JPH10148154A - Fuel injection control device for engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a fuel injection device enabling the engine to operate around a limit of air-fuel ratio or an allowable limit of fluctuation of engine speed due to misfire even immediately after an engine starts. SOLUTION: Misfire of each cylinder is judged by the difference in fluctuation of engine speed for every specified crank angle. When a misfire is detected, a constant of proportionality PKMF is added to a misfire-time correction factor by cylinder KMF #i for a cylinder #1 in question to be skipped in the plus direction (S62). The KMF #i is incorporated in a computing equation for the fuel injection pulse width of the fuel injection quantity for the cylinder #i, and when the misfire occurs, the injection quantity for the cylinder #i is increased to be corrected, and the misfire of the cylinder #i is eliminated. After the misfire is eliminated, the KMF #i is set to be gradually reduced by an integration constant IKMF (S66), and the increased amount of fuel for the cylinder #i by the KMF #i is gradually decreased. Since the fuel injection is increased to be corrected when the misfire is detected, the basic fuel injection quantity can be set, corresponding to the limit of air-fuel ratio around a theoretical air-fuel ratio which corresponds to an allowable limit of fluctuation of the engine speed due to misfire, even immediately after an engine starts.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention enables an engine to be operated near an air-fuel ratio limit, which is an allowable limit of engine rotation fluctuation due to a misfire, even immediately after the start of an engine, thereby achieving both improvement in exhaust emission and improvement in drivability. The present invention relates to a fuel injection control device for an engine.

[0002]

2. Description of the Related Art Conventionally, in order to prevent deterioration of drivability immediately after starting an engine and to stabilize engine rotation, a fuel increase correction coefficient is set based on a cooling water temperature representing an engine temperature and an elapsed time after the engine is started. Then, the fuel injection amount is increased and corrected by the fuel increase correction coefficient to control the air-fuel ratio in the rich direction.

[0003] The fuel increase correction coefficient is set high so as to increase the fuel increase rate in consideration of the variation of each engine, so that the air-fuel ratio becomes more rich.

[0004]

Therefore, during the air-fuel ratio open loop control immediately after the start of the engine, it is not possible to operate around the stoichiometric air-fuel ratio which is effective for improving the exhaust emission.

[0005] Conversely, when the engine is operated near the stoichiometric air-fuel ratio in order to improve the exhaust gas emission immediately after the start of the engine, the engine speed fluctuates due to misfiring, engine stall and the like, and drivability deteriorates.

Japanese Unexamined Patent Publication No. Hei 4-66750 discloses that during idling, the combustion state of each cylinder is determined based on the engine speed fluctuation difference for each predetermined crank angle. The combustion state is improved by increasing the fuel injection amount, and by distributing and decreasing the fuel injection amount for the other cylinders in accordance with the increased fuel amount, the combustion state of each cylinder is made uniform and idle rotation is performed. A technique has been disclosed that prevents fluctuation and also prevents overlean or overrich of the total air-fuel ratio at this time.

However, in the above-mentioned prior art, the combustion state of each cylinder is made uniform by continuously and uniquely correcting the fuel injection amount for each cylinder according to the combustion state of each cylinder at the time of idling. However, the present invention does not cope with the operation at the air-fuel ratio limit that causes misfire, because idling rotation fluctuation is prevented.

In view of the above circumstances, the present invention enables operation near the air-fuel ratio limit, which is an allowable limit of engine rotation fluctuation due to misfire, even immediately after engine start, thereby achieving both improvement in exhaust emission and improvement in drivability. It is an object of the present invention to provide an engine fuel injection control device capable of performing the following.

[0009]

According to a first aspect of the present invention, there is provided an engine fuel injection control apparatus for detecting a misfire as shown in FIG. 1 (a). It is characterized by comprising a detecting means and a fuel injection amount correcting means for increasing and correcting the fuel injection amount when a misfire is detected.

As shown in FIG. 1 (b), a fuel injection control device for an engine according to the present invention has a misfire detecting means for detecting misfire, and a correction for increasing fuel quantity when misfire is detected. The set value is added to the misfire correction coefficient for misfire, and the misfire correction coefficient setting means for gradually reducing the misfire correction coefficient when the misfire has been eliminated, and the basic fuel injection amount set based on the engine operating state as described above. Fuel injection amount setting means for setting the fuel injection amount by correcting with the misfire correction coefficient.

According to a third aspect of the present invention, in the fuel injection control apparatus for an engine according to the third aspect, the misfire correction coefficient setting means limits the misfire correction coefficient by an upper limit value. I do.

According to a fourth aspect of the present invention, there is provided a fuel injection control apparatus for an engine, wherein a misfire detecting means for detecting misfire for each cylinder, and a first misfire correction coefficient for each cylinder of the corresponding cylinder when the misfire is detected. In addition to adding the values, the second set value is subtracted from the cylinder-specific misfire correction coefficients of the other cylinders, and when the misfire has been eliminated, the cylinder-specific misfire correction coefficients are gradually increased until no correction is made. Or a misfiring correction coefficient setting means that decreases, and a fuel injection quantity setting means that corrects the basic fuel injection amount set based on the engine operating state by the cylinder-specific misfiring correction coefficient and sets the fuel injection amount for each cylinder. It is characterized by having.

According to a fifth aspect of the present invention, in the fuel injection control apparatus for an engine according to the fourth aspect, the misfire correction coefficient setting means limits the cylinder-specific misfire correction coefficient by an upper limit value and a lower limit value. It is characterized by doing.

According to a sixth aspect of the present invention, in the fuel injection control device for an engine according to the fourth aspect of the invention, the second set value corresponds to the first set value except for the other misfire cylinders. It is characterized in that the fuel injection amount for the cylinder is set to a value that reduces the equal distribution.

That is, according to the first aspect of the present invention, when a misfire is detected, the fuel injection amount is increased and corrected, so that the base fuel injection amount corresponds to the permissible limit of engine rotation fluctuation due to misfire immediately after the engine is started. Can be set corresponding to the limit air-fuel ratio near the stoichiometric air-fuel ratio. Therefore, the exhaust emission can be improved by operating at the limit air-fuel ratio, and at the time of a misfire, the misfire is immediately eliminated by the fuel increase correction, whereby fluctuations in the engine speed due to the misfire are prevented and drivability is improved.

According to the second aspect of the present invention, when a misfire is detected, a set value is added to a misfire correction coefficient for fuel increase correction, and when the misfire is resolved, the misfire correction coefficient is gradually reduced. . Then, the basic fuel injection amount set based on the engine operating state is corrected by the misfire correction coefficient to set the fuel injection amount. Here, when a misfire is detected, the set value is added to the misfire correction coefficient, and the fuel injection amount is immediately increased and corrected by the misfire correction coefficient, thereby eliminating the misfire. By gradually reducing the time correction coefficient, deterioration of controllability due to a sudden change in the air-fuel ratio is prevented.

In this case, according to the third aspect of the invention, by limiting the misfire correction coefficient by the upper limit value, it is possible to prevent the air-fuel ratio from becoming excessively rich due to an excessive increase in fuel, thereby deteriorating controllability and causing misfire. Prevent dissolution from being hindered.

According to the present invention, a misfire for each cylinder is detected, and when a misfire is detected, a cylinder-by-cylinder misfire correction coefficient of the cylinder is set by a value obtained by adding the first set value. The cylinder-by-cylinder misfire correction coefficient is set by subtracting the second set value. When the misfire is eliminated, the cylinder-specific misfire correction coefficient is gradually increased or decreased until no correction is made. Then, the basic fuel injection amount set based on the operating state of the engine is corrected by the cylinder-specific misfire correction coefficient to set the fuel injection amount for each cylinder. Here, when a misfire is detected, the first set value is added to the cylinder-specific misfire correction coefficient for the relevant misfire cylinder, and the fuel injection amount is immediately increased and corrected by the misfire correction coefficient. At this time, the misfire correction coefficient for each cylinder is set to be reduced by the second set value for the other cylinders, and the fuel injection amount is corrected to be reduced. As a result, immediately after the misfire is eliminated, the combustion state of each cylinder is made uniform, and fluctuations in engine rotation caused by uneven combustion state of each cylinder are prevented. Further, after the misfire is eliminated, the cylinder-specific misfire correction coefficient for the relevant misfire cylinder is gradually reduced until the fuel injection amount is not corrected, and the cylinder-specific misfire correction coefficient for the other cylinders is gradually reduced. The sudden change of the air-fuel ratio at the time of cancellation of the correction by the misfire correction coefficient is suppressed, and deterioration of controllability is prevented.

In this case, according to the fifth aspect of the present invention, the cylinder-specific misfire correction coefficient is limited by an upper limit value and a lower limit value, thereby preventing unnecessary overcorrection.

Further, in the invention according to claim 6, in the above-mentioned second aspect,
Is set to a value corresponding to the first set value so that the fuel injection amount to other cylinders other than the corresponding misfire cylinder is equally distributed and reduced, the macro total air-fuel ratio by the fuel correction at the time of misfire To prevent changes in air-fuel ratio controllability.

[0021]

Embodiments of the present invention will be described below with reference to the drawings. 2 to 19 show a first embodiment of the present invention.

First, the overall structure of the engine will be described with reference to FIG. In the figure, reference numeral 1 denotes an engine, which in this embodiment is a horizontally opposed four-cylinder gasoline engine. The cylinder block 1a of this engine 1
In each of the left and right banks, a cylinder head 2 is provided, and an intake port 2a and an exhaust port 2b are formed in each cylinder head 2.

In the intake system of the engine 1, an intake manifold 3 communicates with each intake port 2a, and a throttle chamber 5 communicates with the intake manifold 3 via an air chamber 4 in which intake passages of respective cylinders are gathered. . An air cleaner 7 is attached to the upstream side of the throttle chamber 5 via an intake pipe 6, and the air cleaner 7 is communicated with the air intake chamber 8.

The throttle chamber 5 is provided with a throttle valve 5a linked to an accelerator pedal. The intake pipe 6 is connected to a bypass passage 9 that bypasses the throttle valve 5a.
At the time of idling, the bypass passage 9 depends on the valve opening.
An idle speed control valve (ISC valve) 10 for controlling the idle speed by adjusting the amount of bypass air flowing through the engine is provided.

Further, an injector 11 is disposed immediately upstream of the intake port 2a of each cylinder of the intake manifold 3. The injector 11 has a fuel supply path 12
The fuel tank 13 is provided with an in-tank type fuel pump 14. The fuel from the fuel pump 14 is pressure-fed to the injector 11 and the pressure regulator 16 via the fuel filter 15 interposed in the fuel supply path 12, and is returned from the pressure regulator 16 to the fuel tank 13, where The fuel pressure to the injector 11 is adjusted to a predetermined pressure.

On the other hand, for each cylinder of the cylinder head 2, a spark plug 17 exposing a discharge electrode at the tip to the combustion chamber is provided.
The ignition plug 17 is connected to an igniter 19 via an ignition coil 18 provided for each cylinder.

As an exhaust system of the engine 1, an exhaust pipe 21 is communicated with a collection portion of an exhaust manifold 20 which communicates with each exhaust port 2b of the cylinder head 2, and a catalytic converter 22 is interposed in the exhaust pipe 21. And is communicated with the muffler 23.

Next, sensors for detecting the operating state of the engine will be described. Immediately downstream of the air cleaner 7 in the intake pipe 6, a thermal intake air amount sensor 24 using a hot wire or a hot film is interposed, and a throttle valve 5a provided in the throttle chamber 5 is provided with a throttle valve 5a. A throttle sensor 25 having a built-in opening sensor 25a and an idle switch 25b that is turned on when the throttle valve 5a is fully closed is connected in series.

The cylinder block 1a of the engine 1
A knock sensor 26 is mounted on the cooling water passage 2 which communicates the left and right banks of the cylinder block 1a.
7, a cooling water temperature sensor 28 is provided, and an O2 for detecting an air-fuel ratio state is provided upstream of the catalytic converter 22.
A sensor 29 is provided.

The crankshaft 30 of the engine 1
A crank angle sensor 32 is provided on the outer periphery of a crank rotor 31 that is axially mounted on the cam shaft 33. Further, a cam rotor 34 connected to a cam shaft 33 that makes a half turn with respect to the crank shaft 30 has a cam angle for cylinder identification. A sensor 35 is provided opposite.

As shown in FIG. 17, the crank rotor 31 has projections 31a, 31b and 31c formed on the outer periphery thereof, and these projections 31a, 31b and 31c are connected to the cylinders (# 1 and # 2 cylinders). (# 3, # 4 cylinders) before compression top dead center (BTDC) θ1, θ2, θ3.
In the present embodiment, θ1 = 97 ° CA, θ2 = 65
° CA, θ3 = 10 ° CA.

As shown in FIG. 18, on the outer periphery of the cam rotor 34, projections 34a, 34b,
The protrusion 34a is formed at a position θ4 after the compression top dead center (ATDC) of the # 3 and # 4 cylinders.
b is composed of three protrusions, and the first protrusion is A of the # 1 cylinder.
It is formed at the position of TDCθ5. Further, the protrusion 34c
Is composed of two projections, and the first projection is the AT of the # 2 cylinder.
It is formed at the position of DCθ6. In the present embodiment, θ4 = 20 ° CA, θ5 = 5 ° CA, θ6 = 20 °
CA.

As shown in the time chart of FIG. 9, the crank rotor 31 and the cam rotor 34 are rotated by the rotation of the crankshaft 30 and the camshaft 33 in association with the operation of the engine. Detected by the angle sensor 32 and detected from the crank angle sensor 32 by θ1, θ2, θ3 (BTDC97
°, 65 °, 10 °) of the engine 1
Output every 1/2 rotation (180 ° CA). On the other hand, θ3
Each protrusion of the cam rotor 34 is detected by the cam angle sensor 35 between the crank pulse and the θ1 crank pulse, and a predetermined number of cam pulses are output from the cam angle sensor 35.

Then, an electronic control unit 40 (FIG. 1)
9), the engine speed NE is calculated based on the input interval time Tθ of the crank pulse output from the crank angle sensor 32, and the order of combustion stroke of each cylinder (for example, # 1 cylinder → # 3 cylinder) → # 2 cylinder → # 4 cylinder)
Based on the pattern of the cam pulse from the cam angle sensor 35 and the value counted by the counter, the cylinders such as the fuel injection target cylinder and the ignition target cylinder are determined.

The injector 11, the spark plug 17,
The calculation of control amounts for actuators such as the ISC valve 10 and the output of control signals, that is, engine control such as fuel injection control, ignition timing control, idle speed control, etc.
Is performed by an electronic control unit (ECU) 40 shown in FIG.

The ECU 40 includes a CPU 41, a ROM 4
2, RAM 43, backup RAM 44, counter
A timer group 45 and an I / O interface 46 are mainly configured by a microcomputer connected to each other via a bus line, and are connected to the constant voltage circuit 47 for supplying a stabilized power to each unit, and to the I / O interface 46. And a peripheral circuit such as an A / D converter 49.

The counter / timer group 45 includes various counters such as a free-run counter, a counter for counting the input of a cam angle sensor signal (cam pulse), a fuel injection timer, an ignition timer, and a timer for generating a periodic interrupt. Various timers such as a periodic interrupt timer, a timer for measuring the input interval of a crank angle sensor signal (crank pulse), and a watchdog timer for monitoring a system abnormality are collectively referred to for convenience. Is used.

The constant voltage circuit 47 is connected to a battery 51 via a first relay contact of a power relay 50 having two relay contacts, and a relay coil of the power relay 50 is connected to an ignition switch 52 by the battery 51.
Connected through. The constant voltage circuit 47
When the ignition switch 52 is turned on and the contact of the power supply relay 50 is closed, power is supplied to each unit in the ECU 40 while the ignition switch 52 is turned on and the ignition switch 52 is turned on and off. Instead, the backup power is always supplied to the backup RAM 44. Further, the fuel pump 18 is connected to the battery 51 via a relay contact of a fuel pump relay 53. The power supply relay 50
A power line for supplying power from the battery 51 to each actuator is connected to the second relay contact.

The input port of the I / O interface 46 has an idle switch 25b, a knock sensor 2
6. A crank angle sensor 32, a cam angle sensor 35, a vehicle speed sensor 36 for detecting a vehicle speed, and a starter switch 37 for detecting a starting state are connected, and further, via the A / D converter 49. And the intake air amount sensor 24, the throttle opening degree sensor 25a, the cooling water temperature sensor 2
8 and the O2 sensor 29 are connected, and the battery voltage VB is input and monitored.

On the other hand, an ISC valve 10, an injector 11, and a relay coil of a fuel pump relay 53 are connected to an output port of the I / O interface 46 by the drive circuit 48.
And the igniter 19 is connected.

In accordance with the control program stored in the ROM 42, the CPU 41 processes the detection signals from the sensors and switches, which are input via the I / O interface 46, the battery voltage, and the like, and stores them in the RAM 43. Based on various data, various learning value data stored in the backup RAM 44, fixed data stored in the ROM 42, and the like, a fuel injection amount, an ignition timing, a duty ratio of a drive signal for the ISC valve 10, and the like are calculated. It performs engine control such as injection control, ignition timing control, and idle speed control.

In such an engine control system, EC
In U40, the engine rotational speed NE is used as an example of the engine rotational state amount, and it is determined whether or not a misfire has occurred for each cylinder based on the amount of change in the engine rotational speed in each predetermined cycle. The misfire is eliminated by increasing the fuel injection amount of the fuel tank.

That is, when the misfire is detected, the fuel injection amount is increased and corrected. Even immediately after the engine is started, the base fuel injection amount is changed to the limit air-fuel ratio near the stoichiometric air-fuel ratio corresponding to the engine rotation fluctuation allowable limit due to the misfire. Therefore, exhaust emission can be improved by operating at this limit air-fuel ratio, and at the time of misfire, misfire is immediately eliminated by fuel increase correction, and fluctuations in engine rotation due to misfire are prevented. Drivability is improved. In other words, immediately after the start of the engine, operation near the air-fuel ratio limit, which is the allowable limit for engine rotation fluctuation due to misfire, becomes possible, and both improvement in exhaust emission and improvement in drivability can be achieved.

More specifically, in this embodiment, the misfire of each cylinder is determined based on the difference in the engine speed for each predetermined crank angle. Specifically, the engine speed between the BTDC θ2 and θ3, that is, between the θ2 and θ3 crank pulse inputs is adopted, and the engine speed in the same section last time is subtracted from the engine speed calculated this time to calculate the differential speed. The misfire of the corresponding cylinder is determined based on the differential rotation. When the misfire is detected, the set value is added to the misfire correction coefficient corresponding to the cylinder for fuel increase correction, and when the misfire is eliminated, the misfire correction coefficient is gradually reduced.
Then, the basic fuel injection amount set based on the engine operating state is corrected by the misfire correction coefficient to set the fuel injection amount for the corresponding cylinder. Therefore, when a misfire is detected, the set value is added to the misfire correction coefficient, and the fuel injection amount is immediately increased and corrected by the misfire correction coefficient, thereby eliminating the misfire. Gradual reduction of the correction coefficient prevents deterioration of controllability due to a sudden change in the air-fuel ratio.

That is, the ECU 40 has the functions of a misfire detection unit and a fuel injection amount correction unit according to the present invention.
Functions as misfire correction coefficient setting means and fuel injection amount setting means are also realized.

The control process according to the present invention executed by the ECU 40 will be described below with reference to the flowcharts shown in FIGS.

First, the ignition switch 52 is turned on.
Then, when the power is supplied to the ECU 40, the system is initialized, and each flag and each counter are initialized except for data such as various learning values stored in the backup RAM 44. And the starter switch 37
Is turned on and the engine is started, a cylinder discriminating / engine speed calculating routine shown in FIG. 2 is executed every time a crank pulse is input from the crank angle sensor 32.

In this cylinder discriminating / engine speed calculating routine, when the crank rotor 31 rotates and the crank pulse is input from the crank angle sensor 32 in accordance with the operation of the engine, first, in step S1, the presently input crank is input. Whether the pulse corresponds to the crank angle of θ1, θ2, or θ3 is identified based on the input pattern of the cam pulse from the cam angle sensor 35. In step S2, the current combustion is determined based on the input pattern of the crank pulse and the cam pulse. The cylinders such as the stroke cylinder, the misfire diagnosis target cylinder, the ignition target cylinder, and the fuel injection target cylinder are determined.

That is, as shown in the time chart of FIG. 9, for example, if there is a cam pulse input between the time when the previous crank pulse is input and the time when the current crank pulse is input, the current crank pulse becomes θ1 crank pulse. It can be identified as a pulse, and the crank pulse input next time can be identified as a θ2 crank pulse.

When there is no cam pulse input between the previous and current crank pulse inputs and there is a cam pulse input between the last and previous crank pulse inputs, the current crank pulse can be identified as the θ2 crank pulse, and The input crank pulse can be identified as a θ3 crank pulse. Further, when there is no cam pulse input between the previous and current times and between the last and last crank pulse inputs, the currently input crank pulse can be identified as the θ3 crank pulse, and the next input crank pulse is It can be identified as a θ1 crank pulse.

Further, three cam pulses are input between the previous and current crank pulse inputs (θ corresponding to the projection 34b).
(5 cam pulses), the next compression top dead center is # 3 cylinder, the current combustion stroke cylinder is # 1 cylinder, the misfire diagnosis target cylinder is # 4 cylinder one combustion stroke earlier, and the ignition target cylinder is # 3 cylinder.
It can be determined that the cylinder and the fuel injection target cylinder are the # 4 cylinder that is two cylinders behind the cylinder. When two cam pulses (θ6 cam pulse corresponding to the projection 34c) are input between the previous and current crank pulse inputs, the next compression top dead center is #
There are four cylinders. The current combustion stroke cylinder is # 2 cylinder, the misfire diagnosis target cylinder is # 3 cylinder, the ignition target cylinder is # 4 cylinder, and the fuel injection target cylinder is # 3 cylinder.

One cam pulse is input between the previous and current crank pulse inputs (θ4 corresponding to the projection 34a).
When the previous compression top dead center determination is # 4 cylinder, the next compression top dead center is # 1 cylinder, the current combustion stroke cylinder is # 4 cylinder, and the misfire diagnosis target cylinder is # 2. The cylinder and the cylinder to be ignited can be determined as # 1 cylinder, and the fuel injection cylinder can be determined as # 2 cylinder. Similarly, when one cam pulse is input between the previous and current crank pulse inputs, and the previous compression top dead center determination was # 3 cylinder, the next compression top dead center is # 2 cylinder, and the current compression top dead center is # 2 cylinder. The combustion stroke cylinder can be determined to be # 3 cylinder, the misfire diagnosis target cylinder is # 1 cylinder, the ignition target cylinder is # 2 cylinder, and the fuel injection target cylinder is # 1 cylinder.

In the four-cycle four-cylinder engine 1 of this embodiment, the combustion stroke is in the order of cylinders # 1, # 3, # 2, and # 4. As shown in the ignition and fuel injection time chart of FIG. This combustion stroke cylinder #n at the time of output is #
Assuming that there is one cylinder, the misfire diagnosis target cylinder is #n
-1 cylinder, that is, # 4 cylinder, the cylinder to be ignited reaches the compression top dead center # n + 1 = # 3 cylinder, and the fuel injection target cylinder #i at this time is # n + 3 = # It becomes four cylinders. The cylinder discrimination results of the current combustion stroke cylinder #n and the misfire diagnosis target cylinder # n-1 are referred to in the misfire diagnosis routine shown in FIGS. Misfire diagnosis for n-1 is performed.

Although not described in detail here, the ignition timing for the corresponding cylinder # n + 1 is set in accordance with the determination result of the cylinder # n + 1 to be ignited, and the ignition timing is controlled for each cylinder. Further, the determination result of the fuel injection target cylinder #i (# n + 3 cylinder) is referred to in a fuel injection amount setting routine executed at predetermined intervals, and the fuel injection amount is set for each cylinder. One sequential injection is performed on the relevant cylinder every 720 ° CA (two engine revolutions).

Thereafter, the process proceeds from step S2 to step S3, in which the time from the previous crank pulse input to the present crank pulse input measured by the crank pulse input interval timer, that is, the crank pulse input interval time (θ1 crank pulse And the input interval time Tθ12 of the θ2 crank pulse, the input interval time Tθ23 of the θ2 crank pulse and the θ3 crank pulse, or the input interval time Tθ31 of the θ3 crank pulse and the θ1 crank pulse), and the crank pulse input interval time Tθ is detected.

Next, the process proceeds to step S4, in which the angle between the crank pulses corresponding to the crank pulse identified this time is read, and the current engine speed N is determined based on the angle between the crank pulses and the crank pulse input interval time Tθ.
E is calculated, stored at a predetermined address in the RAM 43, and the routine exits. The angle between the crank pulses is known and is stored in advance in the ROM 42 as fixed data. In the present embodiment, the angle between the θ1 crank pulse and the θ2 crank pulse is 32 ° CA.
The angle between the θ2 crank pulse and the θ3 crank pulse is 55 ° CA, and the angle between the θ3 crank pulse and the θ1 crank pulse is 93 ° CA. Further, in consideration of the engine start, the engine speed NE is calculated at, for example, 150 rpm or more.

In the misfire diagnosis routine shown in FIGS. 3 and 4 executed every time the θ3 crank pulse is input, the misfire diagnosis target cylinder # n-1, the current combustion stroke cylinder #n,
And the respective data of the engine speed NE are read out, and misfire diagnosis is performed for the cylinder # n-1 for misfire diagnosis.

The misfire diagnosis routine is executed in response to the input of the θ3 crank pulse. At this time, the input interval time Tθ2 between the θ2 crank pulse and the θ3 crank pulse is obtained by the cylinder discrimination / engine speed calculation routine.
The engine speed NE between the BTDCs θ2 and θ3 is calculated based on (3), and the misfire diagnosis of the corresponding cylinder is performed based on the engine speed NE between the BTDCs θ2 and θ3.

In this misfire diagnosis routine, first,
In step S11, each data obtained in the previous execution of the routine is stored in the work area. In step S12, the current engine speed NE, that is, the latest BTDC θ2, θ
The engine speed NE between 3 is read and the current cylinder #n
(MNXn ←←)
NE).

Next, the process proceeds to step S13, and the process proceeds to step S12.
From the engine speed MNXn corresponding to the current cylinder #n set in the above, the engine speed MNXn-1 corresponding to the cylinder # n-1 one combustion stroke earlier set in the previous routine and stored in the work area is set. The difference DELNEn is subtracted to obtain the engine speed variation amount for each predetermined crank angle.
Is calculated (DELNEn ← MNXn-MNXn-1).

Then, in steps S14 and S15, the present combustion stroke cylinder data #n and misfire diagnosis target cylinder data # n-1 determined in the cylinder determination / engine speed calculation routine are read, and the current combustion is performed. The stroke cylinder #n and the misfire diagnosis target cylinder # n-1 (= 1 cylinder before the combustion stroke) are specified.

Next, at step S16, the differential rotation DEL
The difference rotation correction value AVEDNXn-4 for the cylinder #n calculated at the time of execution of this routine four times before, that is, one cycle before (720 ° CA before) by the statistical processing in steps S26 and S27 described later is subtracted from NEn. , Correction difference rotation DE
Calculate LNAn (DELNAn ← DELNen-A
VEDNXn-4). This difference rotation correction value AVEDNX is
As will be described in detail later, the corrected differential rotation DELNA of all cylinders is weighted and averaged to calculate an all cylinder differential rotation weighted average value AVEDN, and further, the differential rotation DELNEn of the cylinder #n with respect to this all cylinder differential rotation weighted average value AVEDN. Are weighted averages.

In other words, the crank angle detected by the crank angle sensor 32 includes manufacturing tolerances such as the position and shape of each of the projections 31a, 31b, 31c of the crank rotor 31, and the crank angle sensor 32 to the engine 1. There is a permissible error or the like in the mounting position of each engine. Therefore, the difference rotation DELNE calculated based on the crank pulse from the crank angle sensor 32 includes variations due to these errors. For this reason, especially at the time of high engine rotation, apparently large engine rotation fluctuations occur uniformly as shown in FIG. 10. If misfire diagnosis is directly performed by this differential rotation DELNE, erroneous diagnosis will occur. .

Therefore, by subtracting the differential rotation correction value AVEDNXn−4 for the cylinder #n calculated one cycle earlier by the statistical processing from the differential rotation DELNEn, and calculating the corrected differential rotation DELNAn, the crank rotor 31
Of the projections 31a, 31b, and 31c, and the tolerance of the mounting position of the crank angle sensor 32 to the engine 1 and the like.
, And an accurate difference between the engine speed MNXn-1 for the cylinder # n-1 one combustion stroke before and the engine speed MNXn-1 for the cylinder # n-1 one combustion stroke earlier. Note that the timing chart of FIG. 11 shows the corrected differential rotation DELNA. In contrast to the differential rotation DELNE shown in FIG. 10, as shown in FIG. 11, it can be seen that the corrected differential rotation DELNA provides an accurate differential rotation excluding the influence of the allowable error.

Note that FIGS. 10 and 11 and FIGS.
2 shows differential rotation data calculated by the ECU 40, with one scale on the vertical axis being 50 rpm and one scale (1 div) on the horizontal axis being 720 ° CA.

Here, for example, when this misfire diagnosis routine is executed by inputting the BTDCθ3 crank pulse of the # 3 cylinder, the current combustion stroke cylinder #n is the # 1 cylinder, and the cylinder # n− one cylinder before the combustion stroke. As # 1, cylinder # 4 becomes the misfire diagnosis target cylinder (see FIG. 9). In the current routine, the rotation speed MNX # 1 (= MNXn) of the # 1 cylinder (#n cylinder) in the current combustion stroke calculated based on the BTDC θ2 and θ3 crank pulse input interval times of the # 3 cylinder.
Is subtracted from the rotational speed MNX # 4 (= MNXn-1) of the # 4 cylinder (# n-1 cylinder) one combustion stroke before, based on the BTDC .theta.2, .theta.3 crank pulse input interval time of the # 1 cylinder.
DELNE # 1 (DELNEn = MN)
Xn-MNXn-1) is calculated. And this differential rotation D
Subtract the differential rotation correction value AVEDNX # 1 (AVEDNXn-4) for the # 1 cylinder calculated at the time of execution of this routine four times before, that is, one cycle before, from ELNE # 1.
The correction differential rotation DELNA # 1 (DELNAn) of one cylinder is calculated.

In the processing after step S17, the cylinder # n-1 to be diagnosed one combustion stroke before, which is calculated at the time of execution of the previous routine, is calculated from the corrected differential rotation DELNA # 1 (DELNAn) of the # 1 cylinder. Corrected differential rotation DELN for four cylinders
A # 4 (DELNAn-1) is subtracted to calculate a corrected differential rotation change DDNEA # 1 (DDNEAn) of the # 1 cylinder. Then, the correction differential rotation change DDNEA # 1 (D
DNe An) and the correction difference rotation change DDNEA # 4 (DDNEAn-1) of the # 4 cylinder calculated during the previous execution of the routine.
A misfire diagnosis is performed for (n-1 cylinder).

That is, in step S17, the corrected differential rotation DELNAn-1 of the # n-1 cylinder calculated in the previous execution of the routine and stored in the work area is calculated from the corrected differential rotation DELNAn of the #n cylinder calculated in step S16. Is subtracted to obtain the corrected differential rotation change DDNEA for the #n cylinder.
n (DDNEAn ← DELNAn−DELN)
An-1).

The behavior of the corrected differential rotation change DDNEA is
This is shown in the timing chart of FIG.

Here, if an intermittent rotational fluctuation such as a snatch occurs in the engine 1, the intermittent rotational fluctuation due to the snatch or the like can be regarded as a misfire simply by comparing the corrected differential rotation DELNA with a predetermined misfire determination level. Misdiagnosed,
Accurate misfire diagnosis cannot be performed. Therefore, by diagnosing misfire based on the change in the corrected differential rotation DELNA of the front and rear cylinders, that is, the corrected differential rotation change DDNEA, the differential rotation (corrected differential rotation DELNA) whose level changes according to the engine operating state at the time of the misfire occurs. On the other hand, the influence of intermittent rotation fluctuations such as snatches is eliminated to enable accurate misfire diagnosis.

Next, the process proceeds to step S18, where steps S18 to S18 are performed.
In S20, it is determined whether a misfire diagnosis condition is satisfied.

That is, it is determined in step S18 whether or not the fuel cut is being performed. If the fuel cut is not being performed, in step S19, the basic fuel that is set in a fuel injection amount setting routine described later and determines the basic fuel injection amount is determined. The injection pulse width Tp is compared with the lower limit value TpLWER, and it is determined whether or not the current engine operation region is in the misfire diagnosable region. Then, when Tp ≧ TpLWRE, in a succeeding step S20, the current engine speed NE is compared with an upper limit value NEUPER.

Here, during the fuel cut, no fuel is injected, and each cylinder is in a forced misfire state, so that misfire diagnosis cannot be performed. In addition, when Tp <TpLWRE and the fuel injection amount is small, a rotational fluctuation component other than misfire is large, and misfire diagnosis is performed if misfire diagnosis is performed based on a change in engine speed. Further, in a high rotation range of NE ≧ NEEUPER, the rotation fluctuation is moderated, and a misfire diagnosis cannot be performed in this range.

Therefore, when the fuel is being cut or when Tp
<TpLWRE misfire diagnosable area, or NE> NE
If the UPER is in the high rotation range, the process proceeds from the relevant step to step S21 due to the failure of the misfire diagnosis condition, clears the diagnosis permission flag FLGDIAG (FLGDIAG ← 0), prohibits the misfire diagnosis, and proceeds to step S23.

On the other hand, when fuel injection is performed, Tp ≧ T
When the current engine operating state is in the misfire diagnosable area in pLWRE and is in the area excluding high revolutions with NE <NEEUPER, it is determined that the misfire diagnosing condition is satisfied, and the routine proceeds to step S22, where the diagnosis permission flag FLGDIAG is set (FLGDIAG ← 1) to permit misfire diagnosis, and proceed to step S23.

Then, in step S23, the misfire diagnosis subroutine shown in FIG.
Perform misfire diagnosis for -1.

In this misfire diagnosis subroutine,
In step S31, the diagnosis permission flag FLGDIAG is referred to, and FL
If GDIAG = 0 and misdiagnosis conditions are not satisfied and the misfire diagnosis is prohibited, the process jumps to step S36, and the cylinder-specific misfire flag FL representing misfire of the cylinder # n-1 to be subjected to misfire diagnosis.
Clear GMISn-1 (FLGMISn-1 ← 0),
The process proceeds to step S24 of the misfire diagnosis routine (FIG. 4).

On the other hand, in step S31, FLGDIA
When G = 1 and misfire diagnosis is permitted, the process proceeds to step S32, where a map is referred to based on the current engine speed NE and the basic fuel injection pulse width Tp representing the engine load, and misfire determination is performed by interpolation calculation. Set the level LVLMIS.

The misfire determination level LVLMIS is used for judging misfire by comparing with the above-described corrected differential rotation change DDNEA, and differs depending on the engine characteristics for each engine type and the engine operating state. Therefore, the engine speed N is determined in advance by an experiment or a simulation.
For each operation region based on E and the basic fuel injection pulse width Tp,
The correction difference rotation change DDNEA when a misfire occurs is obtained,
The corresponding misfire determination level LVLMIS is set as a map using the engine speed NE and the basic fuel injection pulse width Tp as parameters, and is stored in a series of addresses in the ROM 42.

As shown in FIG. 13, the misfire determination level LVLMIS is set when the basic fuel injection pulse width Tp is equal to or smaller than the lower limit as a non-diagnosable area. The value is set to a higher value as the injection pulse width Tp increases and the engine load increases.

Then, in step S33, the above-described step S17
Differential rotation change DDNEA for cylinder #n calculated in
n is read and compared with the misfire determination level LVLMIS. And when DDNEAn ≧ LVLMIS,
In a succeeding step S34, the relevant misfire diagnosis target cylinder # calculated in the previous execution of the routine and stored in the work area.
The correction difference rotation change DDNEAn-1 with respect to n-1 is read and compared with a minus misfire determination level -LVLMIS.

Then, DDNEAn-1≤-LVLMIS
At this time, it is determined that the misfire diagnosis target cylinder # n-1 is misfired, and the process proceeds to step S35, where the misfire flag FLG of the corresponding cylinder is determined.
MISn-1 is set (FLGMISn-1 ← 1), and the routine proceeds to step S24 of the misfire diagnosis routine.

In step S33, DDNEAn <
In the case of LVLMIS or DDNEA in step S34
If n-1> -LVLMIS, it is determined that a misfire has not occurred in the relevant cylinder # n-1 for misfire diagnosis, and step S3 is performed.
The program proceeds to S6, where the misfire flag FLGMISn-1 of the corresponding cylinder is cleared, and the procedure proceeds to Step S24 of the misfire diagnosis routine.

Here, as shown in FIG. 14, when the # n-1 cylinder misfires, the effective pressure due to combustion cannot be obtained, so the engine speed M before the combustion stroke of the # n-1 cylinder is lost.
In contrast to NXn-2, the engine speed MNXn-1 at the end of the combustion stroke when the exhaust valve of the # n-1 cylinder opens decreases, and the corrected differential speed change DDNEAn-1 becomes a negative value, resulting in a negative misfire. Indicates a value equal to or lower than the judgment level -LVLMIS. Then, for the engine speed MNXn-1, the next cylinder #
n increases MNXn in the next section, and
The correction difference rotation change DDNEAn for the #n cylinder is a positive value, and indicates a value equal to or higher than the misfire determination level LVLMIS. Therefore, this causes the misfire diagnosis target cylinder # n-1
It is possible to appropriately diagnose misfires on the vehicle.

Also, the correction difference rotation DELNA of the front and rear cylinders
, That is, cylinder #n for misfire diagnosis one combustion stroke before
Using the corrected differential rotation change DDNEAn-1 of -1 and the corrected differential rotation change DDNEAn of the current combustion stroke cylinder #n, the corrected differential rotation change DDNE for the relevant misfire diagnosis target cylinder # n-1.
After An-1 becomes equal to or less than the minus misfire determination level -LVLMIS, the correction difference rotation change DDNEAn of the next cylinder, that is, the current combustion stroke cylinder #n, becomes the misfire determination level LVLMI.
Only when it becomes S or more, the relevant cylinder # n-1 for misfire diagnosis
Is determined to have misfired, the difference rotation (correction difference rotation DE) whose level changes according to the engine operating state at the time of misfire occurrence
LNA), the misfire can be accurately detected by eliminating the influence of intermittent rotation fluctuation such as snatching.

Then, when the misfire diagnosis subroutine ends and the process proceeds to step S24 of the misfire diagnosis routine (see FIG. 4), the processing after step S24 prepares for the next misfire diagnosis, and the statistical processing is performed for the cylinder # n-1. A difference rotation correction value AVEDNXn-1 is calculated.

In step S24, referring to the misfire flag FLGMISn-1 of the current cylinder # n-1 for misfire diagnosis,
FLGMISn-1 = 0, cylinder # n-1 to be misfired this time
If no misfire has occurred, the process proceeds to step S25. Then, the corresponding cylinder # n-1 stored in the work area
AVE of all cylinders at the time of execution of the previous routine in which the differential rotation of DELNEn-1 and the differential rotation of all cylinders are weighted and averaged.
By calculating the difference (DELNEn-1-AVEDNn-1) from DNn-1 and comparing it with the lower limit value MINDN and the upper limit value MAXDN, each of the protrusions 31a, 3 of the crank rotor 31 is calculated.
It is determined whether or not rotational variations due to manufacturing tolerances such as the position and shape of the 1b and 31c, and tolerances for mounting the crank angle sensor 32 to the engine 1 are included in the differential rotation DELNEn-1. .

Then, MINDN <DELNEn-1-A
If VEDNn-1 <MAXDN, the differential rotation DELNEn-1 of the relevant cylinder # n-1 and the average differential weight ADVDNn of all the cylinder differential rotations
When the difference from -1 is within a set range determined by the upper and lower limits MAXDN and MINDN, the differential rotation DELNE is caused by an error relating to the crank rotor 31 and the crank angle sensor 32.
Since it is determined that n-1 has fluctuated, the process proceeds to step S26, where the weighted average of the all cylinder difference rotation weighted average value AVEDNn-1 up to the previous time and the correction difference rotation DELNAn-1 of the corresponding cylinder # n-1 is calculated. To calculate the current all-cylinder differential rotation weighted average value AVEDNn (AVEDNn ← (3/4) × AVEDNn−1).
+ (1/4) * DELNAn-1).

Next, at step S27, the cylinder # n-1 (four cylinder engine in the present embodiment, the cylinder # n-1 and the cylinder # n-5 are the same cylinder) calculated in the present routine five times before. The differential rotation correction value AVEDNXn-5 for the cylinder # n-1 stored in the work area is stored in the work area.
The above-described step S26 is performed when ELNEn-1 and the previous routine are executed.
AVENn-1
Weighted average using the difference (DELNEn-1 -AVEDNn-1) with respect to the difference rotation correction value AVE for the corresponding cylinder # n-1.
DNXn-1 (AVEDNXn-1 ← (7/8) × AV
EDNXn-5 + (1/8) × (DELNEn-1-AVE
DNn-1)), exits the routine.

Then, the next time (when this routine is executed three times later), the corrected differential rotation DE for the corresponding cylinder # n-1 (#n) will be described.
When calculating LNAn-1 (DELNAn), the difference rotation correction value AVEDNXn-1 (AVEDNXn-4) is used, and the projections 31a,
Manufacturing tolerances such as the position and shape of 31b and 31c,
Influences such as a permissible error in the mounting position of the crank angle sensor 32 to the engine 1 are eliminated.

On the other hand, in step S24, the FLGMIS
If n-1 = 1 and a misfire has occurred in the relevant cylinder # n-1, the process proceeds from step S24 to step S28, in which the weighted average value of all cylinder difference rotations AVEDNn-1 at the time of execution of the previous routine is used as the current all cylinder difference rotation. The weighted average value is AVEDnn (AVED
Nn ← AVEDNn-1), and in the following step S29, the differential rotation correction value AVEDNXn-5 calculated in the present routine five times before.
Is the differential rotation correction value AVEDNn for the cylinder # n-1.
It is set to -1 (AVEDNXn-1 ← AVEDNXn-5), and the routine exits.

In step S25, MIN
When DN ≧ DELNEn−1−AVEDnn−1, it is determined that the influence of the error relating to the crank rotor 31 and the crank angle sensor 32 included in the differential rotation DELNEn−1 is negligible, and DELNEn−1−AVEDnn−1 ≧ MAX
If the rotational fluctuation is large in DN, it is determined that the rotational fluctuation is due to snatching or the like, and similarly, the routine exits through steps S28 and S29.

Then, the cylinder-by-cylinder misfire flag FLGMIS obtained by the above-described misfire diagnosis processing is referred to at the time of setting the fuel injection amount, and when misfire has occurred in the relevant cylinder, the fuel injection amount for the relevant cylinder is immediately increased and corrected.

Next, the fuel injection control process according to the present invention will be described with reference to the flowcharts of FIGS.

The flowcharts shown in FIGS. 6 and 7 show a fuel injection amount setting routine. The fuel injection pulse width Ti # i determines the final fuel injection amount supplied to the engine for each cylinder.
Is set. This fuel injection amount setting routine is executed at predetermined intervals (for example, 180 ° CA), and the routine proceeds to step S4.
At 1, the current engine speed NE and the intake air amount sensor 24
The basic fuel injection pulse width Tp that determines the basic fuel injection amount is calculated from the intake air amount Q based on the output signal from the controller (Tp ← K × Q / NE; K is an injector characteristic correction constant).

Next, in step S42, the operation state of the starter switch 37 is determined, and the starter switch 37 is turned on.
In the case of (during cranking), the process proceeds to step S43, in which the starting increase coefficient KST is set by the set value CKST (where CKST> 1.0), and then the process proceeds to step S45. When the starter switch 37 is OFF, In step S44, the starting increase coefficient KST is set to 1.0 (no correction), and then in step S45.
Proceed to. The start increase coefficient KST is used to increase the fuel only during cranking during operation of the starter motor to ensure startability.

In step S45, the mixture ratio allocation coefficient KMR is set by referring to the table based on the engine speed NE and the basic fuel injection pulse width Tp. The mixture ratio allocation coefficient KMR is used to obtain an appropriate air-fuel ratio for each region specified by the engine speed NE and the basic fuel injection pulse width Tp representing the engine load. The optimal value is stored.

Next, at step S46, a full increase coefficient KFULL is set based on the throttle opening THV, the basic fuel injection pulse width Tp, and the engine speed NE by the throttle opening sensor 25a. This full increase coefficient KFULL is
When the throttle opening THV is in the fully opened state or when the engine load determined by the basic fuel injection pulse width Tp is in a high load state, it is set by a table using the engine speed NE as a parameter. Output performance in the area where Note that KFULL is not used in areas other than the throttle fully open area and the high load area.
= 0.

In the following step S47, the cooling water temperature sensor 2
The water temperature increase coefficient KTW is set by referring to the table on the basis of the cooling water temperature TW obtained in step 8. This water temperature increase coefficient KTW
Defines the fuel increase rate for ensuring the drivability when the engine is cold. As shown in step S47, the lower the cooling water temperature TW, the larger the water temperature increase coefficient KTW in the table. Stored.

Next, at step S48, a post-start increase coefficient KAS is set. The post-start increase coefficient KAS is used to secure the stability of the engine speed immediately after the start of the engine. As shown in step S48, when the starter switch 37 is ON, the initial value is determined based on the coolant temperature TW. Is set, and after the starter switch 37 is turned off, it is gradually decreased until KAS = 0.

In a succeeding step S49, an after-idle increase coefficient KAI is set. The post-idle increase coefficient KAI is for preventing the backlash when the idle is released.
As shown in the step, the vehicle speed VSP is equal to or lower than the set value (for example, 15 km / h) and the idle switch 25b is
N (throttle valve fully closed) to OFF (throttle valve open)
Is set to the initial value based on the cooling water temperature TW, and then gradually decreased until KAI = 0.

Then, in step S50, the cylinder discrimination /
The fuel injection target cylinder data #i determined in the cylinder by the engine speed calculation routine is read, and step S51 is performed.
Then, a misfire correction coefficient setting subroutine shown in FIG. 8 is executed, and when a misfire occurs in the corresponding cylinder #i, a fuel-specific misfire correction coefficient KMF for the corresponding cylinder #i for correcting the fuel increase and eliminating the misfire. Set #i.

In the misfire correction coefficient setting subroutine, in step S61, the misfire flag FLGM corresponding to the fuel injection target cylinder #i set by the misfire diagnosis processing.
Refer to IS # i. If FLGMIS # i = 1 and misfiring has occurred in the fuel-injection target cylinder #i, the process proceeds to step S62, in which a cylinder-specific misfire for the corresponding cylinder #i for increasing the fuel injection amount in response to misfire The correction coefficient KMF # i is updated by adding a proportional constant PKMF of proportional integration control (PI control) as a set value (KMF # i ← KMF # i + PK).
MF), the cylinder-specific misfire correction coefficient KMF # i is skipped in the plus direction.

In step S63, the cylinder-specific misfire correction coefficient KMF # i is replaced with the cylinder-specific misfire correction coefficient KMF # i.
Upper limit value KMFMA to prevent overcorrection of fuel increase due to fuel
Compare with X. When KMF # i ≦ KMFMAX and the cylinder-specific misfire correction coefficient KMF # i is equal to or smaller than the upper limit KMFMAX, the process proceeds directly to step S52 of the fuel injection amount setting routine. When KMF # i> KMFMAX, the cylinder-specific misfire correction coefficient KM
The cylinder-specific misfire correction coefficient KMF # i is updated by increasing F # i.
Has reached the upper limit KMFMAX, the routine proceeds to step S64, in which the cylinder-specific misfire correction coefficient KMF # i is limited by the upper limit KMFMAX (KMF # i ← KMFMAX), and the routine proceeds to step S52 of the fuel injection amount setting routine.

On the other hand, in step S61, FLGM
If IS # i = 0 and no misfire has occurred in the relevant fuel injection target cylinder #i, the process proceeds to step S65, in which the cylinder-specific misfire correction coefficient KMF # i of the relevant cylinder #i is set to a reference value corresponding to no correction. Compare with 1.0. When KMF # i> 1.0, the routine proceeds to step S66, in which the cylinder-specific misfire correction coefficient KMF # i is gradually decreased by the PI control integration constant IKMF every time the routine for the corresponding cylinder #i is executed (KMF # i ←
KMF # i-IKMF), Step S of the fuel injection amount setting routine
Go to 52.

In step S65, KMF # i ≦
If it is 1.0, the process proceeds to step S67, and the corresponding cylinder #i
, The cylinder-by-cylinder misfire correction coefficient KMF # i for KMF # i ← 1.
If the value is set to 0, the fuel increase correction by the cylinder-specific misfire correction coefficient KMF # i is not performed, and the process proceeds to step S52 of the fuel injection amount setting routine.

Here, the cylinder-specific misfire correction coefficient KMF # i
Are incorporated into the equation for calculating the various increase coefficients COEF in step S52 of the fuel injection amount setting routine, and the various increase coefficients COEF are used to calculate the fuel injection pulse width Ti # i that determines the final fuel injection amount for the corresponding cylinder #i. By incorporating it into the equation (see step S56), when a misfire occurs in the relevant cylinder #i, the fuel injection amount for the relevant cylinder #i is immediately increased and corrected.

That is, when a misfire occurs in the cylinder #i, the misfire flag FLGMIS # i of the cylinder #i is set by the above-described misfire diagnosis processing (FLGMIS # i).
= 1), by setting the misfire flag FLGMIS # i, as shown in the timing chart of FIG. 15, each time a misfire of the corresponding cylinder #i occurs, the cylinder-specific misfire correction coefficient KMF # i for the corresponding cylinder #i is set to a proportional constant. Skipped in the plus direction by PKMF. As a result, the fuel injection amount for the relevant cylinder #i is increased and corrected, and the misfire of the relevant cylinder #i is eliminated. At this time, the cylinder-specific misfire correction coefficient K
Since MF # i is limited by the upper limit value KMFMAX as described above, unnecessary fuel increase correction by the cylinder-specific misfire correction coefficient KMF # i is prevented, and controllability deteriorates due to over-rich air-fuel ratio. In addition, it is possible to prevent the elimination of the misfire from being hindered.

Then, when the misfire of the corresponding cylinder #i is eliminated, the misfire flag FLGMIS # i is cleared.
After the elimination of the misfire of the relevant cylinder #i, the cylinder-specific misfire correction coefficient KMF # i for the relevant cylinder #i is gradually reduced by the integration constant IKMF. As a result, after the misfire has been eliminated, the misfire correction coefficient K for each cylinder of the fuel injection amount for the corresponding cylinder #i.
The amount of fuel increase by MF # i is gradually reduced, and deterioration of controllability due to a sudden change in the air-fuel ratio is prevented.

Further, if the non-misfire state of the cylinder #i continues, the cylinder-specific misfire correction coefficient KMF # i eventually becomes KMF # i.
≦ 1.0, and then KMF until misfire occurs again
# i = 1.0, and the misfire correction coefficient KMF # i for each cylinder
The fuel increase correction due to is stopped.

Then, upon completion of the misfire correction coefficient setting subroutine, step S52 of the fuel injection amount setting routine is performed.
(See FIG. 7), various increase coefficients COEF are calculated from the above coefficients (COEF ← KST × KMF # i × (1 + K
MR + KFULL + KTW + KAS + KAI)), at step S53,
The air-fuel ratio feedback correction coefficient α is read.

This air-fuel ratio feedback correction coefficient α is
As is well known, it is set in accordance with the output signal of the O2 sensor 29, and when the O2 sensor 29 is inactive immediately after the engine is started, α is fixed to 1.0. Therefore, at this time,
The air-fuel ratio open loop control is performed.

After the O2 sensor 29 is activated, the air-fuel ratio feedback correction coefficient α is determined by the output signal (output voltage) of the O2 sensor 29 and the stoichiometric air (the stoichiometric air-fuel ratio) except during transients such as acceleration and deceleration. It is set by PI control or the like based on the result of comparison with the corresponding slice level. Thus, the air-fuel ratio is controlled so as to converge to stoichiometry.

In the following step S54, the engine speed NE
The air-fuel ratio learning value KLR is searched by referring to the air-fuel ratio learning value table composed of a series of addresses in the backup RAM 44 based on the basic fuel injection pulse width Tp and the air-fuel ratio learning correction coefficient KBLRC is set by interpolation calculation. , Step S5
Proceed to 5. The air-fuel ratio learning value KLR, which is the basis of the air-fuel ratio learning correction coefficient KBLRC, is, as is well known, the engine speed NE.
And the basic fuel injection pulse width Tp representing the engine load is learned in accordance with the deviation of the average value of the air-fuel ratio feedback correction coefficient α from the reference value in a predetermined cycle with respect to the reference value. This is for correcting deterioration with time of the intake air amount measuring system such as 24 and the fuel supply system such as the injector 11.

In step S55, the voltage correction pulse width Ts for compensating the invalid injection time of the injector 11 is set by referring to the table based on the battery voltage VB.

In the subsequent step S56, the basic fuel injection pulse width Tp is multiplied by the various increase coefficients COEF and the air-fuel ratio feedback correction coefficient α to perform air-fuel ratio correction, and is further multiplied by the air-fuel ratio learning correction coefficient KBLRC to perform learning. Amend,
Further, the voltage is corrected by adding the voltage correction pulse width Ts,
The fuel injection pulse width Ti # i for each cylinder that determines the final fuel injection amount to be supplied to the relevant fuel injection target cylinder #i is calculated (Ti # i ← Tp × COEF × α × KBRRC + Ts).

Then, in step S57, the fuel injection pulse width Ti # i is set in the fuel injection timer corresponding to the relevant cylinder #i, and the process exits.

As a result, the fuel injection timer is started at a predetermined timing, and a drive pulse signal of the fuel injection pulse width Ti # i is output to the injector 11 of the corresponding cylinder #i for fuel injection. Metered fuel is injected. Here, misfire diagnosis is performed for each cylinder, and when a misfire occurs, the cylinder-specific misfire correction coefficient KMF # i for the relevant cylinder #i is increased by the set value based on the proportionality constant PKMF, and fuel injection for the relevant cylinder #i is performed. Immediately after the engine is started, the base fuel injection amount is set to the limit air-fuel ratio near the stoichiometric air-fuel ratio corresponding to the permissible limit of engine rotation fluctuation due to the misfire in order to eliminate the misfire of the corresponding cylinder #i by immediately increasing and correcting the amount. Can be set accordingly.

That is, since the fuel injection amount for the corresponding cylinder #i at the time of misfire is increased by the cylinder-by-cylinder misfire correction coefficient KMF # i, the post-start increase coefficient KAS for ensuring the stability immediately after the engine is started, and The water temperature increase coefficient KTW can be set relatively low, and it is necessary to set the post-start increase coefficient KAS and the water temperature increase coefficient KTW to be high so as to increase the fuel increase rate in consideration of the variation of each engine. Disappears.

Therefore, even in the air-fuel ratio open loop control immediately after the engine is started, the limit air-fuel ratio in the vicinity of the stoichiometric air-fuel ratio which is effective for improving the exhaust emission while preventing the rotation fluctuation due to the misfire and improving the drivability. , The exhaust emission and the drivability can be improved at the same time.

Next, a second embodiment will be described with reference to FIGS.

In this embodiment, only the misfire correction coefficient setting subroutine in the first embodiment described above is changed, and instead of the process shown in FIG. 8 of the first embodiment, FIG.
Is adopted.

In the present embodiment, misfire diagnosis is performed for each cylinder, and when misfire is detected, the cylinder-by-cylinder misfire correction coefficient KMF # i of the corresponding cylinder #i is determined by the first proportional constant PKMF.
And the fuel injection amount for the corresponding cylinder #i is immediately increased and corrected to eliminate the misfire of the corresponding cylinder #i, and at this time, the other cylinders # i + 1, # i + 2, # i + Misfire correction coefficient for each cylinder KMF # (i + 1), KMF # (i + 2), KMF (i + 3)
Is reduced by a lean correction proportional constant PDKMF corresponding to a second set value obtained by equally dividing the proportional constant PKMF, and the fuel injection amount for the other cylinders # i + 1, # i + 2, # i + 3 is reduced as described above. By performing the equal distribution reduction correction corresponding to the fuel increase correction of the corresponding cylinder #i, the macro total air-fuel ratio (total cylinder-by-cylinder total air-fuel ratio) change due to the fuel correction at the time of misfire is prevented.
Prevent deterioration of air-fuel ratio controllability.

The lean correction proportional constant PDKMF is
If the number of cylinders of engine 1 is m, PDKMF = PKMF /
In this embodiment, PDKMF = PKMF / 3 because the engine is a four-cylinder engine.

After the misfire has been eliminated, the cylinder-specific misfire correction coefficient KMF # i for the relevant cylinder #i is changed at every predetermined cycle given by the routine execution cycle.
F gradually decreases until KMF # i = 0 without correction, and the fuel increase correction by the misfire correction coefficient KMF # i is gradually eliminated. At this time, the other cylinder # i +
Regarding 1, # i + 2 and # i + 3, cylinder-specific misfire correction coefficients KMF # (i + 1), KMF # (i + 2), and KMF (i
+3) is gradually increased by the rich correction integration constant IUKMF obtained by equally distributing the integration constant IKMF in the predetermined cycle until no correction is made, and the misfire correction coefficient K
Each cylinder # i + by MF # (i + 1), KMF # (i + 2), KMF (i + 3)
The fuel reduction correction for 1, # i + 2 and # i + 3 is gradually eliminated.

As a result, the misfire correction coefficient KMF for each cylinder is obtained.
#i, KMF # (i + 1), KMF # (i + 2), and KMF (i + 3) suppress sudden changes in the air-fuel ratio when canceling the fuel correction for each cylinder to prevent deterioration of controllability. In addition, the above-described processing after the elimination of the misfire also makes it possible to prevent a change in the total air-fuel ratio on a macro scale and prevent the deterioration of the air-fuel ratio controllability. After the elimination of the misfire, the combustion state of each cylinder is immediately made uniform, and it is possible to prevent engine rotation fluctuation caused by uneven combustion state of each cylinder.

The above-mentioned rich correction integration constant IUKMF is
IUKMF = IKMF / (m-1); m is given by the number of cylinders. In this embodiment, since the four-cylinder engine is used, IUKMF =
IKMF / 3.

The misfire correction coefficient setting subroutine of FIG. 20 will be described. The fuel injection target cylinder data #i is read in step S50 (see FIG. 7) of the above-described fuel injection amount setting routine, and then in step S51, A misfire correction coefficient setting subroutine is executed.

In the misfire correction coefficient setting subroutine of FIG. 20, first, in step S71, the misfire flag FLGMIS # i corresponding to the fuel injection target cylinder #i set by the misfire diagnosis processing is referred to.

If FLGMIS # i = 1 and misfiring has occurred in the fuel injection target cylinder #i, step S7
2 and the cylinder-by-cylinder misfire correction coefficient KMF # i for the corresponding cylinder #i for increasing the fuel injection amount in response to misfire
Is updated by adding the proportionality constant PKMF to (KMF # i ← KMF # i
+ PKMF), the cylinder-specific misfire correction coefficient KMF # i for the corresponding cylinder #i is skipped in the plus direction.

Next, at step S73, the other cylinders # i + 1,
Cylinder-specific misfire correction coefficient KMF # (i + 2) for # i + 2 and # i + 3
1), KMF # (i + 2) and KMF (i + 3) are respectively reduced and updated by a lean correction proportional constant PDKMF corresponding to a value obtained by equally dividing the proportional constant PKMF (KMF # (i + 1)). ← KMF # (i + 1) -P
DKMF, KMF # (i + 2) ← KMF # (i + 2) −PDKMF, KMF # (i + 3) ←
KMF # (i + 3) -PDKMF), and these cylinders # i + 1, # i + 2,
Cylinder misfire correction coefficient KMF # (i + 1), KMF for # i + 3
# (i + 2) and KMF (i + 3) are skipped in the negative direction.

In step S74, the cylinder-specific misfire correction coefficient KMF # i of the cylinder #i is set to an upper limit value KMFMAX for preventing the fuel increase from being excessively corrected by the cylinder-specific misfire correction coefficient KMF # i. Compare. If KMF # i ≦ KMFMAX and the cylinder-specific misfire correction coefficient KMF # i is equal to or smaller than the upper limit KMFMAX, the routine proceeds directly to step S52 (see FIG. 7) of the fuel injection amount setting routine described above. Also, if KMF # i> KMFMAX,
When the cylinder-specific misfire correction coefficient KMF # i reaches the upper limit value KMFMAX by increasing and updating the cylinder-specific misfire correction coefficient KMF # i, the process proceeds to step S75, where the cylinder-specific misfire correction coefficient KMF # i is set to the upper limit. Limited by the value KMFMAX (KMF # i
← KMFMAX), and proceeds to step S52 of the fuel injection amount setting routine.

On the other hand, in step S71, FLGM
If IS # i = 0 and no misfire has occurred in the relevant fuel injection target cylinder #i, the process proceeds to step S76, in which the cylinder-specific misfire correction coefficient KMF # i of the relevant cylinder #i is set to a reference value corresponding to no correction. Compare with 1.0. When KMF # i> 1.0, the routine proceeds to step S77, in which the cylinder-specific misfire correction coefficient KMF # i is gradually decreased by the integration constant IKMF every time the routine for the corresponding cylinder #i is executed (KMF # i ← KMF). # i-I
KMF), and the process proceeds to Step S52 of the fuel injection amount setting routine.

In step S76, KMF # i =
At 1.0, the cylinder-specific misfire correction coefficient KMF for the corresponding cylinder #i
When there is no fuel correction by #i, the routine directly exits the routine and proceeds to step S52 of the fuel injection amount setting routine to continue the state without fuel correction by the cylinder-specific misfire correction coefficient KMF # i.

Therefore, after the misfire of the relevant cylinder #i is eliminated, the cylinder-specific misfire correction coefficient KMF # i for the relevant cylinder #i
Is gradually decreased by the integral value IKMF until KMF # i = 0 without correction, and the fuel increase correction for the corresponding cylinder #i by the cylinder-specific misfire correction coefficient KMF # i is gradually eliminated. You.

On the other hand, when KMF # i <1.0, the other cylinders # i + 1, # i + 2, #i are executed when the previous routine is executed.
After the misfire has occurred in any of +3 and the misfire correction coefficient KMF # i for the corresponding cylinder #i has been skipped in the negative direction by the processing in step S73, the process proceeds to step S78.

In step S78, the cylinder-by-cylinder misfire correction coefficient KMF # i of the corresponding cylinder #i is replaced by the cylinder-by-cylinder misfire correction coefficient K
Lower limit K for preventing overcorrection of fuel loss by MF # i
Compare with MFMIN. That is, the lower limit value KMFMIN is set corresponding to the upper limit value KMFMAX. When the cylinder-specific misfire correction coefficient of the misfiring cylinder is limited by the upper limit value KMFMAX, the lower limit value KMFMIN is set corresponding to the cylinder type of the other cylinder. By limiting the misfire correction coefficient by the lower limit value KMFMIN, a macroscopic change in the total air-fuel ratio is prevented in this case as well.
Therefore, the lower limit value KMFMIN is KMFMIN = 1.0− (KMFMA
X−1.0) / (m−1). In the present embodiment, m = 4 for a four-cylinder engine, and KMFMI
N = 1.0− (KMFMAX−1.0) / 3.

If KMF # i <KMFMIN and the cylinder-specific misfire correction coefficient KMF # i has reached the lower limit value KMFMIN due to the skipping of the cylinder-specific misfire correction coefficient KMF # i in the negative direction. The process proceeds to S79, in which the cylinder-specific misfire correction coefficient KMF # i is limited by the lower limit KMFMIN (KMF # i ← KMFMIN), and the process proceeds to step S52 of the fuel injection amount setting routine.

If KMF # i ≧ KMFMIN and the cylinder-specific misfire correction coefficient KMF # i is equal to or greater than the lower limit KMFMIN, the routine proceeds to step S80, where the cylinder-specific misfire correction coefficient KMF # i is calculated by subtracting the integral constant IKMF from the above-mentioned integral constant IKMF. By the rich correction integration constant IUKMF corresponding to the equally distributed value, the value is gradually increased every time the routine for the corresponding cylinder #i is executed (KMF # i ← KMF # i + IUKMF),
The routine proceeds to step S52 of the fuel injection amount setting routine.

Therefore, the other cylinders # i + 1, # i + 2, # i + 3
After the misfire correction coefficient KMF # i for the relevant cylinder #i is skipped in the negative direction by the lean correction proportional constant PDKMF, the cylinder-specific misfire correction coefficient KMF # for the relevant cylinder #i i is gradually increased by the rich correction integrated value IUKMF until KMF # i = 0 without correction, and the fuel loss correction for the corresponding cylinder #i by the cylinder-specific misfire correction coefficient KMF # i is gradually performed. Will be resolved.

Then, the cylinder-specific misfire correction coefficient KMF # i
Is the step S52 of the fuel injection amount setting routine shown in FIG.
Is incorporated into the equation for calculating the increase coefficient COEF,
By incorporating these various increase coefficients COEF into the equation for calculating the fuel injection pulse width Ti # i that determines the final fuel injection amount for the corresponding cylinder #i (see step S56 in FIG. 7),
When a misfire occurs in the relevant cylinder #i, the fuel injection amount for the relevant cylinder #i is immediately increased and corrected.

That is, when a misfire occurs in the cylinder #i, the misfire flag FLGMIS # i of the cylinder #i is set by the above-described misfire diagnosis processing (FLGMIS # i).
= 1), by setting the misfire flag FLGMIS # i, as shown in the timing chart of FIG. 21, the misfiring correction coefficient KMF # i for each cylinder #i for each cylinder #i is proportional to the proportional constant every time a misfire occurs in the corresponding cylinder #i. Skipped in the plus direction by PKMF. As a result, the fuel injection amount for the relevant cylinder #i is increased and corrected, and the misfire of the relevant cylinder #i is eliminated. At this time, the cylinder-specific misfire correction coefficient K
Since MF # i is limited by the upper limit value KMFMAX as described above, unnecessary fuel increase correction by the cylinder-specific misfire correction coefficient KMF # i is prevented, and controllability deteriorates due to over-rich air-fuel ratio. In addition, it is possible to prevent the elimination of the misfire from being hindered.

At this time, the other cylinders # i + 1, # i +
2, # i + 3, these cylinders # i + 1, # i +
2, cylinder-specific misfire correction coefficient KMF # (i + 1) for # i + 3,
KMF # (i + 2) and KMF (i + 3) are obtained by a lean correction proportional constant PDKMF corresponding to a value obtained by equally dividing the proportional constant PKMF.
Each is skipped in the minus direction. This allows
The fuel injection amounts for the other cylinders # i + 1, # i + 2, # i + 3 are equally-distributed and reduced in correspondence with the fuel increase correction for the corresponding cylinder #i. As a result, a change in the total air-fuel ratio on a macro scale due to the fuel correction at the time of misfire is prevented, and deterioration of the air-fuel ratio controllability is prevented. Then, the cylinder-by-cylinder misfire correction coefficient skipped in the minus direction is limited by the lower limit value KMFMIN, so that excessive correction of fuel loss more than necessary is prevented.

After the misfire of the corresponding cylinder #i is resolved, the misfire flag FLGMIS # i of the corresponding cylinder #i is cleared.
As a result, the cylinder-by-cylinder misfire correction coefficient KMF # i for the corresponding cylinder #i is gradually reduced by the integration constant IKMF. As a result, after the misfire of the corresponding cylinder #i is eliminated, the cylinder-by-cylinder misfire correction coefficient KMF of the fuel injection amount for the corresponding cylinder #i
The amount of fuel increase due to #i is gradually reduced, and deterioration in controllability due to a sudden change in the air-fuel ratio is prevented.

At this time, the other cylinders # i + 1, # i +
2, # i + 3, cylinder-specific misfire correction coefficients KMF # (i + 1), KMF # (i + 2), and KMF (i + 3) for these cylinders are given by
Rich correction integration constant IU obtained by equally dividing the above integration constant IKMF
Gradually increased by KMF. As a result, for each of the cylinders # i + 1, # i + 2, and # i + 3, the cylinder-specific misfire correction coefficients KMF # (i + 1), KMF # (i + 2), and KMF (i +3)
The fuel loss correction due to the above is gradually eliminated, and deterioration in controllability due to a sudden change in the air-fuel ratio is prevented. Further, these cylinder-specific misfire correction coefficients KMF # (i + 1), KMF # (i + 2), and KMF (i + 3)
Is gradually increased, the rich correction integration constant IUKMF obtained by equally dividing the integration constant IKMF is used. In this case as well, a macroscopic change in the total air-fuel ratio is prevented, and the deterioration of the air-fuel ratio controllability is prevented. It becomes possible.
After the elimination of the misfire, the combustion state of each cylinder is immediately made uniform, and it is possible to prevent engine rotation fluctuation caused by uneven combustion state of each cylinder.

Further, if the non-misfire state of each cylinder continues, the misfire correction coefficient KMF # i, KMF # (i +
1), KMF # (i + 2), KMF (i + 3) become 1.0, and thereafter,
Until a misfire occurs, the value is maintained at 1.0, and the misfire correction coefficient KMF # i, KMF # (i + 1), KMF # (i + 2), KMF (i
+3), the fuel correction for each cylinder #i, # i + 1, # i + 2, # i + 3 is stopped.

In each of the above embodiments, the engine speed is used as the engine speed state quantity, and the misfire is detected based on the engine speed change amount for each predetermined cycle. Alternatively, the engine rotation cycle, the engine rotation angular velocity, or the engine rotation angular acceleration may be used. Further, the present invention is not limited to this, and is not limited to misfire detection based on the amount of engine rotation. For example, an ion current between discharge electrodes of a spark plug may be detected, and misfire may be detected based on the ion current. Good.

[0148]

According to the first aspect of the present invention, when the misfire is detected, the fuel injection amount is increased and corrected. Therefore, even immediately after the engine is started, the base fuel injection amount is set to the allowable limit of the engine rotation fluctuation due to the misfire. It can be set corresponding to the limit air-fuel ratio near the corresponding stoichiometric air-fuel ratio. Accordingly, exhaust emission can be improved by operating at this limit air-fuel ratio, and at the time of misfire, misfire is immediately eliminated by fuel increase correction, and engine rotation fluctuation due to misfire is prevented, thereby improving drivability. Immediately after the engine is started, it is possible to achieve both improvement in exhaust emission and improvement in drivability.

According to the second aspect of the present invention, in addition to the effects of the first aspect of the present invention, when a misfire is detected,
The set value is added to the misfire correction coefficient, and the fuel injection amount is immediately increased and corrected by the misfire correction coefficient, whereby the misfire is eliminated.After the misfire is eliminated, the misfire correction coefficient is gradually increased. Therefore, the amount of fuel increase by the misfire correction coefficient of the fuel injection amount is gradually reduced, and it is possible to prevent deterioration in controllability due to a sudden change in the air-fuel ratio when the correction is canceled.

In this case, according to the third aspect of the present invention, the misfire correction coefficient is limited by the upper limit value. Correction is prevented, and there is an effect that deterioration of controllability due to over-rich air-fuel ratio and obstruction of elimination of misfire are prevented.

According to the fourth aspect of the present invention, the misfire is detected for each cylinder. When the misfire is detected, the first set value is added to the cylinder-specific misfire correction coefficient for the misfire cylinder, and the misfire is detected. The fuel injection amount is immediately increased by the time correction coefficient, whereby the misfire of the corresponding misfiring cylinder is eliminated, and at this time, the misfiring correction coefficient for each cylinder is reduced by the second set value for the other cylinders. Since the fuel injection amount is set and corrected to be reduced, in addition to the effect of the invention described in claim 1,
Immediately after the misfire is eliminated, the combustion state of each cylinder becomes uniform, and fluctuations in the engine rotation caused by the uneven combustion state of each cylinder can be prevented. After the misfire is resolved, the cylinder-by-cylinder misfire correction coefficient for the corresponding misfire cylinder is gradually reduced until the fuel injection amount is not corrected, and the fuel increase by the misfire correction coefficient of the fuel injection amount is reduced. For other cylinders, the cylinder-by-cylinder misfire correction coefficient for these cylinders is gradually decreased until no correction is made, and the fuel loss correction by the cylinder-by-cylinder misfire correction coefficient is gradually reduced. Therefore, it is possible to prevent deterioration of controllability due to a sudden change in the air-fuel ratio at the time of cancellation of correction for all cylinders.

In this case, according to the fifth aspect of the present invention, the cylinder-specific misfire correction coefficient is limited by an upper limit value and a lower limit value. Unnecessary overcorrection of the fuel injection amount by the different misfire correction coefficient can be prevented, and the air-fuel ratio controllability can be improved.

According to the sixth aspect of the invention, the second set value is set to a value corresponding to the first set value so that the fuel injection amount to the other cylinders except the misfired cylinder is equally distributed and reduced. Therefore, the fuel injection amounts of the other cylinders are equally distributed and reduced in accordance with the amount of fuel increase by the cylinder-by-cylinder misfire correction coefficient for the relevant misfire cylinder. Therefore, it is possible to prevent a change in the total air-fuel ratio on a macro scale due to the fuel correction at the time of misfire, and to prevent deterioration of the air-fuel ratio controllability.

[Brief description of the drawings]

FIG. 1 is a basic configuration diagram of the present invention.

FIG. 2 is a flowchart of a cylinder discrimination / engine speed calculation routine according to the first embodiment of the present invention;

FIG. 3 is a flowchart of a misfire diagnosis routine according to the first embodiment;

FIG. 4 is a flowchart of a misfire diagnosis routine (continued).

FIG. 5 is a flowchart of a misfire diagnosis subroutine according to the first embodiment;

FIG. 6 is a flowchart of a fuel injection amount setting routine;

FIG. 7 is a flowchart of a fuel injection amount setting routine (continued).

FIG. 8 is a flow chart of a misfire correction coefficient setting subroutine;

FIG. 9 is a time chart showing a relationship among a crank pulse, a cam pulse, a combustion stroke cylinder, an ignition timing, and a fuel injection timing.

FIG. 10 is a timing chart showing the behavior of the differential rotation according to the first embodiment;

FIG. 11 is a timing chart showing the behavior of the correction differential rotation;

FIG. 12 is a timing chart showing the behavior of the change in the corrected differential rotation when a misfire occurs

FIG. 13 is an explanatory diagram of a misfire determination level according to the first embodiment;

FIG. 14 is an explanatory diagram showing a relationship between a correction difference rotation change when a misfire occurs and a misfire determination level;

FIG. 15 is a timing chart showing a misfire flag indicating a misfire state of a corresponding cylinder and a setting state of a cylinder-specific misfire correction coefficient of the corresponding cylinder corresponding thereto;

FIG. 16 is an overall schematic diagram of the engine.

FIG. 17 is a front view of the crank rotor and the crank angle sensor according to the first embodiment;

FIG. 18 is a front view of the cam rotor and the cam angle sensor according to the third embodiment;

FIG. 19 is a circuit diagram of an electronic control system according to the embodiment.

FIG. 20 is a flowchart of a misfire correction coefficient setting subroutine according to the second embodiment of the present invention.

FIG. 21 shows a misfire flag indicating a misfiring state of a corresponding cylinder, a misfiring correction coefficient for each cylinder of the corresponding cylinder corresponding thereto,
And a timing chart showing a setting state of a misfire correction coefficient for each cylinder of another cylinder.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Engine 11 Injector 32 Crank angle sensor 40 Electronic control unit (misfire detection means, fuel injection amount correction means, misfire correction coefficient setting means, fuel injection amount setting means) NE Engine rotation speed (engine rotation state amount) KMF # i cylinder Other misfire correction coefficient (misfire correction coefficient) PKMF Proportional constant (set value, first set value) Tp Basic fuel injection pulse width (basic fuel injection amount) Ti Fuel injection pulse width (fuel injection amount) KMFMAX Upper limit value PDKMF Lean correction proportional constant (second set value) KMFMIN lower limit

Claims (6)

[Claims] A fuel injection control device for an engine, comprising: a misfire detecting means for detecting a misfire; and a fuel injection amount correcting means for increasing and correcting the fuel injection amount when the misfire is detected. 2. A misfire detecting means for detecting misfire, a set value is added to a misfire correction coefficient for fuel increase correction when misfire is detected, and the misfire correction coefficient is gradually reduced when misfire is eliminated. And a fuel injection amount setting means for correcting the basic fuel injection amount set based on the engine operating state with the misfire correction coefficient to set the fuel injection amount. Engine fuel injection control device. 3. The fuel injection control device for an engine according to claim 2, wherein said misfire correction coefficient setting means limits said misfire correction coefficient by an upper limit value. 4. A misfire detecting means for detecting a misfire for each cylinder, and when a misfire is detected, adding a first set value to a cylinder-specific misfire correction coefficient of the corresponding cylinder, and a cylinder-specific misfire of another cylinder. When the second set value is subtracted from the hourly correction coefficient, and the misfire is eliminated, the misfire-time correction coefficient setting means for gradually increasing or decreasing the cylinder-by-cylinder misfire-time correction coefficient until no correction is made; A fuel injection amount setting means for correcting the basic fuel injection amount set based on the state by the cylinder-specific misfire correction coefficient to set a fuel injection amount for each cylinder. apparatus. 5. The fuel injection control device for an engine according to claim 4, wherein said misfire correction coefficient setting means limits said cylinder-specific misfire correction coefficient by an upper limit value and a lower limit value. 6. The system according to claim 1, wherein the second set value is set to a value which, in accordance with the first set value, reduces the fuel injection amount to the other cylinders except for the corresponding misfired cylinder. Item 5. An engine fuel injection control device according to item 4.