JPH0629583B2 - Control device for internal combustion engine - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an internal combustion engine, and more particularly to an internal combustion engine in which the fuel injection amount and the ignition timing are controlled based on the measured value of the intake pipe pressure. Regarding the control device.

[Conventional technology and related technology]

Conventionally, the basic fuel injection time is calculated every predetermined time based on the measured value of the intake pipe pressure and the measured value of the engine rotation speed, and the basic fuel injection time is corrected by the intake air temperature, the engine cooling water temperature, etc. BACKGROUND ART There is known an internal combustion engine that obtains an injection time and opens a fuel injection valve for a time corresponding to this fuel injection time to inject fuel. Further, in such an internal combustion engine, in order to improve the responsiveness during acceleration, the rate of change in the measured value of the intake pipe pressure is detected, and the time proportional to this rate of change is corrected so that the basic fuel injection time becomes longer. It is designed to carry out accelerated fuel quantity increase.

In an internal combustion engine that calculates the basic fuel injection time based on the intake pipe pressure as described above, a pressure sensor that measures the intake pipe pressure (absolute pressure) is attached to the intake pipe, and the basic fuel is measured based on the measured intake pipe pressure. Although the injection time is calculated, the measured value fluctuates due to engine pulsation, and this fluctuation may change the basic fuel injection time, which may result in inaccurate fuel injection amount control. Therefore, conventionally, as shown in Japanese Patent Laid-Open No. 59-201938, two filters having different time constants are used to relax the pressure sensor output to completely remove the pulsating component from the pressure sensor output. The filter output with a large time constant is subtracted from the filter output with a small time constant to provide an overshoot characteristic, and the acceleration amount is increased according to this difference. However, in the method using two filter elements as described above, the degree of relaxing the pressure sensor output is increased by using a filter having a relatively large time constant in order to remove the pulsating component, and thus the actual intake pipe pressure The responsiveness and followability of the change in the filter output to the change deteriorates, the increase in acceleration is delayed, the fuel injection amount is insufficient at the beginning of acceleration, and a lean spike occurs.At the end of acceleration, a latch spike occurs due to the overshoot characteristic. In some cases.

Therefore, recently, the pressure sensor output is processed by using a CR filter which is composed of a resistor and a capacitor and has a relatively small time constant such that a pulsating component can be removed.
The R filter output is converted into a digital value every predetermined time, and 2
It has been proposed to use measured values with better responsiveness and trackability than when using two filters. In this case, since the pulsation component cannot be completely removed by the CR filter, two weighted average values having different degrees of relaxation are calculated using the above digital value, that is, the digital filtering process is performed to reduce the degree of relaxation. The acceleration increase value is determined based on the difference obtained by subtracting the second weighted average value having a large degree of relaxation from the weighted average value of 1.

However, in any of the above methods, a value with a large degree of relaxation is used to obtain the acceleration increase value, so responsiveness and followability deteriorate, and in a driving pattern in which acceleration / deceleration is repeated, the phase delay of the acceleration increase is delayed. Occurs, the fuel injection amount may not match the increase request of the engine, and the exhaust emission and the driver's ability deteriorate. To solve this problem, the pressure sensor output is relaxed to such an extent that the engine pulsation component can be removed, and only a relaxation value with a small degree of relaxation is obtained, and the fuel injection amount including the acceleration increase is calculated based on this relaxation value. However, it takes a certain amount of time for the injected fuel to reach the combustion chamber after the fuel injection time is calculated, due to the influence of the calculation time and the flight time of the fuel. Since the difference between the used intake pipe pressure (relaxation value) and the intake pipe pressure corresponding to the actual intake air amount, it becomes impossible to control the air-fuel ratio required by the engine.

The above will be described in more detail with reference to FIG. Fig. 4 shows the amount of fuel required for one intake stroke per revolution of the engine.
FIG. 5 is a diagram showing changes in the calculated basic fuel injection time TP and intake pipe pressure PM during acceleration of a 4-cylinder 4-cycle internal combustion engine that injects / 2. In this example, the fuel is injected once per revolution of the engine, that is, twice in one cycle (points c and b in the figure), and the amount of fuel supplied for one combustion can be understood from the figure. Thus, the amount corresponds to TPc + TPb. However, the intake pipe pressure that represents the actual intake air amount during combustion is the intake pipe pressure at the end of the intake stroke (intake bottom dead center) shown in a in the figure. As described above, since there is a delay of time t _D between the intake pipe pressure at the time of calculating the fuel injection time and the intake pipe pressure representing the actual intake air amount at the time of combustion, fuel is injected according to the actual intake air amount. It becomes impossible to control the air-fuel ratio required by the engine. On the other hand, even if the calculation time is shortened to make the delay time t _D small enough to be ignored (even if the intake bottom dead center coincides with the point b), the internal combustion that injects fuel once per revolution of the engine 2 points at point b
While the amount of fuel corresponding to TPb is required, TPc
Since only the amount of fuel corresponding to + TPb is supplied, the fuel amount corresponding to TPb-TPc (= ΔTP) is insufficient during acceleration.

Therefore, the present applicant has already proposed a technique for correcting the fuel amount shortage ΔTP (Japanese Patent Application No. 61-277019).
No. 61-277020). Next, the principle of this technique will be described. In the following, a four-cylinder, four-cycle internal combustion engine in which fuel is injected once per revolution of the engine will be described as an example.

As described with reference to FIG. 4, the basic fuel injection time TP corresponding to the actual intake air amount is represented by the following equation, ignoring the delay time t _D from the fuel injection time calculation.

TP = TPb + ΔTP (1) On the other hand, as shown in FIG. 5, if acceleration is performed at uniform acceleration, the difference Δ in basic fuel injection time between points b and c is Δ.
Difference in basic fuel injection time between TP, point b and point b ′ ΔTP ′
Is equal to, so the basic fuel injection time TPb 'at the point b'
Is the basic fuel injection time TPb at point b and ΔTP (=
It can be expressed as follows using ΔTP ′).

TP ′ = TPb + ΔTP (2) Here, assuming that the calculation of the basic fuel injection time is performed every 360 ° CA, as can be understood from the above equation (2), 360 ° CA from the point b. This means that the basic fuel injection time was predicted.

Therefore, in general, the calculation of the basic fuel injection time is CY [°
CA] and that performed for each, the delay time t _D between the points a and b of FIG. 4 in terms of the crank angle CA _D and by obtaining the correction value corresponding to the crank angle CA _D , Next, it is possible to predict the basic fuel injection time of a predetermined crank angle CA _D away from point b. Therefore, considering the correction when changing from point c to point b in FIG. 4, CY [°
The basic fuel injection time TP corresponding to the actual intake air amount when the basic fuel injection time is calculated for each CA] is expressed as follows using the basic fuel injection time TP ₀ calculated immediately before.

TP = TP ₀ + k · ΔTP (4) where k is And ΔTP is CY [° from the current basic fuel injection time.
CA] is a difference obtained by subtracting the basic fuel injection time calculated before, and this difference is positive for acceleration and negative for deceleration.

Here, the delay time t _D is often kept at a constant crank angle for control purposes, but considering the flight time of the injected fuel, this flight time is substantially constant regardless of the engine speed, When the engine speed becomes high, the fuel injected immediately before the intake stroke cannot reach the combustion chamber due to the delay due to the flight time, and the fuel is sucked for the first time in the second intake stroke. Therefore, the crank angle C at which the fuel injection time should be predicted
A _D increases as the engine speed increases.

On the other hand, when the CR filter is used, the CR filter output has a good response to the change of the actual intake pipe pressure, and therefore it is considered that the CR filter output shows substantially the actual intake pipe pressure. The weighted average value (corresponding to the relaxation value) lags the actual intake pipe pressure as shown in FIG. This delay (control delay t _D ′) is the detection delay of the pressure sensor,
This is caused by delays in signal transmission of the input circuit, delays in calculation timing due to these delays, delays in calculation time, delays in relaxing CR filter output, and the like. Therefore, the control delay t _D ′ (in crank angle CA _D ′) from PMb ′ for calculating the fuel injection amount at point b in FIG.
It is necessary to predict the actual intake pipe pressure PMb in consideration of the above, calculate the basic fuel injection time based on this predicted value, and make a prediction in consideration of the delay time t _D described above.

Therefore, if the correction of the control delay t _D ′ (= _{C A D} ′) is added to the above equation (4), it is expressed as follows.

TP = TP ₀ + K ₁ · ΔTP (5) Is.

Further, when the basic fuel injection time TP is calculated from the intake pipe pressure PM and the engine speed NE, TP∝PM, so
The above equation (5) is used to calculate the difference in the relaxation value of the intake pipe pressure (the value obtained by subtracting the relaxation value for calculating the basic fuel injection time before CY ° CA from the relaxation value for the current basic fuel injection calculation), that is, the change rate Δ of the relaxation value.
When expressed using PM, it becomes like the following formula (6).

TP = TP ₀ + K ₁ · ΔPM · C (6) where C is a proportional constant for converting the intake pipe pressure into the basic fuel injection time.

Here, the control delay time t _D ′ can be regarded as substantially constant due to the phenomenon of the time period, and therefore, the crank angle CA _D ′ increases as the engine speed increases.

The values of the crank angles CA _D and CA _D ′ at each rotation speed can be calculated, and the K ₁ value at each rotation speed can be obtained without considering the manufacturing error of the test engine. Further, in the above, the predetermined crank angle (CY °
Although the example of calculating the basic fuel injection time for each CA) has been described, the present invention can also be applied to the case of calculating the basic fuel injection time for each predetermined time. In this case, CA _D ′ does not need to be corrected by the engine rotation speed, but since the delay due to the flight time of the injected fuel is affected by the engine rotation speed, K _{1 as a} whole needs to be corrected by the engine rotation speed. Become. Further, although the example in which the fuel is injected once per one revolution of the engine has been described above, even in the independent injection, when the engine rotation speed increases, the basic fuel injection time becomes longer and a region where unsucked fuel occurs may occur. Therefore, it is desirable to predict the intake pipe pressure (intake bottom dead center value) that represents the actual intake air amount when the basic fuel injection time is calculated one time before the current basic fuel injection time is calculated. Can also be applied to.

[Problems to be Solved by the Invention]

However, in the technique of calculating the basic fuel injection time TP using the above formula (5) or formula (6), the change rate ΔPM during sudden acceleration
Becomes a large value, the fuel injection time TAU is overshot as shown in FIG. 2 (1), the air-fuel ratio becomes rich, CO and HC emissions increase, and driver viability deteriorates. There is a possibility that this problem will occur. Further, in the internal combustion engine described above, the basic ignition advance angle is obtained from the relaxation value of the intake pipe pressure and the engine rotation speed, and the change rate ΔPM
Since the basic ignition advance angle at the time of acceleration is also corrected by, the correction of the basic ignition advance angle by the change rate ΔPM is not appropriate at the time of sudden acceleration. Further, even during a sudden deceleration, the correction based on the change rate ΔPM becomes incorrect, and the fuel injection amount and the ignition timing do not match the engine required value, which deteriorates the driver ability and the exhaust emission. To solve these problems, the correction amount K ₁ · ΔTP (or K
₁ · ΔPM · C) should be limited to an appropriate value, but if you try to limit by ΔTP or ΔPM · C,
In a region where these values are equal to or more than the upper limit value, the overshoot can be sharply reduced, but in the region where the upper limit value is not reached, the overshoot described above occurs, and the driver bilility and the emission are deteriorated.

Therefore, according to the present invention, the control amount such as the basic fuel injection time and the basic ignition advance angle is calculated from the relaxation value of the intake pipe pressure, and the control amount is corrected by the change rate of the relaxation value or the control amount to control the internal combustion engine. Another object of the present invention is to provide a control device for an internal combustion engine, which is capable of appropriately performing correction even in the entire range of sudden acceleration and sudden deceleration.

[Means for Solving the Problems]

In order to achieve the above object, the present invention, as shown in FIG. 3, is a pressure sensor A for detecting the intake pipe pressure, and a relaxation means for calculating a relaxation value by relaxing the change in the signal output from the pressure sensor A. B, a control amount calculation means C for calculating a control amount for controlling the engine based on the relaxation value, a change rate calculation means D for calculating a change rate of the relaxation value or the control amount, and the change rate. Coefficient setting means E for setting a correction coefficient that becomes smaller and accordingly smaller in absolute value, correction means F for correcting the control amount based on the product of the change rate and the correction coefficient, and correction by the correction means F. The control means G for controlling the engine on the basis of the controlled amount thus controlled is included.

[Action]

The relaxation means B of the present invention relaxes the change in the signal output from the pressure sensor A that detects the intake pipe pressure to obtain a relaxation value. As the relaxation value, the weight of the weighted average value calculated in the past is weighted, and the current weighted value calculated by the weighted average value calculated in the past and the current level of the signal output from the pressure sensor A. An average value can be used. That is, the weighted average value PMN _i calculated according to the following equation can be used as the relaxation value.

However, PMN _i-1 is the weighted average value calculated in the past, N
Is a coefficient related to weight, PMAD is the current level of the signal output from the pressure sensor, and the value output by directly converting the signal output from the pressure sensor into a digital value or the pressure sensor output processed by the CR filter into a digital value. The converted value can be adopted. Such a weighted average value can be obtained by digital filtering processing.

Further, the control amount calculation means C calculates the control amount for controlling the engine based on the relaxation value. The control amount includes a basic fuel injection time, a basic ignition advance angle and the like, and the control amount calculation means C calculates at least one of the basic fuel injection time and the basic ignition advance angle as the control amount. The change rate calculating means D calculates the change rate of the relaxation value or the change rate of the control amount, and the coefficient setting means E
Sets a correction coefficient that decreases as the absolute value of this change rate increases. Then, the correction unit F corrects the control amount calculated by the control amount calculation unit C based on the product of the change rate and the correction coefficient, and the control unit G corrects the engine amount based on the corrected control amount. To control. As described above, since the correction coefficient is set to decrease as the absolute value of the change rate increases, the correction amount can be reduced during rapid acceleration / deceleration when the absolute value of the change rate is large. The absolute value of the change rate is large in the region where the absolute value of the initial change rate is small, and the absolute value of the change rate is large. be able to. Also,
Even in acceleration / deceleration in which the absolute value of the rate of change is relatively small, the correction coefficient becomes smaller as the absolute value of the rate of change increases from the initial stage to the middle period of acceleration / deceleration, so overshoot can be reduced.

〔The invention's effect〕

As described above, according to the present invention, since the control amount is corrected with the correction coefficient that decreases as the absolute value of the change rate increases, the acceleration / deceleration is gradually increased without deteriorating the rising response during transition. It is possible to reduce the overshoot until the deceleration, and it is possible to obtain the effect that the emission and the driver ability can be improved.

〔Example〕

Embodiments of the present invention will be described in detail below with reference to the drawings.
In addition, below, the example which mainly used the fuel injection time as a control amount is demonstrated. FIG. 7 schematically shows an internal combustion engine (engine) equipped with a fuel injection amount control device to which the present invention can be applied.

This engine is controlled by an electronic control circuit such as a microcomputer, and a throttle valve 8 is arranged downstream of an air cleaner (not shown), and a voltage corresponding to the throttle opening is applied to the throttle valve 8. A linear throttle sensor 10 for outputting the above is attached, and a surge tank 12 is provided on the downstream side of the throttle valve 8.
A semiconductor pressure sensor 6 is attached to the surge tank 12. The pressure sensor 6 has a small time constant for removing the pulsating component of the intake pipe pressure (for example, 3 to
It is connected to a filter (FIG. 8) composed of a CR filter or the like having a good response for 5 msec. The filter may be built in the pressure sensor. Further, a bypass passage 14 is provided so as to bypass the throttle valve 8 and connect the upstream side of the throttle valve and the surge tank 12 on the downstream side of the throttle valve. An ISC (idle speed control) valve 16B whose opening is adjusted by a pulse motor 16A having a 4-pole stator is attached to the bypass passage 14. The surge tank 12 communicates with the combustion chamber of the engine 20 via the intake manifold 18 and the intake port 22.
A fuel injection valve 24 is attached to each cylinder so as to project into the intake manifold 18.

The combustion chamber of the engine 20 is connected to a catalyst device (not shown) filled with a three-way catalyst via an exhaust port 26 and an exhaust manifold 28. An O ₂ sensor 30 that outputs a signal inverted at the stoichiometric air-fuel ratio is attached to the exhaust manifold 28. A cooling water temperature sensor 34 is attached to the engine block 32 so as to penetrate the engine block 32 and project into the water jacket. This cooling water temperature sensor 34
Detects the engine cooling water temperature and outputs a water temperature signal, and the water temperature signal represents the engine temperature. The engine oil temperature may be detected to represent the engine temperature.

A spark plug 38 is attached to each cylinder so as to penetrate the cylinder head 36 of the engine 20 and project into the combustion chamber. The ignition plug 38 is connected via a distributor 40 and an igniter 42 to an electronic control circuit 44 composed of a microcomputer or the like.
Inside the distributor 40, a cylinder discriminating sensor 46 and a rotation angle sensor 48, each of which is composed of a signal rotor fixed to the distributor shaft and a pick-up fixed to the distributor housing, are mounted. The cylinder discrimination sensor 46 is, for example, 720 °.
A cylinder discrimination signal is output for each CA, and the rotation angle sensor 48 outputs an engine speed signal for each 30 ° CA, for example.

The electronic control circuit 44 includes a micro processing unit (MPU) 60, a read only memory (ROM) 62, a random access memory (R) as shown in FIG.
AM) 64, Back-up Plum (BU-RAM) 66,
Input / output port 68, input port 70, output port 72,
74 and 76, and a bus 75 such as a data bus or a control bus connecting these components. An analog-digital (A / D) converter 78 and a multiplexer 80 are sequentially connected to the input / output port 68. To the multiplexer 80, the pressure sensor 6 is connected via a CR filter 7 and a buffer 82 each composed of a resistor R and a capacitor C, and the cooling water temperature sensor 34 is connected via a buffer 84. Further, the linear throttle sensor 10 is connected to the multiplexer 80. The MPU 60 controls the multiplexer 80 and the A / D converter 78 to sequentially convert the pressure sensor 6 output, the linear throttle sensor 10 output, and the cooling water temperature sensor 34 output, which are input via the CR filter 7, into digital signals. And store it in the RAM 64. Therefore, the multiplexer 8
0, the A / D converter 78, the MPU 60 and the like act as sampling means for sampling the pressure sensor output at predetermined time intervals. The input port 70 has a comparator 8
8 and the buffer 86 to connect the O ₂ sensor 30 and the waveform shaping circuit 90 to the cylinder discrimination sensor 46.
And a rotation angle sensor 48 are connected. Output port 7
2 is connected to the igniter 42 via the drive circuit 92,
The output port 74 is a drive circuit 94 having a down counter.
To the fuel injection valve 24, and the output port 76 is connected via a drive circuit 96 to the pulse motor 16A of the ISC valve. In addition, 98 is a clock and 9
9 is a timer. The ROM 62 stores in advance programs and the like for control routines described below.

Next, the control routine of the embodiment of the present invention when the present invention is applied to the above engine and the relaxation value is detected by the weighted average value by calculation will be described. It should be noted that the following description will be given using numerical values that do not hinder the present invention, but the present invention is not limited to these numerical values.

FIG. 9 shows an A / D conversion routine executed every 4 msec. In step 100, the signal output from the pressure sensor 6 is sent to the A / D converter via the CR filter 7, buffer 82 and multiplexer 80. Enter in 78,
The intake pipe pressure PM digitally converted by the A / D converter 78 is taken in as a digital value PMAD. Next step 10
2, the coefficient N relating to the weight of the above equation (7) is set to n (for example, by using the digital value PMAD of the intake pipe pressure and the weighted average value PMN _{i-1 of} the intake pipe pressure calculated 4 msec before).
By setting 4), the weighted average value PMN _i of the current intake pipe pressure is calculated according to the equation (7). And step 1
In order to calculate the weighted average value of the next intake pipe pressure in 04, the current weighted average value PMN _i of the intake pipe pressure is set to 4 msec.
The weighted average value PMN _i-1 of the previous intake pipe pressure is stored in the register.

FIG. 1 shows a fuel injection time calculation routine executed at every fuel injection time calculation timing (every 360 ° CA in the case of a 4-cylinder 4-cycle engine). In step 110, the coefficient K ₁ is calculated and the coefficient C is calculated. Take in. This coefficient K ₁ is obtained by taking in the engine speed NE at step 106 as shown in FIG.
Is calculated by calculating the coefficient K ₁ corresponding to the current engine speed NE from the map shown in FIG. The coefficient K ₁ is calculated in advance and stored in the ROM as a map, but as shown in FIG. 12, it is represented as an increasing function that increases from 1.0 as the engine speed NE increases. The coefficient C may be a constant value or a variable.

At the next step 112, the weighted average value of the current intake pipe pressure is fetched as PMN. In step 104 of FIG. 9, the weighted average value PMN _i of the current intake pipe pressure is set to PMN _i .
Since it is stored in the register as _i-1 , the weighted average value of the current intake pipe pressure can be taken in as PMN by reading the value of this register. In the next step 114, the current basic fuel injection time TP ₀ is calculated from the current weighted average value PMN of the intake pipe pressure and the engine rotational speed NE fetched in step 128 by a method similar to the conventional one. The next step 116 is to subtract the weighted average value PMNO of the past intake pipe pressure used to calculate the basic fuel injection time 360 ° CA from the current weighted average value PMN of the intake pipe pressure by subtracting the weighted average value PMNO of the past intake pipe pressure. The change rate ΔPM of the weighted average value of pressure is calculated. At the following step 120, calculates the eleventh correction coefficient K ₀ corresponding current change rate ΔPM from Matsupu correction factor K ₀ was expressed as a function of the rate of change ΔPM of FIG. This correction coefficient K ₀ is ΔPM ≧
In the region of 0, it decreases as ΔPM increases,
In the region of ΔPM <0, it is set to be smaller as ΔPM becomes smaller, and it is set to be smaller as | ΔPM | becomes larger as a whole.
Further, the curve showing the correction coefficient K ₀ is asymmetric with respect to the vertical axis, and the correction coefficient K ₀ in the region of ΔPM < _0.
Is larger than the change rate in the region of ΔPM ≧ 0. This is because the pumping action of the engine generally occurs during deceleration, and therefore the change in the intake pipe pressure is larger than that during acceleration. Therefore, in the region of ΔPM <0, the correction coefficient K ₀ changes more than in the region of ΔPM ≧ 0. It is getting bigger. The correction coefficient K ₀ is set to an optimum value for various engines and may be set to be symmetrical with respect to the vertical axis. The broken line in FIG. 11 is ΔPM.
This shows a change in the correction coefficient K ₀ equivalent to when ΔPM = β when ≧ β (for example, 50 mmHg / 1 rotation). As can be seen from the figure, in the present embodiment, the correction coefficient can be smoothly lowered, the overshoot can be appropriately reduced in any acceleration / deceleration, and the correction coefficient is mapped so that it is suitable. You have more freedom.

Next, at step 126, the coefficient K ₁ calculated at step 108 and the correction coefficient K ₀ calculated at step 120.
And the rate of change ΔPM of the weighted average value of the intake pipe pressure calculated in step 116 and the coefficient C for converting the intake pipe pressure into the basic fuel injection time are multiplied to increase the value TPACC (
(Corresponding to the value obtained by multiplying the second term on the right side of the equation (6) by the correction coefficient K ₀ ), and in step 128, the increase value TPACC is added to the current basic fuel injection time TP ₀ to obtain the current basic value. Correct the fuel injection time TP ₀ . Then, in step 130, the current weighted average value PMN of the intake pipe pressure is stored in the register as the weighted average value PMNO of the intake pipe pressure before 360 ° CA, and in step 132 the basic fuel injection time TP is set to the intake temperature or the engine cooling water temperature. The fuel injection time TAU is calculated by correcting the fuel injection time TAU. Then, in a fuel injection amount control routine (not shown), fuel is injected once per engine revolution.

The basic fuel injection time TP used to calculate the fuel injection time TAU in the above step 132 is
Since the correction coefficient K ₀ is used for correction in accordance with the equation (6) described above, the control delay and the delay due to the flight time of the fuel are prevented and the amount of fuel adhered to the inner wall of the intake manifold is prevented. The influence due to is prevented and is corrected to a value corresponding to the actual intake air amount, so it is possible to prevent fluctuations in the air-fuel ratio during transition. The change in the combustion injection time TAU at this time is
As shown in Figure (2), it is possible to reduce overshoot with almost no loss of transient rise response.

Although an example in which the coefficient K ₁ is changed according to the engine rotation speed has been described above, the amount of fuel adhering to the inner wall of the intake manifold increases when the engine cooling water temperature is low and the engine cooling water temperature is high. It is necessary to increase the amount of fuel more than when it is high. Therefore, even if the coefficient K ₁ is represented by a function of the engine speed and the engine cooling water temperature, the coefficient K ₁ is increased as the engine speed increases and the coefficient K ₁ is decreased as the engine cooling water temperature increases. good.
In addition, the coefficient K ₁ is a function f (PMW) of the weighted average value PMN.
The engine speed NE, the engine coolant temperature THW, and the function f (NE, T
HW, PMW). In the above embodiment, the increase value TPACC is set to the change rate ΔPM of the weighted average value of the intake pipe pressure.
From the correction coefficient K ₀ , the second term of the formula (6) is multiplied by the correction coefficient K ₀ to calculate, but the second term of the formula (5) may be multiplied by the correction coefficient K ₀ to calculate, Therefore, it may be calculated from the change rate ΔTP of the basic fuel injection time and the correction coefficient K ₀ . Also,
The correction coefficient K ₀ is the change rate ΔPM of the weighted average value of the intake pipe pressure.
Is set to be smaller as the absolute value of is increased, but may be set to be smaller as the absolute value of the change rate ΔTP of the basic fuel injection time is increased.

Further, the basic fuel injection time may be corrected by the following equation.

K ₂ · DLPMI _i · C (8) However, K ₂ is the second coefficient, which is shown in FIGS.
As shown in the figure, it can change according to the engine rotation speed, the engine cooling water temperature, the intake pipe pressure, etc., and the DLPMI _i is the current relaxation value expressed by the following equation (9) and one cycle before. It is the integrated value of the attenuation value of the difference from the detected relaxation value. Here, it is considered that when the engine rotational speed NE becomes higher, the intake flow velocity becomes faster, the amount of fuel adhering to the inner wall of the intake manifold decreases, and most of the fuel is supplied to the combustion chamber. Therefore, the coefficient K ₂ increases the engine rotational speed. It is set to be smaller according to. Further, as the engine cooling water temperature increases, the amount of fuel evaporated on the inner wall of the intake manifold increases, and the amount of fuel attached to the inner wall of the intake manifold decreases, so the coefficient K ₂
Is set to be smaller as the engine cooling water temperature becomes higher. As the intake pipe pressure increases, the amount of fuel evaporation decreases and the amount of fuel adhering to the inner wall of the intake manifold increases. Therefore, the coefficient K ₂ should be set to increase as the weighted average intake pipe pressure increases. You can

DLPMI _i = ΔPM + K ₃ · DLPMI _i−1 ....
(9) where K ₃ is a positive damping coefficient less than 1, DLPMI
_i-1 is the integrated value calculated last time. This damping coefficient K
₃ may use a constant value, but like the coefficient K ₂ ,
It may be determined according to the engine rotation speed NE, the weighted average value PMN of the intake pipe pressure, the engine cooling water temperature THW, and the like. When the coefficient K ₃ is changed, the damping speed is slowed down by increasing the coefficient K ₃ in the transient operation state in which the amount of fuel adhering to the inner wall of the intake manifold increases, and the fuel adhering to the inner wall of the intake manifold is increased. In the transient operation state where the amount decreases, the coefficient K ₃ is decreased to increase the damping speed.

In the above equation (9), the initial value of the integrated value is set to 0, and the difference ΔPM is ΔPM ₁ , ΔPM ₂ , ... ΔPM during the calculation i times.
_If it is changed to _i , the i-th DLPMI _i is expressed as follows.

_{DLPMI i = ΔPM i + K 3} · ΔPM i-1 + K 3 2
・ ΔPM _i-2 + ・ ・ ・ + K ₃ ^i-2・ ΔPM ₂ + K ₃ ^i-1・ ΔPM ₁ (10) Therefore, the integrated value gradually increases from the start of acceleration and the damping coefficient after the end of acceleration. It takes a certain value until it approaches 0 by K ₃ .

When the correction for predicting the basic fuel injection time corresponding to the actual intake air amount, the correction of the equation (8) and the correction by the correction coefficient K ₀ are simultaneously performed, the basic fuel injection time TP becomes
It becomes like the formula (11) or the formula (12).

TP = TP ₀ + K ₀ · K ₁ · ΔPM · C + K ₂ · DLPMI _i · C …… (11) TP = TP ₀ + K ₀ · K ₁ · ΔTP + K ₂ · DLPMI _i …… (12) However, the above ( DLPMI _{i in the} equation (12) is an integrated value of the attenuation values of the difference between the current basic fuel injection time and the basic fuel injection time one cycle before, which is expressed by the following equation.

DLPMI _i = ΔTP + K ₃ · DLPMI _i-1 (1
3) In addition, K ₁ , K ₂ , and K ₃ used in the above equations (11) and (12) are set so as to cover a wide range of transient operating conditions, such as engine speed, engine cooling water temperature or intake pipe absolute pressure. It may be determined according to the parameters, but the coefficient at which the required value of the fuel injection amount hardly changes in the transient operating state even if each parameter is changed may be defined as a constant value.

When the basic fuel injection time was corrected as described above when the engine was cold, the changes in the acceleration increase value and the air-fuel ratio were adjusted as the value of K ₁ during warm time when the current basic fuel injection time TP ₀ was not corrected. When the value K _H is used, the experimental results in the case of performing the experiment will be described in comparison with the case where the value K _C (> K _H ) which is adapted to the cold state is used as the value of K ₁ . Note that in the following, K ₀ = 1.
It was set to 0. As shown in FIG. 15 (1), if the fuel is injected only for the current basic fuel injection time TP ₀ in the acceleration operation state where the intake pipe pressure PM during cold engine changes from PM ₁ to PM ₂ , The value becomes 0 and the air-fuel ratio is shown in Fig. 15.
As shown in (3), a large amount of lean spikes occur, resulting in exhaust emission and poor driver viability. This basic fuel injection time TP ₀ is corrected to TP ₀ + K
_When the fuel is injected based on _H · ΔPM · C, the lean spike is halved, but the change in the air-fuel ratio may still be large. It is considered that this is because the amount of fuel adhering to the inner wall of the intake manifold changes greatly when cold. The value K _C which further increase the value of K ₁ adapted to the cold
When fuel is injected based on TP ₀ + K _C · ΔPM · C using, the lean spike at the initial stage of acceleration can be almost eliminated as shown in Fig. 15 (3), but the lean spike remains at the end of acceleration or at the end of acceleration. There is. It is considered that this is because the intake pipe pressure increases and the amount of fuel evaporation decreases at the latter stage of acceleration and the end of acceleration, so that the amount of injected fuel adhering to the inner wall of the intake manifold increases.

In consideration of the above phenomenon, in the formulas (11) and (12), the difference between the current basic fuel injection time and the basic fuel injection time calculated one cycle before or the current relaxation value and one cycle before detection On the basis of the product of the rate of change represented by the difference between the relaxation value and the first coefficient that is changed according to the engine speed, and the product of the integrated value of the damping value of the rate of change and the second coefficient. This is to correct the current basic fuel injection time. Since the integrated value of the above-mentioned damping value takes a certain value even at the end of acceleration and after the end of acceleration, the lean spike at the end of acceleration and at the end of acceleration that occurs when the basic fuel injection time is corrected with K ₁ as K _C is used. This can be prevented, and the air-fuel ratio at the time of transition such as acceleration can be made substantially constant as shown by the solid line in FIG. 15 (3).

Further, although the case where the fuel injection amount is controlled has been described above, the present invention can be applied to a case where the ignition timing is controlled and a case where the fuel injection amount and the ignition timing are simultaneously controlled.

Further, the present invention is effective for all phase advance control using the change rate ΔPM. That is, even when a higher-order differential element such as PM + K ₀ · K ₄ · ΔPM PM + K ₀ · K ₄ · ΔPM + K ₅ · ΔΔPM PM + K ₀ · K ₄ · ΔPM + K ₅ · ΔΔPM + K ₆ · ΔΔPM is used, the overshoot reduction effect by K ₀ can be achieved. If the ignition timing is determined, overcorrection of the ignition timing due to overshoot can be prevented. In this case, ΔΔPM and ΔΔΔPM are also correction factors K.
You may correct by ₀ .

[Brief description of drawings]

FIG. 1 is a flow chart showing a fuel injection time calculation routine of one embodiment of the present invention, FIGS. 2 (1) and 2 (2) are diagrams showing changes in fuel injection time, etc., and FIG. 3 is a patent of the present invention. A block diagram corresponding to the scope of claims and FIG. 4 are diagrams for explaining the delay of the fuel injection amount when fuel is injected once per one revolution of the engine, and FIG.
FIG. 6 is a diagram showing changes in intake pipe pressure and basic fuel injection time in a constant acceleration state, FIG. 6 is a diagram for explaining supplement of fuel amount due to control delay, and FIG. 7 is applicable to the present invention. FIG. 8 is a schematic diagram showing an engine equipped with various fuel injection amount control devices, and FIG. 8 is a block diagram showing details of the control circuit in FIG.
9 is a flow chart showing the A / D conversion routine of the above embodiment, FIG. 10 is a flow chart showing a calculation routine of the coefficient K ₁ of the above embodiment, and FIG. 11 is a diagram showing a map of the correction coefficient K ₀ . FIG. 12 is a diagram showing a map of the coefficient K ₁ , and FIGS. 13 and 14 are diagrams showing a map of the coefficient K ₂ .
Figures (1), (2), and (3) are diagrams showing changes in the increase value and the air-fuel ratio. 6 ... Pressure sensor, 7 ... CR filter, 10 ... Slot sensor 24 ... Fuel injection valve, 48 ... Rotation angle sensor.

Claims (1)

[Claims] 1. A pressure sensor for detecting an intake pipe pressure, a relaxation means for obtaining a relaxation value by relaxing a change in a signal output from the pressure sensor, and a control amount for controlling an engine based on the relaxation value. A control amount calculation means for calculating, a change rate calculation means for calculating the change rate of the relaxation value or the control amount, and a coefficient setting means for setting a correction coefficient that decreases as the absolute value of the change rate increases, A control device for an internal combustion engine, comprising: a correction unit that corrects the control amount based on a product of the change rate and the correction coefficient; and a control unit that controls the engine based on the control amount corrected by the correction unit. .