JPH0751914B2 - Cylinder combustion state identification method for each engine - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a cylinder-by-cylinder combustion state determination method for determining the combustion state of each cylinder from the amount of movement change between cylinders.

[Prior Art] In general, it is ideal that combustion in a multi-cylinder engine is performed through the same process every cycle in order to obtain a stable output. However, in a multi-cylinder engine, the intake pipe has a complicated shape, Non-uniform intake distribution ratio due to intake interference between the cylinders, slight difference in combustion temperature between cylinders due to cooling path, manufacturing chamber volume of each cylinder, manufacturing variation in piston shape, injector manufacturing error, etc. Due to a synergistic effect such as a slight variation in the air-fuel ratio of each cylinder caused by a difference in the fuel injection amount due to, the variation in combustion is likely to occur.

If the variation in combustion in each cylinder can be controlled almost uniformly, the engine can be smoothly rotated. In particular, in the idle speed region where there is a tendency to keep the rotation speed low due to demands such as low fuel consumption and quietness, it is stable to know the combustion state of each cylinder because combustion fluctuation has a great influence on the rotation fluctuation. It is extremely important in obtaining idle rotation.

Therefore, various means for grasping the combustion state of each cylinder have been conventionally developed. For example, in Japanese Unexamined Patent Publication No. 59-82534, each cylinder #i (for four cylinders, i = 1 to 4) A differential rotation ΔN # i (ΔN # i = NH # i-NL # i, which is the difference between the instantaneous engine speed NL # i before the combustion stroke and the instantaneous engine speed NH # i after the combustion stroke. ) Is obtained for each cylinder, and then the average value ΔNA of all cylinders of the differential rotation ΔN # i of each cylinder #i is compared with the differential rotation ΔN # i of each cylinder #i to determine each cylinder #i. The differential rotation speed ΔN # i is controlled to be equal to the average rotation speed difference ΔNA of all cylinders to obtain smooth idle rotation.

[Problems to be Solved by the Invention] In this prior art, paying attention to the correlation between the combustion state of each cylinder and the differential rotation ΔN # i, the average differential rotation ΔNA of all cylinders is used as a reference. Then, the combustion state of the cylinder #i is attempted to be estimated from the deviation from this reference, but the combustion of each cylinder is performed in the ignition order (for example, # 1 → # 3 → # 2).
→ Since it is continuously repeated according to # 4), even if control is performed to bring the differential rotation ΔN # i of each cylinder #i closer to the average differential rotation ΔNA of all cylinders, Between the differential rotation ΔN # i−1 and the differential rotation ΔN # i of the current combustion stroke cylinder #i, or between the differential rotation ΔN # i of the current combustion stroke cylinder #i.
If there is a difference between the differential rotation speed ΔN # i + 1 of the cylinder # i + 1 after one combustion stroke and the differential rotation speed ΔN # i + 1, the engine rotation tends to be uneven and the stability is significantly impaired.

Further, since the reference value is the average differential rotation of all cylinders ΔNA, the average differential rotation of all cylinders ΔNA itself is likely to always fluctuate depending on the combustion state. The combustion state of each cylinder estimated based on the average differential rotation ΔNA includes the combustion state factors of the other cylinders when the average differential rotation ΔNA of all cylinders is set. It becomes difficult to grasp it accurately.

[Object of the Invention] The present invention has been made in view of the above circumstances, and it is possible to accurately grasp the combustion state of another cylinder without including the combustion state factors of the other cylinders. It is an object of the present invention to provide a cylinder-by-cylinder combustion state determination method capable of continuously grasping the combustion state and obtaining smooth engine rotation.

[Means for Solving the Problems] In order to achieve the above object, the first cylinder combustion state determination method for an engine according to the present invention is an engine rotation state in a section in which no work is performed by combustion of a cylinder before one combustion stroke. Quantity and
2 combustion stroke From the difference between the average value of the engine rotation state quantity of the section not working by the combustion of the cylinder before combustion and the engine rotation state quantity of the section not working by the combustion of the current combustion stroke cylinder, The combustion state of the front cylinder is determined.

Further, in order to achieve the above object, a second engine combustion state discriminating method for each cylinder of the present invention is such that an engine rotation state quantity of a section in which no work is performed due to combustion of a cylinder before the present combustion stroke and one combustion stroke before. From the difference between the average value of the engine rotation state in the section where the work is not performed due to combustion of the cylinder and the engine rotation state amount of the section where the work is not performed due to the combustion of one cycle before the cylinder after one combustion stroke, the current combustion stroke is calculated. The combustion state of the cylinder is determined.

[Operation] According to the above-described first cylinder-by-cylinder combustion state determination method, the combustion state of the cylinder before the first combustion stroke indicates the amount of rotation state of the engine in a section where there is no work due to combustion of the cylinders before and after the combustion stroke. Since the average value is used as a reference, the combustion state can be continuously grasped in the ignition order, and smooth engine rotation can be obtained.

Further, according to the above-mentioned second cylinder-by-cylinder combustion state determination method, the engine rotation state quantity and one combustion stroke in the section where the combustion state of the current combustion stroke cylinder is not working due to the combustion of the combustion stroke cylinders before and after it Since it is judged based on the average value of the engine rotation state quantity of the section where there is no work due to combustion of one cycle before of the rear cylinder, the combustion state can be continuously grasped in the ignition order as a result. You can get smooth engine rotation.

Embodiments of the Invention Embodiments of the present invention will be described below with reference to the drawings.

1 to 9 show a first embodiment of the present invention, FIG. 1 is a flowchart showing a cylinder-by-cylinder combustion state determining procedure, FIG. 2 is a flowchart showing a cylinder-by-cylinder fuel injection pulse width setting procedure, and FIG. FIG. 4 is a conceptual diagram showing a basic cylinder-by-cylinder combustion state determination method, FIG. 4 is a conceptual diagram showing a dead zone region, FIG. 5 is a schematic diagram of an engine control system, and FIG. 6 is a front view of a crank rotor and a crank angle sensor. FIG. 7 is a front view of the cam rotor and the cam angle sensor, FIG. 8 is a time chart showing cylinder pressure fluctuation, crank pulse, cam pulse, and engine rotation fluctuation. FIG. 9 is a combustion state comparison value for each cylinder. 6 is a time chart showing a cylinder-specific average combustion state determination value and a cylinder-specific invalid injection pulse width.

(Structure) Reference numeral 1 in FIG. 5 denotes an engine, and in the drawing, a 4-cylinder horizontally opposed engine is shown. An intake manifold 3 is communicated with an intake port 2a formed in a cylinder head 2 of the engine 1, a throttle chamber 5 is communicated upstream of the intake manifold 3 through an air chamber 4, and an intake air is upstream of the throttle chamber 5. An air cleaner 7 is attached via a pipe 6.

An intake air amount sensor (a hot wire type air flow meter in the figure) 8 is provided immediately downstream of the air cleaner 7 in the intake pipe 6, and a throttle valve 5a provided in the throttle chamber 5 is provided with a throttle valve 5a. An opening sensor 9a and an idle switch 9b that detects the fully closed throttle valve are connected in series.

A multi-point injector (hereinafter abbreviated as “MPI”) 10 is arranged immediately upstream of each intake port 2a of each cylinder of the intake manifold 3. Also, reference numeral 11
Is a fuel tank, and 12 is a fuel pump for supplying fuel to the MPI 10.

A crank rotor 15 is attached to the crank shaft 13 of the engine 1, and the outer periphery of the crank rotor 15 is
A crank angle sensor 16 including an electromagnetic pickup or the like for detecting a protrusion (or a slit) corresponding to a predetermined crank angle is provided as a pair, and the crankshaft is further provided.
Cam rotor 17 on camshaft 14 that rotates 1/2 of 13
The cam angle sensor 18 is attached to the outer circumference of the cam rotor 17.
Are opposite to each other.

As shown in FIG. 6, protrusions (or slits) 15a, 15b, 15c are formed on the outer circumference of the crank rotor 15. These protrusions 15a, 15b, 15c are located before the compression top dead center (B
TDC) formed at the positions of θ1, θ2, θ3, and the protrusions 15a, 1
From the transit time between 5b to the period f1.2 (where f = 1 / ω
ω: angular velocity) is calculated, and the period f2.3 is calculated from the passage time between the protrusions 15b and 15c. Furthermore, the projection 15b indicates a reference crank angle when setting the ignition timing.

In general, the ignition timing during idle operation is around 20 ° C BTDC, and even if ignition is performed at this crank angle, the combustion pressure does not rise sharply up to about 10 ° A.

Further, as shown in FIG. 8, in the embodiment, the opening timing of the exhaust valve of each cylinder is set to be slightly retarded from the ignition reference crank angle BTDCθ2 of the next combustion stroke cylinder.
Generally, the combustion pressure immediately after the exhaust valve is opened is drastically reduced, so that the crank angle BTDCθ3 has almost no effect on the combustion pressure.

Therefore, the crank angle θ3 of the protrusion 15c should be BTDC10 ° C.
If it is set to the advance side from A, the section between the crank angles BTDCθ2, θ3 of the protrusions 15b, 15c is hardly affected by the combustion between the cylinders, that is, the work by the combustion of the combustion stroke cylinder is performed. It becomes a section that is not done.

In addition, as shown in FIG. 7, cylinder discrimination protrusions (slits) 17a, 17b, 17c are provided on the outer periphery of the cam rotor 17.
Are formed. The protrusion 17a is formed at the position of θ4 after the compression top dead center of the # 3 and # 4 cylinders (ATDC), and the protrusion 17b is composed of three protrusions, and the first protrusion is the compression top dead of the # 1 cylinder. It is formed at the position of the point (ATDC) θ5, and the protrusion 17c is composed of two protrusions, and the first protrusion is # 2.
It is formed at a position of θ6 after compression top dead center (ATDC) of the cylinder.

In the illustrated embodiment, θ1 = 97 ° C. A, θ2 = 65 ° C. A, θ
3 = 10 ° C, θ4 = 20 ° C, θ5 = 5 ° C, θ6 = 20 ° C
A, θ (2−3) = 55 ° C. A, this arrangement allows the cam angle sensor 18 to rotate by θ5 as shown in FIG.
When the cam pulse of (protrusion 17b) is detected, it can be determined that the crank pulse detected by the crank angle sensor 16 thereafter is a signal indicating the crank angle of the # 3 cylinder.

When the cam pulse of θ4 (protrusion 17a) is detected after the cam pulse of θ5, the crank angle sensor after that is detected.
It can be determined that the crank pulse detected in 16 indicates the crank angle of the # 2 cylinder. Similarly, θ6 (protrusion 17
Crank pulse after detecting cam pulse of c) is # 4
When the cam pulse of θ4 (protrusion 17a) is detected after the cam pulse of θ6, the crank pulse detected thereafter indicates the crank angle of the # 1 cylinder. Can be determined.

Further, after detecting the cam pulse by the cam angle sensor 18, it can be determined that the crank pulse detected by the crank angle sensor 16 indicates the reference crank angle (θ1) of the corresponding cylinder.

The crank angle sensor 16 and the cam angle sensor 18 constitute a crank angle detecting means, and the cam angle sensor 18 alone may constitute the crank angle detecting means by changing the cam pulse pattern.

On the other hand, a cooling water temperature sensor 20 is exposed to a cooling water passage (not shown) forming a riser formed in the intake manifold 3 of the engine 1, and an exhaust pipe 21 communicating with an exhaust port 2b of the cylinder head 2 is provided. The O ₂ sensor 22 is exposed. Reference numeral 23 is a catalytic converter and 24 is a vehicle speed sensor.

(Circuit configuration of control device) On the other hand, reference numeral 31 is a control device including a microcomputer, etc., and a CPU (central processing unit) 3 of the control device 31 is used.
2, ROM33, RAM34, backup RAM (nonvolatile RAM) 3
5, and the I / O interface 36 is connected to each other via the bus line 37, and a predetermined stabilizing voltage is supplied from the constant voltage circuit 38.

The constant voltage circuit 38 has a battery 41 via a control relay 39.
The key switch 40 is turned on and the control relay
When the relay contact of 39 is closed, control power is supplied to each part, and the backup power is supplied to the backup RAM 35 directly connected to the battery 41 even when the key switch 40 is turned off.

Further, to the input port of the I / O interface 36, each sensor 8, 9a, 16, 18, 20, 22, 24, and the idle switch 9b is connected, the positive terminal of the battery 41 is connected, Its terminal voltage is monitored and
The MPI 10 is connected to the output port of the I / O interface 36 via the drive circuit 42.

The ROM 33 stores a control program, fixed data and the like. The fixed data includes an initial set value described later.

Further, the RAM 34 stores data after processing the output signals of the respective sensors, data processed by the CPU 32, and the like. Further, the backup RAM 35 is always energized regardless of the key switch 40, and even if the engine is stopped by turning off the key switch 40, the stored contents are not lost. Is stored.

Further, the CPU 32 calculates a fuel injection pulse width Ti for the MPI 10 for each cylinder based on various data stored in the RAM 34 and the backup RAM 35 according to a control program stored in the ROM 33.

In the control device 31, in addition to the general fuel injection control, the combustion state of each cylinder #i (i = 1 to 4) is individually judged.

The basic concept of this cylinder-by-cylinder combustion state determination method is shown in FIGS.
It will be described with reference to the drawing.

FIG. 3 shows engine speed fluctuations during idle rotation. For example, in the case of a 4-cylinder engine, combustion stroke cylinder #i
18 according to the ignition order (for example, # 1 → # 3 → # 2 → # 4)
Since the combustion stroke cylinders are not overlapped before and after the combustion stroke cylinder #i, the combustion stroke cylinders do not overlap each other before and after the combustion stroke cylinder #i finishes combustion and before the next combustion stroke cylinder # i + 1. There is a section that is not affected by combustion, that is, does not perform work due to so-called combustion.

For example, as shown in FIG. 3, when the engine speeds, which are the momentums of the moments in which the work is not performed due to combustion of the cylinders # 1 to # 4, are N # 1 to N # 4, respectively, a smooth engine is obtained. In order to obtain rotation, it is desirable that the engine speeds N # 1 to N # 4 of the cylinders # 1 to # 4 are all uniform.

In this embodiment, focusing on the fact that there is a very strong correlation between the combustion state of each cylinder #i and the engine speed N # i, the combustion state is determined for each cylinder.

That is, in the combustion state determination method according to this embodiment, the average value of the engine speed N # i of the current combustion stroke cylinder #i and the engine speed N # i-2 of the cylinder # i-2 two combustion strokes before (N # i-2 # I-2 + N # i) / 2 and cylinder # i- before one combustion stroke
The engine speed N # i-1 of 1 is compared, and this comparison value (this is the combustion state determination value here) Is in the minus section, it is determined that the combustion state of the cylinder # i-1 is bad, and when it is on the plus side, the combustion state is good.

Then, based on this determination result (combustion state determination value), the fuel injection pulse width in the next cycle of the cylinder # i-1 or the ignition timing is controlled to set the combustion state of each cylinder # i-1 to the ignition order. In accordance with the above, the correction is performed so as to smoothly change continuously, and finally the combustion state of each cylinder # i-1 is controlled to be substantially uniform, that is, to converge to ΔN # i-1 = 0.

In the following, the comparison value ΔN of each cylinder # i-1 before one combustion stroke
The calculation method of # i-1 will be specifically described.

Further, in order to prevent control hunting and the like, as shown in FIG. 4, a constant dead zone region (+) ΔNU, (−) ΔNL is provided, and the comparison value ΔN # i-1 is the dead zone region (+). △
The controllability is further improved by controlling so as to be within NU, (-) ΔNL.

In the following, a correlation equation between the engine speed N in the section where combustion is not performed and the combustion state of the cylinder, that is, the indicated mean effective pressure Pi is shown.

First, when expressing the state where the engine is rotating, I: Moment of inertia N: Engine speed Ti: Instruction torque Tf: Friction torque If we replace it with pressure, Pi: Mean effective pressure indicated Pf: Friction loss effective pressure.

According to the experiment, the engine speed N # after combustion of each cylinder #i
i and the temporal change ΔT in the section for detecting the engine speed N # i (for example, corresponding to the section (θ2-θ3) in FIG. 8)
As a result of obtaining dN / dt in the above equation (3) based on and, a very strong correlation was obtained.

Therefore, by obtaining the engine speed after combustion of each cylinder, the indicated mean effective pressure Pi, that is, the combustion state can be estimated, and by comparing with the engine speed of the adjacent combustion stroke cylinder, It is possible to judge whether the combustion state of the stroke cylinder is good or bad.

(Operation) Next, a concrete procedure for setting the average combustion state determination value for each cylinder and a procedure for setting the fuel injection pulse width, which are executed by the control device 31, will be described with reference to the flowcharts of FIGS.

As shown in FIG. 1, the cylinder-by-cylinder combustion state determination procedure is as follows. First, in step (hereinafter abbreviated as “S”) 101, the vehicle speed S and the idle switch set based on the output signal of the vehicle speed sensor 24 are set.
The output of 9b is read and it is determined in S102 whether it is idle.

When the vehicle speed S = 0 and the idle switch is ON (throttle valve is fully closed), it is determined to be idle, and the process proceeds to S103. When the vehicle speed S ≠ 0 or the idle switch is OFF (throttle valve is open), it is determined that the vehicle is running. And then I leave the routine.

When it is determined that the engine is idle and the routine proceeds to S103, the current combustion stroke cylinder #i (i =
1,3,2,4) is discriminated. Next, in S104, the crank pulse for detecting BTDCθ2, θ3 output from the crank angle sensor 16 is determined by the interruption of the cam pulse.

Then, in S105, the elapsed time t2,3 between the crank pulses for detecting the BTDC θ2, θ3 determined in S104, and the θ2, θ
The period f2,3 is calculated from the angle of inclination (θ2-θ3) of 3 (f2,3 ← dt2,3 / d (θ2-θ3)).

Next, in S106, the engine speed N # i of the section in which the combustion of the current combustion stroke cylinder #i is not working is calculated from the cycle f2,3 calculated in S105, and is stored in a predetermined address of the RAM 34. (N # i → 60 / f2,3).

Then, in S107, the engine speeds N # i-1 and 2 in the section where no work is performed due to the combustion of the cylinder # i-1 before one combustion stroke, which is set in the routines of the previous time and the time before the last time and stored in the predetermined address of the RAM 34. Engine speed N # i-2 in a section not performing work due to combustion of cylinder # i-2 before the combustion stroke
Read and. As shown in FIG. 8, the ignition order is # 1 → # 3 →
In the case of # 2 → # 4, if the current combustion stroke cylinder #i is # 3, the cylinder # i-1 before one combustion stroke is # 1, and the cylinder # i-2 before two combustion strokes is # 4.

Further, the engine speed N # i in the initial routine
-1, N # i-2 is set to N # i-1 = N # i-2 = N # i.

Next, at S108, the engine speed N # i- in the section where no work is performed due to the combustion of the cylinder # i-1 before the first combustion stroke.
1, the engine speed N # i in the section where the combustion of the current combustion stroke cylinder #i is not working and the cylinder #i before the combustion stroke #i
State of the cylinder # i-1 before one combustion stroke from the difference of the average value (N # i + N # i-2) / 2 from the engine speed N # i-2 in the section where no work is performed due to the combustion of -2 Comparison value △ N #
Find i-1.

ΔN # i-1 ← N # i-1 − {(N # i + N # i-2) / 2} Then, in 109, the combustion state comparison value ΔN obtained in S108 above
# I-1 and the previous average combustion state determination value ΔNA # i-1 of the cylinder # i-1 stored at a predetermined address of RAM 34.
Based on (-1), the average combustion state determination value ΔNA # i-1 of the cylinder # i-1 at this time is calculated from the weighted average of the weight r shown in the following expression, and stored in a predetermined address of the RAM 34. The average combustion state determination value ΔNA # i-1 (-
1) is set to 0.

ΔNA # i-1 ← {(2 ^r -1) x ΔNA # i-1 (-1) + ΔN # i-1} / 2 ^r Then, in S110, it is stored in a predetermined address of the RAM 34. Previous average combustion state determination value ΔNA of the cylinder # i-1
# I-1 (-1) is the current average combustion state determination value ΔNA #
Update with i-1 (△ NA # i-1 (-1) ← △ NA # i-
1) Then, the routine goes out.

As described above, in this embodiment, the combustion state comparison value ΔN # i-1 obtained in S108 is not directly used as the combustion state determination value,
Since the combustion state of the cylinder # i-1 is discriminated by the value subjected to the weighted average processing in S109, a temporary output fluctuation,
It is possible to obtain stable controllability in the fuel injection pulse width setting described below without rapidly responding to the measurement error.

Next, the procedure for setting the fuel injection pulse width for each cylinder will be described with reference to FIG.

Note that the battery 41 may be removed due to vehicle repair or the like, and the data of the invalid injection pulse width TS # i for each cylinder stored in the backup RAM 35 may be corrupted to a meaningless value. Therefore, when initializing, the battery
Normal backup to detect if 41 was removed
Store a fixed constant at a specific address from RAM3, and check whether the value of this constant is broken or not by using a reference value (backup RAM35
If the constant value stored in is not broken, it is determined by comparing it with the constant value and the reference value. If the constant value is broken, the battery 41 should be removed. It is assumed that the invalid injection pulse width TS for each cylinder
#I is TS / n (TS: For example, the total value ΣTs of the invalid injection pulse widths of each cylinder when the battery voltage is 14V, n: the number of cylinders, n = 4 in the case of a 4-cylinder engine, the value of TS / n is stored in ROM33. (Stored in advance) and backup RAM35 above
Reset the constant value stored in the specified address of the reference value. If the constant value of the backup RAM35 is not broken at the next startup, the invalid injection pulse width Ts
#I is not initialized.

Normally, the following processing is repeatedly executed for each cylinder according to the control program.

First, in S201, the engine operating state is read from the output signals of the sensors 8, 9a, 16, 18, 20, 22, 24, and the idle switch 9b, and in S202, the fuel injection cylinder # based on the crank pulse and the cam pulse is read. Determine i.

Next, in S203, it is determined from the output signals of the idle switch 9b and the vehicle speed sensor 24 whether the current operating state is idle.

If the vehicle speed S = 0 and the idle switch is ON (throttle fully closed), it is determined that the vehicle is idle, and the process advances to S204. If the vehicle speed S ≠ 0 or the idle switch is OFF (throttle is open), the vehicle is determined to be in the idle release state S205. Go to.

When proceeding to S205, the fuel injection pulse width Ti of the fuel injection corresponding cylinder #i is obtained by the following equation as usual, and the routine jumps to S213.

Ti ← Tp × α × COEF + Ts where Tp = K · Q / N Tp: Basic fuel injection pulse width (basic injection amount) α: Air-fuel ratio feedback correction coefficient COEF: Various increase correction coefficient Ts: Based on battery terminal voltage VB Voltage correction pulse width for correcting the invalid injection pulse width set by: Q: intake air amount N: engine speed K: correction constant based on theoretical air-fuel ratio, injector injection characteristics, number of cylinders, etc. When the determination is made and the process proceeds to S204, the invalid injection pulse width T is set as the injection correction amount of the corresponding cylinder #i stored in the predetermined address of the backup RAM 35.
S # i is read, and in step S206, the average combustion state determination value ΔNA # i stored in the RAM 34 at the address corresponding to the cylinder #i is read out by the cylinder combustion state determination program. In the above-mentioned cylinder combustion state determination program, the average combustion state determination value ΔNA # i is ΔNA # i−1.
, The average combustion state determination value ΔNA # i read by the fuel injection pulse width setting program this time is a value obtained based on the data in the previous cycle of the cylinder #i.

Next, in S207, the average combustion state determination value ΔNA # i is compared with a preset allowable upper limit value ΔNu, and ΔNA # i ≦ ΔN
If u, proceed to S208 and determine the average combustion state determination value ΔNA #
i is compared with a preset allowable lower limit value ΔNL, and ΔNA #
When i ≧ NL, the average combustion state determination value Δ of the cylinder #i concerned
It is determined that NA # i is within the allowable range, that is, the ideal combustion state (ΔNu ≧ ΔNA # i ≧ NL), and S212
Go to.

The allowable limit values ΔNu and ΔNL are, as shown in FIG. 4, a dead zone for the average combustion state determination value ΔNA # i.

On the other hand, if it is determined in step S207 that ΔNA # i> ΔNu, the combustion of the cylinder #i is too good.
Invalid injection pulse width TS # of the cylinder #i read in S204
i is updated with a value obtained by subtracting the preset pulse width ΔTS (TS # ← T # Si−ΔTs).

Further, if it is determined in S208 that ΔNA # i <ΔNL, the combustion state of the cylinder #i is bad, so in S210, the pulse width ΔTs in which the invalid injection pulse width TS # i is preset.
Update with the value added in (TS # i ← T # Si + ΔTs).

The pulse width ΔTs is set in consideration of the output fluctuation amount due to the fuel injection correction, the response characteristic, and the like.

Then, from S209 or S210 to S211, RAM3
Each cylinder # (i, i−1, i stored in the predetermined address of 4
−2, i-3 ...) Average combustion state determination value ΔNA # (i, i−1, i
--2, i-3 ...) are all cleared (△ NA # (i, i-1, i-2, i
-3 ...) ← 0), and proceeds to S212.

Since the average combustion state determination value ΔNA # i of each cylinder is obtained by the weighted average, the invalid injection pulse width TS
Average combustion state determination value ΔNA # i even if #i is properly corrected
Does not always fall within the allowable range (ΔNu ≧ ΔNA # i ≧ NL), but the average combustion state determination value ΔNA # i of all cylinders # (i, i-1, i-2, i-3) If (i, i-1, i-2, i-3) is not cleared, erroneous determination may occur in the subsequent operation cycles.

Then, when proceeding from S208 or S211 to S212,
Basic fuel injection pulse width (basic injection amount) Tp set based on the engine operating state read in S201, air-fuel ratio feedback correction coefficient α, various increase correction coefficient COEF, and
The fuel injection pulse width Ti is set from the following equation based on the invalid injection pulse width Ts # i read in S204 or set in S209 and S210.

Ti ← Tp × α × COEF + TS # i Then, in S213, a drive signal corresponding to the fuel injection pulse width Ti set in S205 or S212 is output to the injector 10 of the corresponding cylinder #i at a predetermined timing, and the routine exits.

Thus, the average combustion state determination value ΔNA # of each cylinder #i
When i exceeds the allowable limit value (ΔNA # i> ΔNu,
Alternatively, ΔNA # i <ΔNL), learning can be achieved by reducing or increasing the invalid injection pulse width TS # i of the cylinder #i by a predetermined pulse width ΔTs to obtain a substantially uniform combustion state in all cylinders. As a result, smooth idle rotation can be obtained even at low output.

Next, the invalid injection pulse width TS # i according to the above program
An example of the setting state will be described with reference to the time chart of FIG.

For example, the combustion state comparison value ΔN # i of the cylinder # 1 obtained at the elapsed times t1, t2, and t3 is on the negative side as shown in (a), and the weighted average combustion average value as shown in (b) is used. The state determination value ΔNA # 1 is lower than the allowable lower limit value NL (ΔNA #
When # 1 <△ NL), the combustion condition is bad, so (c)
As shown in, the invalid injection pulse width TS # of the cylinder # 1
Add the set pulse width ΔTs to 1 (TS # 1 ← TS # 1 +
ΔTs), the fuel injection pulse width Ti of the cylinder # 1 is increased.

At this time, since the fuel injection pulse width Ti of the cylinder # 1 is increased, the total air-fuel ratio of all the cylinders shifts to the rich side. Therefore, as shown in (c) to (f), the O ₂ sensor output is changed. The air-fuel ratio feedback correction coefficient α based on this serves to perform correction control to the lean side, and the fuel injection pulse width Ti of all cylinders is distributed and reduced substantially evenly as the cylinder # 1 is increased. When the O ₂ sensor is inactive after the start, the amount of increase in the start and the water temperature are increased, and since the increase in the pulse width ΔTs is relatively small with respect to the increase rate, the invalid injection pulse width Even if TS # 1 is increased, there is no problem.

Then, after the calculation of the increase correction of the cylinder # 1 is completed, the average combustion state determination value ΔNA # (1,3,2,4) of each cylinder # (1,3,2,4)
4) Reset.

Further, for example, cylinder # 3 obtained at the elapsed times t4 and t5
The combustion state comparison value ΔN # 3 is on the plus side as shown in (a), and as shown in (b), the weighted average combustion state determination value ΔNA # 3 of the cylinder # 3 is the allowable upper limit. When the value becomes higher than the value Nu (ΔNA # 3> ΔNu), the cylinder # 3 concerned
It is judged that the combustion state of is too good, and as shown in (d),
The set pulse width ΔTs is subtracted from the invalid injection pulse width TS # 3 (TS # 3 ← TS # 3-ΔTS), and the fuel injection pulse width Ti of the cylinder # 3 is reduced.

At this time, since the fuel injection pulse width Ti of the cylinder # 3 is reduced, the total air-fuel ratio of all the cylinders shifts to the lean side. Therefore, as shown in (c) to (f), the O ₂ sensor output is changed. Based on this, the air-fuel ratio feedback correction coefficient α works to correct to the rich side, and the fuel injection pulse width Ti of all cylinders is evenly distributed and increased by the increased amount of the cylinder # 3.

Then, after the calculation of the increase correction of the cylinder # 3 is completed, the average combustion state determination value ΔNA # (1,3,2, of each cylinder # (1,3,2,4)
4) Reset.

In this way, the combustion state comparison value ΔN # (1,3,
2, 4) converge to 0.

(Second Embodiment) FIG. 10 is a flow chart showing a cylinder combustion state determination procedure according to the second embodiment of the present invention.

The same steps as those in FIG. 1 of the first embodiment are designated by the same reference numerals.

In this embodiment, the engine speed N # i-1 in the section where the combustion is not performed after the combustion of the cylinder # i-1 before the first combustion stroke and the work is performed by the combustion of the cylinder # i + 1 after the first combustion stroke. The engine speed N # i + 1 (-that is used as an estimated value of the engine speed in the non-running section is the engine speed N # i + 1 (-
(1) is compared with the engine speed N # i in the section where the combustion of the current combustion stroke cylinder #i is not working to determine the combustion state of the current combustion stroke cylinder #i. is there.

The combustion state determination procedure will be described below with reference to the flowchart.

First, proceed from S101 to S106 in the same procedure as in FIG.
Then, the cylinder # i-1 before one combustion stroke stored in RAM34
Engine speed N in the section not working due to combustion of
# I-1 and the engine rotation of the section not performing the work by the combustion of the cylinder # i + 1 one cycle before, which is adopted as the estimated engine rotation number of the section not performing the work by the combustion of cylinder # i + 1 after one combustion stroke The number N # i + 1 (-1) is read out, and in S302, the engine speed N # i of the combustion stroke cylinder #i obtained in S106 and the engine speed N # i-1, N # read in S301. From the difference from the average value of i + 1 (-1), the combustion state comparison value ΔN # of the combustion stroke cylinder #i is calculated.
Find i.

ΔN # i ← N # i-{(N # i-1 + N # i + 1 (-1)) / 2} After that, in S303, the combustion state comparison value ΔN # obtained in S302.
i and the previous average combustion state determination value ΔNA # i (-of the combustion stroke cylinder #i stored in a predetermined address of the RAM 34
1) and the average combustion state determination value ΔNA # i of the current combustion stroke cylinder #i is calculated from the weighted average of the weight r shown in the following equation, R
Store at the specified address of AM34.

ΔNA # i ← {(2 ^r −1) × NA # i (−1) + ΔN # i} / 2 ^r Then, in S304, the combustion stroke cylinder #i stored in the RAM 34 at the predetermined address. Of the previous average combustion state determination value ΔNA # i (-1) to the current average combustion state determination value ΔNA
Update with #i (ΔNA # i (−1) ← ΔNA # i) and exit the routine.

According to this embodiment, after the combustion stroke cylinder #i is determined, the combustion state determination value ΔN # i of the cylinder #i can be set.

Then, based on the average combustion state determination value ΔNA # i set by the combustion state determination program, in the next cycle, the fuel injection pulse width Ti of the cylinder #i shown in FIG. 2 of the first embodiment is set. Increase or decrease the specified amount.

In this embodiment, cylinder # i + after one combustion stroke
The engine speed N # in the previous cycle as an estimated value of the engine speed in the section in which there is no work due to combustion
i + 1 (-1) is adopted, but not limited to this, for example,
A simple average value or a weighted average value of the engine speed # i + 1 up to several cycles before the cylinder # i + 1 may be adopted as the estimated value.

(Third Embodiment) FIG. 11 is a flow chart showing the procedure for setting the fuel injection pulse width for each cylinder according to the third embodiment of the present invention.

The same steps as those in FIG. 2 of the first embodiment are designated by the same reference numerals and the description thereof will be omitted.

In this embodiment, the fuel injection pulse width Ti is first obtained, and at the time of idling, the fuel injection pulse width Ti is corrected by the correction pulse width Tc # i as the injection correction amount to perform the idle control.

The correction pulse width Tc # i is stored in a predetermined address of the backup RAM 35 corresponding to each cylinder, and the initial set value is 0 when data is corrupted. The procedure for calculating the average combustion state determination value ΔNA # i is the same as that in the first embodiment (FIG. 1) or the second embodiment (FIG. 10), and therefore the description thereof is omitted.

First, when proceeding from S201 and S202 to S401, the fuel injection pulse width Ti is obtained based on the engine operating state read in S201 (Ti ← Tp × α × COEF + Ts).

Next, in S203, it is determined whether or not the current operation state is idle, and if it is in the idle release state, the process jumps to S213. On the other hand, if it is determined that the cylinder is in the idle state, the process proceeds to S402, and the corresponding cylinder # stored in the predetermined address of the backup RAM 35 #
The correction pulse width Tc # i for i is read, then the average combustion state determination value ΔNA # i for the cylinder #i is read at S206, and S20
In step 7, the average combustion state determination value ΔNA # i is compared with a preset allowable upper limit value ΔNu. If ΔNA # i ≦ ΔNu, S20
Proceed to step 8 and compare the allowable lower limit value ΔNL with ΔNA # i ≧ Δ
If NL, that is, ΔNu ≧ ΔNA # i ≧ ΔNL, S405
Go to.

Further, in S207, the average combustion state determination value Δ of the cylinder #i concerned
NA # i has reached the allowable upper limit value ΔNu (ΔNA # i> Δ
Nu) and proceed to S403, the above backup RAM35
The correction pulse width Tc # i of the cylinder #i stored in the predetermined address of is updated by a value obtained by subtracting the preset pulse width ΔTc (Tc # i ← Tc # i-ΔTc), and the process proceeds to S211. .

Further, in S208, the average combustion state determination value ΔNA of the cylinder #i concerned
#I has reached the allowable lower limit value ΔNL (ΔNA # i <ΔN
L) and the process proceeds to S404, the correction pulse width Tc # i of the cylinder #i is updated with a value obtained by adding a preset pulse width ΔTc (Tc # i ← Tc # i + ΔTc), and S211 Go to.

Then, when proceeding to S211, the average combustion state determination value ΔNA # (i, i−1, for each cylinder # (i, i−1, i−2, i−3) stored at a predetermined address of the RAM 34 is obtained. i-2, i-3) are all cleared (ΔNA # (i, i-1, i-2, i-3) ← 0), and the process proceeds to S405.

Then, the fuel injection pulse width Ti set in S401 is read in S402, or is corrected by the correction pulse width Tc # i set in S403 or S404 (Ti ← Ti + Tc # i),
In S213, a drive signal corresponding to the fuel injection pulse width Ti is output to the injector 10 of the cylinder #i at a predetermined timing, and the routine is exited.

(Fourth Embodiment) FIGS. 12 and 13 show a fourth embodiment of the present invention, FIG. 12 is a flowchart showing a cylinder-by-cylinder combustion state determination procedure, and FIG. 13 is a cylinder-by-cylinder fuel injection pulse width setting procedure. It is a flowchart showing.

In this embodiment, the engine speed N # i of the section in which the combustion stroke cylinder #i is not working due to combustion is weighted and averaged for each cylinder to obtain the average engine speed NA # i. -1 average engine speed NAi-1 and
From the difference between the average values of the average engine speeds NA # i-2, NA # i of the cylinder # i-2 before the two combustion strokes and the cylinder #i of the present combustion stroke, the average of the cylinder # i-1 before the one combustion stroke is obtained. The combustion state determination value ΔNA # i-1 is obtained.

According to this embodiment, since the engine speed is obtained as a value smoothed by the weighted average, it is possible to effectively prevent erroneous determination especially during a period with a small sampling number, such as immediately after shifting to idle. it can.

Hereinafter, description will be given according to the flowchart.

S101 to S106 are executed through the same routine as in the first embodiment (FIG. 1), and in S501, the engine speed N # in the section where no work is performed by the combustion of the current combustion stroke cylinder #i obtained in S106. i and the cylinder #i (-
The average engine speed NA # i of the current combustion stroke cylinder #i is calculated from the weighted average of the weight r based on the average engine speed NA # i (-1) calculated in the combustion state determination flowchart of 1).

NA # i ← {(2 ^r -1) x NA # i (-1) + N # i} / 2 ^r Then, in S502, the cylinder #i stored in the RAM 34 at the predetermined address up to one cycle before is stored. The average engine speed NA # i (-1) is updated with the average engine speed NA # i obtained in S501 (NA # i (-1) ← NA #.
i).

Next, in S503, set in the routine of the last time before, the RAM 3
Cylinder # i-1 before one combustion stroke stored at a predetermined address of 4
And the average engine speed NA # i−1, NA # i− in the section where there is no work due to combustion between the cylinder # i-2 and the cylinder # i-2 two combustion strokes before
2 is read, and in S504, the average value of the average engine speed NA # i of the current combustion stroke cylinder #i obtained in S501 and the average engine speed NA # i-2 read in S503, and the S
1 from the average engine speed NA # i-1 read in 503
Average combustion state determination value of cylinder # i-1 before combustion stroke ΔNA #
i-1 is set, stored at a predetermined address in RAM 34, and the routine is exited.

ΔNA # i-1 ← NA # i-1 − {(NA # i + NA # i-2) / 2} Then, in the cylinder-by-cylinder fuel injection pulse width setting program of FIG. Based on the set average combustion state determination value ΔNA # i, S207,
In S208, the combustion state of the cylinder #i is determined in the same manner as in the first embodiment (FIG. 2) described above, and in S601, each cylinder # (i, i-1, i) stored in the predetermined address of the RAM 34. -2, i-3 ...)
Of the average engine speed NA # (i, i-1, i-2, i-3 ...) are cleared (NA # (i, i-1, i-2, i-3 ...) ← 0).

The other steps are the same as those in the first embodiment (FIG. 2), and the description thereof will be omitted.

It should be noted that the present invention is not limited to the above-described embodiments, and for example, the momentum of a section in which work is not performed due to combustion of each cylinder is not limited to the engine speed, and may be angular acceleration, cycle, angular velocity, or the like. In particular, by using the angular velocity obtained by differentiating the engine speed with respect to time, it is possible to accurately measure the momentum because a time element is added.

[Effects of the Invention] As described above, according to the present invention, the combustion state of the cylinder can be accurately grasped without including the combustion state factors of the other cylinders, and the combustion state of each cylinder can be determined. Excellent effects such as continuous grasping and smooth engine rotation can be obtained.

[Brief description of drawings]

1 to 9 show a first embodiment of the present invention, FIG. 1 is a flowchart showing a cylinder-by-cylinder combustion state determining procedure, FIG. 2 is a flowchart showing a cylinder-by-cylinder fuel injection pulse width setting procedure, and FIG. FIG. 4 is a conceptual diagram showing a basic cylinder-by-cylinder combustion state determination method, FIG. 4 is a conceptual diagram showing a dead zone region, FIG. 5 is a schematic diagram of an engine control system, and FIG. 6 is a front view of a crank rotor and a crank angle sensor. FIG. 7 is a front view of the cam rotor and the cam angle sensor, FIG. 8 is a time chart showing cylinder pressure fluctuation, crank pulse, cam pulse, and engine rotation fluctuation. FIG. 9 is a combustion state comparison value for each cylinder. Cylinder average combustion state determination value, time chart showing cylinder invalid injection pulse width, FIG. 10 is a flow chart showing the cylinder combustion state determination procedure according to the second embodiment of the present invention, FIG. 11 is the present invention Qi according to three examples 12 and 13 show a fourth embodiment of the present invention, FIG. 12 is a flowchart showing a cylinder-by-cylinder combustion state determination procedure, and FIG. 13 is a by-cylinder chart. It is a flow chart which shows a fuel injection pulse width setting procedure. #I ... Current combustion stroke cylinder, # i-1 ... 1 combustion stroke before cylinder, # i + 1 ... 1 combustion stroke after cylinder, # i-2 ... 2 combustion stroke before cylinder, N # i, N # i -1, N # i-2 ... Momentum of the section not working due to combustion, N # i + 1 (-1)
…… Estimated momentum.

Claims (3)

[Claims] Claim: What is claimed is: 1. An engine rotation state quantity in a section where no work is performed due to combustion in a cylinder before one combustion stroke, and an engine rotation state quantity and a current combustion stroke cylinder in a section where no work due to combustion in a cylinder before two combustion strokes are performed. The cylinder-by-cylinder combustion state determination method, wherein the combustion state of the cylinder before the first combustion stroke is determined from the difference between the average value and the engine rotation state amount in the section where no work is performed due to combustion. 2. An engine rotational state quantity in a section not working by combustion in the cylinder before the current combustion stroke, an engine rotational state quantity in a section not working by combustion in the cylinder before one combustion stroke, and after one combustion stroke. Cylinder combustion state discrimination of the engine characterized by discriminating the combustion state of the present combustion stroke cylinder from the difference of the average value with the engine rotation state amount of the section not performing work due to combustion of the cylinder one cycle before Method. 3. The engine rotational state quantity is any one of an engine rotational speed, an angular acceleration, a cycle, an angular velocity, and an average engine rotational speed obtained by performing a weighted average process on the engine rotational speed.
3. The cylinder-by-cylinder combustion state determination method according to claim 1, wherein one of the two is used.