JPH0417756A - Misfire diagnostic device of internal combustion engine - Google Patents

Abstract

PURPOSE:To diagnose misfire with accuracy by diagnosing both of two cylinders of stroke phase difference corresponding to one cycle of engine rotation as misfiring if misfire is not diagnosed and exhaust component density shows a variation characteristic of misfire. CONSTITUTION:A misfire diagnostic means diagnoses misfire according to the deviation per one cycle of engine rotation or the deviation of this deviation and so cannot diagnose misfire when two cylinders of stroke phase difference corresponding to one cycle of engine rotation both misfire. However, the density of exhaust components must show a variation characteristic of misfire and an exhaust component density detecting means is interposed in the collected portion of an exhaust system and is hard to specify a misfiring cylinder but is capable of detecting generation of misfire. When the exhaust component density shows a variation corresponding to misfire though misfire is not diagnosed according to the cycle of rotation, the two cylinders of stroke phase difference corresponding to one cycle of engine rotation neither of which can be diagnosed by the misfire diagnostic means are therefore both regarded as misfiring.

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to a misfire diagnosis device for an internal combustion engine, and more specifically, a device for diagnosing a misfire based on a deviation of the engine rotation period per engine revolution or a deviation of the deviation. The present invention relates to a technique for preventing omissions in misfire diagnosis.

<Prior Art> As a device for diagnosing a misfire in an internal combustion engine, devices that directly detect the in-cylinder pressure for each cylinder and diagnose a misfire have been proposed. Various devices have been previously proposed that diagnose misfires based on engine rotation fluctuations. For example, a crank angle sensor (180' for 4 cylinders, 6 A device has been proposed that is configured to measure the period of a detection signal output every 120' in the cylinder and diagnose a misfire based on the deviation of this period for each engine revolution or the deviation of this deviation. Ganpei 1-27
(See No. 5046).

That is, for example, the latest measured value of the TDC cycle is Tφ2, and the TDC cycle measured one engine revolution ago is T2, and the TDC cycle measured two engine revolutions ago is T4, and the deviation of the TDC cycle per engine revolution is calculated according to the following formula. Misfire determination value LU based on deviation
A misfire diagnosis is performed by calculating the misfire determination value LU and comparing it with a predetermined slice level.

<Problems to be Solved by the Invention> However, when diagnosing a misfire based on the deviation of the engine rotation period for each engine revolution or the deviation of this deviation as described above, for example, the firing order in a 4-cylinder internal combustion engine must be adjusted. #1
→ #3 → #4 → #2 When two cylinders with a stroke phase difference equivalent to one revolution of the engine misfire, such as cylinder #3 and cylinder #2, as shown in Fig. 11, Changes caused by a misfire no longer appear in the cycle of each engine revolution, and as a result, the diagnosis was made that there was no misfire, even though two cylinders were actually misfiring.

In other words, when a misfire occurs and the cylinder pressure decreases, this appears as a decrease in engine rotational speed, so it is necessary to compare the period corresponding to a normally combusting cylinder with the period corresponding to a misfiring cylinder. However, as shown in Fig. 11, if cylinder #3 and cylinder #2 both misfire, when calculating the periodic deviation for each engine revolution, the two cylinders that both correspond to misfiring cylinders, or , both of which correspond to cylinders in which no misfire has occurred are compared, and even if one attempts to diagnose a misfire based on the magnitude of the periodic deviation, the misfire diagnosis cannot be made.

The present invention has been made in view of the above problems, and includes a device for diagnosing a misfire based on the deviation of the engine rotation period for each engine revolution or the deviation of this deviation.
The first object is to be able to diagnose the occurrence of a misfire even if two cylinders corresponding to the same rotation are both misfired.

<Means for Solving the Problems> Therefore, in the present invention, as shown in FIG. A misfire diagnosis means for diagnosing the occurrence of a misfire based on a deviation for each rotation or a deviation of the deviation; an exhaust component concentration detection means for detecting the concentration of exhaust components by being installed in a collection part of an exhaust system of an internal combustion engine; and a misfire diagnosis. When the occurrence of a misfire is not diagnosed by the means and the detected exhaust component concentration shows a change corresponding to the occurrence of a misfire, both of the two cylinders with a stroke phase difference equivalent to one engine rotation misfire. A misfire diagnostic device is configured to include a two-cylinder misfire diagnostic means for diagnosing that a two-cylinder misfire is occurring.

<Operation> According to this configuration, the misfire diagnosis means diagnoses the occurrence of a misfire based on the deviation of the engine rotation period for each engine rotation or the deviation of the deviation, so that the stroke phase difference 2 corresponding to one engine rotation is detected.
When both cylinders misfire, it becomes impossible to diagnose the misfire as described above. However, the concentration of exhaust components should show a change peculiar to the occurrence of a misfire due to unburned gas being discharged into the exhaust system due to the occurrence of a misfire, and therefore the exhaust component concentration detection means is installed in the collecting part of the exhaust system. Although it is difficult to identify the misfiring cylinder from the above, it is possible to detect the occurrence of a misfire. Therefore, even though a misfire has not been diagnosed based on the rotation period, if the concentration of exhaust components shows a change corresponding to the occurrence of a misfire, the stroke phase difference corresponding to one engine revolution cannot be diagnosed by the misfire diagnostic means. It is assumed that both cylinders are misfiring.

<Example> The present invention will be described in detail below.

In FIG. 2 showing one embodiment, an internal combustion engine 1 includes an air cleaner 2. Intake duct 3. Air is taken in via the throttle channel 4 and the intake manifold 5.

An air flow meter 6 is provided in the intake duct 3 to detect the intake air flow rate Q. The throttle valve 7 is provided in the throttle valve 4 and is connected to an accelerator pedal (not shown) to control the intake air flow rate Q. The intake manifold 5 is provided with an electromagnetic fuel injection valve 8 for each cylinder (four cylinders in this embodiment), which injects fuel that is fed under pressure from a fuel pump (not shown) and controlled to a predetermined pressure by a pressure regulator. It is injected and supplied into the intake manifold 5.

The fuel injection amount is controlled by a control unit 9 with a built-in microcomputer based on the intake air flow tQ detected by the air flow meter 6 and the distributor 1.
The basic fuel injection amount Tp corresponding to the opening driving time of the fuel injection valve 8 is calculated based on the engine rotational speed N calculated based on the signal from the crank angle sensor lO built into the engine 3, and the basic fuel injection amount Tp=KXQ/N(
K is a constant), and by correcting this basic fuel injection amount Tp based on the cooling water temperature Tw detected by the water temperature sensor 14, the air-fuel ratio feedback correction coefficient α, etc. described later, the final fuel injection amount is determined. By calculating Ti and outputting a drive pulse signal with a pulse width corresponding to the fuel injection amount Ti to the fuel injection valve 8 in synchronization with the engine rotation, the required amount of fuel is injected and supplied to the engine 1. It looks like this.

Further, each cylinder of the engine 1 is provided with an ignition plug 11, to which a high voltage generated by an ignition coil 12 is sequentially applied via a distributor 13 (the ignition order in this embodiment is #1 → #3 → #4 → #2), thereby igniting a spark to ignite and burn the air-fuel mixture. Here, the timing of generation of high voltage in the ignition coil 12 is controlled via an attached power transistor 12a. Therefore, ignition timing (ignition advance value)
The ADV is controlled by controlling the on/off timing of the power transistor 12a using an ignition control signal from the control unit 9.

The control unit 9 stores the ignition timing ADV in advance for each of a plurality of operating regions divided by the basic fuel injection amount Tp and the engine rotational speed N, and selects the ignition timing ADV corresponding to the operating condition from a map. is searched and obtained, and an ignition control signal is output based on the ignition timing ADV to control the ignition timing.

The throttle valve 7 is equipped with a throttle sensor 15 that detects its opening with a TVO tensionometer. So BTD
A reference angle signal REF (every 70 degrees) and a unit angle signal PO3 every 1 degree or 2 degrees are output. The reference angle signal REF is transmitted to the control unit 9.
For example, the reference angle signal REF that corresponds to the ignition reference for the #1 cylinder can be distinguished from the others, so that the reference angle signal REF can be applied to each cylinder. The ignition can be controlled individually for each cylinder.

Furthermore, an oxygen sensor 17 corresponding to an exhaust component concentration detection means for detecting the oxygen concentration in the exhaust gas as the exhaust component concentration is installed in the collecting part of the exhaust manifold 16 of the engine 1. Based on the oxygen concentration, it is possible to detect whether the air-fuel ratio of the actual engine intake air-fuel mixture is in a rich state or a lean state with respect to the stoichiometric air-fuel ratio.

Here, the control unit 9 has a function as a misfire diagnostic device that performs the misfire diagnosis according to the present invention and displays the cylinder in which the misfire has occurred near the driver's seat of the vehicle. The misfire diagnosis control is shown in Figures 3 to 8.
The explanation will be made according to the programs shown in the flowcharts shown in the figure.

In this embodiment, the functions of the misfire diagnosing means and the two-cylinder misfire diagnosing means are provided in the form of software as shown in the flowcharts of FIGS. 3 to 8, respectively. Furthermore, in this embodiment, the rotation period measuring means is
It is composed of the crank angle sensor 10 and the control unit 9.

The program shown in the flowchart of FIG. 3 is 15 degrees after compression TDC (ATDCl,
This is executed at every angular position of 5″′).

First, step 1 (indicated as Sl in the figure).

(The same applies hereafter) Then, according to another program (not shown),
From TDC206 to ATDc18o6 (angular range where the crank angular speed is particularly affected by changes in cylinder pressure)
The results of measuring the period of are stored in memory in chronological order.

That is, as shown in FIG. 9, the control unit 9 detects the positions of the ATDC 20@ and ATDC 180' based on the signals output from the crank angle sensor 10, and measures the elapsed time (cycle) during this time. In this step 1, the cycle measured just before is set to TO as the latest measurement cycle, and the data of TO where the latest cycle was set when this program was executed last time is set as the previous cycle. Set it to T1, and in the same way, check the cycles T"1, 2, and 3 times before the previous time T"1.T2゜T
3 and one more old data, respectively, 2 times ago (1 rotation ago) period T2, 3 times ago period T3. 4 times ago (2 rotations ago)
Each is set to period T4.

Then, in the next step 2, the period T2 before the latest period TO11 rotations (1/2 cycle before) set in step 1
．． Using the period T4 before two revolutions (before I cycle), calculate the deviation of the period T for each engine revolution (T2-T4) according to the formula below.
)-(To-72) is calculated.

(T2-74) - (To-72) The above misfire determination value LU is a value that approximately corresponds to the amount of change in the average effective pressure. Changes in the average effective pressure of the cylinders are estimated.

In step 3, the discrimination (t
f! Determine whether LU is greater than or equal to zero, and set the determination value LU
is a negative value, the process advances to step 4 and the discriminant value L
A flag F plus for determining whether U is positive or negative is set to zero.

In the next step 5, the slice level SL is set as a positive value by the program shown in the flowchart of FIG.
is subtracted from zero to convert it into a negative value (0-3L), and this negative slice level SL is compared with the discrimination value LU.

The program shown in the flowchart of FIG. 4 is processed in the background, and in step 81 of the program, each operating region is divided into a plurality of regions based on the basic fuel injection amount Tp representing the engine load and the engine rotational speed N. From a map that stores slice levels SL, a slice level SL that matches the latest operating conditions is searched and set. As mentioned above, the reason why the slice level SL is variably set according to the engine load and the engine rotation speed N is because the level of the discrimination value LU changes greatly depending on the operating conditions of the engine load and the engine rotation speed. The absolute value of the discrimination value LU calculated at the time of occurrence is set so as to exceed this slice level SL.

If it is determined in step 5 that the discriminant value LU is less than or equal to the negative slice level SL, step 6 sets the count value cnt to 4 and then proceeds to step 27; If it is determined that there is (positive), jump step 6 and proceed to step 2.
Proceed to step 7.

In step 27, it is determined whether or not the count value cnt is zero, and if it is not zero, the count value cnt is decreased by 1 in step 28.

Therefore, when the discriminant value LU continues to be in a negative state, when the discriminant value LU becomes less than the negative slice level SL, 4 is added to the count value cnt, but at the same time it is decreased by 1, and then the discriminant value LU is a negative slice level S
If it does not fall below L, it counts down to zero and maintains zero.

On the other hand, when it is determined in step 3 that the discriminant value LU is greater than or equal to zero, the flag Fplus is determined in step 7. When the value is negative, that is, when the discriminant value LU is inverted from negative to positive, first, in step 8, the flag Fplus is
is set to 1, so that the flag F plus is determined to be 1 in step 7 next time as well when the determination value LU is greater than or equal to zero.

Then, in the next step 9, it is determined whether the count value cnt is 3 or not. When the discriminant value LU is reversed from negative to positive for the first time and the count value cnt is 3, this is the case when the discriminant value LU is reversed from the negative slice level SL or lower to the positive discriminant value LU. Sometimes, in step 10, the currant value cnt is set to 2.

In step 11, it is determined whether or not the count value cnt is zero, and if it is not zero, the process proceeds to step 12, where the engine load (basic fuel injection amount Tp) and engine speed are determined by the program shown in the flowchart of FIG. A positive slice level SL and a discrimination value LU that are variably set according to the speed N.
Compare with.

When the determination value LU is larger than the positive slice level SL, it is determined that a misfire has occurred, and steps 13 to 2 are performed.
In step 5, the cylinder in which the misfire has occurred is identified, and the number of misfire detections for each cylinder is counted up.

First, in Step E3, in order to identify the cylinder in which the misfire is occurring, the cylinder counter Cyl indicating the ignition cylinder is used.
Determine the value of cnt.

The cylinder counter Cylcnt is set according to the program shown in the flowchart of FIG. The program shown in the flowchart of FIG. 5 is executed every time the reference angle signal REF is output from the crank angle sensor 10 (BTD
C70'), the cylinder number whose ignition timing ADV (ignition advance value) will be controlled next is determined based on the current reference angle signal REF, and the cylinder number is sent to the cylinder counter Cylcnt. (See Figure 9).

That is, first, in step 91, the current reference angle signal R
It is determined whether or not EF is the ignition reference for the #1 cylinder. If the current reference angle signal REF is the ignition reference for the #1 cylinder, the process advances to step 792 and the cylinder counter Cylcnt is set. Set 1 to indicate the #l cylinder.

When it is determined in step 91 that it is not the ignition standard for the #1 cylinder, the process proceeds to step 93, where it is determined whether or not it is the ignition standard for the #2 cylinder. If the current reference angle signal REF is the ignition reference for #2 cylinder, step 9
At step 4, the cylinder counter Cylcnt is set to 2, which indicates the #2 cylinder.

Also, in step 93, the current reference angle signal REF is set to #2.
If it is determined that it is not the ignition standard for the cylinder, it is determined in step 95 whether or not it is the ignition standard for the #3 cylinder, and if it is the ignition standard for the #3 cylinder, 3 indicating the #3 cylinder is set in Cylcnt in step 96. If the ignition standard is not the ignition standard for the #3 cylinder, the ignition standard remains for the #4 cylinder, so step 97
Set Cylcnt to 4, which indicates cylinder #4.

Returning to the flowchart of FIG. 3 again, the Cylcnt set as described above is determined in step 13, and if it is, for example, Cylcnt-2, the process proceeds to step 17, where the count value cnt is 2. Determine whether or not. If the count value cut is 2, it is determined that the #1 cylinder is misfiring, and step 1
At step 8, the LSTI that counts the number of misfire detections in the #1 cylinder is increased by one.

An actual example of how the #1 cylinder misfire is determined in this way is as follows:
This is shown in FIG. That is, as shown in FIG. 9, #1
When a cylinder misfires, the discriminant value LU corresponding to cylinder #1, which is the misfiring cylinder, first shows a large decrease and becomes negative, and the discriminant value LU corresponding to cylinder #3, which is the next firing order, also becomes negative at the same level. value, and the subsequent misfire determination values LU for the #4 cylinder and #2 cylinder become positive as they receive the reaction from the #1 and #3 cylinders (misfire in the #2 cylinder after the misfire in the #1 cylinder). (Discriminant value LU is not shown in Fig. 9)
.

Here, when the discrimination value LU corresponding to the #1 cylinder is calculated to be less than or equal to the negative slice level SL, the count value cn
After 4 is added to t, it is immediately decreased by 1 to 3, but since the discrimination value corresponding to the next #3 cylinder and U is also less than the negative slice level SL, the currant direct cnt is again set to 3.
is set.

Next, the slice level S for which the discrimination value LU corresponding to #4 cylinder is positive
When the value exceeds L and is reversed from negative to positive, the count value cnt is set to 2, so the process proceeds to step 13 since the count value cnt is not zero and the discrimination value LU exceeds the positive slice level SL. . In step 13, it is determined that Cylcnt=2, and as a result, it is finally determined that the #1 cylinder has misfired.

Here, for example, the first discriminant value LU corresponding to the #4 cylinder that is reversed to positive may not exceed the positive slice level SL, and the next discriminant value LU corresponding to the #2 cylinder may exceed the positive slice level SL. , in this case, the process proceeds from step I2 to step 27 at the first inversion from negative to positive, and the count value c
It is determined that nt is not zero (count value cnt=2
), the count value cnt is decreased by 1 to 1 in step 28, and when the next discrimination value LU exceeds the positive slice level SL, the process advances from step 7 to step 11, and the count value cnt is 1. The process then proceeds to step 12, where a misfire is detected. At this time, Cylc
Since nt is set to 1, the process proceeds to step 14, but since the count value cnt is 1, the process proceeds from step 14 to step 16, where it is finally determined that the #1 cylinder has misfired, and the #1 cylinder is The number of misfire detections LSTI is incremented by one.

Furthermore, in the example shown in FIG. 9, only one of the discrimination value LU corresponding to the #1 cylinder and the discrimination value LU corresponding to the #3 cylinder may be less than or equal to the negative slice level SL; Also, the first time that the discriminant value LU is reversed to positive, the count value cnt is set to 2, and #1
A cylinder misfire can be determined.

After identifying the misfiring cylinder in steps 13 to 25 and incrementing the number of misfire detections LSTI to LST4 for each cylinder, the process proceeds to step 26, where the count value cnt is reset to zero. As a result, if the discriminant value LU is positive next time and the process proceeds from step 7 to step 11, the count value cnt=o will be determined in step 11.
Disable misfire detection.

Then, in the next step 27, it is determined whether or not the count value cnt is zero, and if it is not zero, the count value cnt is decreased by 1 in step 28.

In the next step 29, a predetermined multiple (for example, 1.5 times) of the slice level SL is compared with the discrimination value LU calculated in step 2.

Here, the reason why the slice level SL is multiplied by a predetermined value as described above is that when two consecutive cylinders such as #1 cylinder and #3 cylinder in the firing order #1 → #3 → #4 → #2 both misfire. Since there is a characteristic that the discriminant value LU drops significantly compared to the normal case where only one specific cylinder misfires, it is determined whether the discriminant value LU shows a large drop that is recognized as two consecutive cylinder misfires. It's for a reason.

Therefore, for a value obtained by multiplying the slice level SL by a predetermined value,
A coefficient for multiplying the slice level SL by a predetermined value is determined in advance through experiments so that the discrimination value LU does not exceed when one cylinder misfires alone, but exceeds only when two consecutive cylinder misfires occur.

If it is determined in step 29 that the discrimination value LU is smaller than a predetermined multiple of the slice level SL, and it is predicted that two consecutive cylinders will misfire, the process will proceed to step 30 to identify the cylinder in which two consecutive cylinders have misfired. For cylinder counter Cy
Determine cnt.

For example, when two cylinders, #3 and #4, are misfiring consecutively, the discrimination value LU corresponding to cylinder #4 is calculated to be significantly negative, and it is determined that two consecutive cylinders are misfiring. At this time, the cylinder counter cycnt is set to 2, so if Cycnt is 2 when it is determined that two consecutive cylinder misfires have occurred, the two cylinders #3 and #4 are This means that there are continuous misfires. Therefore, at this time, the process proceeds to step 32 where a counter c is incremented to count up the number of times it has been detected that two cylinders #3 and #4 are consecutive.
Increase nt34 by 1.

Similarly, when the cylinder counter Cycnt is 1, the counter cnt42 is set to the cylinder counter Cycnt42.
When nt is 3, counter cnt12 is set, and when cylinder counter Cycnt is 4, counter cn is set.
Count t13 respectively.

In addition, with the above discriminant value LU, if two cylinders with a stroke phase difference equivalent to one engine revolution both misfire as shown in Fig. 11, the periodic deviation for each engine revolution becomes approximately zero. It is impossible to diagnose a misfire.

In the next step 35, the latest value of the period is set to T0φ in the same manner as the period T O-T 4, and the period data of T0φ, to which the latest measurement result was set during the previous execution of this program, is set to 7.1. Similarly, the cycle data T, 1 from the previous execution of this program is set to 7.2.

In step 36, the absolute value of the deviation between T0φ and T,1 of the periodic data set in step 35 is calculated, and the result is D! Set to t.

In step 37, the weighted average value Dftav of the period deviation Dft calculated in step 36 is determined according to the following formula.

In the next step 38, the deviation of the TDC period measured based on the output of the crank angle sensor 10 is corrected by the weighted average value DJ2tav, and the second differential value of the period T0 is calculated according to the following formula, and the result is is set as the misfire discrimination value Zu.

As described above, the deviation between the period Tel and T, 2 and the period T.

If the deviation between φ and Tel is corrected by the above DILaν,
Even if there is a deviation between Tol and To2 and between T0φ and To1 due to the angular error of the crank angle sensor 10, the period difference due to the difference in detection angle can be corrected in the direction of decreasing. , rotational fluctuations can be detected with high accuracy.

In step 39, the Zu calculated in step 38 this time is
Find the smallest value among the multiple data calculated below zero among the recently calculated misfire discrimination values Zu including is weighted averaged, and the result is set as a new minimum value m1nZu to update the minimum value m1nZu.

In the next step 40, the minimum value m1nZu obtained as described above is compared with the Zu calculated in step 4 this time, and if the latest calculation result is smaller than the average minimum value m1nZU, step The process advances to step 41 and sets the currently calculated Zu to the minimum value m1nZu. Therefore, if Zu calculated in step 38 falls below the average level, the minimum value m1nZu will immediately follow this change.

After setting the minimum value m1nZu of the misfire discrimination value Zu as described above, in the next step 42, the currently calculated Zu
The absolute value of is corrected to increase according to the increase in the engine rotation speed N, while it is corrected to decrease according to the increase in the engine load, and the result of this correction is set in DZu. The engine rotational speed N can be calculated based on a cycle, and the engine load can be calculated using a basic fuel supply amount TP when electronically controlling the fuel supply amount to the engine to an amount commensurate with the intake air amount. Good.

The above-mentioned Zu tends to be calculated to a smaller value as the engine rotational speed N is higher, and to a larger value as the engine load is higher. It is possible to set a value that indicates the true degree of engine rotation variation that is unrelated to the conditions of .

In the next step 43, a coefficient ga is used to set the slice level SL1 for determining the negative level of the misfire discrimination value Zu as a predetermined multiple of the minimum value m1nZu.
in is searched from the map based on the DZu calculated in step 42 above.

Here, the larger the DZu, the larger the coefficient gain
is set to a small value, and the DZu
When the engine rotational fluctuation is greater than a certain level, the coefficient gain is set to a value of 1 or less, and the slice level SL on the side different from the minimum value m1nZu to zero is set.
conversely, when the DZu is above a certain level and the engine rotational fluctuation is small, the coefficient gain is set to a value of 1 or more, so that the slice level SLI is lower than the minimum value m1nZu. It is set to an even smaller value.

As a result, when the engine speed fluctuation is large, the misfire determination value Zu is determined to be smaller than the slice level SLI, and the misfire diagnosis is made easier. Conversely, when the engine speed fluctuation is small, the slice level SLI is more likely to be diagnosed. This is to avoid the misfire discrimination value Zu from becoming too small, thereby making it difficult to diagnose a misfire.

In step 44, the minimum value m1nZu is multiplied by the coefficient gain obtained by searching from the map in step 43, and the result is set as the slice level SLI.

In the next step 45, it is determined whether or not the misfire discrimination value Zu calculated in the step 38 is greater than or equal to zero. If it is not greater than or equal to zero, that is, if the misfire discrimination value Zu is less than zero, the process proceeds to step 46. The misfire determination value Zu is compared with the slice level SLI set in step 44 above.

Here, if it is determined that the misfire discrimination value Zu is a value smaller than the slice level SLI, it is assumed that the engine rotational fluctuation has exceeded a predetermined level due to the occurrence of a misfire, and a misfire is detected. Proceeding to step 47, LSTD, which counts the number of misfire detections, is incremented by one.

Incidentally, when one cylinder misfires independently, the misfire determination value Zu can be used to diagnose the occurrence of a misfire by increasing the period in which that cylinder is affected. In addition, if two cylinders with a stroke phase difference equivalent to one engine rotation both misfire, only when the period corresponding to one of the misfiring cylinders is set to T0φ, or when misfires occur occasionally. Zu changes significantly to the negative side and a misfire diagnosis can still be performed, but if the cycles of cylinders other than the misfire are set to T0φ or if the misfire continues as shown in Figure 11, a misfire can be detected. Cannot be diagnosed with certainty.

On the other hand, when two consecutive cylinders in the firing order are both misfired, it is possible to detect the occurrence of such two-cylinder consecutive misfires. Therefore, the misfire detection number LSTD includes the number of misfire detections in one cylinder alone and the number of misfire detections in two consecutive cylinders in the firing order. When both misfires occur, the number of misfire detections cannot be accurately reflected as in the discrimination based on the discriminant value LU.

In the next step 48, it is determined whether total, which is a counter that counts up the number of times this program has been executed, has counted up to a predetermined value (for example, 1000). If it has not reached the predetermined value, step 49 is performed. The total is counted up.

On the other hand, if it is determined in step 48 that the total has been counted up to a predetermined value, the process proceeds to step 50 and subsequent steps to diagnose the occurrence of a misfire by determining the number of misfire detections within the predetermined number of times.

In step 50, in a rich state of the air-fuel ratio detected based on the oxygen concentration in the exhaust gas detected by the oxygen sensor 17,
The number of misfire detections 1 s t Oz, which is detected when the output changes in the lean direction, is corrected to include the lean state in which misfires cannot be detected, and the final number of misfire detections is 0□ Let it be LST.

That is, as will be described later, in the internal combustion engine of this embodiment, the air-fuel ratio of the engine intake air-fuel mixture is detected by the oxygen sensor 17, and feedback correction is performed so that this air-fuel ratio approaches the target air-fuel ratio (stoichiometric air-fuel ratio). , the average air-fuel ratio is controlled to the target air-fuel ratio by repeating rich and lean around the target air-fuel ratio. Here, if there is a misfiring cylinder, unburned gas is discharged from that cylinder, resulting in a temporary lean state, and in a rich state, as shown in FIG. This appears as a sudden change in lean direction. Therefore, when the air-fuel ratio is in a rich state, the occurrence of a misfire can be predicted by detecting a sudden change in the output of the oxygen sensor 17 in the lean direction, but the counter total may not count up to a predetermined value. Since rich and lean cycles are repeated until the engine reaches a lean state, misfires in a lean state cannot be detected. This is to ensure that the final number including the number of times is set.

A program related to setting the number of misfire detections based on the output of the oxygen sensor 17 will be described below.

The flowchart in FIG. 6 is a setting program for the air-fuel ratio feedback correction coefficient α used for basic fuel injection amount TPO correction, and is executed at predetermined minute intervals (for example, 10+ms).

First, in step 101, a timer time for measuring lean time and rich time is incremented by 1,
In the next step 102, the output of the oxygen sensor 17 is
Set to . The oxygen sensor 17 detects whether the actual air-fuel ratio is rich or lean with respect to the stoichiometric air-fuel ratio by utilizing the fact that the oxygen concentration in the exhaust gas changes suddenly after the stoichiometric air-fuel ratio. In this case, the output is about 1 V, and at a lean air-fuel ratio where the oxygen concentration is high, the output is around 0 ■, which is a known method.

In step 103, the oxygen sensor 17 output VOz set in step 102 is compared with a slice level equivalent to the stoichiometric air-fuel ratio (for example, 50Qmv), and it is determined whether the actual air-fuel ratio is rich or lean with respect to the stoichiometric air-fuel ratio. Determine if there is.

If it is determined that the output exceeds the slice level and the air-fuel ratio is rich, step 104 occurs.

The chi/lean determination flag FR is determined.

As will be described later, the flag FR is set to zero when the air-fuel ratio is at 1, but the first time the air-fuel ratio is reversed from lean to rich, the flag FR is determined to be zero in step 104. That will happen.

When it is determined in step 104 that the flag FR=φ in the first determination from lean to rich, the process proceeds to step 105 and the flag FR is set to 1.

In the next step 106, the value of the timer time counted up in the previous air-fuel ratio restoration state is set to TmontL indicating the lean time, and
07, the timer t is set to measure the rich time.
Reset ime to zero.

Then, in step 108, the newly detected lean time TmontL and the weighted average value AvcontL of the lean time TtaontL calculated up to the previous time are weighted averaged, and the result is newly set in AvmontL. .

In step 109, the air-fuel ratio feedback correction coefficient α
By subtracting a predetermined proportional control amount PL from the previous value of , and multiplying the basic fuel injection amount Tp by the air-fuel ratio feedback correction coefficient α to reduce the final fuel injection amount Ti, Control is performed in a direction that eliminates the rich state.

- Furthermore, when the rich state of the air-fuel ratio continues next time, the flag FR is determined to be 1 in step 104, and the process proceeds from step 104 to step 110.

In step 110, the fuel injection amount Ti is calculated from the previous air-fuel ratio feedback correction coefficient α to a predetermined integral control amount I.
The correction coefficient α is gradually decreased by subtracting the product multiplied by .

Similar control is performed when it is determined in step 103 that the air-fuel ratio is lean, and when the air-fuel ratio is lean, the flag FR is set to zero, and the rich time A
vmont R settings are performed (steps 111-1
17).

On the other hand, after integrally controlling the air-fuel ratio feedback correction coefficient α, the process proceeds to step 118.

In step 118, the maximum value ma of the oxygen sensor 17 output
The previous values and predetermined values of x and the minimum value min are each weighted averaged so that may and min gradually approach the median value, but in the next steps 119 to 12
2, the latest oxygen sensor 17 output VO□ is compared with max and min, and when the latest value exceeds may and min of the weighted average result, the latest value is
and set to min.

The program shown in the flowchart of FIG.
This is executed every 30' crank angle based on TDC60°, and the output of the oxygen sensor 17 at every 30' crank angle is 0□, and whether the air-fuel ratio at that time was rich or lean. It is designed to memorize.

First, in step 131, the output of the oxygen sensor 17 is set to v
O2, and in the next step 132, the counter i
The data of V Oz is stored in the address determined by .

In the next step 133, it is determined whether the current air-fuel ratio is rich or lean based on the flag FR,
When the air-fuel ratio is in a rich state where the flag FR is 1, the process proceeds to step 134t\ and stores 1 as the flag FR data at the address determined by the same counter i as described above, and when the flag FR is in a lean state where the flag is φ, the step 13
5, memorize φ.

Therefore, the output ■0 of the oxygen sensor 17 is determined by the counter i.
□ is memorized in chronological order, and each output ■
It is possible to determine whether 0□ is in a rich air-fuel ratio state or in a lean state.

The program shown in the flowchart of FIG.
This is a program for diagnosing the occurrence of a misfire based on the data sampled in the flowchart shown in the figure.
It is executed every time the is output.

First, in step 151, the current reference angle signal REF
It is determined whether or not corresponds to 70 degrees before the compression top dead center of the #l cylinder.

Here, if it is determined that the reference angle signal REF is not for the #1 cylinder, that is, if it is the reference angle signal REF for the other #2 to #4 cylinders, the process proceeds to step 152;
As will be described later, 5LO2, which is a slice level for detecting that the output of the oxygen sensor 17 has changed in the lean direction when the air-fuel ratio is in a rich state, is calculated according to the following formula.

In the formula for calculating the slice level SLO, above, m i
n and max are the minimum and maximum values of the output of the oxygen sensor 7 determined in the flowchart of FIG. ，
is set, and when a misfire occurs in a rich air-fuel ratio state and unburned gas from that cylinder reaches the oxygen sensor 7,
- Set 5LO2 so that the output temporarily falls below the slice level SLO□ (see Figure 10)
.

On the other hand, in step 151, the current reference angle signal REF is #
If it is for one cylinder, the process advances to step 153.

In step 153, the counter i used for sampling the output of the oxygen sensor 17 at every 30° crank angle in the flowchart of FIG. F. Further, a counter Z for counting the number of misfire occurrence determinations based on the counter F is reset to zero.

In the next step 154, FR[i] is first determined, and whether the oxygen sensor 17 output ■0□ (i) sampled at counter i is in a rich air-fuel ratio state or It is determined whether the fuel ratio is in a lean state.

Since the flag FR is set to 1 in the air-fuel ratio rich state, when it is determined in step 154 that it is 1, the oxygen sensor 17 output ■0° [i) is data sampled in the air-fuel ratio rich state. , at this time, the process advances to step 155 and the oxygen sensor 17 output VOz(
i) and the slice level 5LO2 calculated in step 152 above.

Here, when ■0□ [i)≧SLO, the air-fuel ratio is in a stable rich state, and in this case, the process proceeds to step 156 and the counter F is reset to zero.

On the other hand, in step 155 ■0□ (i) < S L
When it is determined that the air-fuel ratio is 02, it is assumed that the air-fuel ratio is being reversed by air-fuel ratio feedback control, but it is also possible that the output of the oxygen sensor 17 has decreased due to the occurrence of a misfire, so proceed to step 157 l. Determine whether or not counter F is zero. When it is zero, the counter F is incremented by 1 in step 160, and if the state of ■0□(i)<5LO2 continues, the counter F is incremented each time.

Then, when it is determined in step 157 that the counter F is not zero, and in step 158 that the counter F is 3, that is, ■0□(i ) <
When it is determined that the air-fuel ratio is S L○2, it is assumed that the continuous trend toward lean air-fuel ratio (see FIG. 10) in the rich state of the base air-fuel ratio is due to the occurrence of a misfire, and in step 159, the occurrence of a misfire is determined. Increase the number of determinations Z by 1.

In order to repeat such determination until it is determined in step 161 that the counter i is 23 or more, step 16 is performed.
If it is determined that the counter i is less than 23 at step 162, the counter i is incremented by 1 and the step processing from step 154 is performed again.

23 of the counter i is the reference angle signal R for the #1 cylinder.
Oxygen sensor 1 sampled while EF is output
This corresponds to 24, which is the output number of 7, and by counting up the counter i from φ to 23, all of the 24 oxygen sensor 17 outputs are at the slice level SLO□
Make a comparison with

All sampled outputs of the oxygen sensors 17 are determined while the reference angle signal REF for the #1 cylinder is being output, and when it is determined in step 161 that the counter i is 23 or more, the process proceeds to step 163. The number of misfire occurrence determinations Z that has been counted up up to that point is added to 1sto2.

In the next step 164, the counter i is reset to zero, and the counter i is again counted up every 30 degrees of the crank angle, so that the output of the oxygen sensor 17 is sampled each time.

Misfire occurrence determination count 42 set in step 163 above
s t O2 is used in step 50 of the flowchart of FIG. 3 above.

Here, to continue the explanation by returning to the flowchart in FIG. 3, the misfire occurrence determination count 42 s t O2 detected based on the output of the oxygen sensor 17 as described above is detected only in the rich state of the air-fuel ratio. As a result, when the air-fuel ratio is in a lean state, since the air-fuel ratio is originally in a lean state, even if unburned gas is discharged due to a misfire, it does not appear as a change in the output of the oxygen sensor 17, so it cannot be detected. be.

However, if we try to determine whether or not a true misfire has occurred based on the ratio of the number of misfire determinations per total number of combustions, it is difficult to determine the number of misfire determinations based on the output of the oxygen sensor 17 per total number of combustions. Therefore, in step 50, it is assumed that misfires occur in the lean state at approximately the same rate as the misfire occurrence determination number 1 stO The misfire occurrence per number of occurrences is estimated and set based on the data obtained from the output of the oxygen sensor 17 in the rich state according to the following formula.

In this way, when the number of misfire occurrence determinations 0□LST from the output of the oxygen sensor 17 until the counter total is counted up to a predetermined value is set, in the next step 51, the number of misfire occurrence determinations 0□LST is set. By comparing LST with a predetermined value, it is determined whether misfire occurrence is determined at a rate higher than a predetermined rate based on the output of the oxygen sensor 17.

Then, when the misfire occurrence determination number 0zLST is equal to or greater than a predetermined value, in step 52, the consecutive two cylinder misfires counted up based on the determination result in step 29 (for example, the firing order #1→#3→#4→ It is determined whether there is any data exceeding a predetermined value in any one of the detection times of cylinder #1 and cylinder #3 at the time of #2.

Here, the number of consecutive two cylinder misfire detections cnt42cnt3
4. cnt12. If any one of cnt13 exceeds a predetermined value (for example, 12), the process proceeds to step 53, and a display panel near the driver's seat displays, for example, that a two-cylinder misfire has occurred.

On the other hand, when the step 51 determines that the misfire occurrence determination number 02LST is less than the predetermined value, the oxygen sensor 1
Since the occurrence of a misfire at a problematic level has not been confirmed from the output No. 7, the process proceeds to step 67 and subsequent steps without making any determination of the occurrence of a misfire and displaying based on the determination.

Further, if it is determined in step 52 that the number of consecutive two-cylinder misfire detections is less than or equal to a predetermined value, and no misfire has occurred in two consecutive cylinders, the process proceeds to step 54 and subsequent steps.

In steps 54 to 61, each of LSTI to LST4, in which the number of misfire determinations for each cylinder has been counted up, is compared with a predetermined value (for example, 25), and for cylinders that have been misfired the number of times exceeding the predetermined value and are 0. displays the misfire occurrence along with the cylinder number.

After each of LSTI to LST4 is compared with a predetermined value, the process proceeds to step 62, where it is determined whether the misfire determination number LSTD counted up by the steering knob 47 is greater than or equal to a predetermined value. The misfire determination number LSTD is
It includes the number of misfires when two consecutive cylinders misfire in the firing order, and the number of misfires when one single cylinder misfires.
This includes uncertainly the number of misfires when two cylinders misfire due to the stroke phase difference corresponding to the rotation. Therefore, when the value is equal to or greater than the predetermined value and the individual misfire of one cylinder is not determined and displayed in steps 54 to 61 (step 63), 2
Since it can be considered that the occurrence of a misfire in a cylinder is being counted, the occurrence of a misfire in two cylinders is displayed in step 64.

Furthermore, as a result of comparing each misfire determination number with a predetermined value as described above, if it is not possible to identify the occurrence of a misfire from any of the determination times and no misfire display is performed (step 65), in step 51 the exhaust gas is It has been confirmed that changes in the oxygen concentration in the engine indicate a misfire, so if there is a stroke phase difference equivalent to one revolution of the engine that cannot be accurately determined from any number of misfire determinations (for example, ignition order # 1 → #3
This can be considered to be a case where cylinders #1 and #4 (when → #4 → #2) have misfired, and the misfire occurrence state of the two cylinders is displayed in step 66.

That is, if two cylinders with a stroke phase difference equivalent to one engine revolution both misfire, it is not possible to determine the misfire based on the discriminant value LU, and depending on the discriminant value Zu, it is possible that the misfire is a one-shot event. Although it can be detected depending on the periodic sampling timing or when the misfire occurs, it cannot be reliably detected when the misfire occurs continuously. However, since it is possible to detect that a misfire has occurred in some form from the output of the oxygen sensor 17, the oxygen sensor If the 17 output indicates that a misfire has occurred, the engine 1
It can be determined that both cylinders have misfired due to the stroke phase difference corresponding to the rotation.

Therefore, the misfire determination based on the discriminant value Zu is omitted, and the oxygen sensor When the output No. 17 indicates the occurrence of a misfire, it may be determined that two cylinders, which have a stroke phase difference equivalent to one revolution of the engine, are both misfiring.

When the misfire display is controlled as described above, step 67
- At step 72, each misfire discrimination number is reset to zero, and the counter total is reset to zero, so that each misfire discrimination number is incremented until the counter total is counted up again to a predetermined value.

<Effects of the Invention> As explained above, according to the present invention, when diagnosing a misfire based on the deviation of the rotation period for each rotation of the engine or the deviation of this deviation, misfire occurrence is determined based on the concentration of exhaust components such as oxygen concentration. When a misfire occurs in two cylinders with a stroke phase difference equivalent to one revolution of the engine that cannot be detected from the period, the occurrence of a misfire can be determined based on the exhaust component concentration. It is possible to detect misfires in two consecutive cylinders, and even misfires in two cylinders with a stroke phase difference equivalent to one rotation of the engine, thereby improving the accuracy of misfire diagnosis.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the configuration of the present invention, Fig. 2 is a system schematic diagram showing an embodiment of the invention, and Figs. 3 to 8 are flowcharts showing control details of misfire diagnosis in the above embodiment. , Fig. 9 is a time chart showing the characteristics of the discriminant value using the deviation of the deviation for each rotation period of the engine in the above embodiment, and Fig. 10 is a time chart showing the characteristics of the discriminant value using the deviation of the deviation for each engine rotation period in the above embodiment. FIG. 11 is a time chart for explaining the problems of the conventional device. 1... ill control unit 9..., control unit 10
... Crank angle sensor 17 ... Oxygen sensor patent applicant Japan Electronics Co., Ltd. Agent Patent attorney Fujio Sasashima Figure 2 Figure 3 Figure 5

Claims (1)

[Scope of Claims] A rotation period measuring means for measuring the rotation period of an internal combustion engine, and a misfire diagnosis based on a deviation of the rotation period detected by the rotation period measuring means for each rotation of the engine or a deviation of the deviation. misfire diagnosis means for detecting the concentration of exhaust components, which is installed in a collection part of an exhaust system of an internal combustion engine to detect the concentration of exhaust components; two-cylinder misfire diagnostic means for diagnosing that both of the two cylinders, which have a stroke phase difference equivalent to one engine rotation, are misfiring when the exhaust component concentration detected by shows a change corresponding to the occurrence of a misfire; A misfire diagnostic device for an internal combustion engine, comprising: