JPH03156340A - Fault detecting apparatus for engine - Google Patents

Abstract

PURPOSE:To make it possible to detect a defective cylinder even for a multiple-cylinder engine using a crank-angle sensor for the small number of pulses by dividing one ignition period into two rotary angles, and inspecting the defective cylinder in response to the times of the rotary angles corresponding to the respective rotary angles. CONSTITUTION:One ignition period is divided into two angles of the first rotary angle and the second rotary angles at a specified angular interval with a rotary-angle detecting means 1 and the angles are detected. Then, the time in response to the first rotary angle and the time in response to the second rotary angle are detected in a rotary-angle-time detecting means 2. The abnormality of the engine is detected in correspondence with the first rotary angle time and the second rotary angle time in an abnormality detecting means 3 at the next stage. Since the abnormal cylinder is detected based on the rotary angle time, the abnormal cylinder can be detected even for the faults on the secondary side of an ignition coil and in a fuel system. Since one-ignition period is divided into the two angles of the rotary angles having the large change in rotating speed and the rotary angle having the small change and the abnormal cylinder is detected in response to the corresponding rotary angle times, this apparatus can be applied to an engine using a crank-angle sensor for a small number of pulses.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an engine failure detection device for detecting a failed cylinder in a multi-cylinder engine, and particularly to an engine failure detection device for a multi-cylinder engine using a small number of pulse crank angle sensors. The present invention relates to a failure detection device.

[Conventional technology]

Conventionally, a device has been disclosed that detects a failed cylinder in a multi-cylinder engine according to the maximum and minimum values of the rotational speed in one ignition cycle so that failures in the secondary side of the ignition coil, the fuel system, etc. can also be detected. (For example, Japanese Patent Laid-Open No. 61-258955).

[Problem to be solved by the invention]

In the above-mentioned device, in order to detect the maximum value and minimum value of the rotational speed in one ignition cycle, it is necessary to generate a signal for each small rank angle. Therefore, there is a problem that it cannot be applied to an engine using a crank angle sensor with a small number of pulses.

The present invention has been made to solve the above-mentioned problems, and its purpose is to provide an engine that can detect failures even in multi-cylinder engines using a small number of pulse crank angle sensors. An object of the present invention is to provide a failure detection device for use in a vehicle.

[Means to solve the problem]

As shown in FIG. 1, the present invention comprises: a rotation angle detection means that detects one ignition cycle of the engine divided into two rotation angles, a first rotation angle and a second rotation angle, at predetermined angular intervals; rotation angle time detection means for detecting a time corresponding to one rotation angle and a time corresponding to the second rotation angle; The gist of the present invention is an engine failure detection device characterized by comprising an abnormality detection means for detecting an abnormality in the engine.

Here, the rotation angle detection means is configured to detect one ignition cycle of the engine by dividing it into a first rotation angle where the rotation speed change is large and a second rotation angle where the rotation speed change is small when the engine is in an accelerated state. It is preferable to do so.

The abnormality detection means is configured to detect the difference between the first rotation angle time and the second rotation angle, assuming that in the acceleration state of the engine, the rotation speed changes with constant acceleration during the first rotation angle. a rotation angle time estimating means for estimating the first rotation angle time in the next ignition according to the angular time, the first rotation angle, and the second rotation angle; and the detected first rotation. The engine may further include a first abnormality detection means for detecting an abnormal cylinder of the engine according to a deviation between the angular time and the estimated first rotational angular time.

Alternatively, the abnormality detection means includes: unit rotation angle time calculation means for calculating a first unit rotation angle time for the first rotation angle and a second unit rotation angle time for the second rotation angle; The engine may further include a second abnormality detection means for detecting an abnormal cylinder of the engine according to the first unit rotation angular time and the second unit rotation angular time.

Further, the second rotation angle is 1/1 of one ignition cycle before top dead center.
2, and the first rotation angle is preferably an angle obtained by subtracting the second rotation angle from one ignition period.

Furthermore, the second rotation angle is approximately ⅓ of one ignition period before top dead center, and the second rotation angle is an angle obtained by subtracting the second rotation angle from one ignition period. It is better to do this.

On the other hand, the rotation angle detecting means detects the rotation angle from the top dead center of each cylinder to two points after the top dead center.
0' and a second angular position that divides the first angular position of each cylinder into two at a predetermined angular ratio, and the abnormality detecting means detects the first angular position between the two cylinders. It is also possible to determine a misfire from the asymmetry between the first rotation angle time and the second rotation angle time.

Here, it is preferable to provide a deceleration determining means that substantially invalidates the misfire determination by the abnormality detecting means when determining the deceleration state of the engine.

Furthermore, this deceleration detection means can also determine that the deceleration state is occurring when 1/2 or more of all the cylinders continuously misfire by the abnormality detection means.

Further, the abnormality detection means is a correction coefficient adding means for multiplying or dividing either the first rotation angular time or the second rotation angular time by a correction coefficient to make the required time of the two sections asymmetrical when the engine is normal. It is also possible to provide a comparison means for determining a misfire by comparing the magnitude relationship between the required time to which the correction coefficient is added and the required time in the section to which the correction coefficient is not added.

Here, it is preferable to include a correction coefficient variable means for changing the correction coefficient according to at least one of engine rotational speed and load.

[Effect]

As a result of the above, first, the rotation angle detection means
The ignition period is divided into two rotation angles, a first rotation angle and a second rotation angle, and detected at predetermined angular intervals.

Next, the rotation angle time detection means detects the time corresponding to the first rotation angle and the time corresponding to the second rotation angle.

The abnormality detection means detects an abnormality in the engine according to the first rotation angle time and the second rotation angle time.

〔Example〕

EMBODIMENT OF THE INVENTION Hereinafter, the present invention will be explained based on a first embodiment in which the present invention is applied to a four-cylinder four-stroke engine.

FIG. 2 shows a configuration diagram of the embodiment. IO is a rotating shaft that rotates at an angular velocity of 1/2 of the rotation of a crank mounted on a 4-cylinder 4-cycle engine (hereinafter referred to as the engine). 11
is a disk that is fixed to the rotating shaft lO and rotates in synchronization with the rotation of the rotating shaft 10. On the circumference of this disk 11, a number of slits 11a corresponding to the number of cylinders (four in the embodiment) are formed. , are provided at equal intervals. Reference numeral 12 denotes a rotation angle sensor (for example, an optical sensor or a Hall sensor) as a rotation angle detection means that determines the presence or absence of 11a and outputs a pulse signal. 13 is an electronic control unit M (ECU), which is a well-known central fluorescing unit (CPU).
13a, read only memory (ROM) 13
b. Random access memory (RAM) 13
c, bank amplifier RAM 13d, etc., as a calculation logic operation circuit, and is interconnected via a bus 13g with an input port 13e that receives input from various sensors, an output port 13f that outputs control signals to various actuators, etc. It is connected to the. Reference numeral 14 indicates power transistors 14a to 14d according to the ignition signal output from the ECU 13.
15a to 15d are ignition coils whose primary current is switched on and off by power transistors 14a to 14d, and 16a to 16d are spark plugs connected to the secondary side of the ignition coil 15.

The electronic control unit 13 inputs the intake air amount, intake air temperature, cooling water temperature, rotation angle, etc. through the input port 13e, calculates the ignition timing in terms of angle based on these, and converts the calculated ignition timing in terms of time. The control signal is converted and outputted to the igniter 14 via the output port 13f.

Hereinafter, estimation of the rotation angle time in this embodiment will be explained.

When the inventors measured changes in the instantaneous rotational speed of various engines when the engine was in an acceleration state, the third
The characteristics shown in the figure were obtained.

As is clear from this characteristic diagram, the rotational speed of each cylinder starts to increase from almost top dead center P, , P immediately after ignition of each cylinder, and P3 before ignition. From Pi to P2°P immediately after ignition, a characteristic characteristic of almost constant rotational speed was obtained.

Here, the second rotation angle time is Tθ for the second rotation angle θ where the rotation speed change is small (almost constant).

Let Tθ be the first rotation angle time for the first rotation angle θ, which has a large rotational speed change.

Therefore, in Fig. 3, the second rotation angle θ8 in the previous ignition cylinder, that is, the unit rotation angle time Tθ□, , /θ per 1” CA between P, -P, and the second rotation angle in the current ignition cylinder. The difference between the rotation angle Pj Pa and the unit rotation angle time TθA(+-+)/θ per l″CA is the rotational speed change that occurred between the first rotation angle θ, that is, P and −P. caused by Here, the average value of unit rotation angle time per BiCA between P2 and P1 is Ta2. , /θ3. Here, from Fig. 3(a), it can be concluded that the rotational speed is increasing with constant acceleration between P and -P, so Tθl
(il/θ1 is the midpoint P between P2-P. Therefore, the following equation holds true.

Ta08. /θa　TθII　(i)　/θ.

ζTθ1□-, /θo-TθA (i) /θ.

Transforming the above formula, we get ・・・・・・　■ becomes.

Therefore, the rotation angular time between P t P 1, that is, the first rotation angular time Tθ, (1, is expressed as the point P in FIG. 3(a)
! It can be estimated as follows.

In the 4-cylinder engine of this embodiment, P2°P4 is the top dead center of each cylinder, θA = 60'CA, 0m = 1
By setting 20° CA, it is possible to divide the rotational speed into a region where the change in rotational speed is large and a region where the change is small in the accelerated state.

Therefore, since the disk 11 rotates once for every two revolutions of the engine, the pickpockets 7) 11a are each 30'.

Therefore, this θ1. Substituting θ into equation 0, we get Tθ
The first rotation angular time TθII(i) can be estimated using a simple calculation formula: m+a+=4XTθa+t+Ta2(i-11).

Here, when an abnormality occurs in the accelerating state of the engine and a misfire occurs, the characteristic of the rotational speed is as shown by the broken line in FIG. 3(a). As is clear from this characteristic, the rotation speed does not increase when an abnormality occurs. Therefore, as shown in FIG. 3(C), the first rotation angle time Tθ0. ．． becomes larger. Therefore, the first rotation angular time T'θ, 3 estimated from equation (2)
．． ．． and the actual first rotation angle time Tθ0. ．． Abnormality can be detected based on the deviation from the

Here, even when the engine is in a steady state, the rotational speed may decrease due to a failure. However, this does not occur as significantly as under engine acceleration conditions.

Therefore, in this embodiment, a failed cylinder is detected only when the engine is in an accelerated state.

Next, the operation of ignition timing control in the electronic control unit 13 will be explained based on the flowchart of FIG. 4.

In this routine, the input signal from the rotation speed sensor 12 rises. Alternatively, it is an interrupt handling routine that is executed every time the signal falls.

First, in step 1 (10 to step 102), the rotation angle time TθA111+Ta2., is detected according to the input signal. In step 1 (10), it is determined whether the interrupt is caused by the rising edge or the falling edge of the input signal. Step 1 (If it is determined in step 10 that the interrupt is due to a falling edge, then in step 101 the second rotation angle time Tθ
11. ．． Measure and store.

The second rotation angle time Tθ□, , is measured because the time when the interrupt occurs is stored in a predetermined register.
It can be measured by subtracting the time when the previous interrupt occurred from the time when the current interrupt occurred.

Further, if it is determined in step 1 (10 that the interrupt is caused by a rising edge), in step 102, the first rotation angular time Tθ13. is measured and stored.This first rotation angular time T
θ3. ．． ) also for the measurement of the second rotation angle time T
This is similar to the measurement of θA, .

If the interrupt is caused by the fall of the input signal in step 1 (10), then in step lO3 the second rotation angle time Tθ□, measured in step 101, and the second rotation angle time Tθ□, measured in the previous interrupt 1 rotation angle time Tθ1
(Using i-11, the first rotation angular time T
An estimated value T'θ1,1) of fl s t=> is determined and stored. In this example, as mentioned above, θ, = 60°
CA, θm=120°CA, the estimated value T'θ3. ．． ) can be obtained.

T' θ+i+ = 4 X Tθ^(,, -Ta2
N-1) Also, if the interrupt is caused by the rise of the input signal in step 1 (10), step 104 to
Failure detection is performed in step 114.

First, in step 104, Ta2. i, and the estimated value T'θ3.i estimated in the previous interrupt processing. ．． ) is calculated from the following equation and stored.

Δ(i) = T' θm+=+ Ta2(,.

In this embodiment, the acceleration state is detected by continuously changing the rotation angle time of a predetermined rotation angle between ignition times, where k is an arbitrary constant. It is preferable to set an integer multiple of the number of cylinders (for example, 8 in this embodiment) so that failure detection for each cylinder is performed several times. Therefore, steps 107 to 1 described later are
The failure detection process No. 15 is also performed every second ignition. Therefore, in steps 105 and 106, the second ignition from the previous failure detection process is detected. First, in step 105, 1 is added to the variable i. Step 105 is processed once per ignition. That is, the variable i indicates the number of ignitions. Next, in step 106, the variable i is detected. Then, when the variable i is k, that is, when the second ignition has been performed since the previous failure detection process, the process proceeds to step 107.

Then, in step 107, it is determined whether or not the engine was in an accelerated state between previous ignitions. As a method for determining the acceleration state of the engine, for example, in this embodiment, a predetermined rotation angle before a predetermined ignition (for example, 180°C A=Tθ, +T
The change between the rotation angle time of θ, ) and the rotation angle time of the predetermined rotation angle in the current ignition is the predetermined value! In the above cases, it is determined that the engine is in an accelerating state.

In other words, the rotation angle time of the predetermined rotation angle before the predetermined ignition is ToA(
. , +Tθ, (.), and the rotation angle time of the predetermined rotation angle in the current ignition is -n+Tθ, in Tθ□. -0, it is determined whether the engine is in an accelerating state based on whether the following equation holds true.

! ≦(Tθato+ + Tθg+., l −(TθA
fk-11+Tvllfk-1)] If L2 does not hold, that is, if the engine is not in an accelerating state, the process proceeds to step 115. Further, if the above formula holds true, that is, if the engine is in an accelerating state, the process proceeds to step 108.

Step 10B, Step 109. Step l14 is a process for performing steps 110 to 113, which will be described later, a predetermined number of times.

First, in step 108, variable j is reset (j-〇). Then, in step 109, it is determined whether the variable j is equal to a predetermined number of times.

Here, if the variable j is not equal to the predetermined value in step 109, failure detection is performed in step 110 according to Δ(j) determined and stored in step 104. Specifically, as described above, failure detection is performed based on whether the absolute value of Δ(j) is larger than a predetermined value. Therefore, failure detection is performed depending on whether the following equation holds true or not.

Δ(j) l > K Here, if the above equation does not hold, it is considered normal.
ζProceed to step 114, and if the above formula holds true, it is determined that there is a failure and the process proceeds to steps 111 to 11.
At step 3, it is detected whether a failure has occurred in the cylinder.

First, in step 111, the sign of Δ(j) obtained in step 104 is detected. If Δ(j) is negative, this is because the rotational speed did not increase between P and =P in FIG. That is, it is determined that the cylinder corresponding to the j-th ignition is malfunctioning, and in step 112, a flag ERR(j) indicating a malfunction in the cylinder corresponding to the j-th ignition is set (ERR(j)- 1). Moreover, if Δ(j) is positive, it is because the rotational speed did not increase between P0P8 in FIG. 3.

That is, it is determined that the cylinder corresponding to the (j-1)th ignition is malfunctioning, and in step 113, the flag ERR (j-1) indicating the failure of the cylinder corresponding to the (j-1)th ignition is set to cent ( ERRN-1) -1).

Subsequently, in step 114, 1 is added to the variable i. Then, the processes of steps 109 to 114 described above are performed.

Furthermore, if the variable j is equal to the predetermined number of times in step 109, that is, if the processes of steps 110 to 114 described above have been performed, the variable l is reset (i=0) in step 115 and the process is deprecated. Winter ends.

Then, processing such as prohibiting fuel supply to the abnormal cylinder detected in the above processing is performed.

In this embodiment, a failed cylinder is detected only when the engine is in an accelerating state. Since the engine frequently accelerates during normal driving, there is no problem in detecting a failed cylinder only when the engine is accelerating. However, even during normal running, a decrease in rotational speed occurs due to a failure, and the failure cylinder can be detected in the same way, so step 107 in FIG. 5 for detecting the acceleration state of the engine can be omitted.

As described above, abnormal cylinders are detected based on rotation angle time. Therefore, it is possible to reliably detect abnormal cylinders even when there is a failure in the secondary side of the ignition coil, the fuel system, or the like.

In addition, each ignition is divided into two rotation angles: a first rotation angle with a large change in rotation speed and a second rotation angle with a small change in rotation speed, and the corresponding first rotation angle time and second rotation angle are Abnormal cylinders are detected according to the rotation angle time. Therefore,
It can also be applied to engines using a crank angle sensor with a small number of pulses.

Furthermore, in this embodiment, failure cylinder detection is performed in steps 109 to 114 based on data for each ignition, but it is also possible to detect a failure in a cylinder corresponding to the current or previous ignition using data for one ignition. Therefore, a failed cylinder may be detected based on data every two ignitions.

Furthermore, in this embodiment, the estimated first rotation angle time T'
θ, 1 and the actual first rotation angle time Tθ0. .

Abnormality is detected based on the deviation Δ(i) between the first rotation angle θ and the first unit rotation angle time Toar++/θ per I″CA and the second rotation angle θ4 Second unit rotation time Tθ□.

It is also possible to detect an abnormality based on the deviation from /θ.

That is, as is clear from FIG. 3(a), the following equation holds true in the accelerated state.

Δ1(i)=TθA (il /θ, -Tθm (il
/θ, ≧M where M is a predetermined value. However, if a misfire occurs due to a malfunction, as shown in Figure 3 (C),
First rotation angle time Tθ8.1. becomes large, so the above equation no longer holds true.

Therefore, it is possible to detect that a failure has occurred in the ignition cylinder when Δ1 (i) < M holds true.

The flowchart of the second embodiment of the present invention in this case is shown in the fifth embodiment.
As shown in the figure. At step 204, Δ1(1) is calculated and stored using the following equation.

Δ1 (i) = Tθ□, , /θaTθ□i)/θ.

Then, in step 210, Δ1(i) is set to a predetermined value M
If Δ1(i) is smaller than a predetermined value M, it is determined that the cylinder is malfunctioning.

Furthermore, since the rotational speed characteristic shown in FIG. 3(a) is noticeable at low speeds, accuracy can be improved by detecting a failed cylinder only at a predetermined rotational speed or less.

Even in engines other than 4-cylinder engines, it is good to set the ratio of θ and θ8 to l:2 for the rotation angle of one ignition cycle.However, as the number of cylinders increases, θ is the maximum advance angle. If it becomes smaller than , it will not be possible to realize it, so
In order to ensure the maximum advance angle, for example, in a 6-cylinder engine, θ and θ.

The ratio is 1:! Divide into As a result, although the accuracy of rotation angle time estimation is lower than when the ratio of θ^ and θ1 is divided into 1=2, it does not affect the detection of a failed cylinder.

In addition, in the above-described embodiment, a rotation angle sensor having the same number of slits as the number of cylinders was used, but if the ignition cycle can be divided into the first and second rotation angles, the number of slits can be reduced to the number of cylinders. It may be more than the number.

Next, a third embodiment of the present invention will be described with reference to FIG.

11A is a rotary shutter that rotates in conjunction with the crankshaft of an internal combustion engine (not shown); the figure shows a case where it is attached to the camshaft 10 of a 4-stroke, 4-cylinder engine, and has a shielding effect using the Hall effect.

12A is a rotation angle sensor consisting of a magnet and a Hall element that sandwich the shutter 11A. Then, the closing edge of this shutter 11A is set after the top dead center (ATD) of each cylinder.
C) 10"CA (preferably an angular position between the top dead center (TDC) and ATDC20'CA of each cylinder), and the opening edge is set at the midpoint of each cylinder's ATDC1 (10"CA). It is set to .

131 is a buffer that amplifies the angle sensor signal and sends it to the first
．． It is connected to second timer counters 132 and 133. The first timer counter 132 is ATDC1 (10
'(BTDC80') CA to ATDCIOoC
The time required for the opening section (first rotation angle) up to A (first rotation angle time) is counted. The second timer counter 133 changes from ATDC10°C to ATDC1 (10°C).
The time required for the closing section (second rotation angle) to ℃A (second rotation angle)
The rotation angle time) is counted. A latch circuit 134 updates the count value of the first timer counter 132 at the timing of a strobe signal generated from the arithmetic unit 13A. At this time, the contents of the first timer counter 132 are cleared. 135 is a comparator, and the count value of the second timer counter 133 is latched times! 134, the level becomes high, and the output of the comparison result is connected to the W generator 13A. Furthermore, the arithmetic unit 13A is configured to clear the second timer counter 133 at a predetermined timing.

Note that these calculation processes can also be performed as program processing using a microcomputer, in which case the calculation unit 1
3A is a microcomputer, and the output of the buffer 131 is directly connected to the interrupt generation terminal of the microcomputer 13A. The second timer latches 132, 133, latch circuit +34 and comparator 135 can also be omitted.

Figures 7 (A) and (B) show examples of measuring the instantaneous rotational speed of a 4-cylinder engine in a steady state, and the marks indicate the TDC of each cylinder.
It shows the location. Figure 7 (A) shows combustion including misfire;
Figure 7 (B) shows the change in rotational speed during normal combustion,
The lowest speed is reached near ATDCIO'CA.2. After rising sharply, it has again fallen sharply. That is, it can be seen that during normal combustion, almost symmetrical movements occur in one ignition cycle.

The operation of the third embodiment will be explained using FIG. 8. 8th
FIG. 8(a) shows the output waveform of the rotation angle sensor 12A, and its rising edge indicates the timing of ATDC1 (10″CA). FIG. 8(b) shows the output waveform of the rotation angle sensor 12A.
2, after being initialized after a predetermined time has elapsed from ATDC1 (10°CA), counting of clocks with a constant frequency is started from the falling edge of ATDClOoCA, and A
TDC1 (10° CA (7) Counting is stopped at the rising edge. The arithmetic unit 13A generates a strobe signal to the latch circuit 134 after a predetermined time after detecting the rising edge, and transfers the contents of the first timer counter 132 to the latch circuit 134. Immediately thereafter, the contents of the first timer counter 132 are initialized. FIG. 8(C) shows the contents of the launch circuit 134.
This corresponds to the 320 section count value. FIG. 8(d) shows the contents of the second timer counter 133,
ATD after initialization after a predetermined time has elapsed from 10°CA
Counting of a clock with a constant frequency starts at the rising edge of C1 (10° CA) and stops at the falling edge of ATDC10° CA. This clock frequency is used for the first timer counter 132 and the second timer counter 1.
33 may have a predetermined ratio, or the comparator 13 may have a predetermined ratio.
You may change the gain ratio of the output connected to 5, but
The counting output of the timer counter 132 is adjusted to be larger so that when the rotation speed is completely symmetrical, the first timer counter 132 always has a larger counting result than the second timer counter 133. It's Nishide.

Here, when a misfire occurs in a specific cylinder as shown in FIG. 7(A), the count result of the second timer counter 133 becomes larger than that of the first timer counter 132,
The output is inverted and becomes a high level as shown in FIG. 8(e).

When the performance X14W13A detects the falling edge of the output of the rotation angle sensor 12A, it determines whether there is a misfire by checking the output of the comparator 135, and then initializes the contents of the second timer counter 133.

Repeat the above steps.

The above has been described under steady state conditions, but when accelerating, Figure 7 (A)
The comparator output always remains at a low level because the engine speed increases as if after a misfire occurs, but even if a misfire occurs during acceleration, it can be detected.

However, if fuel supply is stopped and engine braking is applied during deceleration, a misfire will be determined, so it is necessary to prohibit misfire determination during deceleration.

For this reason, when misfires are detected continuously for 1/2 or more of all the cylinders, it is determined that the engine is in a deceleration state, and the misfire determination is canceled, thereby preventing erroneous detection.

The calculator 13A calculates a misfire ratio based on the number of misfires during a predetermined number of ignitions. The flowchart in FIG. 9 shows the processing when a microcomputer is used as the arithmetic unit 13A. The flow shown in FIG. 9 is executed every time the output of the buffer 131 is inverted. When the output of the buffer 131 is inverted, depending on whether the output is at a low level or a high level, step 1 (step 10 checks whether it is a rising edge of the rotation angle sensor 12A; if YES, the process branches to step 130, and a latch strobe signal is sent to the latch circuit 134. The process proceeds to the next step 131 and generates a clear signal to the first timer counter 132. Next, the process proceeds to step 132 to count up the number of ignitions, and the number of ignitions is set to a predetermined value m in the next step 133. Determine if it is smaller than m, and if YES, end without doing anything. If it is greater than or equal to m, step 134
Then, the misfire rate is calculated from the number of misfires relative to the number of ignitions.

After this calculation is completed, in step 135, the counters for the number of ignitions and the number of misfires are cleared.

On the other hand, if there is no rising edge in step 1 (10), the process branches to step 101, and when the output of the comparator 135 is checked and it is not a misfire (low level potential), the process branches to step 110, where the counter for the number of consecutive misfires is cleared. After that, Ste.
The process branches to the amplifier 107, generates a clear signal to the second timer counter 133, and ends.

If a misfire occurs in step 101 (high level potential), the process branches to step 102, where the counter for the number of consecutive misfires is incremented, and then the process goes to step 104, and if the counter value is 2 or more (for 4 cylinders), the process branches to step 102. If the answer is YES, it is judged as deceleration and is not counted as a misfire. On the other hand, step 1
If the answer is No in 04, the process branches to step 105, in which the number of misfires is incremented, and then the process branches to step 107.

By the above operation, when 1/2 or more of all cylinders misfire consecutively, it can be determined as a deceleration state and not included in the count of misfires.

Note that, of course, an optical sensor or a magnetic pickup sensor may be used as the rotation angle sensor instead of the Hall sensor.

Furthermore, in the embodiment shown in FIG. 6, the entire misfire determination method may be performed by program processing by a microcomputer. A flowchart of the fourth embodiment in this case is shown in FIG.

The flow shown in FIG. 1O is also executed every time the output of the buffer 131 is inverted. When the output of the buffer 131 is reversed, the process proceeds to the next step 201 after calculating the time required each time the output of the buffer 131 is reversed (for each first and second rotation angle) in step 2 (10). 20
In step 1, whether the output of the buffer 131 is at a low level or a high level is checked to see if it is the rising edge of the rotation angle sensor 12A, and if YES, the current required time calculated in step 10 is the time for the first rotation angle. The process proceeds to step 202. In step 202, this first rotation angle time is multiplied by a predetermined coefficient, and then step 203
Proceed to and store this result as TF90.

On the other hand, if No in step 201, step 2 (10
It is determined that the current required time calculated in step 2 is the time for the second rotation angle, and the process proceeds to step 210.

In this step 210, the current second rotation angle time is compared with the first rotation angle time previously stored in the TF90, and if the current second rotation angle time is longer than the time stored in the TF90, a misfire occurs. Then, it is determined that the current second rotation angle time has become longer, and the process proceeds to step 211, where a misfire flag is set. Further, in step 210, when the current second rotation angle time is less than the time stored in the TF 90, it is determined that combustion is normal, and the process proceeds to step 220, where the misfire flag is cleared.

In the embodiment shown in FIG. 1O, the first rotation angle time is multiplied by a predetermined coefficient, but the second rotation angle time may be divided by a predetermined coefficient. It will be more effective if the coefficient is changed depending on the rotational speed, the amount of intake air per rotation, or the load. Further, by changing the angular ratio between the first rotation angle and the second rotation angle, multiplication and division of the coefficients can be omitted.

〔Effect of the invention〕

As detailed above, the present invention divides one ignition period into two, the first rotation angle and the second rotation angle, and sets the corresponding first rotation angle time and second rotation angle time. A faulty cylinder is detected according to the condition. Therefore, there is an excellent effect that detection of failed cylinders including failures in the secondary side of the ignition coil, fuel system, etc. is possible even in a multi-cylinder engine using a crank angle sensor with a small number of pulses.

[Brief explanation of the drawing]

Figure 1 is a claim correspondence diagram, and Figure 2 is a diagram of 4 to which the present invention is applied.
A configuration diagram of a first embodiment of a four-cylinder four-stroke engine, FIG. 3 is a timing chart in an acceleration state of the above embodiment, FIG. 4 is a flowchart for explaining the operation of the above embodiment, and FIG. 5 is a flow chart of the present invention. Flowchart for explaining the operation of the second embodiment, FIG. 6 is a configuration diagram showing the third embodiment of the device of the present invention, and FIGS. 7(A) and (B) are instantaneous rotational speed characteristics of a 4-cylinder engine in a steady state. Figure 8 is the above 3rd figure.
FIG. 9 is a waveform diagram of each part to explain the operation of the embodiment, and a flowchart 1.
The figure is a flowchart for explaining the operation of the fourth embodiment of the device of the present invention. II, lla, 12... Rotation angle detection means, 13...
E.C.U.

Claims (11)

[Claims] (1) Rotation angle detection means that detects one ignition cycle of the engine divided into two parts, a first rotation angle and a second rotation angle, at predetermined angular intervals; and a time corresponding to the first rotation angle. and a rotation angle time detection means for detecting a time corresponding to the second rotation angle; and an abnormality detection means for detecting an abnormality in the engine according to the first rotation angle time and the second rotation angle time. 1. A failure detection device for an engine, comprising: means. (2) a rotation angle detection means that detects one ignition cycle of the engine in a state of acceleration of the engine by dividing it into a first rotation angle where the rotation speed change is large and a second rotation angle where the rotation speed change is small; rotation angle time detection means for detecting a time corresponding to one rotation angle and a time corresponding to the second rotation angle; 1. An engine failure detection device comprising: abnormal cylinder detection means for detecting an abnormal cylinder of the engine. (3) The abnormal cylinder detection means detects the first rotation angle time and the second rotation angle, assuming that in the acceleration state of the engine, the rotation speed changes with constant acceleration during the first rotation angle. rotation angle time estimating means for estimating the first rotation angle time in the next ignition according to the rotation angle time, the first rotation angle, and the second rotation angle; The rotation angle time and the estimated first
3. The engine failure detection device according to claim 2, further comprising: first abnormality detection means for detecting an abnormal cylinder of the engine according to a deviation from a rotation angle time of the engine. (4) The abnormal cylinder detection means includes unit rotation angle time calculation means for calculating a first unit rotation angle time for the first rotation angle and a second unit rotation angle time for the second rotation angle. , a second abnormality detection means for detecting an abnormal cylinder of the engine according to the second unit rotation angular time and the second unit rotation angular time. failure detection device for engines. (5) The second rotation angle is 1/1 of one ignition cycle before top dead center.
2, and the first rotation angle is an angle obtained by subtracting the second rotation angle from one ignition period. failure detection device for engines. (6) The second rotation angle is approximately 1/3 of one ignition period before top dead center, and the second rotation angle is the angle obtained by subtracting the second rotation angle from one ignition period. The engine failure detection device according to any one of claims (2) to (5). (7) The rotation angle detection means detects a first angular position between the top dead center of each cylinder and 20 degrees after the top dead center, and a predetermined angular ratio between the first angular positions of each cylinder. and a second angular position divided into two, and the abnormality detection means determines a misfire from the asymmetry between the first rotation angular time and the second rotation angular time. Claims (
1) The engine failure detection device described above. (8) The engine failure detection device according to claim (7), further comprising deceleration determining means that substantially invalidates the misfire determination by the abnormality detecting means when determining the deceleration state of the engine. (9) The engine failure according to claim (8), wherein the deceleration detecting means determines that the deceleration state is occurring when 1/2 or more of all cylinders continuously misfire by the abnormality detecting means. Detection device. (10) The abnormality detection means adds a correction coefficient to make the required time of the two sections asymmetrical when the engine is normal by multiplying or dividing either the first rotation angular time or the second rotation angular time by a correction coefficient. An engine failure according to claim (7), characterized in that the engine malfunction according to claim (7), further comprising a comparing means for determining a misfire by comparing the magnitude relationship between the required time with the correction coefficient added and the required time of the section without the addition. Detection device. (11) The engine failure detection device according to claim (10), further comprising a correction coefficient variable means for changing the correction coefficient depending on at least one of engine speed and load.