, and a return process is performed.

Unnecessary operation of the A / D converter 31 is prohibited. FIG. 18 is a diagram showing a 4 μs according to the eleventh embodiment of the present invention.
It is a flowchart explaining a program. In the figure, a different step S109 performs A / D conversion processing of another input other than the A / D converter 31, such as an air flow A / D converter (not shown). Effective use of the entire A / D converter.

FIG. 19 is a flowchart illustrating a 4 μs program according to the twelfth embodiment of the present invention.
In steps S111 and S112, the knocking A / D
Activate the converter 31 and capture the A / D value. Step S
At 113, it is determined whether the previous A / D value ≦ the current A / D value is satisfied. If this determination is "NO", the flow proceeds to step S115.

In step S114, the determination "Y
In the case of “ES”, C = 0 is set. In step S115, a count-up operation of C = C + 1 is performed.

In step S116, it is determined whether or not C ≧ 3. If this determination is "NO", the flow proceeds to step S120. In step S117, the previous A / D value = the current A / D value is set.

In step S118, it is determined whether or not MAX <A / D value. If the determination is "NO", a return process is performed. In step S119, if the above determination is "YES", MAX = A / D value is set and return processing is performed.

In step S120, the activation of the 4 μs routine is prohibited for only 40 μs. In step S121, C = −10 is set. In this manner, when the A / D value continuously decreases, the conversion operation of the knocking A / D converter 31 is stopped for a certain period of time.

FIG. 20 is a flowchart illustrating an ATDC10 ° CA program according to the thirteenth embodiment of the present invention. In step S131, MAX = 0 is set.

In step S132, the constant A is set as A = Av × １ ／. Here, Av is step S in FIG.
25. Thus, the constant A = 1 in step S47 in FIG. 8, step S47A in FIG. 9, step S47B in FIG. 10, and steps S54 and 55 in FIG.
00 is set as the constant A. In this manner, the constant A is made variable based on the value calculated from the peak value up to that time.

FIG. 21 is a flowchart illustrating a 4 μs program according to the fourteenth embodiment of the present invention.
In steps S131 and 132, the knocking A / D
Activate the converter 31 and capture the A / D value. Step S
At 133, a plurality of maximum values stored in advance, for example, 2
It is determined whether the A / D value is greater than MAX1 between MAX1 and MAX2 (MAX1> MAX2). If this determination is "NO", the flow proceeds to step S137.

In step S134, if the above determination is "YES", MAX = MAX2 is set. In step S135, MAX2 = MAX1 is updated.

In step S136, MAX = A / D value is updated and return processing is performed. In step S137, it is determined whether the A / D value is greater than MAX2. If this determination is "NO", the flow proceeds to step S140.

In step S138, if the above determination is "YES", MAX = MAX2 is set. In step S139, MAX2 = A / D value is updated and return processing is performed.

In step S140, when the A / D value is M
It is determined whether it is larger than AX, and if this determination is "NO", return processing is performed. In step S141, if the above determination is "YES", MAX = A / D value is set and return processing is performed.

FIG. 22 is a diagram for reinforcing the description of FIG. As shown in FIG. 7A, when there is no noise,
The A / D values are MAX1, MAX2, MAX in descending order.
Is stored. That is, the third largest A / D value is MA
X is adopted. When the next A / D value is input, since the knocking waveform is a sine wave, the input A / D value becomes MAX1, MAX1 becomes MAX2, and MAX2 is updated as MAX.

As shown in FIG. 9B, when there is noise, for example, the A / D value is stored as MAX1 when there is noise, MAX before that, and MAX2 after that. When the next A / D value is input, MAX1 remains unchanged, the input A / D value becomes MAX2, and MAX2 becomes M
The storage is updated as AX. In this way, the A / D value at the time of noise is not adopted as MAX.

As shown in FIG. 10C, when the A / D value passes through the peak of the knocking waveform, the input A / D value becomes M
Only when it is larger than AX, the input A / D value becomes MAX, and MAX1 and MAX2 remain as they are. Next input A
When the / D value is smaller than MAX, MAX is not updated. In the case where there is no noise, there is no problem even if the third largest A / D value is set as MAX, since there is not much difference from the size of MAX1.

FIG. 23 is a flowchart illustrating a 4 μs program according to the fifteenth embodiment of the present invention.
In steps S151 and S152, the knocking A / D
Activate the converter 31 and capture the A / D value. Step S
At 153, it is determined whether A / D value ≧ 512 and whether the A / D value has passed the center value 512 of vibration.

In step S154, if the above determination is "YES", it is determined whether FA = 1 for the flag FA. If the determination is "NO" in step S155, FA = 1 is set.

[0077] In step S156, the process proceeds to step S16 2 starts measuring the time B (a half cycle). In step S157, the determination in step S153 is “NO”
Then, it is determined whether FA = 1. If this determination is "NO", the flow proceeds to step S162.

In step S158, it is determined whether time B (half cycle) ≦ 50 μs. If this determination is “NO”, step S158
Go to 159. In a normal state, the knocking frequency is 6 to 10 kHz, and the cycle is in the range of 100 to 160 μs.

In step S159, it is determined whether time B (half cycle) ≧ 100 μs. If this determination is “NO”, step S1 is executed.
Go to 61. In step S160, step S1
If the determination at 58 or 159 is "YES", it is determined that the knock sensor is abnormal. This is because when the knock sensor is abnormal, a signal having a cycle different from the knocking cycle is output.

In step S161, FA = 0 is set, and the flow advances to step S162. Step S16
At 2, it is determined whether MAX <A / D value. If this determination is "NO", a return process is performed.

In step S163, if the above determination is "YES", MAX = A / D value is set and return processing is performed. Thus, the abnormality of the knock sensor is detected by detecting the cycle of the vibrating knock signal.

FIG. 24 is a drawing showing ATDC 90 ° C. in FIG.
6 is a flowchart illustrating a program A. In step S171, the start of the 4 μs routine is prohibited. In step S172, it is determined whether or not the knock sensor is abnormal. If this determination is "YES", the flow proceeds to step S177.

When the determination is " YES ", steps S173 to S176 are the same as steps S22 to S25 in FIG. In step S177, the most retarded instruction is given to the control of the ignition timing, and the return processing is performed. In this way, 10 ° CA from top dead center
If there is an abnormality in the cycle of the vibrating knock signal in the range of 90 ° CA, it is determined that the knock signal is abnormal.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a high-speed sampling processing apparatus for a knock sensor signal according to the present invention.

FIG. 2 is a diagram for explaining the operation of a time interrupt controller 33 of FIG. 1;

FIG. 3 is a flowchart illustrating a program of ATDC10 ° CA.

FIG. 4 is a flowchart illustrating a 4 μs program.

FIG. 5 is a flowchart illustrating a program of ATDC 90 ° CA.

FIG. 6 is a flowchart illustrating a 4 μs program according to the first embodiment of the present invention.

FIG. 7 is a diagram schematically illustrating formation of an upper limit determination value in step S33 of FIG. 6;

FIG. 8 is a flowchart illustrating a 4 μs program according to the second embodiment of the present invention.

FIG. 9 is a flowchart illustrating a 4 μs program according to the third embodiment of the present invention.

FIG. 10 is a flowchart illustrating a 4 μs program according to a fourth embodiment of the present invention.

FIG. 11 is a flowchart illustrating a 4 μs program according to a fifth embodiment of the present invention.

FIG. 12 shows an ATDC1 according to a sixth embodiment of the present invention.
It is a flowchart explaining a program of 0 degree CA.

13 is an example of a 4 μs program in consideration of the ATDC 10 ° CA program of FIG. 12, which is a modification of FIG. 11;

FIG. 14 is a flowchart illustrating a 4 μs program according to a seventh embodiment of the present invention.

FIG. 15 is a flowchart illustrating a 4 μs program according to an eighth embodiment of the present invention.

FIG. 16 is a flowchart illustrating a 4 μs program according to a ninth embodiment of the present invention.

FIG. 17 is a flowchart illustrating a 4 μs program according to the tenth embodiment of the present invention.

FIG. 18 is a flowchart illustrating a 4 μs program according to an eleventh embodiment of the present invention.

FIG. 19 is a flowchart illustrating a 4 μs program according to a twelfth embodiment of the present invention.

FIG. 20 is an ATDC according to a thirteenth embodiment of the present invention.
It is a flowchart explaining a 10 degree CA program.

FIG. 21 is a flowchart illustrating a 4 μs program according to a fourteenth embodiment of the present invention.

FIG. 22 is a diagram reinforcing the description of FIG. 21;

FIG. 23 is a flowchart illustrating a 4 μs program according to a fifteenth embodiment of the present invention.

FIG. 24 is a flowchart illustrating a program for ATDC 90 ° CA in FIG. 23;

FIG. 25 is a diagram showing an analog peak detection circuit and an analog integration circuit provided at a stage preceding the conventional knock control.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Knock sensor 31 ... A / D converter 32 ... Operation part 33 ... Time interrupt controller 100 ... Engine

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Shuji Kimura 1-2-28, Goshodori, Hyogo-ku, Kobe-shi, Hyogo Prefecture Inside Fujitsu Ten Co., Ltd. (72) Inventor Kazuaki Murota 1, Goshodori, Hyogo-ku, Kobe-shi, Hyogo No. 2-28, Fujitsu Ten Limited (72) Inventor Tomohide Kasame 1-2-2, Goshodori, Hyogo-ku, Kobe City, Hyogo Prefecture Inside Fujitsu Ten Limited (56) References JP-A-6-42440 (JP (A) JP-A-3-194427 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) F02D 43/00-45/00 F02P 5/145-5/155

Claims (19)

(57) [Claims] 1. A signal processing apparatus of the knock signal <br/> for detecting a knock click engine, analog sampling the knock signal - A / D converter for digitally <br/> conversion of the engine certain crank angle in the crank TDC
In at regular time intervals, the start of A / D converter, and the A / D takes the output value of the transducer write No interrupt controller, against the output value of the sampling of the A / D converter Up
Limit and lower limit judgment values, and the output value is determined by the upper limit judgment.
A normal value when a value and in the lower judgment value, positive
The normal value was compared with the normal value at the time of the previous sampling, the notch
Seeking the maximum values in each period of the click signal, characterized by comprising a determining operation unit knocking the maximum value of the each cycle <br/> statistical processing, the signal processing apparatus of the knock sensor. 2. The arithmetic unit according to claim 1, wherein the calculating unit is configured to perform a previous operation of the A / D converter.
The output value at the time of sampling and the lower limit determination value, characterized in that it further the upper threshold value by adding a design excess Hiroshichi a straight line outside the interpolated values from the second preceding sampling time and the output value of the previous sampling, wherein Item 2. A signal processing device for a knock sensor according to Item 1. 3. The arithmetic unit according to claim 1, wherein the upper limit determination value is obtained by adding a constant value considering a noise level to a latest normal value in an output value at the time of sampling of the A / D converter. The signal processing device for a knock sensor according to claim 1, wherein 4. The arithmetic unit according to claim 1, wherein the upper limit determination value is obtained by adding a constant value considering a noise level to the maximum value of the output value at the time of sampling of the A / D converter. The signal processing device for a knock sensor according to claim 1, wherein 5. The arithmetic unit according to claim 1, wherein the upper limit determination value is obtained by sampling a previous output value of the A / D converter at the time of sampling.
The signal processing device for a knock sensor according to claim 1, wherein the signal is obtained by adding a fixed value considering a noise level to an upper limit determination value set at the time of ringing . 6. The arithmetic unit, when the normal value is determined in the determination, the upper limit determination value is obtained by adding a constant value considering a noise level to the latest normal value. 2. The method according to claim 1, wherein, if not determined as a normal value, the upper limit determination value is obtained by adding a fixed value considering a noise level to an upper limit determination value set at the previous sampling. 3. Knock sensor signal processing device. The method according to claim 7, wherein the arithmetic unit, before Symbol constant value is varied by the engine condition, the signal processor of the knock sensor according to any one of claims 3 to 6. 8. The arithmetic and logic unit according to claim 1, wherein :
Output value is passed along with the time after becoming upper amplitude center value of in the knock signal during sampling, pre Symbol constant value is characterized by the variable so as to be smaller, of claims 3 to 6 A signal processing device for a knock sensor according to any one of the preceding claims. The method according to claim 9, wherein the calculating unit, for each period of the knock signal, and performs initial setting of the constant value which is variable, the signal processing apparatus of the knock sensor according to claim 7 or 8 . 10. The arithmetic unit according to claim 1, wherein one of the knock signals is
Only when the normal value judgment is repeated several times in each cycle ,
The maximum value of each cycle of the lock signal is obtained .
The knock sensor signal processing device according to claim 1. 11. The arithmetic unit according to claim 1, wherein:
The output value at the time of sampling is the amplitude of the knock signal .
Characterized by stopping the conversion process of the A / D converter after a certain time from when the lower heart value in the claim 1
3. A signal processing device for a knock sensor according to claim 1. 12. The arithmetic unit according to claim 1, wherein the A / D converter includes:
2. The knock sensor signal processing device according to claim 1, wherein the conversion process of the A / D converter is stopped for a predetermined time when the output value at the time of sampling continuously decreases. 13. While the A / D conversion processing is stopped, another A
The signal processing device for a knock sensor according to claim 11 or 12, wherein conversion processing of a / D input is performed. 14. The knock sensor according to claim 3, wherein the calculation unit changes the constant value based on the magnitude of the maximum value. Signal processing device. 15. knock signal for detecting a knock click engine
In the signal processing apparatus of JP, the samples the knock signal analog - digital <br/> and A / D converter for converting, within a certain crank angle from the crank TDC of the engine
At regular time intervals in, start the A / D converter, an interrupt controller to capture the output value of the A / D converter, previous San the output value of the sampling of the A / D converter
A knock sensor comprising: a plurality of maximum values in one cycle of the knock signal obtained by comparison with an output value at the time of pulling; Signal processing device. The method according to claim 16, wherein the arithmetic unit, before SL A / D converter
The output value at the time of sampling depends on the amplitude of the knock signal.
The knock sensor is determined to be abnormal if the time from when the center value becomes higher than the center value to when the center value becomes lower than the center value is not within a certain range . Knock sensor signal processing device. 17. knock signal for detecting a knock click engine
In the signal processing apparatus of JP, by sampling the knock signal analog - digital
An A / D converter for converting, within a certain crank angle from the crank TDC of the engine
In at regular time intervals, the start of A / D converter, and the A / D takes the output value of the transducer write No interrupt controller, against the output value of the sampling of the A / D converter Up
Limit and lower limit judgment values, and the output value is determined by the upper limit judgment.
If the value is within the lower limit judgment value, it is regarded as a normal value.
Is compared with the normal value at the previous sampling,
No. of seeking the minimum value in each cycle, characterized in that it comprises a determining operation unit knocking by statistically processing the minimum value of the each period, the signal processing apparatus of the knock sensor. The method according to claim 18, wherein the arithmetic unit, before SL A / D converter
The output value at the time of sampling depends on the amplitude of the knock signal.
The knock sensor signal according to claim 17, wherein the knock sensor is determined to be abnormal if the time from when the center value falls below the center value to when the center value rises above the center value is not within a certain range . Processing equipment. 19. knock signal for detecting a knock click engine
In the signal processing apparatus of JP, before sampling the keno click signal analog - digital <br/> and A / D converter for converting, within a certain crank angle from the crank TDC of the engine
At regular time intervals in, start the A / D converter, an interrupt controller to capture the output value of the A / D converter, previous San the output value of the sampling of the A / D converter
A knocking unit that sequentially stores a plurality of minimum values in one cycle of the knock signal obtained by comparison with an output value at the time of pulling , and sets a smallest one of the minimum values as a minimum value; Sensor signal processing device.