JP2832997B2 - Combustion abnormality detection device for internal combustion engine - Google Patents

Description

The present invention relates to a combustion abnormality detection device for an internal combustion engine.

[Conventional technology]

A device that detects an abnormal combustion when no more fuel is supplied due to clogging of the fuel injection valve or a large amount of fuel is supplied due to a failure of the fuel injection valve and explosion combustion is no longer performed in a specific cylinder. The rotational angular velocity of the crankshaft is detected in synchronism with the combustion stroke of each cylinder, the fluctuation of this rotational angular velocity is calculated for each cylinder, and the frequency of occurrence of a fluctuation equal to or higher than a predetermined fluctuation level is determined in advance. 2. Description of the Related Art A combustion abnormality detecting device for an internal combustion engine which determines that a combustion abnormality has occurred when the frequency exceeds a reference frequency is known (see Japanese Patent Application Laid-Open No. 61-258955).

[Problems to be solved by the invention]

However, according to a correction coefficient obtained based on an output signal of an oxygen concentration sensor provided in an exhaust passage,
In an internal combustion engine provided with air-fuel ratio control means for controlling the air-fuel ratio to, for example, the stoichiometric air-fuel ratio, for example, a fuel injection valve of one cylinder fails and a considerably large amount of fuel is supplied to the cylinder in excess of the required fuel injection amount. Therefore, when the misfire occurs in the cylinder and the fuel injection amount from the failed fuel injection valve can be controlled to some extent by the air-fuel ratio control means, the following problem occurs. That is, when the air-fuel mixture of one cylinder becomes over-rich, the exhaust gas flowing through the exhaust passage becomes rich as a whole, so that the oxygen concentration sensor continuously outputs a rich signal, so that the correction coefficient is set so that the air-fuel ratio is lean. Can be changed. For this reason, the degree of over-rich of the air-fuel mixture in the cylinder that has been over-rich is reduced, and the frequency of occurrence of misfire in this cylinder decreases. There is a problem that it cannot be detected as abnormal.

[Means for solving the problem]

To solve the above problems, according to the present invention, as shown in the block diagram of the invention in FIG. 1, an air-fuel ratio correction coefficient obtained based on an output signal of an oxygen concentration sensor provided in an engine exhaust passage is used. Air-fuel ratio control means for controlling the air-fuel ratio to be a predetermined air-fuel ratio by correcting
200 and angular velocity detecting means 2 for detecting the angular velocity of the engine output shaft
01, an angular velocity fluctuation calculating means 202 for calculating the fluctuation of the angular velocity between the cylinders in the combustion stroke based on the detection result of the angular velocity detecting means 201, and a state where the fluctuation of the angular velocity calculated by the angular velocity fluctuation calculating means 202 is large is a reference. Abnormality determining means 203 for determining that a combustion abnormality has occurred when the air-fuel ratio is equal to or higher than the level, and changing the reference level according to the air-fuel ratio correction coefficient so that the value of the air-fuel ratio correction coefficient becomes lower than the predetermined air-fuel ratio correction coefficient. A reference level changing means for changing the reference level to a value smaller than the reference level of the predetermined air-fuel ratio correction coefficient when the fuel ratio is a value to be corrected to the lean side; The predetermined value of the air-fuel ratio correction coefficient is, for example, a value near the limit value of the change of the air-fuel ratio correction coefficient, which can be regarded as a normal change with time.

(Operation)

It is determined that a combustion abnormality has occurred when the state in which the angular velocity fluctuation calculated by the angular velocity fluctuation calculating means is large is equal to or higher than the reference level. The reference level is changed to a small value when the air-fuel ratio correction coefficient changes to make the air-fuel ratio lean, so that the reference level can be accurately calculated even when the frequency of misfires decreases due to the change in the air-fuel ratio correction coefficient. An abnormal combustion is detected.

〔Example〕

FIG. 2 shows a four-cylinder internal combustion engine. Referring to FIG. 2, 1 is an engine body, 2 is each cylinder, 3 is an intake port, 4 is an exhaust port, 5 is an intake manifold, 6 is an exhaust manifold, and 7 is a spark plug provided in each cylinder 2. Show. Each branch pipe 5a of the intake manifold 5 is connected to each corresponding intake port 3. A fuel injection valve 8 for injecting fuel into each corresponding intake port 3 is attached to each branch pipe 5a. The intake manifold 5 is connected to an air cleaner 10 via an intake pipe 9. An air flow meter is provided in the intake pipe 9 sequentially from the upstream side.
11 and a throttle valve 12 are provided. Each branch pipe 6 a of the exhaust manifold 6 is connected to a corresponding exhaust port 4, and a collection portion 6 b of the exhaust manifold 6 is connected to an exhaust pipe 13. An O ₂ sensor 14 is attached to the exhaust pipe 13.

The electronic control unit 30 is composed of a digital computer, and a ROM (read only memory) 32, a RAM (random access memory) 33, a CPU (microprocessor) 34, an input port 35 interconnected by a bidirectional bus 31.
And an output port 36. Note that a backup RAM 33a is connected to the CPU 34 via a bus 31a. The air flow meter 11 generates an output voltage proportional to the amount of intake air,
This output voltage is input to the input port 35 via the AD converter 37. The throttle valve 12 is provided with an idle switch 15 that is turned on when the throttle valve 12 is at an idle opening, and an output signal of the idle switch 15 is supplied to an input port.
Entered in 35. A water temperature sensor 16 that generates an output voltage proportional to the engine cooling water temperature is attached to the engine body 1, and the output voltage of the water temperature sensor 16 is input to an input port 35 via an AD converter 38. The output voltage of the O ₂ sensor 14 AD converter 39
Through the input port 35. Crank angle sensor
17 generates an output pulse every time the crankshaft rotates a predetermined crank angle, and the output pulse of the crank angle sensor 17 is input to the input port 35. From this output pulse, CP
At U34, the engine speed is calculated. The vehicle speed sensor 18 generates an output pulse corresponding to the vehicle speed, and the output pulse is input to the input port 35. On the other hand, the output port 36 is connected to each fuel injection valve 8 via the corresponding drive circuits 40 to 43.

FIG. 3 shows a routine for calculating the fuel injection time.
This routine is executed by interruption every fixed crank angle. First, in step 50, it corresponds to the engine load.
The basic fuel injection time TP is calculated from Q / NE. Here, Q is the intake air amount, and NE is the engine speed. Step 51
Then, the fuel injection time TAU is calculated based on the following equation.

TAU = TP ・ FAF ・ KG ・ F where FAF: feedback correction coefficient KG: learning correction coefficient F: other correction coefficients KG is a coefficient for correcting the deviation of the air-fuel ratio caused by aging of the fuel injection valve, etc. Yes, average FAF ▲
▼ is changed to be a value near 1.0.

FIG. 4 shows a calculation routine of FAF and KG.
This routine is executed by interruption every predetermined time. Referring to FIG. 4, in step 60, it is determined whether a condition for feedback control of the air-fuel ratio is satisfied. Step 61 when the feedback condition is not satisfied
The routine is terminated after setting the FAF to 1.0. Therefore, in this case, the air-fuel ratio feedback control is not executed. Step 62 when the feedback condition is satisfied
It is determined whether the air-fuel ratio has been inverted from rich to lean or from lean to rich. When the air-fuel ratio is not inverted, the routine proceeds to step 63, where it is determined whether the air-fuel ratio is rich. When the air-fuel ratio is rich, proceed to step 64,
FAF is subtracted by the integral Ki. On the other hand, when it is determined in step 63 that the air-fuel ratio is lean, the process proceeds to step 65, where the FAF is added by the integral amount Ki.

When it is determined in step 62 that the air-fuel ratio has been inverted, the routine proceeds to step 66, where the average value of FAF is calculated by the following equation.

That is, the average value of FAF
It is obtained as an arithmetic mean of the value FAF in the current processing cycle of the FAF immediately before the AF skips and the value FAF0 in the previous processing cycle. In step 67, the FAF in the current processing cycle is stored in FAF0, and is used as FAF0 in the calculation of ▼ in the next processing cycle.
In step 68, it is determined whether or not ▲ ▼ ≧ 1.02. ▲
If ▼ ≧ 1.02, proceed to step 69. ▲ ▼ ≧
In the case of 1.02, the air-fuel ratio of the air-fuel mixture formed by the fuel injected by the injection times TP, KG, F is leaner than the stoichiometric air-fuel ratio due to factors such as aging. Therefore, by increasing the learning correction value KG by ΔK in step 69, ▲ ▼ approaches 1.0. On the other hand, in step 68, ▲ ▼ <1.02
When it is determined that is, the routine proceeds to step 70, where it is determined whether or not ▼ ≦ 0.98. If ▲ ▼ ≦ 0.98, the routine proceeds to step 71, where the learning correction value KG is decreased by ΔK. ▲
When ▼ ≦ 0.98, the air-fuel ratio of the air-fuel mixture formed by the fuel injected during the injection time TP · KG · F is richer than the stoichiometric air-fuel ratio. Therefore, by decreasing KG by ΔK, ▲ ▼ approaches 1.0. When it is determined in step 70 that ▲> 0.98, KG is maintained at the current value. Thus KG
Is increased or decreased so that the average value of FAF is close to 1.0. This learning correction coefficient KG is stored in the backup RAM3
Stored in 3a. In step 72, it is determined whether the air-fuel ratio is rich. When the air-fuel ratio is rich, the routine proceeds to step 73, where the FAF is subtracted by the skip amount Rs. This skip amount Rs is set to be sufficiently larger than the integral amount Ki, and Rs≫Ki. On the other hand, when it is determined in step 72 that the air-fuel ratio is lean, the process proceeds to step 74, where the FAF is added by the skip amount Rs.

FIG. 5 shows a routine for determining a combustion abnormality. This routine is executed by interruption every 180 crank angles. Referring to FIG. 5, in step 80, the engine speed NE, the vehicle speed V, and the engine cooling water temperature TW are read. In step 81, it is determined whether a combustion abnormality diagnosis condition is satisfied. The combustion abnormality diagnosis conditions include, for example, NE <1000 rpm, V <2.8 km / h, TW> 60 ° C., idle operation, and 120 seconds after the start of engine start. If any one of these conditions is not met,
Proceed to step 82. When all of these conditions are satisfied, the routine proceeds to step 83, where it is determined whether or not the counter CI indicating the elapsed time after the start of the idling operation ≦ 40.

FIG. 6 shows a CI control routine. This routine is executed by interruption every second. In step 100, it is determined whether or not the idle switch 15 is turned on, that is, whether or not the idle operation is being performed. If the vehicle is idling, the process proceeds to step 101, where the CI is incremented by 1. If the vehicle is not idling, the process proceeds to step 102, where it is determined whether the CI is zero. If CI is 0, it is kept at 0. CI
If is not 0, the process proceeds to step 103, where the CI is decremented by 1. As described above, the CI increases by one every second while the idle switch 15 is turned on, and decreases by one every second when the idle switch 15 is turned off.

Referring again to FIG. 5, when CI> 40, the routine proceeds to step 82. In step 82, it is determined whether or not a timer counter CD (described later) indicating the elapsed time from the start of the abnormality diagnosis is 25 or more. At first, since CD is 0, a negative determination is made and the routine proceeds to step 98. Step 98 In the abnormality counter CF ₁ ~CF _4, cylinder discrimination number n, the timer counter CD, and a flag B 0
The routine is terminated without performing the combustion abnormality diagnosis. CI> 40 were to not perform the combustion abnormality diagnosis when the temperature of the O ₂ temperature sensor 14 is lowered O ₂ sensor 14 is not properly detecting the air-fuel ratio becomes below a predetermined temperature when the idle operation is continued Therefore, the combustion abnormality diagnosis is performed only while CI ≦ 40 to prevent erroneous diagnosis. In the routine of FIG. 6, the CI at the time of the idle switch-off is gradually decreased by 1 because the normal operation at a higher speed than the idle operation is performed for a short period after the idle operation and the idle operation is performed again. _{2 The} temperature of the sensor 14 has not been sufficiently raised by the normal operation described above,
Therefore, since the temperature reaches a temperature at which normal operation is not performed in a short time during the idle operation again, erroneous diagnosis is prevented even in such a case.

Step if it is determined in step 83 that CI ≦ 40
Proceeding to 84, the cylinder identification number n is incremented by one. Initially, n is 1 because 0 is stored in n. Note that n does not indicate the cylinder number, but indicates the order in which the combustion stroke starts. In step 85, n is 5
It is determined whether or not. If n = 5, go to step 86 and n
To 1. If n is not 5, the value of n is kept as it is and the process proceeds to step 87. In step 87, the engine speed rotation speed variation ΔNE between cylinders in the combustion stroke is calculated by the following equation.

ΔNE = NEB−NE Here, for example, NE is the engine speed in the combustion stroke of the cylinder identification number n = 1, and NEB is the combustion cycle just before the cylinder of n = 1 (180 degrees before the crank angle). = 4 is the engine speed in the combustion stroke of the cylinder. In step 88, it is determined whether or not the rotation speed fluctuation ΔNE is, for example, 30 rpm or more. ΔN
When E ≧ 30 rpm, the routine proceeds to step 89, where the abnormality counter CFn of the cylinder with the cylinder identification number _n is incremented by one.
[Delta] NE <value of CF _n when 30rpm is maintained at the current value. For example, when the fuel injection valve of the cylinder of n = 2 fails and fuel exceeding the required fuel injection amount is supplied into the cylinder, and thus the air-fuel mixture becomes overrich and misfires, the cylinder of n = 1 normally operates. Since the engine is burning, the engine speed during the combustion stroke is, for example, 650 rpm. Since the misfire has occurred in the cylinder of n = 2 which will be the next combustion stage, the engine speed during the combustion stroke is, for example, 600 rpm. Therefore, the difference ΔNE becomes 50 rpm, and a misfire in the cylinder of n = 2 can be detected. Therefore
CF _n indicates the number of times misfire has been determined in the n-th cylinder.

In step 90, the NE in the current processing cycle is stored in the NEB, and the calculation of ΔNE in the next processing cycle uses the NEB
Used as In step 91, it is determined whether or not the timer counter CD indicating the elapsed time from the start of the abnormality diagnosis is 25 or more.

FIG. 7 shows a control routine of the CD. This routine is executed by interruption every second. Steps 81 and 83 correspond to steps 81 and 83 of the routine shown in FIG.
Same as 83. If both steps 81 and 83 are affirmatively determined, the routine proceeds to step 110, where CD is incremented by one. On the other hand, if a negative determination is made in either step 81 or step 83, nothing is executed and this routine ends. However, in this case, CD is set to 0 in step 98 of the routine of FIG. Thus, the CD indicates the elapsed time since the start of the cylinder abnormality diagnosis.

Referring to FIG. 5 again, when CD <25, that is, when 25 seconds have not elapsed since the start of the abnormality diagnosis, the present routine is terminated without judging a normal abnormality. This is for performing the determination after performing the abnormality diagnosis for at least 25 seconds. When CD ≧ 25, the routine proceeds to step 92, where an abnormality judgment reference value FL and a normal judgment reference value NL are calculated based on FAF · KG.

FIG. 8 shows the relationship between FAF · KG and determination reference values FL and NL. Referring to FIG. 8, the criterion values FL and NL are FA
The value is constant when F · KG is 0.9 or more, and when FAF · KG is smaller than 0.9, FL and NL decrease with the same slope. Further, the abnormality determination reference value FL is always larger than the normal determination reference value.

Returning to Figure 5 again, the abnormality counter CF _n of the n-th cylinder in step 93 it is determined whether or failure determination reference value FL. If any one of CF ₁ , CF ₂ , CF ₃ and CF ₄ becomes FL or more, the routine proceeds to step 94, where an abnormality is determined. Subsequently, in step 95, the flag B is set to 1. In step 93, CF
If it is determined that _n / FL, the process proceeds to step 96, abnormality counter CF _n of the n-th cylinder is determined whether subnormal determination reference value NL. When CF _n ≧ NL, this routine ends. Step 97 if any one of CF ₁ to CF ₄ exceeds NL
Then, the flag B is set to 1 and the routine is terminated. After determining that CD ≧ 25 in step 91 and making a failure determination,
If a negative determination is made in step 81 or step 83 to end the combustion abnormality diagnosis and proceed to step 82,
An affirmative determination is made at 82 and the routine proceeds to step 99. In step 99, it is determined whether or not the flag B is 0.
Proceeding to 98, the counters and flags are cleared, and the routine ends. If B = 0, ie, step
If all of CF ₁ to CF ₄ are below NL in 96, step
Proceeding to 100, a normality determination is made.

In the present embodiment, the criterion values FL and NL are changed according to the value of FAF · KG. However, if the criterion values FL and NL are not changed according to, for example, FAF · KG, the following problem occurs. That is, since the fuel injection valve of one of the four cylinders fails and a considerably large amount of fuel is supplied into the cylinders, the misfire occurs in the cylinders. fuel injection quantity from the fuel injection valve when the can to some extent reduced by a command from the electronic control unit 30, O ₂ sensor 14 is learning correction coefficient KG to continue emitting a rich signal is being caused to decrease for example 0.8. For this reason, the air-fuel mixture in the normal cylinder becomes lean but burns sufficiently, while the amount of fuel supply to the misfired cylinder is also reduced, and the degree of over-lean of the air-fuel ratio is reduced. The frequency of misfires is reduced. For this reason, even when the learning correction value KG is reduced due to the over-rich mixture in the cylinder in which the fuel injection valve has failed, the failure determination is performed using the same determination reference value as when KG is large. No abnormal combustion can be detected. Therefore, in the present embodiment, when the fuel injection valve fails and the air-fuel mixture supplied into the cylinder becomes over-rich and KG becomes smaller than a value (for example, 0.9) that becomes smaller due to normal aging, the criterion is used. The values FL and NL are reduced so that abnormal combustion can be accurately detected. In FIG. 8, the reason why the judgment reference values FL and NL are changed according to the value of FAF / KG at the position where the FAF / KG is 0.9 or less is that the KG changes due to the normal aging as described above. It is considered to be at most 0.9, and it is considered that the reason why the KG becomes 0.9 or less is due to the occurrence of some abnormality. This is to prevent erroneous determination (normal state is determined to be abnormal) within the range of normal change of KG.

FIG. 9 shows a combustion abnormality determination routine for executing another embodiment. This routine is executed by interruption every 180 crank angles. In FIG. 9, in steps denoted by the same reference numerals as those in the routine shown in FIG.
Since the same processing is performed, the description thereof will be omitted. Referring to FIG.
G is read. In step 121, it is determined whether n = 0. At first, since n = 0, an affirmative determination is made and step 12
Proceeding to 2, KG is stored in KGO. In the next and subsequent processing cycles, n changes from 1 to 4, so a negative determination is made in step 121 and step 122 is skipped. Therefore, KG at the start of the abnormality diagnosis is stored in KGO. In step 123, the variation KG of KG is calculated from the following equation.

ΔKG = KG−KGO Here, KGO is a learning correction value in a normal state where no misfire has occurred, and KG indicates a learning correction value in the current processing cycle. In step 124, the misfire determination variable speed NEF and the normal determination variable speed NEN are calculated based on ΔKG.

FIG. 10 shows the relationship between ΔKG and the determination change rotation speed NEF, NEN. Referring to FIG. 10, NEF and NEN indicate that ΔKG is 0
It is the largest and constant value in the vicinity, and becomes smaller as ΔKG moves away from 0 toward the plus side and the minus side. This is because the larger the ΔKG, the larger the amount of correction of the air-fuel ratio by the KG at the time of abnormality.

Referring again to FIG. 9, in step 125, it is determined whether or not the rotation speed fluctuation ΔNE is equal to or greater than the misfire determination fluctuation rotation speed NEF. When [Delta] NE ≧ NEF, that is, when a misfire is determined to have occurred, the process proceeds to step 126, the abnormality counter CF _n of the n-th cylinder is incremented by 1. Step 127
Then, the engine speed NE in the current processing cycle is added to the misfiring determination variable speed NET to obtain NEB in the next processing cycle. This is because NE is reduced in the current processing cycle, and if it is used as NEB in the next processing cycle as it is, there is a risk of erroneous determination in the next processing cycle, so NE in the current processing cycle should be corrected by NEF This prevents erroneous determination. If it is determined in step 125 that ΔNE <NEF, the routine proceeds to step 128, where it is determined whether or not the rotation speed fluctuation ΔNE is equal to or less than the normal determination fluctuation rotation speed NEN. When ΔNE ≦ NEN, that is, when it is determined that no misfire has occurred,
Proceeds to step 129, the abnormality counter CF _n of the n-th cylinder is decremented by 1. If CF _n of [Delta] NE> NEN is maintained at the current value. In step 130, NE is stored in NEB, and in the next processing cycle,
NE is used as NEB. CD ≧ 25 in step 91
If it is determined that the abnormality counter CF _n of the n-th cylinder proceeds to step 131 it is determined whether, for example, 100 or more. CF ₁ to CF ₄
If any one of them is 100 or more, the routine proceeds to step 94, where an abnormality is determined. If CF _n <100, proceed to step 132 CF _n
Is determined to be, for example, 60 or less. If any _{one of} CF ₁ to CF ₄ exceeds 60, go to step 97 and set flag B
Is set to 1. Steps similar to the routine in FIG.
If B = 0 is determined in 99, a normal determination is made in step 100.

As described above, according to the present embodiment, the difference ΔK between the learning correction coefficient KGO in the normal state and the learning correction coefficient KG in the abnormal state
Since the determination change rotation speeds NEF and NEN are changed based on the change in G, the combustion abnormality can be accurately detected as in the embodiment shown in FIG.

In the above-described embodiment, the apparatus for learning control of the air-fuel ratio has been described.In the apparatus for not performing learning control, for example, the determination reference value FL, NL, or the determination change rotation speed NEF, based on the average value of FAF. NEN may be varied.

In the embodiment shown in FIG. 9, although the 1 values allowed to increase or decrease the CF _n in step 126, 129, it may be a value which allowed to increase and decrease allowed to value different values.

〔The invention's effect〕

Since the reference level is changed to a small value when the air-fuel ratio correction coefficient is changed so that the air-fuel ratio is lean, it is possible to accurately detect abnormal combustion.

[Brief description of the drawings]

1 is a block diagram of the invention, FIG. 2 is an overall block diagram of a four-cylinder internal combustion engine, FIG. 3 is a flowchart for calculating a fuel injection time, and FIG. 4 is a diagram showing a feedback correction coefficient FAF and a learning correction coefficient KG. Flowchart for calculation, fifth
Fig. 6 is a flowchart for executing the combustion abnormality determination, Fig. 6 is a flowchart for controlling the counter CI, Fig. 7 is a flowchart for controlling the counter CD, and Fig. 8 is FAF / KG and the reference value FL. , A diagram showing the relationship with NL,
FIG. 9 is a flowchart for executing a combustion abnormality judgment of another embodiment, and FIG. 10 is a graph showing ΔKG and the judgment fluctuation rotation speeds NEF and NEN.
FIG. 4 is a diagram showing the relationship between 2 ...... cylinder, 8 ...... fuel injection valve, 13 ...... exhaust pipe, 14 ...... O ₂ sensor, 17 ...... crank angle sensor.

Continuation of the front page (58) Field surveyed (Int. Cl. ⁶ , DB name) F02D 41/00-41/40 F02D 45/00

Claims (1)

(57) [Claims] An air-fuel ratio for controlling an air-fuel ratio to be a predetermined air-fuel ratio by correcting with an air-fuel ratio correction coefficient obtained based on an output signal of an oxygen concentration sensor provided in an engine exhaust passage. Control means; angular velocity detecting means for detecting an angular velocity of the output shaft of the engine; angular velocity fluctuation calculating means for calculating a fluctuation of the angular velocity between the cylinders in a combustion stroke based on a detection result of the angular velocity detecting means; Abnormality determination means for determining that a combustion abnormality has occurred when the state in which the variation in the angular velocity calculated by the calculation means is greater than or equal to the reference level, further comprising: Change according to the fuel ratio correction coefficient, when the value of the air-fuel ratio correction coefficient is a value for correcting the air-fuel ratio to the lean side from the value of the predetermined air-fuel ratio correction coefficient, Abnormal combustion detecting apparatus for an internal combustion engine, characterized in that the serial reference level provided the reference level changing means for changing to a value smaller than the reference level in the value of the air-fuel ratio correction coefficient which the predetermined.