JP2000227044A - Control device for cylinder injection type engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To raise control accuracy of air-fuel ratio in a stratified combustion mode, completely attain improvement of fuel consumption and decrease of emission, and prevent a torque shock accompanying changeover of operation modes from generating, in a cylinder injection type engine which performs feedback control in a stoichiometric mode based on the output of an O2 sensor. SOLUTION: When feedback control is performed in a stoichiometric mode, actual charging efficiency cecal is computed based on O2 sensor output and injection pulse width Ti (SA9). The ratio of the charging efficiency ce based on output of an air flow sensor to the computed value cecal is obtained as the correcting factor (k) of the charging efficiency (SA10), a charging efficiency correcting map M is rewritten. When feedforward control is performed in a stratified combustion mode, the charging efficiency ce based on the air flow sensor output is corrected by the correcting factor (k) read out from the charging efficiency correcting map M.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an in-cylinder injection type engine control apparatus which controls the amount of fuel injected into a combustion chamber in accordance with the operating state of the engine. In which the fuel injection amount is open-controlled (feed-forward control) based on the above, belongs to the technical field of correction control for appropriately correcting the correspondence between the intake charge amount and the fuel injection amount.

[0002]

2. Description of the Related Art Conventionally, as a control device for a direct injection type engine of this kind, as disclosed in Japanese Patent Application Laid-Open No. Hei 7-301139, for example, fuel is compressed into a cylinder in a predetermined low load and low speed region of the engine. It is known to operate in a stratified combustion state by injecting fuel in a stroke, and to operate in a uniform combustion state by injecting fuel in an intake stroke of a cylinder in other operating regions. That is, in the stratified combustion state, a relatively rich air-fuel mixture field is formed in the vicinity of the spark plug, and a stratified air-fuel mixture is formed therearound, and the air-fuel ratio of the entire combustion chamber is extremely lean and the air-fuel mixture is extremely lean. Is burned. On the other hand, in the uniform combustion state, a substantially uniform mixture near the stoichiometric air-fuel ratio is formed in the entire combustion chamber and burned.

In the conventional control device, the amount of intake air to the engine is estimated by the throttle speed method based on the engine speed and the throttle opening, and the intake charge amount of each cylinder is obtained. The fuel injection amount is controlled to be open so that the air-fuel ratio becomes appropriate for the amount. Further, a pressure sensor for detecting the intake pressure is provided, and the estimated value of the intake air amount is corrected in accordance with the output of the pressure sensor, whereby the calculation accuracy of the intake charge amount and, consequently, the air-fuel ratio Control accuracy has been improved.

[0004] By the way, the estimation of the intake air amount by the throttle speed method is not very accurate even if it is corrected in accordance with the intake air temperature and intake air pressure. With such control of the air-fuel ratio, it is not possible to sufficiently improve the fuel efficiency and reduce the emission of the engine. Therefore, in response to a recent increase in demand for environmental protection, for example, a sensor employing a high-precision sensor such as a hot wire airflow sensor is increasing.

[0005]

However, since the flow of air (intake) in the intake system of the engine is a pulsating flow, the quantitative relationship between the output of the airflow sensor and the actual intake air charge of the cylinder is accurately specified. It is extremely difficult to do. This is because the influence of the pulsation in the intake system of the engine has individual differences, and the degree of the influence changes due to various factors such as clogging of the air cleaner. Furthermore, since there is a measurement error of the airflow sensor itself, after all, no matter how high the accuracy of the airflow sensor is used in the engine actually mounted on the vehicle,
The open air control based on the sensor output cannot control the operating air-fuel ratio of the engine with sufficiently high accuracy.
In fact, there is still room for improvement in both fuel efficiency improvement and emission reduction.

On the other hand, it is conceivable that an O2 sensor is provided in the exhaust passage, and the detection result is fed back to correct the fuel injection amount. According to such feedback control, the control of the air-fuel ratio is sufficiently high. It is known that precision can be achieved. However, the conventional O2 sensor, even if it is a linear O2 sensor having a relatively wide detection range, can obtain excellent detection accuracy only when the air-fuel ratio of the exhaust gas is within a predetermined narrow range. A state in which the air-fuel ratio of the combustion chamber is extremely lean such as a combustion state (for example, A / F
When ≧ 30), feedback control cannot be performed.

That is, in the direct injection type engine in which the operation is switched between the stratified combustion state and the uniform combustion state as in the conventional example, the accuracy of the air-fuel ratio control can be improved by the feedback from the O2 sensor. This is limited to the case where the engine is set to a uniform combustion state and the air-fuel ratio is controlled within a predetermined range. In addition, if feedback control is performed only in the uniform combustion state of the engine, open control similar to the conventional one is performed in the stratified combustion state, so the control mode is switched during operation of the engine. This causes a new problem of occurrence of a torque shock accompanying this switching.

Specifically, as described above, when the engine is operated in a stratified combustion state and the fuel injection amount is controlled based on the output from the air flow sensor, the air-fuel ratio is reduced by the low accuracy of the control. The error increases and the deviation of the engine output from the target value also increases. Then, the state shifts from the stratified combustion state to the uniform combustion state, and when the feedback control is performed based on the signal from the O2 sensor as described above, the error in the air-fuel ratio decreases and the engine output approaches the target value. As a result, the engine output fluctuates, and a torque shock that causes the driver to feel uncomfortable occurs.

The present invention has been made in view of the above points, and has as its object the purpose of controlling the fuel injection amount of the engine based on the intake air charge amount as described above or the feedback control based on the detected air-fuel ratio. In the direct injection type engine, which is controlled by switching to control, the method of detecting the intake charge is devised to improve fuel efficiency and thoroughly reduce emissions by improving control accuracy during open control. The purpose of the present invention is to prevent the occurrence of a torque shock due to the switching of the torque.

[0010]

In order to achieve the above-mentioned object, according to the present invention, at the time of executing the feedback control, the actual intake air charging amount of the combustion chamber is calculated based on the fuel injection amount and the air-fuel ratio. Based on the deviation between the calculated value and the detection value of the filling amount detecting means, the detection value of the filling amount detecting means is corrected at the time of executing the feedforward control.

More specifically, according to the first aspect of the present invention, as shown in FIG.
A fuel injection valve 7 for injecting and supplying, a filling amount detecting means a for detecting a filling amount of intake air in the combustion chamber 4, an air-fuel ratio detecting means b for detecting an air-fuel ratio of the combustion chamber 4, and at least an operating state of the engine 1. Target air-fuel ratio setting means c for setting a target air-fuel ratio in accordance with
According to the detection result by the charging amount detecting means a, the combustion chamber 4
Either a feedforward control operation for performing feedforward control so that the air-fuel ratio of the air-fuel ratio becomes the target air-fuel ratio or a feedback control operation for performing feedback control so that the value detected by the air-fuel ratio detection means b becomes the target air-fuel ratio. It is assumed that the control device A of the direct injection type engine includes the fuel injection control means g for switching to the above.

An injection amount detecting means d for detecting an amount of fuel injection by the fuel injection valve 7 and the air-fuel ratio detecting means b and the injection amount detecting means d when the feedback control operation is performed by the fuel injection control means g. A charging amount calculating means e for calculating the intake charging amount of the combustion chamber 4 based on the detected values of the above, and a charging amount between the calculated value by the charging amount calculating means e and the detection value by the charging amount detecting means a. Of the filling amount detecting means a based on the difference of the filling amount when the fuel injection control means g executes the feedforward control operation.
And a configuration including:

With the above-described configuration, the air-fuel ratio of the combustion chamber 4 is precisely controlled by feedback-controlling the fuel injection amount of the fuel injection valve 7 so that the value detected by the air-fuel ratio detecting means b becomes the target air-fuel ratio. Can control. At this time, based on the air-fuel ratio of the combustion chamber 4 detected by the air-fuel ratio detecting means b and the fuel injection amount detected by the injection amount detecting means d, the actual intake charging amount of the combustion chamber 4 is determined by the charging amount. Since the difference between the calculated value and the detected value of the charged amount detecting means a can be calculated by the charged amount calculating means e, the deviation of the charged amount can be calculated by the intake pulsation of the engine 1 or the sensor. , Which accurately reflects the characteristics of the difference in intake air charge.

When the amount of fuel injected by the fuel injection valve 7 is controlled in a feed-forward manner in accordance with the value detected by the charged amount detecting means a, the detected value is used as the charged amount correcting means f.
By performing the correction based on the filling amount deviation, the detection accuracy of the intake filling amount can be greatly improved. As a result, the control accuracy of the air-fuel ratio during the feedforward control is greatly improved, so that the fuel economy is improved and emission is reduced as a whole in the operating region of the engine 1. Also,
Since the control accuracy at the time of the feedforward control is greatly improved, the deviation of the engine output is also reduced, so that the engine output does not fluctuate greatly even at the time of switching between the feedforward control and the feedback control. Therefore, occurrence of torque shock can be prevented.

According to a second aspect of the present invention, the injection amount detecting means includes:
It is assumed that the fuel injection amount is detected based on the pulse width of the control signal input to the fuel injection valve. This makes it possible to accurately detect the actual fuel injection amount by the fuel injection valve.

According to the third aspect of the present invention, the fuel injection control means performs the feedforward control operation when the target air-fuel ratio is set to a value larger than the stoichiometric air-fuel ratio by the target air-fuel ratio setting means. Is set near the stoichiometric air-fuel ratio, a feedback control operation is performed.

That is, in general, when the air-fuel ratio is larger than the stoichiometric air-fuel ratio even when the engine output is the same, the amount of intake air becomes larger than when the air-fuel ratio is near the stoichiometric air-fuel ratio. The detected value of the filling amount is likely to greatly deviate from the actual value. Therefore, while the feed-forward control is performed when the air-fuel ratio is larger than the stoichiometric air-fuel ratio, the feedback control is performed when the air-fuel ratio is close to the stoichiometric air-fuel ratio. That the control accuracy of the control can be greatly improved is particularly effective in preventing torque shock.

According to a fourth aspect of the present invention, the target air-fuel ratio setting means includes: a target load of the engine corresponding to the accelerator operation amount;
The target air-fuel ratio is set based on at least one of the engine speed and the intake charge amount. With this,
The target air-fuel ratio can be appropriately set so as to correspond to the accelerator operation amount and the operating state of the engine. Further, when the target air-fuel ratio is set based on the intake air charge, the target air-fuel ratio itself becomes an inappropriate value due to a decrease in the detection accuracy of the intake air charge. It is particularly effective to increase the detection accuracy of.

According to the fifth aspect of the present invention, the target air-fuel ratio setting means sets the target air-fuel ratio when the engine is in a predetermined low-load operation range.
The target air-fuel ratio is set to a value larger than the stoichiometric air-fuel ratio, and the fuel injection control means injects fuel in the compression stroke of the cylinder by the fuel injection valve when the engine is in the predetermined operation range, and performs stratified combustion. On the other hand, in other operation regions, the fuel is injected in the intake stroke of the cylinder to perform uniform combustion.

According to this configuration, the engine is operated in the stratified combustion state in the predetermined operation region on the low load side, and the feedforward control is performed so that the air-fuel ratio becomes larger than the stoichiometric air-fuel ratio. On the other hand, in other operation regions, the engine is operated in a uniform combustion state, and in at least a part of the operation region, feedback control is performed so that the air-fuel ratio becomes close to the stoichiometric air-fuel ratio. That is, when the control mode is switched between the feedforward control and the feedback control, at the same time, the target air-fuel ratio greatly changes, and the combustion mode of the engine also changes. There is a very strong risk of doing so. Therefore, in such a system, the fact that the control accuracy of the air-fuel ratio during the feedforward control can be greatly improved has a particularly effective effect in preventing the occurrence of torque shock when the control mode is switched.

According to a sixth aspect of the present invention, an exhaust gas recirculation means for recirculating a part of the exhaust gas of the engine to the intake system is provided. Thus, the generation of nitrogen oxides (NOx) accompanying the combustion of the engine can be suppressed by the recirculation of the exhaust gas. On the other hand, since the influence of the flow and pulsation of the intake air changes due to the recirculated exhaust gas, it is considered that the detection accuracy of the intake air charge is also reduced. Correcting the value detected by the filling amount detecting means in this way has a particularly effective action.

In the invention according to claim 7, the air-fuel ratio detecting means includes:
The air-fuel ratio is detected only when the air-fuel ratio of the exhaust is near the stoichiometric air-fuel ratio. Based on the output from the air-fuel ratio detecting means, feedback control can be performed so that the air-fuel ratio of the combustion chamber becomes a value near the stoichiometric air-fuel ratio. Also,
The air-fuel ratio detecting means can be constituted by a relatively inexpensive lambda O2 sensor.

According to the present invention, the fuel injection control means performs a feedforward control operation when the engine is in a high load range. That is, in general, when the engine is in a high load range, the intake air amount increases,
The influence of the intake pulsation becomes large, and the detected value of the intake charge is easily shifted from the actual value. Therefore, in such a case, when feedforward control is performed based on the detected value of the intake air charge, it is particularly effective that the detected value can be corrected to improve the control accuracy.

According to a ninth aspect of the present invention, the filling amount correcting means of the eighth aspect includes a deviation characteristic learning means for learning a characteristic of a deviation of the filling amount when the engine is in a low load range, and the learned deviation characteristic is provided. The correction value detected by the filling amount detecting means is corrected based on the above. Because of this, usually, the engine mounted on the vehicle is in a so-called low load to medium load operation state for a long time, and therefore, it is necessary to learn the characteristics of the deviation of the filling amount in such an operation state. Then, the deviation characteristic can be learned at an early stage.

According to the tenth aspect, the air-fuel ratio detection means detects the air-fuel ratio only when the air-fuel ratio of the exhaust gas is within a predetermined range including the stoichiometric air-fuel ratio. The feedback control operation is performed when the target air-fuel ratio is set within the predetermined range by the target air-fuel ratio setting means, and the feedforward control operation is performed when the target air-fuel ratio is set outside the predetermined range.

According to this configuration, the same function and effect as the third aspect of the invention can be obtained. Further, based on the output from the air-fuel ratio detecting means, feedback control can be performed so that the air-fuel ratio of the combustion chamber becomes a value within a predetermined range, so that feedback control is performed only near the stoichiometric air-fuel ratio. Thus, the operation range in which the feedforward control is performed can be expanded, so that the characteristics of the deviation of the intake air charge amount can be detected over a wider operation range.

According to the eleventh aspect of the present invention, the charging amount correcting means includes a deviation characteristic learning means for learning a characteristic of a charging amount deviation in at least a plurality of engine operating states having different engine speeds. Based on this, a configuration is adopted in which the detection value of the filling amount detection means is corrected. According to this configuration, the characteristics of the deviation can be accurately learned based on the deviation of the charged amount corresponding to various operating states of the engine.

According to a twelfth aspect of the present invention, the deviation characteristic learning means according to the eleventh aspect learns the characteristic of the deviation of the intake air charge amount each time the number of times of starting the engine exceeds the set number of times. Accordingly, even if the influence degree of the intake pulsation changes due to various factors such as clogging of the air cleaner, and the characteristic of the deviation of the intake air charge changes, the learning content can be updated so as to correspond to the change. it can. On the other hand, the content already learned is sufficient until the number of times of starting of the engine exceeds the set number, so that the load on the control device can be reduced by not performing unnecessary learning.

According to a thirteenth aspect of the present invention, the deviation characteristic learning means according to the eleventh aspect learns the characteristic of the deviation of the intake air charging amount when the battery terminal of the vehicle equipped with the engine is disconnected once and then connected again. Shall be. Thus, even if the battery terminal of the vehicle is disconnected and the learning result of the deviation characteristic is lost from the memory, the learning is immediately performed thereafter. On the other hand, the load on the control device can be reduced by not performing unnecessary learning until that time.

[0030]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (Overall Configuration of Control Device) FIG. 2 shows a control device A for a direct injection engine according to an embodiment of the present invention.
1 is, for example, a multi-cylinder engine mounted on a vehicle. This engine 1 has a plurality of cylinders 2, 2,
(Only one is shown), and a piston 3 is reciprocally fitted into each cylinder 2, and a combustion chamber 4 is defined in the cylinder 2 by the piston 3. This combustion chamber 4
An ignition plug 6 connected to an ignition circuit 5 is mounted at a position on the cylinder axis on the upper wall so as to face the combustion chamber 4. An injector (fuel injection valve) 7 is attached to a side wall of the combustion chamber 4 at a position not interfering with the moving piston 3 so as to directly inject and supply fuel to the combustion chamber 4.

Although not shown, the injector 7
A high-pressure fuel pump, a fuel supply circuit having a pressure regulator and the like are connected, and while the fuel from the fuel tank is adjusted to an appropriate pressure by the fuel supply circuit,
The power is supplied to the injector 7. Further, a fuel pressure sensor 8 for detecting the fuel pressure is provided.
When fuel is injected by the injector 7 in the latter half of the compression stroke of the cylinder 2, the fuel spray is trapped in a cavity (not shown) formed in the top surface of the piston 3, and near the ignition plug 6. A relatively dense mixture is formed. On the other hand, when fuel is injected by the injector 7 in the intake stroke of the cylinder 2, the fuel spray
And is mixed with the intake air (air) to form a uniform mixture in the combustion chamber 4.

The combustion chamber 4 is connected to an intake passage 10 via an intake valve 9 via an intake port (not shown).
The intake passage 10 supplies the intake air filtered by the air cleaner 11 to the combustion chamber 4 of the engine 1, and detects the amount of intake air taken into the engine 1 in order from the upstream side to the downstream side. Wire type air flow sensor 12 and electric type throttle valve 13 for restricting intake passage 10
And a surge tank 14 are respectively provided. The electric throttle valve 13 is not mechanically connected to an accelerator pedal (not shown), and is driven by a motor 15 to open and close. Further, a throttle opening sensor 1 for detecting the opening of the throttle valve 13 is provided.
6 and an intake pressure sensor 17 for detecting the intake pressure in the surge tank 14 are provided.

The intake passage 10 on the downstream side of the surge tank 14 is an independent passage branching for each cylinder 2. The downstream end of each independent passage is further branched into two and connected to the intake port. The swirl control valve 18 is provided on one of the branches. The swirl control valve 18 is driven by an actuator 19 to open and close. When the swirl control valve 18 is closed, intake air is supplied to the combustion chamber 4 only from the other branch passage, and strong intake air is supplied to the combustion chamber 4. While swirl is generated, intake swirl is reduced as the swirl control valve 18 opens. A swirl control valve opening sensor 2 for detecting the opening of the swirl control valve 18
0 is provided.

In FIG. 2, reference numeral 22 denotes an exhaust passage for discharging combustion gas from the combustion chamber 4. The upstream end of the exhaust passage 22 branches off for each cylinder 2 and is connected to an exhaust port (not shown) via an exhaust valve 23 through an exhaust valve 23. 4 is connected. An O2 sensor (air-fuel ratio detecting means) 24 for detecting the oxygen concentration in the exhaust gas is provided in the exhaust passage 22 in order from the upstream side to the downstream side.
And a catalyst 25 for purifying the exhaust gas. The O2 sensor 24 is used to detect the air-fuel ratio based on the oxygen concentration in the exhaust gas, and a so-called lambda O2 sensor whose output is inverted in a stepwise manner at the stoichiometric air-fuel ratio is used. (See FIG. 15).

The catalyst 25 has a cordierite carrier (not shown) having a honeycomb structure in which a large number of through holes extending in parallel with each other along an axial direction (flow direction of exhaust gas) are formed. The catalyst layer is formed on the wall surface of the through hole. While the catalyst 25 adsorbs NOx in a lean state where the air-fuel ratio is higher than the stoichiometric air-fuel ratio, when the air-fuel ratio becomes rich near or below the stoichiometric air-fuel ratio, the adsorbed NOx is released and reduced and purified. In particular, in the vicinity of the stoichiometric air-fuel ratio, a NOx adsorption reduction type is used, which exhibits high exhaust gas purification performance similar to a so-called three-way catalyst.

Further, an upstream end of an EGR passage 26 is branched and connected to the exhaust passage 22 upstream of the O2 sensor 24, and a downstream end of the EGR passage 26 is provided between the throttle valve 13 and the surge tank 14. And a part of the exhaust gas is recirculated to the intake system. An electric EGR valve 27 whose opening degree can be adjusted is disposed near the downstream end of the EGR passage 26.
Exhaust gas recirculation amount through the R passage 26 (hereinafter referred to as EGR amount)
Is to be adjusted. The EGR passage 26 and the EGR valve 27 constitute an exhaust gas recirculation unit.
Further, a lift sensor 28 for detecting a lift amount of the EGR valve 27 is provided.

The ignition circuit 5 of the ignition plug 6, the injector 7, the drive motor 15 of the electric throttle valve 13, the actuator 19 of the swirl control valve 18, the electric EGR
The operation of the valve 27 and the like is controlled by a control unit 40 (hereinafter, referred to as an ECU). On the other hand, the ECU 40 includes the air flow sensor 12,
The output signals of the throttle opening sensor 16, the intake pressure sensor 17, the swirl control valve opening sensor 20, the O2 sensor 24, and the lift sensor 28 of the EGR valve 27 are input, and in addition, the cooling water temperature of the engine 1 ( Engine water temperature)
Temperature sensor 30, an intake air temperature sensor 31, which detects an intake air temperature, an atmospheric pressure sensor 32, which detects an atmospheric pressure, a rotation speed sensor 33, which detects an engine speed, and an opening degree of an accelerator pedal (accelerator operation amount). Each output signal of the accelerator opening sensor 34 to be detected is input.

(Outline of Engine Control) The engine 1 according to this embodiment has an injector 7 according to its operating state.
(Fuel injection timing and air-fuel ratio, etc.) are switched to operate in different combustion states. That is, as shown in FIG. 3A, when the engine 1 is warm (for example, when the engine water temperature is 45 ° C. or higher), the predetermined region (a) on the low-load low-rotation side is a stratified combustion region. Then, the fuel is injected by the injector 7 in the latter stage of the compression stroke, and the combustion mode is set in which the mixture is burned in a stratified state in which the air-fuel mixture is unevenly distributed near the ignition plug 6. In this stratified combustion mode, the opening of the throttle valve 13 is increased in order to reduce the pump loss of the engine 1, whereby the average air-fuel ratio of the combustion chamber 4 is significantly lean (for example, A / F). = About 30).

On the other hand, the other operating regions (b) and (c) are defined as uniform combustion regions, in which fuel is injected by the injector 7 in the first half of the intake stroke, mixed sufficiently with the intake air, and uniformly mixed in the combustion chamber 4. The combustion mode is a mode in which the mixture is formed and then burned. In the low load side region (b) in the uniform combustion mode, the fuel injection amount and the throttle opening are set so that the air-fuel ratio of the air-fuel mixture in the combustion chamber 4 becomes substantially the stoichiometric air-fuel ratio (A / F = 14.7). (Hereinafter referred to as stoichiometric mode). In the high-load or high-rotation-side operation region (c) in the uniform combustion region, the air-fuel ratio is made richer than the stoichiometric air-fuel ratio (for example, A / F = 13 to 14) (hereinafter, referred to as enrichment mode). . Further, in a region shown by hatching in FIG. 5A, a part of the exhaust gas in the exhaust passage 22 is recirculated to the intake passage 10 by the EGR passage 26, whereby the combustion in the combustion chamber 4 The accompanying generation of NOx can be suppressed.

In addition, when the engine is cold as shown in FIG. 3B, the entire operating region of the engine 1 is set as a uniform combustion region in order to improve combustion stability. In the low-load operation region (d) corresponding to the combination of (b) and (d), the engine 1 is operated in the stoichiometric mode, while the operation region (e) corresponding to the warm operation region (c). In), the engine 1 is operated in the enrichment mode.

FIG. 4 is a functional block diagram showing basic processing of engine control in the ECU 40. That is, the ECU 40 includes the intake density state detecting means 41 for detecting the intake density state based on the signals from the intake air temperature sensor 31 and the atmospheric pressure sensor 32, and converts the signals from the rotation speed sensor 33 and the accelerator opening sensor 34 into signals. Based on
Further, a target load setting means 42 for setting a target load of the engine 1 in consideration of the intake air density state is provided.

In the target load setting means 42, as shown in FIG. 5, first, the virtual volume efficiency calculating section 42a calculates the virtual volume efficiency veimg based on the accelerator opening accel and the engine speed ne. Specifically, the accelerator opening degree accel and the engine speed ne are determined by a bench test or the like in advance so that the required output performance can be obtained under standard operating conditions in which the air-fuel ratio is kept at the stoichiometric air-fuel ratio under standard atmospheric conditions. A correspondence relationship with the virtual volume efficiency veimg is obtained, and this correspondence relationship is stored in the memory of the ECU 40 as a map. From this map, the virtual volumetric efficiency v corresponding to the actual accelerator opening degree accel and the engine speed ne is obtained.
eimg is read. The correspondence between the accelerator opening accel and the engine speed ne and the virtual volumetric efficiency veimg is, for example, as shown in FIG. It increases as the accelerator opening accel increases, and increases as the engine speed ne decreases.

Subsequently, the virtual filling efficiency calculating unit 42b calculates the virtual filling efficiency ceimg by adding the intake volume obtained by the intake density state detecting means 41 to the virtual volume efficiency veimg obtained as described above. I do. The virtual charging efficiency ceimg is a charging efficiency (intake charging amount) corresponding to the output required of the engine 1 under the standard operating conditions. The delay processing means 42c performs a temporary delay correction as shown in the following equation. That is, delay processing is performed on the virtual filling efficiency ceimg.

[0044]

Ceimgd = (1−α) × ceimg + α × ceimgd [i-1] where ceimgd [i-1] is the previous value of ceimgd, and α is the coefficient (0
<Α <1).

Then, based on the virtual filling efficiency ceimg calculated by the virtual filling efficiency calculating section 42b or the virtual filling efficiency ceimgd delayed by the smoothing processing section 42c, the target load calculating section 42d calculates the respective values. The corresponding indicated average effective pressure (Pi) is calculated as the target load. That is, the virtual filling efficiency that has not been annealed
A first target load Piobjd is calculated based on ceimg, and a second target load Piobjd is calculated based on the simulated virtual filling efficiency ceimgd.

[0046]

## EQU2 ## Piobjd = K1.times.ceimg + K2 Piobjd = K1.times.ceimgd + K2 When an external load such as an air conditioner is applied to the target load setting means 42 while the engine 1 is idling, the external load is reduced. In order to increase the engine output to an appropriate degree, an idling load correction unit 42e that corrects the virtual filling efficiency ceimg, ceimgd prior to the calculation of the target load is provided.

The ECU 40 sets an operation mode setting means 4 for setting a basic operation mode mods based on the first target load Piobj and the engine speed ne obtained as described above.
3 is provided. That is, for example, when the engine is warm, as shown in FIG.
Is lower than a predetermined low load side threshold value Piobj * and the engine speed ne is low, the stratified combustion mode is set in the operation region (a), while the other operation regions (b, c) are set to the uniform combustion mode. , The first target load Piobj and the engine speed n
Either the stoichiometric mode or the enriched mode is set according to e.

The ECU 40 determines the values of various control parameters related to the engine output. Specifically, the ECU 40 adjusts the intake air amount adjusted by the throttle valve 13 and the EGR valve 27. EGR amount,
The intake swirl intensity adjusted by the swirl control valve 18, the fuel injection amount by the injector 7, the injection timing, and the ignition timing by the spark plug 6 are used as control parameters.
The values of these control parameters are changed to the first and second target loads Pi.
It is determined according to obj, Piobjd and the engine speed ne.

Here, the intake air amount, the EGR amount and the swirl strength of the control parameters are low-speed response systems having relatively low responsiveness to the operation of the throttle valve 13, the EGR valve 27 and the swirl control valve 18, respectively. These control amounts such as throttle opening tvoobj, EGR valve opening,
The swirl control valve opening is determined according to the first target load Piobj and the engine speed ne. On the other hand, since the fuel injection amount, the injection timing, and the ignition timing are all of a high-speed response system that responds quickly to the control signal, they are determined according to the second target load Piobjd after the delay processing and the engine speed ne. .

(Throttle control) More specifically, the ECU
40 is a first target air-fuel ratio setting means 44, a target charging efficiency calculating means 45, and a throttle opening degree, as means for controlling the throttle valve 13 according to the first target load Piobj set by the target load setting means 42. A calculation means 46 is provided. The first target air-fuel ratio setting means 44 sets a target air-fuel ratio afwb for controlling the intake air amount for each operation mode set by the operation mode setting means 43, as shown in FIG. In the stratified combustion mode or the enrichment mode, the target air-fuel ratio afwb is obtained from a map prepared in advance according to the first target load Piobj and the engine speed ne. In the stoichiometric mode, the target air-fuel ratio afwb is calculated based on the theoretical air-fuel ratio. Fuel ratio.

The target filling efficiency calculating means 45 includes:
The first target load Piobj or the virtual filling efficiency cei corresponding thereto
Based on mg and the target air-fuel ratio afwb, the target charging efficiency ceobj is calculated according to, for example, the following equation.

[0052]

[Equation 3] ceobj = ceimg × ｛(afwb + K3) /14.7｝ × K4 The equation of this (Equation 3) is obtained from the virtual charging efficiency ceimg by the excess air ratio of the target air-fuel ratio when the engine is operated in the lean state ( afwb / 14.7) and the fuel efficiency improvement effect are taken into account to calculate the target charging efficiency ceobj, and the coefficients K3 and K4 are values that reduce the target charging efficiency to an extent commensurate with the fuel efficiency improvement effect. It has been.

That is, since the virtual charging efficiency ceimg is a value corresponding to the target load when the engine 1 is operated under the standard operating conditions, it is necessary to ensure the same fuel injection amount at the time of lean operation by using the above-described air. Although it is necessary to take into account the excess ratio, if the fuel injection amount is secured in the same manner as in the case of the stoichiometric air-fuel ratio, the engine output will increase due to high thermal efficiency during lean operation (this Fuel efficiency improvement effect). Therefore, in order to obtain the engine output corresponding to the target load, in addition to taking into account the excess air ratio as described above, the fuel consumption improvement effect is also taken into account.

From the above (Equation 2), ceimg = (Piobj
−K2) / K1, and this may be substituted into the above (Equation 3) to determine the target charging efficiency ceobj from the first target load Piobj.

Further, the throttle opening calculating means 46
Then, as shown in FIG. 6, the target filling efficiency ceobj determined as described above is corrected by the target volume efficiency calculating unit 46a in accordance with the intake air density, and the target volume efficiency veobj is determined. The throttle opening tvoobj is calculated according to the engine speed ne. At this time, since the correspondence between the volumetric efficiency and the engine speed and the throttle opening differs depending on the presence or absence of the EGR, a map showing the correspondence is prepared in advance for each case, and the presence or absence of the EGR by the EGR determination unit 46c is determined. In accordance with the determination result, the throttle opening tvoobj corresponding to the target volumetric efficiency veobj and the engine speed ne is read from any of the maps.

Here, the relationship between the volumetric efficiency, the engine speed and the throttle opening is, for example, as shown by a solid line in FIG. 11 when EGR is not performed.
When GR is performed, it is indicated by a broken line in FIG. That is, the throttle opening tvoobj is equal to the target volumetric efficiency.
The larger the veobj, the larger the engine speed ne
Is higher when the EGR is higher, and is made larger than when there is no EGR.

In the stratified charge combustion mode, the exhaust gas has an extremely lean air-fuel ratio, so that not only the burned gas but also a large amount of air (oxygen) is contained in the EGR gas. Therefore, if there is EGR, the burned gas volume ratio in the EGR gas is obtained, and the throttle opening tvoobj and the EGR valve control amount are corrected according to the result. Further, the EGR amount and the swirl strength are also controlled for each operation mode mods according to the operation state of the engine 1.

(Fuel Injection Control) The ECU 40 includes second target air-fuel ratio setting means 47 for setting a target air-fuel ratio according to a target load and an operating state of the engine 1, and controls fuel injection by the injector 7. Operation mode setting means 48, split ratio setting means 49, injection amount calculating means 50, injection timing setting means 51, and injection control means 52.

The second target air-fuel ratio setting means 47 obtains a target air-fuel ratio used for controlling the fuel injection amount and the like. More specifically, as shown in FIG.
Based on objd or the virtual charging efficiency ceimgd corresponding thereto and the actual charging efficiency (intake charging amount) ce, the calculation unit 47a calculates the target air-fuel ratio afw0 mainly used during the transient operation of the engine 1.

[0060]

Afw0 = 14.7 × K1 × ce / ｛K4 × (Piobjd−K2)｝ − K3 [= 14.7 × ce / (K4 × ceimgd) −K3] The equation of (Formula 4) is the theoretical air-fuel ratio. And the actual charging efficiency ce, the second target load Piobjd (or the virtual charging efficiency ceimgd), and the coefficients K3 and K4 that take into account the above-described fuel efficiency improvement effect.
The air-fuel ratio is determined so that the engine torque corresponding to the target load can be obtained under the actual charging efficiency.

The setting unit 47b sets a target air-fuel ratio afwbd used mainly during steady operation of the engine 1 for each operation mode modf set by the operation mode setting means 48. That is, as shown in FIG. 13A, in the stratified combustion mode or the enrichment mode, the second
Based on the target load Piobjd and the engine speed ne, the target air-fuel ratio afwbd is read from a map created in advance, while in the stoichiometric mode, the target air-fuel ratio afwb is read.
Let d be the stoichiometric air-fuel ratio.

The deviation da between the target air-fuel ratio afwb for controlling the intake air amount set by the first target air-fuel ratio setting means 44 and the target air-fuel ratio afw0 calculated by the calculator 47a.
fwb is calculated by the deviation calculation unit 47c, and this deviation dafwb is calculated.
When the engine 1 is in a transient operation, the target air-fuel ratio afw0 calculated by the calculation unit 47a is set as the final target air-fuel ratio afw, while the engine 1 whose deviation dafwb is small.
During the steady operation, the target air-fuel ratio afwbd set by the setting unit 47b is set as the final target air-fuel ratio afw.

The second target air-fuel ratio setting means 47 is configured in this way in order to simultaneously satisfy the requirement on engine output and emission. The unit 47b and the deviation calculation unit 47c may be omitted, and the target air-fuel ratio afw0 always obtained by the calculation unit 47a may be set as the final target air-fuel ratio afw in the fuel injection control.

In FIG. 7, reference numeral 60 denotes a means for calculating air-fuel ratio deviations dafwbd and dafw0 for correcting the ignition timing at the time of transition as described later. The operation mode setting means 48 sets the stratified combustion mode or enrich mode. When the stoichiometric mode is set, dafwbd = afwbd−afw is calculated, while dafw0 = afw0−afw is calculated.

The operation mode setting means 48 is an operation mode setting means for determining the control parameters of the high-speed response system.
modf is set based on the target air-fuel ratio afw0 for controlling the fuel injection amount and the engine speed ne. That is, as shown in FIG. 12, the target air-fuel ratio af calculated by the calculation unit 47a.
If w0 is smaller than the lower limit reference value afw0 * of the stratified combustion mode and is equal to or higher than the stoichiometric air-fuel ratio, the stoichiometric mode is set, and the target air-fuel ratio afw0 is a predetermined value smaller than the stoichiometric air-fuel ratio.
If it becomes afw0 **, it will be in enriched mode. Conversely, if the target air-fuel ratio afw0 is equal to or more than the lower limit reference value afw0 *, the stratified combustion mode is set. By switching the operation mode modf, the fuel injection mode is switched, and the operation mode of the engine 1 is finally switched. Incidentally, when the operation mode modf is changed between the stoichiometric mode and the stratified combustion mode, the fuel injection may be temporarily performed by dividing the fuel injection into the intake stroke and the compression stroke. Then, a rapid change in the combustion state can be avoided.

The split ratio setting means 49 sets the fuel split ratio between the intake stroke injection and the compression stroke injection according to the operation mode modf set by the operation mode setting means 48. In the stratified combustion mode, Intake stroke injection ratio is 0
%, While in the stoichiometric mode or the enriched mode, the intake stroke injection ratio is set to 100%. When performing the split injection, the split ratio may be set according to the target air-fuel ratio afw and the engine speed ne.

The feature of the present invention lies in how the fuel injection amount is determined by the fuel injection amount calculating means 50. The fuel injection amount calculating means 50 mainly determines the actual charging efficiency ce obtained from the output of the air flow sensor 12. The fuel injection amount is calculated based on the target air-fuel ratio afw set by the second target air-fuel ratio setting means 47 and the injection ratio set by the split ratio setting means 49. As is well known, the actual filling efficiency ce is calculated based on the output of the air flow sensor 12, the engine speed ne, and the like. Therefore, in this embodiment, the air flow sensor 12 constitutes a filling amount detecting means. Have been.

Specifically, as shown in FIG. 8, first, based on the actual charging efficiency ce, the target air-fuel ratio afw and the conversion coefficient KGKF, the basic injection amount calculating section 50a calculates the intake stroke in accordance with the following equation. And the total basic injection amount qbas of the compression stroke injection
Calculate e.

[0069]

[Mathematical formula-see original document] qbase = KGKF * ce * k / afw where k is a correction coefficient for correcting the filling efficiency ce.
It is read from a filling efficiency correction map M created in advance based on the engine speed ne and the throttle opening tvoobj. This is because the charging efficiency ce detected based on the output of the airflow sensor is different from the actual charging efficiency ce mainly due to the influence of the intake air pulsation in the intake passage 10, and therefore, a coefficient for correcting this is used. k is set in advance as a map in association with the operating state of the engine.

Next, a correction value cdp corresponding to the fuel pressure is added to the basic injection amount qbase by the final injection amount calculating section 50c.
f and the feedback correction value cfb from the O2 sensor 24
And other various correction values ctotal,
qinj is calculated, and the final injection amount qinj and injection ratio
Based on rqbasep, the divided injection amount calculation unit 50d calculates the intake stroke and the compression stroke injection amount qinjp, qinjd, respectively.

[0071]

Qinj = qbase × cdpf × (1 + cfb + ctotal) qinjp = qinj × rqbasep qinjd = qinj−qinjp Then, the valve opening time of the injector 7 which is proportional to each of the injection amounts qinjp and qinjd, that is, the injection pulse width Ti is calculated by The pulse width calculation unit 50e reads and sets the injector flow rate characteristic map created in advance.

Further, an actual injection amount calculating means (injection amount detecting means) 55 for calculating an actual fuel injection amount based on the injection pulse width Ti, and a fuel injection amount calculated by the actual injection amount calculating means 55 A charging efficiency calculating means 56 for calculating the actual charging efficiency cecal of the combustion chamber 4 based on the air-fuel ratio detected by the O2 sensor 24, and a charging efficiency cecal based on the calculated charging efficiency cecal and the air flow sensor output.
And a correction coefficient calculating means 57 for calculating a ratio (filling amount deviation) with the correction coefficient k.

Then, the correction coefficient k thus calculated
Correction coefficient learning means (deviation characteristic learning means) 58 using
Accordingly, the corresponding value on the filling efficiency correction map M is rewritten. With this, Engine 1
Is rewritten the value of the correction coefficient k when the vehicle is in various operating states with different throttle opening and engine speed,
The filling efficiency correction map M is updated according to the characteristics of the deviation of the filling efficiency ce.

The injection timing setting means 51 sets the fuel injection timing for each of the operation modes modf set by the operation mode setting means 48. As shown in FIG. (2) While the injection timing thtinjd for the compression stroke injection is obtained from a map created in advance according to the target load Piobjd and the engine speed ne, from the map preset in accordance with the engine speed ne in the uniform combustion mode. Injection timing for intake stroke injection thtinjp
Ask for.

For convenience of the arithmetic processing, some value is always given to both thtinjd and thtinjp as the injection timing data. In the stratified combustion mode, the injection timing thtinjd for the compression stroke injection is given by a map. A fixed value is set for the injection timing thtinjp for the intake stroke injection (however, the intake stroke injection is not actually performed because the intake stroke injection ratio rqbasep is 0%). In the stoichiometric mode or the enriched mode, the injection timing thtinjp for the intake stroke injection is given by a map, and the injection timing thtinjd for the compression stroke injection is set to a fixed value (for example, a constant time at the beginning of the compression stroke). Injection only is used for additional injection when the fuel injection amount is insufficient. Further, when performing the split injection, the data in the stratified combustion mode is used as the injection timing thtinjd for the compression stroke injection, and the intake stroke is determined from a map prepared in advance according to the target air-fuel ratio afw and the engine speed ne. The injection timing thtinjp for the injection may be obtained.

The injection control means 52 controls the injection amount calculated by the injection amount calculation means 5 based on the injection timing set by the injection timing setting means 51.
0 so that the injector 7 is operated only for a time corresponding to the injection pulse width Ti calculated by 0.
To output a pulse signal.

(Flow of Fuel Injection Control) Next, the processing procedure of the above-described fuel injection control will be described with reference to the flowchart shown in FIG.

First, in step SA1 after the start, the engine speed ne and the accelerator opening accel are read based on the signals from the rotation speed sensor 33 and the accelerator opening sensor 34, respectively, and the signal from the air flow sensor 12 is read. Is read, and various sensor signals such as the O2 sensor 24 and the water temperature sensor 30 are received. In the following step SA2,
It is determined whether the operation mode modf is either the stoichiometric mode or the enriched mode, that is, whether it is the uniform combustion mode. If this determination is NO and the engine 1 is in the stratified combustion mode, the process proceeds to step SA12, while the determination is YE
If the engine 1 is in the uniform combustion mode in S, the process proceeds to step SA3.

In step SA3, the O2 sensor 24
It is determined whether the feedback control of the air-fuel ratio can be performed based on the output signal from the controller (if the F / B condition is satisfied). For example, if the engine water temperature is equal to or higher than the predetermined temperature, the O2 sensor 24 is in a state of operating correctly, and the operation mode modf is in the stoichiometric mode, it is determined that the feedback condition is satisfied and YES is determined. Proceeding to SA4, a feedback correction value cfb of the fuel injection amount is calculated based on the output signal from the O2 sensor 24, and step S4 is performed.
Proceed to A5. On the other hand, if the engine water temperature is low or the operation mode modf is in the enrichment mode, it is determined that the feedback condition is not satisfied, and the process proceeds to step SA5 (cfb = 0). The procedure for calculating the feedback correction value cfb in step SA4 will be described later.

Subsequently, at step SA5, the basic fuel injection amount qbase is calculated according to the above (Equation 5). In step SA6, the final injection amount qinj is calculated according to the above (Equation 6), and further, the intake stroke and compression stroke injection amounts qinjp, qinjd are calculated based on the injection ratio rqbasep. That is, if the stoichiometric mode is set, qbase = KGKF × ce × k / 14.7 qinj = qbase × cdpf × (1 + cfb + ctotal) according to (Equation 5) and (Equation 6), and the fuel injection amount of the injector 7 becomes Feedback control is performed based on the signal from the O2 sensor 24 so that the air-fuel ratio of the combustion chamber 4 becomes the stoichiometric air-fuel ratio.

On the other hand, in the enrichment mode, the feedback correction value cfb becomes zero. Therefore, according to (Equation 5) and (Equation 6), qbase = KGKF × ce × k / afw: where afw = 13 to 14 qinj = qbase × cdpf × (1 + ctotal), and the fuel injection amount of the injector 7 is increased from that in the stoichiometric mode, and is open-controlled based on the charging efficiency ce.

Subsequently, at step SA7, the intake stroke and the compression stroke injection amount qinjp, calculated at step SA6,
The injection pulse width Ti of the injector 7 that is proportional to qinjd is read, and in a subsequent step SA8, a control signal is output to the injector 7 of each cylinder 2 to execute fuel injection.
In this way, if the engine 1 is in the stoichiometric mode, the air-fuel ratio of the combustion chamber 4 is corrected by the feedback correction of the basic injection amount qbase obtained based on the output of the airflow sensor based on the output of the O2 sensor. Can be controlled with high precision. In the enrich mode, the fuel injection amount is controlled to be open based on the output of the airflow sensor.

Step SA9 following step SA8
Then, the fuel injection amount by the injector 7 is calculated backward based on the injection pulse width Ti, and the actual fuel injection amount is accurately obtained. True filling efficiency) cecal
Is calculated. The signal from the O2 sensor 24 is input only in the stoichiometric mode, and in the enriched mode, this step SA9 and the following steps SA10 and S10 are performed.
The control of A11 is not performed.

Subsequently, in step SA10, the charging efficiency ce based on the output of the airflow sensor is determined in step SA9.
Is divided by the actual charging efficiency cecal calculated in (1), and the divided value (filling amount deviation) is set as a correction coefficient k corresponding to the operating state of the engine 1 at that time. As a result, the correction coefficient k becomes an accurate value corresponding to the current operating state of the engine 1 and reflecting the influence of the intake pulsation, the individual difference of the air flow sensor 12, and the like. Then, in subsequent step SA11, the corresponding portion on the filling efficiency correction map M is rewritten by the calculated correction coefficient k, and the process returns thereafter.

That is, in the stoichiometric mode, O
2 The actual air-fuel ratio detected by the sensor 24 (A / F = 1
4.7) and the actual fuel injection amount corresponding to the injection pulse width Ti, the actual charging efficiency can be accurately calculated. Therefore, the calculated value cecal and the charging efficiency ce based on the output of the air flow sensor are calculated. By rewriting the correction coefficient k on the filling efficiency correction map M according to the deviation from the above, the filling efficiency correction map M is updated so as to reflect the characteristic of the deviation of the charging efficiency ce.

On the other hand, in step SA12, in which the determination is NO in the stratified combustion mode in step SA2, the final injection amounts qinj, qinjp, qinjd are obtained in accordance with (Equation 5) and (Equation 6) as in steps SA5 and SA6. , And
In the following step SA13, the injection pulse width Ti of the injector 7 is read in the same manner as in step SA7, and then the routine proceeds to step SA8 to execute the fuel injection.
In this stratified combustion mode, the feedback correction value cfb becomes zero as in the enrichment mode. Therefore, according to (Equation 5) and (Equation 6), qbase = KGKF × ce × k / afw: where afw ≧ 30 qinj = qbase × cdpf × (1 + total), and the fuel injection amount of the injector 7 is open-controlled based on the charging efficiency ce.

By repeatedly executing the control procedure shown in the flow chart of FIG. 14 in the stoichiometric mode of the engine 1, the operation region (b) shown in FIGS.
The value of the correction coefficient k is rewritten to an accurate value over a wide area (d). That is, since the charging efficiency correction map M is updated to reflect the influence of the individual difference of the intake pulsation and the air flow sensor 12, the fuel injection amount is charged in the stratified combustion region (a) shown in FIG. Even when the open control is performed based on the efficiency ce, the filling efficiency ce can be extremely accurately corrected by the correction coefficient k on the map M, thereby significantly controlling the air-fuel ratio in the combustion chamber 4. Can be improved. Since the signal from the O2 sensor 24 is not input even in the stratified combustion mode, the correction coefficient k in steps SA9 to SA11 is set.
Is not calculated.

In the flow of FIG. 14, step S
Each step of A5 and SA12 is a basic injection amount calculation unit 50a.
And corresponding to the filling efficiency correction section 50b,
Each step of SA12 corresponds to the final injection amount calculation unit 50c and the divided injection amount calculation unit 50d. Step SA9 corresponds to the actual injection amount calculating means 55 and the charging efficiency calculating means 56. Further, step SA10 corresponds to the correction coefficient calculating means 57, and step SA11 corresponds to the correction coefficient learning means 58. I have.

(Calculation of Feedback Correction Value) Next, the calculation procedure of the feedback correction value cfb in step SA4 in FIG. 14 will be described in detail with reference to FIGS.

First, as shown in FIG. 15, the output (electromotive force) of the O2 sensor disposed in the exhaust passage 22 of the engine 1 is such that the oxygen concentration in the exhaust gas substantially corresponds to the stoichiometric air-fuel ratio. The reference value E1 is obtained when the value is present, but when the density is higher (rich side), it rapidly increases, and when it is lower (lean side), it rapidly decreases.

Therefore, as shown in the flow chart of FIG. 16, first, in step SB1 after the start, O
2 The output E from the sensor 24 is compared with a reference value E1, and E> E
If NO, the process proceeds to step SB5, while E> E1
If YES in step SB2, the flow advances to step SB2. In this step SB2, it is determined whether or not the sensor output E is equal to or less than the reference value E1 in the previous control cycle. If the determination is YES (E ≦ E1), the flow advances to step SB3 to determine whether the feedback correction value cfb A relatively large control gain Cp is subtracted from the value to calculate the current value. On the other hand, if the output E> E1 is NO, the process proceeds to step SB4, in which a relatively small control gain CI is subtracted from the previous value of the feedback correction value cfb, and the current value is calculated.

On the other hand, in step SB5, in which it is determined that E> E1 is not satisfied in step SB2, the air-fuel ratio feedback correction value cfb is calculated in the same manner as in steps SB2 to SB4. That is, it is determined whether or not the sensor output E is larger than the reference value E1 in the previous control cycle, and if this determination is YES, the process proceeds to step SB6.
Then, the control gain Cp is added to the previous value of the feedback correction value cfb to calculate the current value, while if E> E1 is not NO, the process proceeds to step SB7, where the previous value of the feedback correction value cfb is calculated. Is added to the control gain CI.

That is, as shown in FIG. 17, while the air-fuel ratio of the exhaust gas is rich and the output E of the O2 sensor 24 is larger than the reference value E1, the control gains Cp and CI are subtracted from the feedback correction value cfb. By decreasing the value, the fuel injection amount by the injector 7 is corrected to be reduced. On the other hand, while the sensor output E is smaller than the reference value E1, the control gains Cp and CI are added to the feedback correction value cfb. By increasing the value, the fuel injection amount is increased and corrected. Accordingly, the air-fuel ratio of the combustion chamber 4 is maintained at a value near the stoichiometric air-fuel ratio while periodically changing to both the rich side and the lean side with respect to the stoichiometric air-fuel ratio.

(Ignition Timing Control) The ECU 40 includes, as means for controlling the ignition timing of the engine 1, a setting means 53 for setting a basic ignition timing and a correction value, and an ignition timing calculation means 54. The setting means 53 includes an operation mode modf set by the operation mode setting means 48.
Separately, the basic ignition timing thtigb and various ignition timing correction values are set. Specifically, as shown in FIG. 13C, in the stratified combustion mode, according to the second target load Piobjd and the engine speed ne. In addition to obtaining the basic ignition timing thtigb from a map created in advance,
A correction value thtigwd corresponding to dafwbd is obtained from a table created in advance. The correction according to the target air-fuel ratio deviation dafwbd is performed when the basic ignition timing thtigb is set to the target air-fuel ratio during steady-state operation.
While it is determined according to the second target load Piobjd corresponding to afwbd and the engine speed ne, afw0
Is set as the final target air-fuel ratio afw, and there is a deviation in the air-fuel ratio from the steady state, so the ignition timing is adjusted to match the deviation.

In the stoichiometric mode or the enriched mode, the basic ignition timing thtigb is determined from a map prepared in advance according to the charging efficiency ce and the engine speed ne, and the correction value thtigwe at the time of EGR is calculated as the charging efficiency ce. The correction value thtigwd is obtained from a map created in advance according to the engine speed ne, and is corrected according to the target air-fuel ratio deviation dafw0.
And a cold-time correction value thtigwc corresponding to the engine coolant temperature thw is obtained from a table created in advance. The correction according to the target air-fuel ratio deviation dafw0 (= afw0-afw) is performed as described below. The air-fuel ratio at which the NOx generation amount increases when the target air-fuel ratio afw0 falls below a predetermined value leaner than the stoichiometric air-fuel ratio. When the final target air-fuel ratio afw is set to the stoichiometric air-fuel ratio, the ignition timing is adjusted so as to match the change in the air-fuel ratio in order to avoid the passage. When performing the split injection, the basic ignition timing may be obtained from a table created in advance according to the target air-fuel ratio afw.

In this way, based on the basic ignition timing thtigb and various correction values set by the setting means 53, the ignition timing calculating means 54 calculates the ignition timing thtig as follows.

[0097]

(7) thtig = thtigb− (thtigwd + thtigwe + thtigwc) When the ignition timing thtig is reached for each cylinder 2, a control signal is output to the ignition circuit 5 and ignition is performed by the ignition plug 6.

(Effects of Embodiment) In the direct injection type engine 1 of this embodiment equipped with the control device A as described above, any one of the stratified combustion mode, the stoichiometric mode and the enriched mode is selected according to the operating state. In the stratified charge combustion mode, the stratified charge combustion is performed in a state where the air-fuel ratio is significantly leaner than the stoichiometric air-fuel ratio, so that the fuel efficiency is greatly improved. In the stoichiometric mode, the fuel injection amount is set to O
By performing feedback correction based on the output from the two sensors 24, it is possible to control the air-fuel ratio of the combustion chamber 4 with extremely high accuracy so as to be the stoichiometric air-fuel ratio. Further, in the enrich mode, the combustion chamber 4 can be appropriately cooled while increasing the output of the engine 1 by increasing the amount of fuel.

In the stratified combustion mode, for example, the operation mode and the target load are set such that the throttle opening is increased to increase the intake air amount in order to make the air-fuel ratio lean while ensuring the required torque. The intake air amount is controlled on the basis of the fuel injection amount, the injection timing,
The ignition timing and the like are controlled, and at the same time, the EGR valve 27 and the swirl control valve 18 are also controlled. These various control parameters are appropriately controlled in consideration of the control response and the like.

That is, as for the control of the intake air amount, the target charging efficiency ceobj is obtained based on the target air-fuel ratio afwb for intake air amount control set according to the target load and the like. Target volumetric efficiency veobj by correction
Is calculated, and the calculation of the throttle opening is performed based on the calculated value. As a result, the throttle opening is accurately controlled. On the other hand, based on the air-fuel ratio for controlling the injection amount obtained from the target load and the actual charging efficiency by the second target air-fuel ratio setting unit 47, the operation mode setting by the operation mode setting unit 48 and the fuel injection amount, injection timing, Control of the ignition timing and the like is performed, whereby the operation mode, the air-fuel ratio, and the like are appropriately controlled even during the transient operation of the engine in which the charging efficiency changes.

The first target load piobj is used for controlling the intake air amount having a low response speed to the control signal, while the control of the fuel injection amount or the like having a high response speed to the control signal is performed after the smoothing process. By using the second target load piobjd based on the virtual filling efficiency ceimgd, the operation timing of each control parameter can be appropriately adjusted.

That is, in a general gasoline engine in which the throttle opening changes in accordance with the accelerator operation amount while maintaining the standard operating condition in which the air-fuel ratio is the stoichiometric air-fuel ratio in most of the operating range, For example, even if the accelerator operation amount and the throttle opening corresponding to it suddenly change during the acceleration operation of the engine, the change in the intake air amount has a delay, and the change in the engine output corresponds to the change in the intake air amount. become. Therefore, if the output control simulating such a general gasoline engine is to be performed, the second target load Piob based on the simulated virtual filling efficiency ceimgd is obtained.
It is appropriate to consider jd as the actual target load.

Accordingly, the control of the fuel injection amount and the like, which is a high-speed response system, is performed by the second target load Piobjd regarded as the actual target load.
Based on the above, torque characteristics similar to those of a general engine can be obtained, and a good engine feel can be secured. On the other hand, regarding the control of the throttle opening and the like, which is a low-speed response system, since the change in the intake air amount and the like has a certain large delay and the change tends to be slow, the virtual filling without smoothing process is performed. By controlling the opening of the throttle valve 13 and the like according to the first target load Piobj based on the efficiency ceimg, it is possible to suppress a control response delay.

As described above, when the engine 1 is operated while being switched to the stratified charge combustion mode, the stoichiometric mode or the enriched mode, the fuel injection amount is controlled by an air flow sensor or the like in the stratified charge combustion mode or the enriched mode. Open control based on this signal. In that case, the influence of the intake pulsation and the airflow sensor 12
Since there is an individual difference of the fuel cell itself, the detection accuracy of the charging efficiency is insufficient as it is, and the air-fuel ratio of the combustion chamber 4 cannot be controlled with sufficiently high accuracy. In particular, as in the engine 1 of this embodiment, a part of the exhaust gas is recirculated to the intake passage 10 so as to suppress the generation of NOx accompanying the combustion of the engine 1. Changes, the detection accuracy of the charging efficiency ce also decreases.

That is, since the control accuracy of the air-fuel ratio is not so high in the stratified combustion mode or the enrichment mode, the improvement of the fuel efficiency and the reduction of the emission are not sufficiently achieved. However, there is still room for improvement in both fuel efficiency improvement and emission reduction.

In addition, for example, when the mode is switched from the stratified combustion mode to the stoichiometric mode, a torque shock that makes the driver feel uncomfortable due to the difference in air-fuel ratio control accuracy between the respective modes. There was a risk of occurrence. Hereinafter, this point will be specifically described with reference to FIG.

First, when the engine 1 is operated in the stratified combustion mode with a low load and a low rotation speed, as shown in the lowermost position (a) of FIG. Accordingly, the throttle opening tvoobj increases as shown in FIG. At this time, as shown by a broken line in FIG.
If ce is detected to be larger than the actual value, the influence of such a pulsation usually increases as the flow rate of the intake air increases, so that the difference between the true value of the charging efficiency ce and the throttle opening degree also increases.
It grows as tvoobj increases. Then, the fuel injection amount is increased in accordance with the erroneous charging efficiency ce (see Equation 5 and FIG. 5F), and as shown by the broken line in FIG. The fuel ratio temporarily becomes excessively rich, and the output torque of the engine 1 also temporarily becomes excessively large (see FIG. 3H).

On the other hand, when the first target load Piobj exceeds the threshold value Piobj * with the increase of the accelerator opening accel, the basic operation mode mods is changed to the stoichiometric mode as shown in FIG. The throttle switch is switched (t = t1), and the throttle opening tvoobj sharply decreases as shown in FIG.
The throttle valve 13 is closed. Then, slightly later than the valve closing operation, the actual charging efficiency ce in the combustion chamber of each cylinder 2 rapidly decreases as shown in FIG. Since the deviation from the true value also decreases rapidly, the fuel injection amount rapidly decreases and converges to an appropriate amount. As a result, as shown by a broken line in FIG. Output torque decreases. That is, when the operation mode mods is switched, the error in the charging efficiency ce temporarily increases, so that the output torque of the engine 1 fluctuates and a torque shock occurs.

On the other hand, in this embodiment, in FIG.
As shown in the flow of FIG. 4, when the engine 1 is in the stoichiometric mode, deviations (deviations) of the charging efficiency caused by intake pulsation and individual differences between sensors are learned in various operating states, and based on the learning results, Therefore, the filling efficiency ce based on the output of the air flow sensor can be accurately corrected based on the filling efficiency correction map M even in the stratified combustion mode. As a result, the air-fuel ratio of the combustion chamber 4 can be controlled with a high degree of accuracy even in the stratified combustion mode so as not to be so different from that in the stoichiometric mode, so that the torque shock occurs when the operation mode mods is switched as described above. It does not occur.

That is, according to the control device A of this embodiment, as shown by the solid lines in FIGS. 18F and 18G, the fuel injection amount (injection pulse width Ti) before and after the operation mode mods is switched. The air-fuel ratio of the combustion chamber 4 changes smoothly, and during that time, the output torque of the engine 1 gradually increases as shown by the solid line in FIG. When a predetermined time elapses after the switching of the operation mode mods, the final operation mode modf is switched to the stoichiometric mode as shown in FIG. 9E (t = t2), and the fuel injection timing is compressed. The process is switched from the stroke to the intake stroke. Further, in the stoichiometric mode, the fuel injection amount is feedback-corrected based on the signal from the O2 sensor 24, so that the air-fuel ratio of the combustion chamber 4 is maintained near the stoichiometric air-fuel ratio as shown in FIG. Is done.

In short, in this embodiment, the engine 1
In the stratified combustion region, the charging efficiency ce of the cylinder 4 is obtained based on the output signal from the air flow sensor 12, and this charging efficiency
While the fuel injection amount of the injector 7 is controlled to be open based on the ce, the O2 sensor 2 is controlled in the stoichiometric mode.
In the engine control device A configured to switch to feedback control based on the output signal from No. 4, the characteristic of the deviation of the charging efficiency ce is learned during the feedback control,
By correcting the charging efficiency ce at the time of the open control based on the learning result, it is possible to greatly improve the control accuracy of the air-fuel ratio at the time of the open control, that is, in the stratified combustion region.

As a result, in the stratified combustion region, the fuel efficiency and the emission are greatly improved, and the fuel economy can be improved and the emission can be sufficiently reduced even in the entire operation region of the engine 1. In addition, the stratified combustion mode and the stratified combustion mode can be achieved. It is possible to prevent the occurrence of torque shock when the operation mode is switched between the stoichiometric mode and the stoichiometric mode.

This means that, like the engine 1 of this embodiment, the engine 1 is operated by switching between the stratified combustion mode and the stoichiometric mode, and the fuel injection amount is feed-forward controlled in the stratified combustion mode. This is effective in the case where feedback control is performed in the stoichiometric mode. That is, since the combustion state itself of the engine 1 is different between the stratified combustion mode and the stoichiometric mode, there is a possibility that the output torque may fluctuate only by switching. In addition, in the stoichiometric mode, the air-fuel ratio of the combustion chamber 4 is substantially the stoichiometric air-fuel ratio, whereas in the stratified combustion mode, the target air-fuel ratio is made much larger even if the target load is the same.
Since the amount of intake air is significantly increased, the influence of pulsation is considerably increased. Therefore, in the stratified combustion mode, the detection error of the charging efficiency ce based on the output signal from the air flow sensor 12 becomes considerably large, and as a result, problems such as a decrease in the control accuracy of the air-fuel ratio and a torque shock at the time of switching the operation mode occur. Since the fear is particularly large, the operation and effect of the present invention are extremely effective.

(Other Embodiments) The present invention is not limited to the above embodiments, but includes other various embodiments. That is, in the first embodiment,
By rewriting the numerical value on the filling efficiency correction map M in the corresponding operation state with the correction coefficient k calculated in the stoichiometric mode, the filling efficiency correction map M
Although the characteristic of the deviation of ce is reflected, the whole numerical value on the map M may be corrected based on the calculated correction coefficient k. That is, the characteristic of the deviation between the calculated correction coefficient k and the numerical value on the map M may be obtained, and the entire numerical value on the map M may be rewritten according to this characteristic. Regarding the enriched region in the form,
The filling efficiency ce can be accurately corrected.

Further, according to the present invention, the present invention performs feedback control in a relatively low load operation state (for example, an operation region (d) shown in FIG. The present invention can also be applied to an engine in which feedforward control is performed in a load operation state (for example, an operation region (e) shown in FIG. 4B). In general, when the engine is in a high-load operation state, the influence of the intake pulsation is large, and the charging efficiency ce
In the case where the feedforward control is performed in the high load operation state, it is particularly effective that the filling efficiency ce can be accurately corrected by the learned correction coefficient k, as in the present invention, because the detection error of tends to increase. . On the other hand, since the learning of the correction coefficient k itself is performed in a relatively low-load operation state, an engine mounted on a vehicle usually has a long period of time in a so-called low-load to medium-load operation state. Therefore, the characteristics of the deviation of the charging efficiency ce can be quickly learned.

In the above-described embodiment, the lambda O2 sensor is used as the air-fuel ratio detecting means. Alternatively, when the air-fuel ratio of the exhaust gas is within a predetermined range including the stoichiometric air-fuel ratio, the air-fuel ratio is detected. A so-called linear O2 sensor that can obtain a linear output corresponding to the above may be used. In this case, the feedback control can be performed when the air-fuel ratio of the exhaust gas is within a predetermined range in which a sufficiently high detection accuracy can be obtained by the linear O2 sensor. The control area can be expanded. Thus, the correction coefficient k can be obtained over a wider operation range, and the characteristics of the deviation of the charging efficiency ce can be learned over a wider operation range. Further, the air-fuel ratio can be controlled with high accuracy by feedback control in a high-load operation region where the influence of pulsation is particularly strong. Therefore, the operation and effect of the present invention are further enhanced.

Further, in the above-described embodiment, the calculation of the correction coefficient k, that is, the learning of the characteristic of the deviation of the charging efficiency ce is performed, for example, every time the number of times of turning on and off the engine 1 exceeds the set number of times. May be performed when the battery terminal is disconnected once and then connected again.

In the former case, even if the influence degree of the intake pulsation changes due to various factors such as clogging of the air cleaner, the filling efficiency correction map M is updated so as to correspond to the change. Can be. In addition, according to the latter, even if the battery terminal of the vehicle is disconnected and the update result of the charging efficiency correction map M is lost from the memory,
Thereafter, the map M can be immediately updated. On the other hand, in both cases, the updated map M is sufficient until the number of on / off times of the engine exceeds the set number of times or the battery terminal is disconnected, so that unnecessary learning is not performed, thereby reducing the load on the control device. be able to.

[0119]

As described above, according to the first aspect of the present invention, control is performed by switching the fuel injection amount of the in-cylinder injection engine between open control based on the intake charge and feedback control based on the detected air-fuel ratio. When performing feedback control, the control device calculates the actual intake air charge of the combustion chamber based on the detected air-fuel ratio and the fuel injection amount, and calculates the calculated value and the value detected by the charge amount detecting means. Is calculated by the filling amount calculating means. When the feedforward control is performed, the detection value of the charged amount detection means is corrected based on the deviation of the charged amount, so that the detection accuracy of the intake charged amount can be significantly increased. As a result, the control accuracy of the air-fuel ratio at the time of feed forward control is greatly improved, and as a whole it is possible to thoroughly improve fuel efficiency and reduce emissions.
Further, it is possible to prevent the occurrence of torque shock when switching between the feedforward control and the feedback control.

According to the second aspect of the present invention, the fuel injection amount can be accurately detected based on the pulse width of the control signal input to the fuel injection valve.

According to the third aspect of the present invention, the feedforward control is performed when the air-fuel ratio is higher than the stoichiometric air-fuel ratio, while the feedback control is performed when the air-fuel ratio is close to the stoichiometric air-fuel ratio. In some cases, the influence of intake pulsation increases, and the deviation of the detected value of the intake air charge tends to increase.Therefore, it is particularly effective to correct the detected value to improve control accuracy, which is particularly effective in preventing torque shock. Become.

According to the fourth aspect of the invention, when the target air-fuel ratio is set based on the intake air charge, it is particularly effective to increase the detection accuracy of the intake air charge.

According to the fifth aspect of the present invention, when the engine is operated in the stratified combustion state, the feedforward control is performed so that the air-fuel ratio becomes larger than the stoichiometric air-fuel ratio.
In at least a part of the region where the operation is performed in the uniform combustion state, the feedback control is performed so that the air-fuel ratio becomes close to the stoichiometric air-fuel ratio. This is particularly effective in preventing torque shock at the time of mode switching.

According to the invention of claim 6, the generation of NOx accompanying the combustion of the engine can be suppressed by the recirculation of the exhaust gas. In addition, since the accuracy of detecting the intake air charge decreases due to the recirculated exhaust gas, it is particularly effective to correct the intake air charge detection error in such a case.

According to the seventh aspect of the present invention, feedback control can be performed using a relatively inexpensive lambda O2 sensor so that the air-fuel ratio of the combustion chamber becomes a value near the stoichiometric air-fuel ratio.

According to the eighth aspect of the present invention, it is particularly effective to improve the accuracy of the feedforward control in a high load operation range where the influence of the intake pulsation is large.

According to the ninth aspect of the present invention, since the engine mounted on the vehicle is operated for a long period of time from a low load to a medium load, it is necessary to learn the characteristic of the deviation of the intake air charge in the low load side operation state. Thus, the characteristics of the deviation can be learned early.

According to the tenth aspect of the present invention, the feedback control can be performed not only in the vicinity of the stoichiometric air-fuel ratio but also in a wider operation range, so that the characteristic of the deviation of the intake air charge can be detected over a wider operation range.

According to the eleventh aspect of the present invention, the learning accuracy is improved by learning the characteristics of the deviation of the intake air charge amount in a plurality of operating states of the engine.

According to the twelfth aspect, even if the characteristic of the deviation of the intake air charge changes due to various factors, the learning content can be updated so as to correspond to the change, but no unnecessary learning is performed. The load on the control device can be reduced.

According to the thirteenth aspect, even when the battery characteristic of the deviation is lost due to the disconnection of the battery terminal of the vehicle,
Thereafter, while learning can be performed promptly, unnecessary learning is not performed, and the load on the control device can be reduced.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing a configuration of the present invention.

FIG. 2 is a schematic configuration diagram showing an embodiment of an engine control device according to the present invention.

FIG. 3 is a map (a) in which the stratified combustion mode, the stoichiometric mode, and the enrich mode are set when the engine is warm, and a map (b) in which the stoichiometric mode and the enrich mode are set when the engine is cold. FIG.

FIG. 4 is a functional block diagram of an ECU.

FIG. 5 is a functional block diagram showing a specific configuration of a target load setting unit in FIG.

6 is a functional block diagram showing a specific configuration of a throttle opening calculating means in FIG. 4;

FIG. 7 is a functional block diagram showing a specific configuration of a second target air-fuel ratio setting means in FIG. 4;

8 is a functional block diagram showing a specific configuration of an injection amount calculating means in FIG.

FIG. 9 is a diagram showing a correspondence relationship between an accelerator operation amount, an engine speed, and virtual volume efficiency.

FIG. 10 is a diagram showing a map in which a target air-fuel ratio for controlling an intake air amount is set for each operation mode.

FIG. 11 is a diagram showing a correspondence relationship between a target volume efficiency and a throttle opening.

FIG. 12 is a diagram showing setting of an operation mode used for calculation of a fuel injection amount and the like.

FIG. 13 is a diagram showing a map in which a target air-fuel ratio (a), an injection timing (b), and an ignition timing (c) for controlling a fuel injection amount and the like are set for each operation mode.

FIG. 14 is a flowchart showing a schematic procedure of fuel injection control and a learning procedure of a correction coefficient k.

FIG. 15 is a diagram showing output characteristics of the O2 sensor with respect to changes in the air-fuel ratio.

FIG. 16 is a flowchart illustrating a procedure for calculating a feedback correction value.

FIG. 17 is a time chart showing a change in a feedback correction value during feedback control and an O2 sensor output in association with each other.

FIG. 18 shows how the operation mode is switched at the time of slow acceleration of the engine, the accelerator opening, the basic injection mode, the throttle opening, the actual charging efficiency, the final injection mode,
FIG. 5 is a time chart showing the injection pulse width, the actual air-fuel ratio, and the change state of the output torque of the engine in association with each other.

[Explanation of symbols]

A Control device of in-cylinder injection engine 1 engine 2 cylinder 4 combustion chamber 7 injector (fuel injection valve) 12 air flow sensor (filling amount detecting means) 24 O2 sensor (air-fuel ratio detecting means) 26 EGR passage (exhaust recirculation means) 27 EGR valve (exhaust gas recirculation means) 40 control unit (ECU) 47 second target air-fuel ratio setting means (target air-fuel ratio setting means) 48 operation mode setting means (fuel injection control means) 49 split ratio setting means (fuel injection control means) Reference Signs List 50 injection amount calculating means (fuel injection control means) 50b filling efficiency correction means (filling amount correction means) 51 injection timing setting means (fuel injection control means) 52 injection control means (fuel injection control means) 55 actual injection amount calculation means ( Injection amount detecting means) 56 filling amount calculating means 57 correction coefficient calculating means (deviation characteristic learning means) 58 correction coefficient learning means (Deviation characteristic learning means)

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) F02D 41/14 310 F02D 41/14 310H 330 330A 41/18 41/18 FG 43/00 301 43/00 301G 301N 45/00 340 45/00 340B (72) Inventor Michihiro Imada 3-1, Fuchu-machi, Shinchu, Aki-gun, Hiroshima Mazda Co., Ltd. (72) Inventor Keiji Araki 3-1, Shinchi, Fuchu-cho, Aki, Hiroshima F-term in Mazda Co., Ltd. (reference) 3G084 AA04 BA05 BA09 BA13 BA15 BA17 BA20 BA21 CA03 CA04 DA02 DA10 DA25 EA07 EA11 EB08 EB12 EB13 EB16 EB20 EB25 EC02 EC03 FA03 FA07 FA10 FA13 FA18 FA29 FA33 FA36 3G301 JA04 KA08 KA09 KA11 KA21 LA03 LA05 LB04 LC03 MA01 MA11 MA19 MA26 NA01 NA08 NB02 NC02 ND02 ND05 ND12 ND33 ND42 NE13 NE14 NE15 NE16 NE 17 NE19 NE22 NE23 PA01Z PA07Z PA09Z PA11Z PA17Z PB03Z PB08Z PC02Z PD03A PD15Z PE01Z PE08Z PF03Z PF13Z PF16Z PG01Z

Claims (13)

[Claims] A fuel is directly injected into a combustion chamber in a cylinder of an engine.
A fuel injection valve for injecting and supplying, a filling amount detecting means for detecting a filling amount of intake air in the combustion chamber, an air-fuel ratio detecting means for detecting an air-fuel ratio of the combustion chamber, and a target air-fuel ratio at least according to an operating state of the engine. A target air-fuel ratio setting means for setting a fuel injection amount by the fuel injection valve, and a feed for performing feed-forward control so that an air-fuel ratio of a combustion chamber becomes the target air-fuel ratio in accordance with a detection result by the filling amount detection means. A control device for a direct injection engine, comprising: a forward control operation; and a fuel injection control unit that switches to a feedback control operation for performing feedback control so that a value detected by the air-fuel ratio detection unit becomes the target air-fuel ratio. In the above, an injection amount detecting means for detecting a fuel injection amount by the fuel injection valve, and a feedback control by the fuel injection control means A charge calculating means for calculating an intake charge amount of the combustion chamber based on the detected values of the air-fuel ratio detecting means and the injection amount detecting means during execution of the operation; A filling amount deviation is calculated from the detected value by the detecting means, and the detected value of the charged amount detecting means is corrected based on the difference in the charged amount when the fuel injection control means executes the feedforward control operation. A control device for a direct injection type engine, comprising: a quantity correcting means. 2. The in-cylinder injection system according to claim 1, wherein the injection amount detecting means detects the fuel injection amount based on a pulse width of a control signal input to the fuel injection valve. Engine control device. 3. The fuel injection control device according to claim 1, wherein the fuel injection control means performs the feedforward control operation when the target air-fuel ratio is set to a value larger than the stoichiometric air-fuel ratio by the target air-fuel ratio setting means. A control device for a direct injection engine, which is configured to perform a feedback control operation when is set near the stoichiometric air-fuel ratio. 4. The target air-fuel ratio setting unit according to claim 3, wherein the target air-fuel ratio setting means sets the target air-fuel ratio based on a target load of the engine corresponding to the accelerator operation amount and at least one of an engine speed and an intake charge amount. A control device for a direct injection engine, comprising: 5. The target air-fuel ratio setting means according to claim 3, wherein the target air-fuel ratio is set to a value larger than the stoichiometric air-fuel ratio when the engine is in a predetermined low load operation region. The fuel injection control means performs stratified combustion by injecting fuel in the compression stroke of the cylinder by the fuel injection valve when the engine is in the predetermined operation range, and performs fuel injection in the intake stroke of the cylinder in other operation ranges. A control device for a direct injection engine, wherein the control device is configured to perform uniform combustion by injecting fuel. 6. The control apparatus for a direct injection engine according to claim 3, further comprising an exhaust gas recirculation means for recirculating a part of the exhaust gas of the engine to an intake system. 7. The control of a direct injection engine according to claim 3, wherein the air-fuel ratio detecting means detects the air-fuel ratio only when the air-fuel ratio of the exhaust gas is near the stoichiometric air-fuel ratio. apparatus. 8. The control device for a direct injection engine according to claim 1, wherein the fuel injection control means is configured to perform a feedforward control operation when the engine is in a high load range. . 9. The charging amount correcting means according to claim 8, further comprising a deviation characteristic learning means for learning a characteristic of a deviation of the charging amount when the engine is in a low load range,
A control device for a direct injection engine, wherein the control unit is configured to correct a value detected by a charged amount detecting means based on the learned deviation characteristic. 10. The air-fuel ratio detecting means according to claim 1, wherein the air-fuel ratio detecting means detects the air-fuel ratio only when the air-fuel ratio of the exhaust gas is within a predetermined range including a stoichiometric air-fuel ratio. A feedback control operation is performed when the target air-fuel ratio is set within the predetermined range by the air-fuel ratio setting means, and a feedforward control operation is performed when the target air-fuel ratio is set outside the predetermined range. A control device for an in-cylinder injection engine. 11. The method according to claim 1, wherein the filling amount correction unit includes a deviation characteristic learning unit that learns a characteristic of a deviation of the filling amount at least in a plurality of engine operating states having different engine speeds. A control device for a direct injection type engine, wherein the control value is corrected based on a charged amount detecting means based on the detected value. 12. The in-cylinder according to claim 11, wherein the deviation characteristic learning means is configured to learn a characteristic of a deviation of the intake air charge amount each time the number of engine starts exceeds a set number of times. Control system for injection engine. 13. The deviation characteristic learning means according to claim 11, wherein when the battery terminal of the vehicle equipped with the engine is disconnected once and then reconnected, the characteristic of the deviation of the intake air charge is learned. A control device for an in-cylinder injection engine.