JPH03182655A - Fuel supply controller of internal combustion engine - Google Patents

Abstract

PURPOSE:To improve air-fuel ratio control performance by learningly compensating transient compensation based on deviation between intake air flow calculated by an engine operation condition means at a transient operation condition and a calculation result of intake air flow calculation means. CONSTITUTION:Basic fuel supply is set by a means B based on intake air quantity detected by an engine operation condition detection means A and an engine operation speed. When a transient operation condition is detected by a transient operation detection means C, basic fuel supply is compensated by the transient compensation of a transient compensation set means D according thereto. Intake air flow is calculated by an intake air flow calculation means F based on intake pressure by an intake pressure detection means E and the engine operation speed. The transient compensation set by the transient compensation set means D is learningly corrected by the transient compensation correction means G according to deviation between the calculated value and an intake air flow detection value. Air-fuel ratio control at the transient operation time is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to a fuel supply control device for an internal combustion engine, and more particularly to a technique for improving fuel supply amount correction control during transient engine operation.

<Prior Art> As a fuel supply control device for an internal combustion engine, there are conventionally known devices such as those shown below (Japanese Patent Laid-Open No. 59-203828).
(Refer to the publication number, etc.)

That is, the basic fuel injection amount Tp (=KXQ/N;
K is a constant). Furthermore, various correction coefficients C0E are applied depending on engine operating conditions such as engine temperature represented by cooling water temperature.
In addition to calculating F, the oxygen concentration in the exhaust, which is closely related to the air-fuel ratio of the engine intake air-fuel mixture, is detected by an oxygen sensor installed in the exhaust passage, and based on this detection result, the actual air-fuel ratio is set to the target air-fuel ratio. A feedback correction coefficient LAMBDA is set to approximate the fuel ratio, and a correction numerator s and the like are calculated to correct changes in the effective opening time of the electromagnetic fuel injection valve due to battery voltage. Then, the basic fuel injection amount 'rp is corrected based on the various correction terms described above to obtain a final fuel injection amount Ti (=TpXCOEFXLAMBDA
+Ts).

Here, by outputting a drive pulse signal with a pulse width corresponding to the fuel injection amount Ti set as described above to the t magnetic fuel injection valve provided in each cylinder at a predetermined timing, the engine is It was designed to inject and supply a fixed amount of fuel.

<Problems to be Solved by the Invention> By the way, according to the fuel supply control device as described above, as shown in FIG. There is a problem in that an amount smaller than the air amount is set as the detected intake air flow rate Q, and the air-fuel ratio at the initial stage of acceleration becomes over lean with respect to the target. If the air-fuel ratio becomes over-lean in the early stages of acceleration, a misfire may occur and acceleration shock may occur, and the nitrogen oxide NOx conversion efficiency of the three-way catalyst installed in the exhaust system for exhaust purification may decrease due to the air-fuel ratio. This situation worsens as the trend toward leaner conditions increases, and the amount of nitrogen oxide NOx emissions increases.

In order to solve this problem, the applicant set a correction amount for the detected value of the intake air flow rate Q based on the operating conditions during acceleration, and based on this correction amount, the intake air flow detected by the air flow meter is adjusted. We have previously proposed a device that can cope with the delay in detection response by an air flow meter by increasing iQ (see Japanese Patent Laid-Open No. 63-57834).

However, when using a hot wire flow meter as an air flow meter, if the detection characteristics change over time due to dust adhering to the hot wire, etc. (see Figure 15), the responsiveness will change from the initial state. Therefore, with the correction amount of the intake air flow rate Q set by matching the response of the air flow meter in the initial state, it becomes difficult to accurately correct when the response changes (first
(See Figure 4). That is, even if the correction amount of the intake air flow rate Q is matched so that the air-fuel ratio during acceleration becomes the desired value in the initial state, if the responsiveness of the air flow meter changes, the correction amount will be too large or too small. This makes it impossible to perform desired air-fuel ratio control.

Additionally, in order to cope with lean air-fuel ratio during acceleration due to a delay in the detection response of the air flow meter, in addition to the increase correction of the intake air flow rate Q, normal twisted fuel injection of a set amount of fuel according to the acceleration operating conditions is performed. The present applicant has also previously proposed a fuel supply control device configured to inject and supply fuel by interrupting the fuel injection (see Japanese Patent Laid-Open No. 63-65152 and Japanese Patent Laid-Open No. 63-9646).

However, even if such an interrupt injection is used to increase the amount of fuel to compensate for the delayed response of the airflow meter, if the responsiveness of the airflow meter changes from the initial state, it will be difficult to match the response to the initial state. It becomes impossible to obtain the desired air-fuel ratio with the included injection amount, and this point is similar to the case where the intake air flow IQ is corrected to increase the amount described above.

In this way, the present author has proposed various devices that can cope with the response delay during acceleration of air flow meters, but even if it takes a long development time to set the correction amount based on the initial response of the air flow meter, If the responsiveness of the air flow meter changes over time due to changes over time, such as dirt sticking to the hot wire, it becomes impossible to control the air-fuel ratio during acceleration to a desired value using transient correction control.

The present invention has been made in view of the above-mentioned problems, and it is possible to perform the desired correction even if the responsiveness changes due to aging when correcting the response delay of the detected value of the intake air flow rate, or to correct the initial response delay. It is an object of the present invention to provide a fuel supply control device having a learning function that can easily set a corresponding optimum transient correction amount.

<Means for Solving the Problems> Therefore, in the present invention, as shown in FIG. 1, first, as shown in FIG. a basic fuel supply amount setting means for setting a basic fuel supply amount based on the intake air flow rate and engine rotational speed; a transient operation detection means for detecting a transient operating state of the engine; transient correction amount setting means for setting a transient correction amount for correcting the basic fuel supply amount according to the engine operating state when the operating state is detected; A fuel supply control device for an internal combustion engine, comprising: a fuel supply means for supplying an amount of fuel corrected based on the amount of fuel to the engine;
an intake pressure detection means for detecting the intake pressure of the engine; an intake air flow rate calculation means for calculating the intake air flow rate based on the detected intake pressure and the engine rotational speed; Transient correction amount correction that calculates the deviation between the detected intake air flow rate and the intake air flow rate calculated by the intake air flow rate calculation means, and learns and corrects the transient correction amount to be set by the transient correction amount setting means based on this deviation. We decided to provide means and.

In addition to the above-mentioned configuration, there is also a steady operation detecting means for detecting the steady operating state of the engine as shown by the dotted line in FIG. Calculating the deviation between the intake air flow rate calculated by the calculation means and the intake air flow rate detected by the engine operating state detection means, and correcting the calculation constant of the intake air flow rate in the intake air flow rate calculation means based on this deviation. It is preferable to provide a flow rate calculation constant correction means.

On the other hand, as shown in FIG. 2, there is an engine operating state detection means for detecting an engine operating state including at least one intake air flow rate and one engine rotational speed, and based on the detected intake air flow rate and engine rotational speed, basic fuel supply amount setting means for setting the fuel supply amount; transient operation detection means for detecting the transient operation state of the engine; transient correction amount setting means for setting a transient correction amount for correcting the basic fuel supply amount based on the set basic fuel supply amount or an amount of fuel corrected based on the transient correction amount based on the set basic fuel supply amount, In a fuel supply control device for an internal combustion engine, the fuel supply control device includes an intake pressure detection means for detecting intake pressure of the engine, and an intake pressure detection means for detecting the intake pressure of the engine, and a means for supplying fuel to each cylinder at a predetermined timing. and an intake air flow rate calculation means that calculates an intake air flow rate based on the rotational speed, and a pair of the intake air flow rate detected by the engine operating state detection means and the intake air flow rate calculated by the intake air flow rate calculation means. Transient intake air flow rate data storage means for storing data for a predetermined period after the start of transient operation detected by the operating state detection means, and a timing adjustment for setting the amount of fuel actually supplied to each cylinder. The detected intake air flow rate stored in the transient intake air flow rate data storage means is stored in the transient intake air flow rate data storage means corresponding to a time that differs by a predetermined period from the supply amount setting timing. a timing difference transient correction amount correcting means for determining a deviation from the calculated intake air flow rate and correcting and setting a transient correction amount corresponding to the engine operating state at the supply amount setting timing in the transient correction amount setting means based on the deviation; .

Here, as shown by the dotted line in Figure 2, the intake air flow rate calculated by the intake air flow rate calculation means is simply averaged every predetermined crank angle rotation, and the simple average value is used as the intake air flow rate calculated by the intake air flow rate calculation means. It is preferable to provide simple averaging means for calculating the calculated intake air flow rate to be stored in the transient intake air flow rate data storage means instead of the air flow rate.

Further, as shown by the dotted line in FIG. 2, each cylinder is actually supplied at a timing that is different by a predetermined period from the supply amount setting timing, which is the timing at which the calculated intake air flow rate is picked up in the timing difference transient correction M correction means. It is preferable to provide a deviation detection timing setting means for setting the deviation detection timing based on the crank angle between the fuel supply amount setting timing and the predetermined timing of the intake stroke.

<Operation> According to the fuel supply control device for an internal combustion engine shown in FIG. 1, the basic fuel supply amount setting means sets the basic fuel supply amount based on the intake air flow rate and engine rotational speed detected by the engine operating state detection means. do. Furthermore, when the transient operation detecting means detects that the engine is in a transient operating state, the transient correction amount setting means sets a transient correction amount for correcting the basic fuel supply amount according to the engine operating state. The fuel supply means supplies the engine with the basic fuel supply amount or the amount of fuel corrected by the transient correction amount.

Furthermore, intake pressure detection means for detecting intake pressure is provided, and intake air flow rate calculation means calculates the intake air flow rate based on the intake pressure detected by this and the engine rotational speed detected by the engine operating state detection means. . Then, the transient correction amount correcting means determines the deviation between the detected value and the calculated value of the intake air flow rate in the transient operating state, and based on the magnitude of this deviation, the i! ! The J-transit correction/transient correction amount setting means learns and corrects the transient correction amount.

That is, in addition to directly detecting the intake air flow rate of the engine, the intake air flow rate is also determined indirectly from the intake pressure, and the excess or deficiency of the basic fuel supply amount based on the detected intake air flow rate is set based on the deviation between the two during transient operation. By learning and correcting the transient correction amount in order to eliminate the excess or deficiency, the transient correction amount set by the transient correction amount setting means is optimized. As a result, for example, the characteristics of the intake air flow rate detection change, which adversely affects the setting of the basic fuel supply amount, and even if the initially set transient correction amount cannot perform the desired air-fuel ratio control, the By learning and correcting as described above, it is possible to respond to changes in the intake air flow rate detection characteristics and maintain good air-fuel ratio controllability.

In addition, in the correction of the transient correction amount as described above, the intake air flow rate calculated from the intake pressure is used as the standard, so the intake air flow rate calculation constant correction means is used to correct the intake air flow rate at the steady state of the engine, which generally has good detection accuracy. In the operating state, the deviation between the calculated intake air flow rate and the detected value is determined, and the calculation constant for the intake air flow rate is corrected so that the intake air flow rate calculated based on this deviation matches the detected value.

In other words, since the detected value of the intake air flow rate is highly reliable during steady engine operation, the accuracy of the intake air flow rate calculated from the intake pressure and engine rotational speed is determined based on this highly accurate detected value of the intake air flow rate. This makes it possible to accurately capture the delay in the intake air flow rate detection value based on the calculated intake air flow rate during transient operation when the accuracy of the intake air flow rate detection value decreases due to response delay, etc.

Furthermore, in the fuel supply control device for an internal combustion engine shown in FIG.
The basic fuel supply amount setting means sets the basic fuel supply amount based on the intake air flow rate and the engine rotational speed detected by the engine operating state detection means. Further, when a transient operating state of the engine is detected by the transient operating state, the transient correction amount setting means sets a transient correction amount for correcting the basic fuel supply amount according to the engine operating state. The cylinder-specific fuel supply means supplies the basic fuel supply amount or the amount of fuel corrected by the transient correction amount to each cylinder of the engine at a predetermined timing.

Furthermore, intake pressure detection means for detecting intake pressure is provided, and intake air flow rate calculation means calculates the intake air flow rate based on the intake pressure detected by this and the engine rotational speed detected by the engine operating state detection means. . The transient intake air flow rate data storage means stores the intake air flow rate data for a predetermined period after the transient operation is detected by the transient operation detection means by pairing the detected value and the calculated value of the intake air flow rate.

Further, the timing difference transient correction amount correcting means picks up the data of the detected intake air flow rate at the supply amount setting timing of the fuel actually supplied to each cylinder from the storage means, and only for a predetermined period with respect to the setting timing. The calculated intake air flow rate data at different times is picked up from the storage means to find the deviation between the two, and based on this deviation, the transient correction amount setting means performs transient correction corresponding to the engine operating state at the supply amount setting timing. Correct the amount.

That is, during transient operation of the engine, there is a delay in the response of the intake air flow rate detection value, so the required fuel amount for each cylinder of the engine becomes a value corresponding to the intake air flow rate that is earlier than the fuel supply amount setting timing. Therefore, the deviation between the detected intake air flow rate when the fuel supply amount is set and the calculated intake air flow rate at a time that is different (in the future) by a predetermined period in which the intake air flow rate corresponding to the engine's required fuel amount appears is calculated. Based on this deviation, it is possible to set a transient correction amount corresponding to the timing at which the required supply amount appears after a predetermined period based on the engine operating state at the set timing.

In addition, in order to improve the accuracy of the intake air flow rate calculated based on the intake pressure and engine speed, the calculated intake air flow rate simple averaging means simply averages the calculated intake air flow rate every predetermined crank angle rotation, The simple average value is stored in the transient intake air flow rate data storage means in place of the intake air flow rate calculated by the intake air flow rate calculation means.

Furthermore, since the intake air flow rate corresponding to the required supply amount of the engine appears during the intake stroke of each cylinder, the deviation detection timing setting means is configured to adjust the intake air flow rate corresponding to the fuel supply amount actually supplied to each cylinder and the intake stroke. Based on the crank angle between the stroke and the predetermined timing, the storage timing of the calculated intake air flow rate to be compared with the detected intake air flow rate in the supply amount setting timing is set.

When supplying fuel to each cylinder, it is actually necessary to set the fuel supply amount corresponding to the maximum intake air flow rate during the intake stroke of each cylinder, but during transient operation, the intake air flow rate is detected. Because of the delay, a deviation occurs between the detected intake air flow rate at the timing when the fuel supply amount is set and the true intake air flow rate during the intake stroke in which fuel is drawn into the cylinder. In order to accurately correct the detection delay of the intake air flow rate and optimize the fuel supply amount, it is possible to compare the detected intake air flow rate at the supply amount setting timing with the calculated intake air flow rate at a predetermined timing of the intake stroke. , to optimize the amount of transient correction.

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

In FIG. 3, an engine l has an air cleaner 2 to an intake duct 3. Air is taken in via the throttle valve 4 and the intake manifold 5. A fuel injection valve 6 is provided in a branch portion of the air manifold 5 for each cylinder as a fuel supply means and a fuel supply means for each cylinder. fuel injection valve 6
1 is an electromagnetic fuel injection valve that opens when the solenoid is energized and closes when the energization is stopped; the valve opens when energized by a drive pulse signal from the control unit 12, which will be described later, and is fed under pressure from a fuel pump (not shown). Then, the pressure is adjusted to a predetermined pressure by the pressure regulator, and the fuel is injected into the intake manifold 5.

An ignition plug 7 is provided in the combustion chamber of the engine 1, which ignites a spark to ignite and burn the air-fuel mixture in the combustion chamber.

From engine 1, exhaust manifold 8. Exhaust duct9. Exhaust gas is discharged via the three-way catalyst lO and the muffler 11. The three-way catalyst 10 is an exhaust purification device that oxidizes Go and HC in the exhaust components, and also reduces NOx and converts it into other harmless substances.The conversion efficiency is closely related to the air-fuel ratio of the intake air-fuel mixture. In a relationship.

The control unit 12 includes a CPU and a ROM.

It is equipped with a microcomputer including a RAM, an A/D converter, and an input/output interface, and receives input signals from various sensors, performs arithmetic processing as described below, and controls the operation of the fuel injection valve 6. .

As the various sensors described above, a hot wire type air flow meter 13 is provided in the intake duct 3, and outputs a voltage signal Us corresponding to the intake air flow rate Q.

In addition, a crank angle sensor 14 is provided, and in the case of a 4-cylinder engine, a reference signal Ref and a reference signal Ref every 180 degrees of crank angle corresponding to the injection interval and 1 degree of crank angle are provided.
(or 2") unit signal (position signal) Pos. Here, by measuring the cycle of the reference signal Ref or the number of occurrences of the unit signal Pos within a predetermined time, the engine rotation speed N can be determined. It is possible to calculate.

Further, a water temperature sensor 15 is provided in the water jacket of the engine 1 to detect a cooling water temperature Tw representing the engine temperature. Further, the throttle valve 4 is provided with a throttle sensor 17 that detects the opening degree TVO of the throttle valve 4 using a potentiometer.

These air flow meters 13. crank angle sensor 14,
The water temperature sensor 15 and the like correspond to engine operating state detection means, and the throttle sensor 17 corresponds to transient operation detection means and steady operation detection means.

Further, an oxygen sensor 16 is provided at the gathering part of the exhaust manifold 8, and detects the air-fuel ratio of the intake air-fuel mixture via the oxygen concentration in the exhaust gas. The oxygen sensor 16 is made of zirconia (ZrO□), which is an oxygen ion conductor used as a solid electrolyte for concentration batteries, and has a cylindrical shape with a bottom, and its closed end faces the exhaust gas. The electromotive force is generated by the ratio of the oxygen concentration to the oxygen concentration. A platinum catalyst layer is provided on the outer surface of the zirconia tube by vapor-depositing platinum that functions as an oxidation catalyst. By combining unburned components such as CO and reducing the outside oxygen concentration to almost zero, the ratio of inside and outside oxygen concentrations is increased and a large electromotive force is generated.In contrast, in a lean mixture, the difference in oxygen concentration is small and there is almost no difference in oxygen concentration. No voltage is generated,
This is a well-known phenomenon in which the electromotive force changes suddenly near the stoichiometric air-fuel ratio.

Further, downstream of the throttle valve 4, an intake pressure sensor 18 is provided as an intake pressure detection means 7 for detecting the intake pressure PB, and an intake air temperature sensor 19 for detecting the intake air temperature T- is also provided. It is provided.

Here, the CPU of the microcomputer built into the control unit 12 performs arithmetic processing according to the program on the ROM shown as the flowcharts in FIGS. 4 to 10, and controls fuel injection (fuel supply amount). In this embodiment, the basic fuel supply amount setting means and the transient correction amount setting means.

Each means of the intake air flow rate calculation means, the transient correction amount correction means, the intake air flow rate calculation constant correction means 2 time difference transient correction amount correction means, the deviation detection timing setting means, and the calculated intake air flow rate simple averaging means is shown in FIG. - Software is provided as shown in the flowchart of FIG. Further, the transient intake air flow rate data storage means corresponds to a RAM provided in a microcomputer built into the control unit 12.

Here, various controls based on the flowcharts of FIGS. 4 to 10 will be explained.

The routine shown in the flowchart of FIG.
This is a routine for setting Ti, and is executed every minute time of about 10 m5. First, in step 1 (indicated as Sl in the figure; the same applies hereinafter), the opening degree TVO of the throttle valve 4 detected by the throttle sensor 17 is input.

Then, in the next step 2, the opening degree TVO input manually in the previous execution of this routine is subtracted from the opening degree TVO input in step 1 above, and the execution cycle of this routine (10m5
) Set the throttle valve opening change rate ΔTVO.

In the next step 3, the opening degree change rate Δ obtained in step 2 is
By comparing TVO and a predetermined value, it is determined whether the engine 1 is in a transient operating state (the opening degree TVO of the throttle valve 4 is changing at a predetermined rate or more).In step 3, the engine 1 If it is determined that the is in a transient operation state, the process proceeds to step 4, where the transient flag Ftr is determined. This transient flag Ftr has an initial value of zero, and is set to 1 in step 6, which will be described later, when the transient operation is determined for the first time. Therefore, when the current transient determination is the first time, it is determined to be zero, and the process proceeds to step 5.

In step 5, the counter cnt for picking up the operation data at the first time of transient detection is reset to zero,
During subsequent learning, the operation data at the first time of transient detection is picked up based on this counter cnt.In the 0th step 6, the transient flag Ftr is set to 1 indicating that the transient operation is in progress. In this way, when the transient flag Ftr is set to 1 at the first time of transient determination, when the next time this routine is executed, the transient flag Ftr is determined to be 1 in step 4. Steps 5.6 are not executed and the process proceeds to step 7.

In steps 7 to 10, an increase correction coefficient 1 ( Q
Set acc(%). Note that the correction coefficient for correcting the detection error of the air flow meter 13 during deceleration operation due to response delay can be set in the same way, but in this embodiment, the intake air flow rate Q during acceleration, which greatly affects drivability, can be set. Only the correction of the detection response delay is described.

First, in step 7, the acceleration correction coefficient KQI is searched from the map based on the opening change rate ΔTVO of the throttle valve 4 calculated in step 2. As shown in the graph in the flowchart, this acceleration correction coefficient KQI is set to a large value during sudden acceleration when the rate of change in opening degree ΔTVO is large, so that the air flow increases during sudden acceleration. This is so that a negative error due to a delay in the detection response of the meter 13 can be corrected.

In the next step 8, the detected intake air flow I during acceleration
Calculation delay correction coefficient KQ3 for correcting calculation delay of Q
is determined by searching from a map based on the engine rotational speed N calculated based on the output signal from the crank angle sensor 14. As shown in the graph in the flowchart, the calculation delay correction coefficient KQ3 is set to a larger value as the engine rotational speed N increases, so that the detected value of the intake air flow rate Q is corrected to increase. This is because the calculation of the intake air flow 1tQ based on the voltage signal Us outputted from the air flow meter 13 corresponding to the intake air flow rate Q is performed every predetermined minute time (for example, 4 ass) as described later. This is to correct the increase in error in the intake air flow rate Q due to a calculation delay at high engine speeds.

In the next step 9, the last time this routine was executed (IoIl)
The load correction coefficient KQ2 is searched from the map based on the basic fuel injection Tp calculated in step 19, which will be described later. As shown in the graph in the flowchart, this load correction coefficient KQ2 is set to a larger value when the basic fuel injection amount Tp a short time ago is smaller, and when accelerating from a state with a small engine load (initial acceleration), This is so that the detected value of the intake air flow rate Q is corrected to increase.

Then, in the next step 10, an increase correction coefficient KQacc during acceleration is calculated according to the following formula.

KQacc←(KQ1+KQ3) Here, the transient flag Ft
When it is determined that r is 1, it indicates that the vehicle was in a transient operating state before this time.

If it is determined that the transient flag Ftr is 1, the process proceeds to step 12, where it is determined whether the counter value cnt is greater than or equal to a predetermined value (for example, 400). The counter value an
t is reset to zero in step 5 at the first time of acceleration determination, and is incremented by 1 each time injection is started in each cylinder (every 180 degrees of crank angle in the case of 4 cylinders) in a routine shown in the flowchart of FIG. 8, which will be described later. If it is determined in step 12 that the counter value cnt is less than the predetermined value, it means that fuel has not been injected the number of times corresponding to the predetermined value since the end of the transient operation. Even if the change in the throttle valve opening degree TVO is minute, it cannot be recognized that the engine 1 has shifted to a stable steady operating state immediately after the transient operation.

On the other hand, if it is determined in step 12 that the counter value cnt is greater than or equal to the predetermined value, the engine 1 is operated for a period long enough to stabilize the operating state of the engine 1 after the transient operation ends, and the engine 1 is brought into a steady operating state. Assume that you have returned, and proceed to step 13.
After proceeding to step 14 and setting the transient flag Ftr to zero indicating that there is no transient state, the process proceeds to step 14.

Further, if it is determined in step 11 that the transient flag Ftr is zero, that is, the engine 1 is not in a transient operating state,
However, even when it is not immediately after the end of the transient operation, the engineer assumes that the operation is in a steady state and proceeds to step 14.

In step 14, the intake air flow rate QB calculated based on the intake pressure PB, engine rotational speed N, etc. at every minute time in the routine shown in the flowchart of FIG. 6 is calculated using the routine shown in the flowchart of FIG. crank angle 18
The ratio (Q a / Q m□〉 is determined, and the value is set as the correction coefficient ηvj!rn of the volumetric efficiency η, which is used when calculating the intake air flow rate based on the intake pressure PB.

That is, since the air flow meter 13 detects the intake air flow rate Q with high accuracy when the engine I is in steady operation, the intake air flow rate Q detected by the air flow meter 13 during steady operation is
Intake air flow rate Q calculated based on a and intake pressure PB.

Compared with V, the intake air flow rate Q based on the intake pressure PB
This is to check the calculation accuracy of s^1, and Quay is
When it is smaller (larger) than a, the correction coefficient ηv
lrn is set to increase (decrease) in It approaches the actual value detected by the meter 13.

In step 15, the volumetric efficiency η9 corresponding to the current operating state is retrieved from a map in which the volumetric efficiency η is stored in accordance with the intake pressure PB and the engine rotational speed N.

Then, in the next step 16, by multiplying the correction coefficient ηvlrn calculated in step 14 by the volumetric efficiency η found by searching from the map in step 15, the detected value of the air flow meter 13 and the intake pressure during steady operation are calculated. Based on the calculation error obtained by comparing the intake air flow rate QIAv, which is a calculated value based on PB, the volumetric efficiency η is corrected and newly set, and the map value of the corresponding area is rewritten to the corrected value. .

In this way, each time the engine 1 is operated steadily, the volumetric efficiency η is corrected based on the highly accurate intake air flow rate Qa detected by the air flow meter 13 at that time, so that the intake pressure PB and the accuracy of the intake air flow rate QB calculated using the volumetric efficiency η9 based on the engine rotational speed N.

Here, the detected intake air flow rate Qa setting routine shown in the flowchart of FIG. 5 will be explained.

This routine is executed at predetermined minute intervals (for example, 4 ms), and first, in step 31, the voltage Us output from the air flow meter 13 to the intake air flow rate Q of the engine 1 is input.

Then, in the next step 32, the intake air flow rate Q is set based on the voltage US input in step 31 by searching or calculating from the map.In the next step 33, the intake air flow rate Q set in step 32 is set. The average value (Q+Qa)/2 of the previous average intake air flow rate Qa is calculated and is set as the final detected intake air flow rate Qa. That is, instead of using the detected value by the air flow meter 13 as it is, the detected intake air flow rate Qa is obtained by averaging the previous average value at every predetermined minute time.

Further, to explain the intake air flow rate QB calculation routine based on the intake pressure PB shown in the flowchart of FIG. 6, this routine is executed every predetermined minute time (for example, 1 as), and first, in step 41, In calculating the average value of the intake air flow rate QB between the reference signals Ref, a counter value cnt1ms for counting the number of samples of the intake air flow rate QB is incremented by 1.

In the next step 42, the intake air flow rate QB is calculated according to the following formula.
Calculate.

Here, PB is the intake pressure (absolute pressure) detected by the intake pressure sensor 18, v6 is the total displacement of the engine l, and ηV is the intake pressure PB and engine rotation as shown in step 15 of the flowchart in FIG. The volumetric efficiency is retrieved from the map based on the speed N, where N is the engine rotational speed [rρ] and is converted to (rph) by multiplying by 60. Further, Rs is a gas constant, and T- is an intake absolute temperature detected by the intake air temperature sensor 19. Incidentally, the aforementioned volumetric efficiency ηV is learned and corrected during steady operation of the engine 1 according to the flowchart of FIG.

In this way, based on the intake pressure PB, the intake air flow rate QB
is calculated, in the next step 43, the sum of the intake air flow rate QB up to the previous time (i S Q B is added to the current step 42).
The calculated intake air flow rate QB is added up, and the intake air flow rate QB is integrated according to the routine shown in the flowchart of FIG. 7 until the reference signal Ref from the crank angle sensor 14 is output and reset to zero.

Returning to the flowchart of FIG. 4 again, the explanation will be given during steady operation where the volumetric efficiency ηV is learned and corrected in steps 14 to 16, and during transient operation where the counter value cnt is determined to be less than a predetermined value in step 12. Immediately after operation, there is no need to correct the intake air flow rate Qa detected by the air flow meter 13, so the process proceeds to step 17 and the increase correction coefficient KQacc used during engine acceleration is set to zero.

If the intake air flow rate Q is to be corrected even during deceleration operation of the engine I, the Q correction amount for deceleration is also set to zero in step 17.

Then, after the engine l is in a transient operating state AI (accelerating operating state) and the increase correction coefficient KQacc is set based on the operating state in step 10, and when the engine l is in a steady operating state or immediately after transient operation, the amount is increased in step 17. Correction coefficient KQac
After setting c to zero, proceed to step 18-1 and multiply the detected intake air flow rate Qa, which is averaged at predetermined minute intervals as described above, by (10 increase correction coefficient KQacc) to perform increase correction. In step 19, the final intake airflow IQACC is set as the final intake airflow 1tQACC, or when KQacc is set to zero in step 17, the detected intake airflow IQa is set as the final intake airflow IFQACC. Based on the intake air flow IQACC set in step 1 and the engine rotational speed N calculated based on the output signal from the crank angle sensor 14, the basic fuel injection amount 1゛p (←-K N
; K is a constant). In addition, this step 19
If the basic fuel injection 1]Tp calculated in step 1 is for transient (acceleration) operation at the next execution of the routine, it becomes the map value search data for the load correction coefficient KQ2 in step 9.

In the next step 20, the final fuel injection amount T4 is set according to the following formula. The basic fuel injection amount Tp and the fuel injection amount Ti are calculated as the opening time (IIs) of the fuel injection valve 6, that is, the pulse width of the drive pulse signal.

T i = T p XLAMBDA)lBLRcXc
OEF+ T s Here, the LAMBDA is an air-fuel ratio feedback correction coefficient LAMBDA, which is adjusted, for example, proportionally so that the actual air-fuel ratio detected by the oxygen sensor 16 approaches the target air-fuel ratio (theoretical air-fuel ratio) when a predetermined control condition is satisfied.・It is set by integral control.

Further, the KBLRC is a learning correction coefficient, which is set by learning the deviation of the air-fuel ratio feedback correction coefficient LAMBDA from the reference value, and the base air-fuel ratio without correction by the air-fuel ratio feedback correction coefficient LANBD^ is the target air-fuel ratio. The fuel injection amount Ti is corrected so that Furthermore, the COI'! F is various correction coefficients, and these various correction coefficients C0EF include a water temperature correction coefficient KTw set based on the cooling water temperature Tw detected by the water temperature sensor 15, a starting increase correction coefficient KAS, etc. .

Also, Ts depends on the battery voltage! This is for correcting changes in the effective valve opening time (responsiveness) of the magnetic fuel injection valve 6.

After calculating and setting the final fuel injection amount Ti in L7, in the next step 2I, the fuel injection amount Ti set as the opening time of the fuel injection valve 6 is converted to the crank angle Tia according to the following formula. Convert.

The mcntlIIIs is a count (ffic) when the reference signal Ref is output from the crank angle sensor 14.
ntims (set in step 60 described later), and the count value cnt1ms is the reference signal Ref.
Since it is a value that is reset to zero in step 61 of FIG. 7, which will be described later, for each input, and is incremented by 1 every IIIs according to the flowchart of FIG. 6, mcntl+s is the output interval time (+es) of the reference signal Ref, that is, the crank This is the time required for the shaft to rotate 180 degrees, which is the injection interval in a four-cylinder engine.

The crank angle conversion value Tia of the fuel injection tTi determines the injection start timing from the fuel injection amount Ti so that the injection end timing is at a predetermined crank angle position (for example, the intake valve opening timing of each cylinder) in accordance with the intake stroke of each cylinder. This is calculated backwards and used to variably control the injection start timing.

Next, the injection timing setting routine shown in the flowchart of FIG. 7 will be explained.

This routine is executed every time the crank angle sensor 14 outputs the reference signal Ref for every 180° rotation of the crankshaft. First, in step 51, the reference signal Ref is calculated from the intake pressure PB set in step 43. The integrated value SQB of the intake air flow rate QB is calculated as the counter value cnt at that time.
The average value Q, ^V of QB is calculated by dividing by lss. The counter value cnt1ms is a value that is assisted by 1 in step 4 every time the intake air flow rate QB is integrated in step 43, so it is the same number as the number of samples of the integrated data of the intake air flow rate QB, and will be described later. In step 61, the counter value cnt1ms and the integrated value SQB are both reset to zero for each reference signal Ref. This is an average value, and is updated and set every time the reference signal Ref is output.

In the next step 52, the crank angle Ti from the reference signal Ref to the injection end timing is determined based on the engine rotational speed N.
Set a+. In this embodiment, the injection end timing is set to be a constant time regardless of the engine rotational speed N, so the crank angle is fixed at 600@ in the graph shown in the flowchart. 600' is when the intake valve of each cylinder is opened based on the reference signal Ref corresponding to each cylinder! This is the crank angle up to tlI (opening start timing). In other words, the crank angle sensor 14 is set so that the reference signal Ref is output 600' before the intake valve opening timing of each cylinder (70 degrees before the intake TDC). The detected crank angle position is set by
In this embodiment, the injection end timing of each cylinder is fixed to the intake valve opening timing of each cylinder.

In the next step 53, it is determined whether the reference signal Ref outputted from the crank angle sensor 14 this time is the reference signal Ref for the #4 cylinder in the 4-cylinder engine, that is, as shown in FIG. When the intake valve is open, it is determined whether it is time to open the intake valve of the #1 cylinder again at a crank angle position rotated by 600° (Tits) from the current reference signal Ref.

In the case of this embodiment, when the reference signal Ref for the #4 cylinder is output, the #1 cylinder is in the intake stroke, and similarly the #2 reference signal R is output.
When ef is output, #3 cylinder is in the intake stroke, #1 reference signal Re
When f is output, cylinder #4 is in the intake stroke, and when #3 reference signal Ref is output, cylinder #2 is in the intake stroke. In this way, in order to make the reference signal Ref output from the crank angle sensor 14 correspond to each cylinder, one of the reference signals Ref is made larger in pulse width than the other so that they can be distinguished.

If it is determined in step 53 that the current reference signal Ref is the reference signal for the #4 cylinder, the process proceeds to step 54 to determine whether the current reference signal Ref is the reference signal for the #1 cylinder for measuring the period from the current reference signal Ref to the injection start timing in the #1 cylinder. counter critl
Reset cyf to zero and set this #l cylinder counter cnt
By counting up lcyf to a predetermined value corresponding to the injection start time, the injection start time is set so that the injection start time can be controlled so that the injection ends at the intake valve opening time of the #l cylinder.

On the other hand, if it is determined in step 53 that the current reference signal Ref is not the reference signal Ref of #4 cylinder, step 55
It is determined whether or not the reference signal Ref is for the #2 cylinder. When the reference signal Ref is for the #2 cylinder this time, the intake valve of the #3 cylinder is open as shown in Fig. 11, and as with the reference signal for the #4 cylinder, this Since the intake valve opening timing for the #3 cylinder is reached again at a crank angle position rotated by 600 degrees from the reference signal Ref for the #2 cylinder, the process proceeds to step 56 and measures up to the injection start timing for the #3 cylinder. Reset the 3 cylinder counter cnt3cyl to zero.

Similarly, in step 57, it is determined whether the current reference signal Ref is for the #1 cylinder, and the reference signal Re for the #l cylinder is determined.
When it is f, the process proceeds to step 58, and the #4 cylinder counter cnt4cyi, which measures up to the injection start timing in the corresponding #4 cylinder, is reset to zero.

If it is determined in step 57 that the current reference signal Ref is not for the #1 cylinder, the current reference signal Ref is for the #1 cylinder.
．． #2. Since it is in a state that does not correspond to any of #4, the reference signal Ref is inevitably for the #3 cylinder, so the process proceeds to step 59 and the #2 cylinder counter cnt2cyf is set to measure up to the injection start timing in the corresponding #2 cylinder. Reset to zero.

In this way, when the counters for measuring the injection start timing of each cylinder are sequentially reset to zero based on the reference signal Ref outputted from the crank angle sensor 14 this time, in the next step 60, Average value Q of intake air flow rate QB between reference signals Ref. In order to obtain V, the counter value cnt1ms, which is incremented by 1 every IIls between the reference signals Ref, is set to scntlms), and this cnt, which is the output interval time of the reference signal Ref, is
Fuel injection I in step 21 described above using l*s
It is used for converting Ti into a crank angle.

In the next step 61, a counter (!cnt11
1s and the integrated value SQB of the intake air flow IQB are reset to zero.

The injection start timing measurement counters cntlcy l to cnt4cy are reset to zero as described above.
Il is counted up every unit crank angle according to the routine shown in the flowchart of FIG. 8, and the injection start timing is detected based on this counted up value.

The routine shown in the flowchart of FIG. 8 is executed every time the unit signal Pos output from the crank angle sensor 14 for each unit crank angle (lo) is input. An injection start timing measurement counter (cntlcyl-cnt4c) is reset to zero for each reference signal Ref corresponding to each cylinder according to the routine shown in the flowchart in FIG.
y l ) are all increased by l. That is, the injection start timing measurement counter receives the corresponding reference signal Ref.
Each signal is reset to zero at the time of input, and then air-filled every time the unit signal Pos is output from the crank angle sensor 14, and the injection start timing of each cylinder is detected from each reference signal Ref.

In the next step 72, the fuel injection set in the above step 21 is determined from the crank angle Tit■ from the reference signal Ref set in the above step 52 to the injection end timing (intake valve opening timing in the case of this embodiment). The value obtained by subtracting the crank angle conversion value Tia from the amount Ti, that is, the crank angle from the reference signal Ref to the injection start time (Tile-Tia)
When the reference signal Ref corresponding to the #4 cylinder is output from the crank angle sensor 14, the #1 cylinder injection start tie 4 is reset to zero and then incremented by 1 for each unit signal Pos.
and the value of the counter cntlcy1 for measurement measurement.

The counter cntlcyj for measuring the injection start timing
! ~cnt4cyffi is zero-reset with the corresponding reference signal Ref and is 17 every l° unit crank angle.
Counter c for measuring injection start timing
The values of ntlcy l to cnt4cy l indicate the crank angle from the predetermined reference signal Ref, and in step 72 cntlcyjl! = T i tm - T i
When it is determined as a, it indicates that it is the injection start timing for the #1 cylinder.

Therefore, in step 72 cntlcyj! = T i
If it is determined that it is tm-T i a, the process proceeds to step 73, and starts outputting a drive pulse signal with a pulse width corresponding to the fuel injection ITi to the fuel injection valve 6 provided in the #l cylinder, Fuel is injected and supplied to the #1 cylinder so that the injection ends at the opening timing of the intake valve.

Meanwhile, in step 72, cntlcyj! ≠Tits-T
If it is determined that ia is the injection start timing for the #1 cylinder, this indicates that the injection start timing for the #1 cylinder has not yet arrived, so the process proceeds to step 74 and similarly sets the injection start timing counter cnt2cyl and Tits-Tia for the #2 cylinder. By comparison, it is determined whether or not it is the injection start timing for the #2 cylinder.

Here, cnt2cyi! = Ti tm - Ti a
If it is determined that it is the injection start timing for the #2 cylinder, the process proceeds to step 75 and starts outputting a drive pulse signal having a pulse width equivalent to the fuel injection amount Ti to the fuel injection valve 6 of the #2 cylinder. , the fuel is injected and supplied to the #2 cylinder so that the injection ends at the opening timing of the intake valve.

Also, in step 74, cnt2cyf * T i t
If it is determined that the timing is s - T i a, and it is determined that it is not the injection start timing for the #2 cylinder, then step 7 is performed.
Proceed to step 6 and determine whether it is the injection start timing for #3 cylinder cnt3c)/! cnt3cy
If It = Tits-Tia, it is determined that it is the injection start timing of the #3 cylinder, and in step 77, the output of the drive pulse signal is started to the fuel injection valve 6 of the #3 cylinder, and the intake air of the #3 cylinder is started. Make sure that the injection ends when the valve opens.

Furthermore, in step 76 cnt3cy 1≠Tits-T
When it is determined that the timing is ia, the process proceeds to step 78, and by comparing cnt4cyl and Tits-Tia, it is determined whether or not it is the injection start timing for the #4 cylinder. Here, cnt4cyf −T i tm −
If T ia, it is the injection start timing for the #4 cylinder, so the process advances to step 79 and the fuel injection valve 6 of the #4 cylinder is activated.
Then, output of a drive pulse signal having a pulse width corresponding to the fuel injection amount T1 is started, and fuel is injected and supplied to the #4 cylinder so that the injection ends at the opening timing of the intake valve.

Also, in step 78, cnt4cy It≠Tits+
-Tia, since the injection start timing has not yet reached in all cylinders, this routine is ended as is.

On the other hand, it is determined that it is the injection start timing of the predetermined cylinder, and the above-mentioned step? When the output of the drive pulse signal to the corresponding fuel injection valve 6 is started at any time of 3.75.77.79, the process proceeds to step 8o.

In step 80, it is determined whether the counter (Iicnt), which is reset to zero at the first time of transient operation determination in step 5 described above, is zero.

The counter value cnt is reset to zero at the first time of transient operation determination, and is incremented by 1 at step 83, which will be described later, each time injection is started.
When it is determined that the counter value cnt is zero, this indicates that injection has started for the first time after transient operation has been determined.

Here, if it is determined that the counter (icnt) is zero, the process proceeds to step 81, and each time injection is started, the address ADH(i) of the map on the RAM that stores various operating data at that time as one set is set. Among them, the address & ADR (i) indicating the part where the current driving data is stored is ADH.
Set to PICK.

Therefore, the address ADR set in ADHPICK
By (i), it is possible to pick up operational data at the start of the first injection in a transient operation.

On the other hand, if it is determined in step 80 that the counter value cnt is not zero, step 81 is jumped and step 82
Proceed to. In step 82, various types of operational data for the current injection are stored in order by item at address ADH(i) of the operational data storage map. The various operating data include counter value cnt, air flow meter 1
Intake air flow rate Qa detected by 3, increase correction coefficient K
Qacc, the average of the intake air flow rate QB based on the intake pressure PB between the reference signal Ref (+IQ, , v, fuel injection amount T
Crank angle conversion of i (1iTia, crank angle Tim from reference signal Ref to injection end time, calculation delay correction coefficient KQ3゜acceleration correction coefficient KQI, load correction coefficient KQ
2. Engine rotation speed N, throttle valve opening change rate ΔTVO
11 data (0 to 10) are stored in this order.

When the operation data at the start of injection in a predetermined cylinder is stored in the predetermined address ADH (il) in step 82,
In the next step 83, the counter value cnt is incremented by 1 so that the counter (icnt) represents the number of injections from the start of the transient operation.

Then, in the next step 84, the address ADR(i
) is incremented by 1, so that the operation data at the time of the next injection start is stored in the area (address number) next to the address ADR[i) where the operation data was stored in step 82 this time.

In the next step 85, the address ADR(i) incremented by 1 in step 84 becomes the maximum address ADR(i).
It is determined whether the value has reached 500, which exceeds =499. Here, when it is determined that the address ADR(t)=500, the process advances to step 86 and the address ADH(i
) to zero and set the map's first address ADH.
(i) The operation data is stored sequentially starting from the position.

That is, the operational data map whose storage portion is specified by address ADH(i) stores the operational data sequentially from zero of address ADR(i) every time injection is started, and i) is maximum 4
99 (up to 499 sets of operating data can be stored), the operating data is updated and stored from the zero position of the address again from the next time, and the operating data of the address where the counter cnt is zero is transient. This is the operation data at the time of the first injection during operation, and the data at the time of transient operation is stored for a predetermined period after the address.

Next, learning and correction of the increase correction coefficient KQacc for correcting the intake air flow rate Qa during acceleration operation will be explained according to the routine shown in the flowchart of FIG.

This routine is executed as a background Jijibu (BGJ), and first, in step 101, the transient flag Ft set in the flowchart of FIG.
r is determined, and when the transient flag Ftr is 1 and the engine 1 is in the transient operating state 71M (or immediately after the transient operating state), the process proceeds to step 102.

In step 102, the learning permission flag Ftrfrn is set to 1 indicating a learning permission state, and in the next step 103, the transient learning experience flag Ftrlrnex is set to 1 indicating that the learning has been completed. The transient learning experience flag F tr A rnex has an initial value of zero when the engine 1 is new (at the time of shipment) or when the microcomputer has been completely reset, but when the engine 1 experiences transient operation, the above step In the ROM constant transition control routine shown in the flowchart of FIG. 10, the increase correction coefficient KQac stored in the ROM is
It is used to determine whether to execute control to transfer the map of various correction coefficients that determine c to the RAM and make it rewritable.

That is, the routine shown in the flowchart of FIG. 10 is an initialization routine executed when the ignition switch (IGSW) of the engine 1 is turned on.

Here, first, in step 91, the transient learning experience flag F
tr 1 rnex is determined, and if the transient operation is completely inexperienced and the transient learning experience flag F tr Il rnex is zero, the process advances to step 92 and an initial map of the acceleration correction coefficient KQI stored in the non-rewritable ROM is determined. In addition to migrating to rewritable RAM,
In the next step 93, the initial map of the computation delay correction coefficient KQ3 stored in the non-rewritable ROM is transferred to the rewritable RAM.

This is the KQI initially set at the shipping stage of engine 1.
, KQ3 and other various control constants, it is necessary to use a non-rewritable ROM as the memory. , the map values of KQI and KQ3 cannot be directly rewritten, so the KQI and KQ3 correction amounts are stored separately, and the KQI and KQ3 retrieved from the map are stored each time the acceleration increase correction of the intake air flow rate Qa is performed. etc. and these correction amounts to calculate the final acceleration increase correction coefficient KQa.
It is necessary to set cc, which may lead to an increase in calculation time.

Therefore, in this embodiment, the acceleration increase correction coefficient KQa C,
After learning the error amount of C, the initial map of KQI and KQ3 stored in the ROM can be transferred to a rewritable RAM in advance so that the MA knob value can be directly rewritten based on this learning. The data is rewritten and the KQI and KQO values are read from the RAM.

Returning to the flowchart of FIG. 9 again, if it is determined in step 101 that the transient flag Ftr is zero, the process advances to step 104 and the learning permission flag F t is set.
The learning permission flag F t for determining r i rn
Since r l rn is also set to 1 when a transient operation state of engine l is detected and the transient flag Ftr is set to 1, the learning permission flag F is set here immediately after the transient operation ends. It is determined that tr 1 rn is 1, and the process proceeds to the next step 105.

In addition, since the learning permission flag F tr l rn is reset to zero at the end of learning as described later, even after a transient operation, when a series of learning to be described later has been completed, the flag F tr lrn is reset in step 104. j! rn
is determined to be zero, and this routine is terminated.

In step 104, the learning permission flag F tr Il rn
is determined to be 1 and the process proceeds to step 105, when the first determination of transient operation is made, the ADRPICK
The address ADH(i) where the first transient operation data set in is stored is set to a.

That is, in step 105, a indicates the address number where the first transient operation data is stored.

In the next step 106, maxb indicating the maximum correction error amount of the increase correction amount KQacc and the address counter m are each reset to zero.

Then, in the next step 107, from the start of the predetermined injection corresponding to the operation data stored in the address ADR(a) of the operation data storage map, the intake valve is opened in the cylinder into which the injected fuel is supplied. The number of bones (Z) corresponding to the interval of the reference signal Ref up to the timing at which the maximum intake air flow rate Q occurs is calculated according to the following formula.

This is how long after the set timing of the fuel injection amount Ti (injection start), when the intake valve actually opens and the intake air flow rate Q becomes the largest (the timing at which the true fuel demand amount appears). It detects
Thereby, it is detected how much increase correction amount is truly necessary for the intake air flow 11Q when the fuel injection lTt is set.

Note that in the case where the process advances from step 106 to step 107, when step 107 is executed for the first time, the address A
DH(a) indicates the address where the operation data at the time of the first injection after the transient operation determination is stored, as set in step 105.

Here, the ADR (a, 5) is the address number a
ADH(a, 5) represents the fifth data out of the plurality of operating data stored in the address ADH(
This is the value of Tit+w (crank angle from the reference signal Ref to the injection end time) stored in a) (see step 82 in the flowchart of FIG. 8). Similarly, ADR (a.

4] indicates the crank angle equivalent fuel injection amount Tia, which is the fourth operation data stored at address ADR(a).

Furthermore, in the case of this embodiment, as shown in FIG.
Reference signal R from the crank angle sensor 14 at 70° before DC
Since the setting is such that ef is output, 540' + after the reference signal Ref for injection timing control is output.
The intake TDC is 70@=610@, and 610° in the above equation is the crank angle from the reference signal Ref to the intake TDC. Furthermore, in the case of this embodiment, the intake T D
The intake valve is set to start opening 10 degrees before C, and the timing where the intake air flow 1tQ is the largest with the intake valve open is BDC after 180 degrees rotation from intake TDC. .

Therefore, the numerator of the above equation is from the injection start time to the intake B
It shows the crank angle up to DC, and the denominator is 180
By dividing by °, the reference signal R is calculated from the injection start time.
It shows how many ef intervals (180'' intervals) after which the intake BDC, which is the timing for requesting an increase, is reached (see FIG. 12).

Furthermore, in step 107, Z obtained as above
Y (%) which is the true increase correction coefficient KQacc based on
is calculated according to the formula below.

Here, the ADH(t) is the fuel injection interval of 180
” This is an instruction to store operation data for each
is the increase request timing for a predetermined injection (intake BDC
), and the simple average value Qmav of the intake air flow rate QB between the reference signal Ref, which is the third data at this intake increase timing, is ADRC.
a+Z, 3). Further, ADH (a, 1) indicates the intake air flow IQa which is the first data (see step 82 of the flowchart in FIG. 8) at the address where the operational data at the time of the predetermined injection is stored.

Therefore, if the above equation is rewritten, it becomes as follows, and the numerator is the intake air flow rate corresponding to the true required fuel appearing at the intake BDC and the intake air flow IQa that is the basis for setting the fuel injection amount Tt stage at the start of injection. represents the deviation of Qa
Since KQacc)=QACC, this value of Y is also the truly required increase correction coefficient KQacc.

Qa when Ti is set is Qmav (a + Z) - Qa [a] (1 + true Q ace), and at the start of injection (final fuel injection 1
The intake air flow rate Qa (a) at Ti setting) is
The value approximated to the intake air flow rate Qmav [a + Z) corresponding to the true required fuel stem that appears during the subsequent intake stroke becomes the true increase correction coefficient KQacc.

In the next step 108, the required increase correction coefficient Y (%) (true increase correction coefficient K
Qacc) from the actually used increase correction coefficient KQac
c(ADR(a, 2)) is subtracted to obtain the correction coefficient KQacc
Find the error 1b("Y-ADR(a,2).

Note that when the current setting value of the increase correction coefficient KQacc can increase the amount of Qa when the fuel injection amount Ti is set until it substantially matches QllAV at the intake BDC, which is the increase request timing, the error amount is approximately zero. It should be.

Then, in the next step 109, the maximum error amount maxb up to the previous time is compared with the absolute value of the error amount S obtained in step 108 above, and the current error 1b is determined to be the maximum of the error amount S up to the previous time. If (maxb<1b
l), proceed to step 110 and this time step 108
Set the error amount obtained in step to the maximum error amount maxb,
In the next step 111, correction coefficient data (KQI, KQ3.KQ2) and operation data (N) in the current ADH (a) are stored.

ΔTVO), furthermore, the true increase correction coefficient Y is m as the data at the time of picking up the maximum error amount maxb.
axKQl, maxKQ3. maxKQ2. maxN,
Set ΔTVOm to maxY.

When it is determined in step 109 that the error amount calculated this time is smaller than the previous maximum error amount maxb, and after setting the data for picking up the maximum error finaxb in step 111, the process advances to step 112 to address Increase counter m by 1.

Then, in the next step 113, the number of verifications of the error quantity set according to the engine speed N is compared with the address counter m incremented by 1 in step 112, and the error quantity is detected a predetermined number of times. Determine whether or not you have done so. Here, if it is determined that the error amount has not been detected a predetermined number of times, the process proceeds to step 114, where the address counter m that was incremented by 1 in step 112 is added to the current a, and a is newly set. , return to step 107 again.

That is, when the transient operation ends, first, the correction error amount for the first injection in the transient operation is calculated based on the operation data storage of the first injection in the transient operation determined by ADHPICK, and then the correction error amount is calculated based on the engine rotational speed N. The correction error 1b is calculated for the number of injections, and the correction coefficient when the maximum error 1maxb is obtained,
It picks up the driving state and the required correction coefficient Y.

The number of times the correction error amount is picked up is set to a high value at high engine speeds based on the engine speed N as described above, but this is because the lean time of the air-fuel ratio is Although it is approximately constant depending on the engine rotation speed N, since the correction error amount is data for each injection, it is necessary to pick-amplify the correction error amount for a larger number of injections as the engine rotation speed N is higher. This is because there is.

As described above, when a predetermined number of errors have been picked up in the increase correction control of the intake air flow rate Qa, the process proceeds from step 113 to step 115, where the speed and speed of the acceleration operation during the transient operation learned this time is determined by the maximum error. The determination is made by comparing the throttle valve opening change rate ΔTVOm, which is the data when the amount maxb is obtained, with a predetermined value.

Here, if the acceleration is slow acceleration, step 11
6, the membership value mKQ3 representing the appropriateness of the increase correction coefficient KQacc (calculation delay correction coefficient KQ3) based on the fuzzy theory is set to the maximum error amount maxb.
Search and find from the map based on. That is, the calculation delay correction coefficient KQ3 is determined in advance based on the maximum error amount maxb.
A membership function is set to indicate the degree of appropriateness as human subjective ambiguity. For example, if the maximum error amount maxb is zero, the membership value mKQ3 is 1, which indicates the highest degree of appropriateness. As the maximum error amount maxb becomes larger, the membership value mKQ3 is set smaller to indicate a state where the degree of appropriateness is lower.

In the next step 117, the weighting used in weighted averaging of the membership values mKQ3 based on the engine rotational speed maxN when the maximum error amount maxb is obtained is determined by searching the constant XKQ3 from the map. In this weighting, the constant XKQ3 indicates the weighting for the weighted average value mKQ3AV up to the previous time, and the maximum error amount maxb
In a state where the obtained engine rotational speed maxN is high and the learning frequency is low, the weighting constant XKQ3 is set smaller to avoid the weighted average value mKQ3AV from changing rapidly with few learning opportunities, while maxN In a state where the learning frequency is low and the learning frequency is high, the weighting is done by increasing the constant XKQ3 to ensure the speed of learning progress.

When the weighting constant XKQ3 is set in step 117,
In the next step 118, the weighted average value mKQ of the membership value mKQ3 is stored in correspondence with the engine rotational speed m a x N when the maximum error tmaxb is obtained.
From the map of 3AV, the weighted average value mKQ3AV corresponding to the current engine rotational speed maxN is searched and obtained.

Then, in the next step 119, the weighted average value mKQ3AV up to the previous time and the current maximum error amount ma are calculated according to the following formula.
A weighted average calculation is performed with the membership value mKQ3 set based on xb, and mKQ3 is calculated corresponding to maxN.
The value corresponding to the current maxN in the map for recording AV is rewritten to mKQ3AV, which is the current weighted average result.

In step 120, mKQ3AV, which is the weighted average result in step 119, is compared with a predetermined value (for example, 0.9), and it is determined whether the degree of appropriateness of the calculation delay correction coefficient KQ3 is equal to or higher than an allowable level. Here, when mKQ3AV is greater than a predetermined value, the calculation delay correction coefficient KQ3 set in the map approximately matches the required amount, and the correction error zb is sufficiently small, the calculation delay correction coefficient KQ
Since there is no need to learn and correct 3, proceed to step 132 and proceed by resetting the learning permission flag Ftrlrn to zero and setting the increase correction coefficient KQacc after transient operation.
This routine ends without any learning correction.

On the other hand, mKQ3A of the weighted average result in step 120
If it is determined that V is less than a predetermined value and the calculation delay correction coefficient KQ3 set in the map does not accurately match the required amount and a correction error amount exceeding the allowable level occurs, the process proceeds to step 121. .

In step 121, m corresponding to the true increase correction coefficient KQacc required when the maximum error amount maxb is obtained.
From the value obtained by dividing axY (%) by the calculation delay correction coefficient maxKQ2 when the maximum error tmaxb is obtained, the acceleration correction coefficient max when the maximum error amount maxb is also obtained.
The value obtained by subtracting xKQ1 is set as the correction correction coefficient C.

Here, since KQacc ← (KQ1 + KQ3)

In the next step 122, the current maximum error amount ma
The true data corresponds to the engine rotational speed maxN when xb is obtained, and the KQ of the corresponding area of the N-KQ3 map is
3 Rewrite the map value to C above.

That is, when the transient operation before learning is in a slow acceleration state, the increase correction is performed in the operating state with the largest correction error amount maxb obtained by learning without correcting the load correction coefficient KQ2 or the acceleration correction coefficient KQI. Coefficient K
Only the computation delay correction coefficient KQ3 is rewritten and corrected so that Qacc matches the required true value (Y). This is because, when the engine 1 is operated at slow acceleration, the calculation delay correction coefficient KQ3 has a greater influence on the set value of the increase correction coefficient KQacc than the acceleration correction coefficient KQI.

After rewriting the map value of the computation delay correction coefficient KQ3 in step 22, the process proceeds to step 132, where the learning permission flag F tr l rn is reset to zero, and this routine ends.

On the other hand, when it is determined in step 115 that the transient operation is a sudden acceleration operation, the process proceeds to step 123, and the engine rotation speed maxN when the maximum error 31maxb is obtained is
The weighted average value m, KQ3 of the membership 4amKQ3a indicating the appropriateness of the calculation delay correction coefficient KQ3 corresponding to
From the map that stores aAV, the current maximum error amount ma
Search and find the value corresponding to the engine rotational speed maxN at the time of xb pickup.

Then, in the next step 124, the membership value mKQ3aAV obtained by searching from the map is compared with a predetermined value (for example, 0.9), and the membership value mKQ3aAV indicating the appropriateness of the calculation delay correction coefficient KQ3 is determined. It is determined whether or not the error amount is a predetermined value or more and is sufficiently small.

Here, the membership value mKQ3aAV is calculated based on the maximum error amount maxb in the slow acceleration operation state after the engine 1 is slowly accelerated, and the map value is rewritten. Since the learning correction of the calculation delay correction coefficient KQ3 is performed at a certain time, the calculation delay correction coefficient KQ3 in slow acceleration operation
This indicates the degree of appropriateness.

Therefore, in step 124 the membership value mKQ3a
When it is determined that AV is less than a predetermined value, the calculation delay correction coefficient K corresponding to the current engine rotation speed maxN
This indicates a state in which learning correction of Q3 has not progressed sufficiently in a slow acceleration state, and if learning correction of acceleration correction coefficient KQI, which will be described later, is executed in this state, the balance of learning correction will be lost and K
Since there is a risk that learning of QI and KQ3 will not progress together, it is desirable to execute learning of KQ3 after learning correction of the calculation delay correction coefficient KQ3 after the slow acceleration operation has sufficiently progressed.

Therefore, when it is determined in step 124 that mKQ3aAV<predetermined value, the process proceeds to step 132 without executing the learning correction of the increase correction coefficient KQacc, after which the learning permission flag F tr l rn is reset to zero. Terminate the routine.

Meanwhile, in step 124, m K Q 3 a A V
When it is determined that the predetermined value is greater than or equal to the predetermined value, learning of the computation delay correction coefficient KQ3 after the slow acceleration operation has progressed sufficiently;
Since the maximum error amount maxb is suppressed to a sufficiently small level in the slow acceleration operation, the process proceeds to step 125 to execute learning correction of the acceleration correction coefficient KQI corresponding to the rapid acceleration operation state.

In step 125, the maximum error cover rnaxb of the increase correction coefficient KQacc during the current transient (sudden acceleration) operation is determined.
A membership value mKQ1 representing the appropriateness of the acceleration correction coefficient KQI is searched from the map based on the maximum error amount maxb picked up this time.

That is, a membership function is set in advance to indicate the degree of appropriateness of the acceleration correction coefficient KQI based on the maximum error amount maxb as human subjective ambiguity. For example, if the maximum error 31maxb is zero, it is the most appropriate. 1 indicating a high degree of membership value mKQI
The membership value mKQ1 is set to be smaller as the maximum error amount maxb becomes larger, so as to indicate a state where the degree of appropriateness is lower,
This is similar to membership (a m K Q 3).

In the next step 126, the weighting used in weighted averaging of the membership value mKQ1 based on the throttle valve opening change rate ΔTVOm when the maximum error i1maxb is obtained is determined by searching the constant XKQI from the map.

For this weighting, the constant XKQI is the weighted average value m up to the previous time.
It shows the weighting for KQ I A V,
In a sudden acceleration state that is not regularly used and where the maximum error amount maxb is obtained, the opening change rate ΔTVOm is large and the learning frequency is low, the weighting constant XKQI is set smaller and the weighted average (amKQIAV changes rapidly with fewer learning opportunities) On the other hand, in a state where ΔTVOm is small and learning is frequently performed even during rapid acceleration, the weighting is done by increasing the constant XKQI to ensure the speed of progress of learning.

In the next step 127, the weighted average value mKQI of the membership value mKQ1 is stored in correspondence with the opening degree change rate ΔTVOm when the maximum error 11maxb is obtained.
The weighted average value mKQIAV corresponding to the current opening degree change rate ΔTVOm is searched and obtained from the AV map.

Then, in the next step 12B, the weighted average value mKQIAV up to the previous time and the maximum error amount ma of this time are calculated according to the following formula.
A weighted average calculation is performed with the membership value mKQ1 set based on xb, and mKQ is calculated corresponding to ΔTVOm.
The value corresponding to the current ΔTVOm in the map for recording IAV is rewritten to the current weighted average result mKQIAV.

In step 129, the mKQIAV resulting from the weighted average in step 128 is compared with a predetermined value (for example, 0.9), and it is determined whether the degree of appropriateness of the acceleration correction coefficient KQI is equal to or higher than an allowable level. Here, when mKQIAv is greater than a predetermined value, the acceleration correction coefficient KQI set in the map approximately matches the required amount, and the correction error amount is sufficiently small, the increase correction coefficient KQacc
Since it is not necessary to learn and correct the (acceleration correction coefficient KQI), proceed to step 132, reset the learning permission flag Ftrlrn to zero, and perform learning correction of the increase correction coefficient KQacc after transient operation. This routine ends without any change.

On the other hand, if it is determined in step 129 that mKQIAV<predetermined value, and the degree of appropriateness of the acceleration correction coefficient KQI is less than the allowable level, the process proceeds to step 130;
In step 130, m corresponding to the true increase correction coefficient KQacc required when the maximum error amount maxb is obtained.
The correction correction coefficient C is the value obtained by subtracting the calculation delay correction coefficient KQ3 when the maximum error amount maxb is obtained from the value obtained by dividing axY (%) by the calculation delay correction coefficient maxKQ2 when the maximum error amount i1maxb is obtained. Set.

Here, since KQacc←(KQI+KQ3)
This is the value of KQI to match.

In the next step 131, C, which is the value of the acceleration correction coefficient KQ1 in order to match the true increase correction coefficient KQacc obtained in step 130, is changed to the current maximum error 1max.
Assuming the true data corresponding to the opening change rate ΔTVOm obtained by b, the KQI of the corresponding area of the ΔTVO-KQ1 map
Rewrite the map value to C above.

That is, when the transient operation before execution of learning is a sudden acceleration state, first, it is determined whether the appropriateness of the calculation delay correction coefficient KQ3 has reached a sufficient level in learning after the slow acceleration operation state, and the calculation delay correction is performed. When the degree of appropriateness of the coefficient KQ3 is sufficient, the increase correction coefficient KQacc is calculated based on the map value in the operating state with the largest correction error amount maxb obtained by learning, without modifying the load correction coefficient KQ2 or the calculation delay correction coefficient KQ3. Only the acceleration correction coefficient KQI is rewritten and corrected so that it matches the required true value (Y).

After rewriting the map value of the acceleration correction coefficient KQI in step 131, the process advances to step 132, where the learning permission flag F tr l rn is reset to zero, and this routine ends.

As described above, according to this embodiment, the detected intake air flow rate Qa, which is the basis for setting the fuel injection IT stage, and the intake air flow rate corresponding to the true required fuel amount appearing at BDC in the intake stroke after injection is started. By comparing the value calculated from the intake pressure PB, the increase correction coefficient KQacc is determined.
The true required value of the increase correction coefficient KQacc is determined and the determining factors of the increase correction coefficient KQacc are learned and corrected according to the transient operating condition so that the increase correction coefficient KQacc matches this required value.
Even if the detection characteristics of the air flow meter 13 change due to dust adhering to the hot wire, etc., and the detection response during transient periods changes compared to when Qacc is initially set, the increase correction coefficient KQac
By maintaining and correcting c at an optimum value, it is possible to maintain good air-fuel ratio controllability during transient operation.

In addition, since the increase correction coefficient KQacc can be learned and corrected to the optimum value as described above, if learning as shown in this example is performed at the development stage of the fuel supply control device, correction during transient operation (transient correction This makes it possible to automatically match the fuel supply control device (quantity), and it becomes possible to significantly reduce the development period and development cost of the fuel supply control device.

In this embodiment, the fuel supply control device is configured to correct the intake air flow rate Qa detected by the air flow meter 13 during transient operation, but the detected intake air flow rate Qa
When the basic fuel injection amount Tp set based on the
Even when injecting an amount of fuel according to the transient operating state between 79 and 79, the fuel correction amount and interrupt injection amount can be learned by comparing the detected intake air flow rate Qa and the calculated intake air flow rate Qmav in the same way as in this embodiment. By carrying out the correction, even if the detection response of the air flow meter 13 changes, it is possible to maintain good air-fuel ratio controllability during transient operation.

<Effects of the Invention> As explained above, according to the present invention, during transient operation, the detected intake air flow rate is compared with the intake air flow rate calculated based on the intake pressure and the engine rotation speed, and both are compared. Since the transient correction amount is learned and corrected based on the deviation of Improved controllability.

In addition, by correcting the intake air flow rate calculation constant so that the calculated value of the intake air flow rate approaches the detected value in the steady operating state of the engine where the intake air flow rate detection accuracy is high, detection responsiveness during transient operation can be improved. It is possible to capture changes with high accuracy.

Further, the detected intake air flow rate and the calculated intake air flow rate are stored as a pair for a predetermined period from the start of transient operation, and the detected intake air flow rate at the supply amount setting timing for each cylinder and the set timing are determined. By comparing the calculated intake air flow rates stored corresponding to different times, it is possible to find the deviation of the calculated intake air flow rate at the timing that matches the engine required fuel amount with respect to the detected intake air flow rate at the set timing. Based on this deviation, the transient correction amount can be learned and corrected to a value commensurate with the true required amount.

Here, if the calculated intake air flow rate to be compared with the detected intake air flow rate is a value obtained by simply averaging the data at each predetermined crank angle, the effects of intake pulsation etc. can be avoided and the calculated intake air flow rate can be stabilized. can be compared at the same level as the detected intake air flow rate.

In addition, since the timing at which the intake air flow rate matches the engine required fuel amount generally occurs at a predetermined timing in the intake stroke, the crank angle between the fuel supply amount setting timing and the predetermined timing in the intake stroke for each cylinder is By determining this, a calculated value of the intake air flow rate that matches the true required fuel amount can be specified from the above-mentioned intake air flow rate storage data, which has the effect of improving the accuracy of learning correction.

[Brief explanation of drawings]

1 and 2 are block diagrams showing the configuration of the present invention, and FIG. 3 is a system schematic diagram showing an embodiment of the present invention,
FIGS. 4 to 10 are flowcharts showing control details in the above embodiment, FIGS. 11 and 12 are time charts to explain control characteristics in the above embodiment, and FIGS. 13 to 15 are flowcharts showing control characteristics in the above embodiment. They are time charts for explaining problems of conventional control. l... Engine 4... Throttle valve 6... Fuel injection valve 12... Control unit 13
... Air flow meter 14 ... Crank angle sensor Water temperature sensor 17 ... Throttle sensor Barometric pressure sensor 19 ... Intake temperature sensor 15 ... 18 ... Intake

Claims (5)

[Claims] (1) Engine operating state detection means that detects the engine operating state including at least the intake air flow rate and engine rotational speed, and the basic fuel supply that sets the basic fuel supply amount based on the detected intake air flow rate and engine rotational speed. a transient operation detecting means for detecting a transient operating state of the engine; and a transient operating detecting means for correcting the basic fuel supply amount according to the engine operating state when the transient operating state of the engine is detected by the transient operating detecting means. a transient correction amount setting means for setting a transient correction amount for the engine; and a fuel supply means for supplying the set basic fuel supply amount or an amount of fuel corrected based on the transient correction amount to the engine. The fuel supply control device for an internal combustion engine includes: an intake pressure detection means for detecting the intake pressure of the engine; an intake air flow rate calculation means for calculating the intake air flow rate based on the detected intake pressure and the engine rotational speed; In a transient operating state, the deviation between the intake air flow rate detected by the engine operating state detection means and the intake air flow rate calculated by the intake air flow rate calculation means is determined, and the transient correction amount setting means sets the deviation based on the deviation. 1. A fuel supply control device for an internal combustion engine, comprising: transient correction amount correction means for learning and correcting a correction amount. (2) The fuel supply control device for an internal combustion engine according to claim 1, further comprising steady-state operation detecting means for detecting a steady-state operating state of the engine; A deviation between the intake air flow rate calculated by the air flow rate calculation means and the intake air flow rate detected by the engine operating state detection means is determined, and a calculation constant for the intake air flow rate in the intake air flow rate calculation means is corrected based on the deviation. 1. A fuel supply control device for an internal combustion engine, comprising: intake air flow rate calculation constant correction means. (3) Engine operating state detection means that detects the engine operating state including at least the intake air flow rate and engine rotational speed, and basic fuel supply that sets the basic fuel supply amount based on the detected intake air flow rate and engine rotational speed. a transient operation detecting means for detecting a transient operating state of the engine; and a transient operating detecting means for correcting the basic fuel supply amount according to the engine operating state when the transient operating state of the engine is detected by the transient operating detecting means. a transient correction amount setting means for setting a transient correction amount for each cylinder of the engine; and a cylinder for supplying a set basic fuel supply amount or an amount of fuel corrected based on the transient correction amount to each cylinder of the engine at a predetermined timing. A fuel supply control device for an internal combustion engine configured to include a separate fuel supply means, an intake pressure detection means for detecting the intake pressure of the engine, and an intake air flow rate based on the detected intake pressure and engine rotation speed. and an intake air flow rate calculation means for calculating the transient operating state detected by the transient operating state detection means by pairing the intake air flow rate detected by the engine operating state detection means and the intake air flow rate calculated by the intake air flow rate calculation means. a transient intake air flow rate data storage means for storing data for a predetermined period after the start of operation; and a transient intake air flow rate data storage means for storing data in response to the timing of setting the amount of fuel actually supplied to each cylinder. The deviation between the detected intake air flow rate and the calculated intake air flow rate stored in the transient intake air flow rate data storage means corresponding to a time that differs by a predetermined period from the supply amount setting timing is determined; An internal combustion engine characterized by comprising: timing difference transient correction amount correcting means for correcting and setting a transient correction amount corresponding to the engine operating state at the supply amount setting timing in the transient correction amount setting means based on the deviation. fuel supply control device. (4) The intake air flow rate calculated by the intake air flow rate calculation means is simply averaged every predetermined crank angle rotation, and the simple average value is used in place of the intake air flow rate calculated by the intake air flow rate calculation means during the transient intake. 4. The fuel supply control device for an internal combustion engine according to claim 3, further comprising means for simply averaging the calculated intake air flow rate to be stored in the air flow rate data storage means. (5) The supply amount setting timing of the fuel actually supplied to each cylinder is set at a time that differs by a predetermined period from the supply amount setting timing, which is the timing at which the calculated intake air flow rate is picked up in the timing difference transient correction amount correction means. 5. The fuel supply control device for an internal combustion engine according to claim 3, further comprising a deviation detection timing setting means for setting the deviation detection timing based on the crank angle between the timing and the predetermined timing of the intake stroke.