JPH1136993A - Exhaust gas circulation control device of direct injection engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make compatible a reduction in NOx amount with a reduction in smoke amount. SOLUTION: This exhaust gas circulation control device of direct injection engine is provided with an exhaust gas circulating passage 13 to circulate part of exhaust gas to an intake system, a means 14 to control exhaust gas circulation flow, a means 6 to measure intake air amount for each cylinder, and a means to obtain fuel injection amount. Also a target air-fuel ratio common to all cylinders at which the requirements can be made compatible with each other is determined, intake air amount is detected for each cylinder and, according to the intake air amount, the exhaust gas circulation flow is controlled for each cylinder so that the target air-fuel ratio is obtained. Thus the air-fuel ratio for each cylinder can be uniformed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exhaust gas recirculation control device for a direct injection engine mounted on an automobile or the like.

[0002]

2. Description of the Related Art Japanese Patent Publication No. 63-50544 discloses an exhaust gas recirculation control device for a diesel engine. In order to reduce NOx (nitrogen oxides) in the exhaust gas, the intake air (new) is controlled by adjusting the exhaust gas recirculation amount. It is described that the air ratio λ of the gas is adjusted.

JP-A-6-229322 discloses an exhaust gas recirculation control device for a multi-cylinder engine, which detects an intake air amount for each cylinder and adjusts the exhaust gas recirculation amount in accordance with the intake air amount. ing. This is the EGR between cylinders
(Exhaust gas recirculation) rate (= EGR amount / intake air amount) is not varied, that is, the EGR rate of each cylinder is made equal.

[0004]

Incidentally, a diesel engine is operated in a state in which the air-fuel ratio is extremely lean (lean), so that the amount of NOx emission increases. This problem can be dealt with by increasing the amount of exhaust gas recirculation, thereby reducing NOx. However, when the exhaust gas recirculation amount is increased, the amount of air in the intake air decreases, thereby increasing the amount of smoke in the exhaust gas. The fact that the amount of air during intake decreases
That is, the air-fuel ratio is changing to the rich side.

On the other hand, the present inventor has investigated the relationship between the air-fuel ratio and the amount of smoke, and has found that the amount of smoke increases rapidly when the air-fuel ratio exceeds a certain value. Therefore, in order to achieve both the reduction of NOx and the reduction of smoke, it can be said that it is preferable to control the exhaust gas recirculation amount with the target of the air-fuel ratio as rich as possible before the smoke amount starts to increase rapidly.

However, in a multi-cylinder engine, even if the exhaust gas recirculation amount is adjusted with the target of the air-fuel ratio before the smoke amount starts to increase rapidly, NOx is reduced, but a cylinder with a large smoke amount, and conversely a smoke Some cylinders have a small amount but a large amount of NOx. Due to such inter-cylinder variations, NOx and smoke are not reduced as expected as a whole engine.

[0007]

In this regard, the EGR rate of each cylinder (= EGR amount / total exhaust amount; hereinafter, unless otherwise specified), the EGR ratio is the recirculated exhaust amount of the total exhaust amount. When referring to the ratio of the EGR amount) and the deviation of the intake air amount of each cylinder from the average intake air amount, the results shown in FIG. 49 were obtained. That is, even if the opening degree of the exhaust recirculation passage valve is the same, the EGR rate and the intake air amount deviation of each cylinder vary, and the cylinder with a high EGR rate has a small intake air amount and the cylinder with a low EGR rate has a small amount. The intake air volume is large. This is considered to be due to the fact that not only the distribution of the recirculated exhaust gas to the cylinders varies, but also the air intake characteristics between the cylinders vary.

Accordingly, the present invention basically determines a common target air-fuel ratio for all cylinders, detects an intake air amount for each cylinder, and sets the target air-fuel ratio in accordance with the intake air amount. Instead of controlling the exhaust gas recirculation amount for each cylinder, that is, instead of making the ratio of the EGR amount to the intake air amount of each cylinder uniform, the air-fuel ratio of each cylinder may be made uniform by targeting a predetermined air-fuel ratio. It was made.

That is, the present invention comprises an exhaust gas recirculation passage for recirculating a part of the exhaust gas to the intake system, means for adjusting the amount of exhaust gas recirculated, and means for measuring the amount of intake air for each cylinder. In an exhaust gas recirculation control device for a multi-cylinder direct injection engine, which controls the amount of exhaust gas recirculation based on the amount of intake air, means for determining the fuel injection amount of the engine, and the amount of intake air measured for each cylinder An exhaust gas recirculation amount control means for controlling the operation of the exhaust gas recirculation amount adjusting means for each cylinder based on the fuel injection amount so as to attain a target air-fuel ratio.

Therefore, the exhaust gas can be recirculated until the target air-fuel ratio reaches the target air-fuel ratio in accordance with the intake air amount for each cylinder, which is advantageous in achieving both reduction of NOx and reduction of smoke over all cylinders.

Here, in the direct injection type engine, the fuel injection amount supplied to each cylinder is determined by the accelerator opening (accelerator pedal depression amount), and there is substantially no difference.
If a target air-fuel ratio common to all cylinders is determined and the EGR amount of each cylinder is controlled based on the measured intake air amount, the intake air amount of each cylinder changes accordingly to achieve the target air-fuel ratio. be able to. The target air-fuel ratio common to all cylinders means that the target air-fuel ratio of each cylinder is substantially the same, and includes completely the same.

The target air-fuel ratio is an air-fuel ratio that is as rich as possible before the smoke starts to increase rapidly when the exhaust gas recirculation amount is increased, that is, the air-fuel ratio becomes richer as the air-fuel ratio becomes richer. What is necessary is just to set the air-fuel ratio when the amount of smoke changes from a gradual increase to a sudden increase when the change characteristic in which the amount of smoke increases is observed. For the target air-fuel ratio, an optimal value experimentally determined in advance in accordance with changes in the engine speed, the fuel injection amount, and the like is stored in the memory of the control means so that reduction of NOx and reduction of smoke can be achieved at the same time. What is necessary is just to store it electronically above, and to calculate from there.

<About the intake air amount measuring means> Regarding the means for measuring the intake air amount for each of the cylinders described above, O
_{It is also possible} to provide an O ₂ sensor for detecting the concentration in the exhaust passage, obtain the air-fuel ratio at that time from the output of the sensor, and obtain the intake air amount based on the air-fuel ratio and the fuel injection amount. Preferably, a sensor for detecting the air flow rate or the intake pipe pressure is provided in the intake passage, and the intake air amount is determined for each cylinder based on the output from the sensor.

That is, in the measurement by the O ₂ sensor, the exhaust gas recirculation amount is controlled by, for example, obtaining the intake air amount about two cycles before. However, in the case of the sensor in the intake passage, the intake air amount before combustion is measured. Quantity can be measured,
The exhaust gas recirculation amount can be controlled with good responsiveness in response to a change in the intake air amount, which is advantageous for reliably reducing the NOx and smoke. When the intake pipe pressure is detected, the intake air amount is determined based on the detected intake pipe pressure and the engine speed.

As a sensor for detecting the air flow rate, it is preferable to employ a constant temperature type hot film type air flow sensor. This is because the amount of heat dissipated by the hot film heated by energization depends on the mass of the air passing therethrough, so the air flow rate is determined based on the amount of energization necessary to keep this hot melt at a constant temperature.
That is. According to this, even if there is a variation in the flow velocity, the air flow rate can be reliably detected.

The constant-temperature hot-film type air flow sensor includes a heater disposed in an intake passage so as to be orthogonal to the direction of flow of the intake air, and hot films disposed upstream and downstream of the heater. A backflow detection type that detects backflow based on the level of the temperature of both hot films is preferable, whereby only the forward flow of air flowing through the cylinder can be measured, and the backflow is controlled to control the exhaust gas recirculation amount. Can be avoided.

<Regarding Control of Exhaust Gas Recirculation Amount for Each Cylinder> In controlling the exhaust gas recirculation amount for each cylinder by the control means, the intake air of each cylinder is determined based on the difference in the measured intake air amount between the cylinders. It is preferable to obtain a parameter representing the amount characteristic and control the operation of the exhaust gas recirculation amount adjusting means for each cylinder based on the parameter.

That is, the present invention does not exclude control of the exhaust gas recirculation amount based on the deviation of the actual intake air amount of each cylinder from the average intake air amount, but the deviation is generally small. And is affected by disturbances such as engine torque fluctuations. Therefore, if a parameter representing the intake air amount characteristic of each cylinder is obtained based on the difference in the measured intake air amount, and the individual difference (characteristic difference) related to the air intake of each cylinder is clarified, the parameter becomes Control based on the individual difference between the cylinders can be performed based on this, which is advantageous in achieving both the reduction of NOx and the reduction of smoke.

In particular, there is provided means for measuring the intake stroke time for each cylinder, and the intake air amount measured for each cylinder is reduced to reduce the difference in intake air amount due to the length of the intake stroke time for each cylinder. It is preferable to obtain the above parameters by performing the above processing.

That is, when the torque fluctuation of the engine temporarily increases, the intake stroke time of one cylinder becomes longer and that of the other cylinder becomes shorter, so that the variation of the intake stroke time becomes large. In this case, even if there is a difference in the amount of air actually sucked into each cylinder, there is a portion caused by disturbance such as a short intake stroke time. Therefore, even if the characteristic parameter is obtained based on the difference in the actual intake air amount, the characteristic parameter does not reflect the actual individual difference between the cylinders, and the exhaust gas recirculation control that is significant in the present invention can be performed. become unable. For this reason, it is preferable to obtain the parameters so as to exclude the variation in the intake stroke time.

For example, the change rate .DELTA.Qi of the intake air amount of the cylinder with respect to the cylinder whose intake stroke is immediately before is determined by the intake stroke of the cylinder with respect to the cylinder whose intake stroke is immediately before. ΔQti = ΔQi / divided by the rate of change of time ΔTi
ΔTi is determined, and as shown in the following equation, ΔQti is obtained by reflecting the previous value of the cylinder at a predetermined ratio to the current value of ΔQti.
t '(i) may be calculated as a parameter representing the intake air amount characteristic of the cylinder.

ΔQt ′ (i) = ΔQti × r + ΔQti ′ (1-r) where 0 <r ≦ 1, that is, ΔQti ′ is the previous value of ΔQti and the current value ΔQti.
The previous value is reflected in Qti at a predetermined ratio. As a result, the individual difference between cylinders with respect to the intake air amount (property of each cylinder relating to air intake) becomes gradually clearer, which is advantageous for controlling the exhaust gas recirculation amount according to such individual difference. In other words, since the value of ΔQti itself is not so large, it is used after emphasizing it.

Specifically, based on the average intake air amount obtained by averaging the intake air amount of each cylinder measured by the intake air amount measuring means over all cylinders, the exhaust gas recirculation amount for obtaining the target air-fuel ratio is set. What is necessary is just to determine the basic control amount of the adjusting means, and to correct the basic control amount using the parameters of the respective cylinders.

The exhaust gas recirculation utilizes the pressure difference between intake and exhaust during the intake stroke of the cylinder.
This differential pressure differs between cylinders even in the same engine operating state, and the differential pressures of the respective cylinders change in different modes according to changes in the engine operating state. Therefore, even when the exhaust gas recirculation amount adjusting means is controlled based on the required exhaust gas recirculation amount, the amount of exhaust gas actually recirculated to each cylinder or the EGR rate becomes different.

Therefore, a means for detecting a predetermined parameter for specifying the operating state of the engine and a correction for reducing a variation in the recirculation amount of the exhaust gas among the cylinders due to the difference in the differential pressure based on the parameter. It is preferable to provide a correction means for giving the control amount of the reflux amount adjusting means. As a result, it is possible to avoid the occurrence of a control error in the exhaust gas recirculation amount due to the individual difference in the exhaust gas recirculation characteristics between the cylinders, and to achieve both a reduction in NOx and a reduction in smoke in the air-fuel ratio of each cylinder. The target air-fuel ratio can be reliably controlled.

[0026]

Therefore, according to the present invention, the target air-fuel ratio is determined, and the exhaust gas recirculation amount is adjusted for each cylinder based on the fuel injection amount and the intake air amount measured for each cylinder so as to achieve the target air-fuel ratio. Since the operation of the means is controlled, the exhaust gas can be recirculated until the target air-fuel ratio reaches the target air-fuel ratio in accordance with the intake air amount of each cylinder, thereby achieving both the reduction of NOx and the reduction of smoke over all cylinders. It is more advantageous.

[0027]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS <Engine Configuration> In FIG. 1, reference numeral 1 denotes an engine body of a four-cylinder diesel engine as a direct injection engine mounted on an automobile, 2 denotes an intake passage, 3 denotes an exhaust passage, and 4 denotes a cylinder of each cylinder. A fuel injection valve 5 for injecting fuel into the combustion chamber is a control unit (computer control means). The intake passage 2 is provided with an air flow sensor 6, a supercharger (VGT) 7, and an intercooler 8 in this order from the upstream side, and is branched at the downstream end and connected to each cylinder. In the exhaust passage 3, an O ₂ sensor 9, an exhaust pressure sensor 11, the supercharger 7, and a catalytic converter (catalyst) are provided on the downstream side of the collecting portion.
12 are provided.

The intake passage 2 and the exhaust passage 3 are connected to the exhaust passage 3
Is connected by an EGR passage 13 extending from the downstream side of the turbocharger 7 and reaching the gathering portion of the intake passage 2. The EGR passage 13 is provided with a negative-pressure operated EGR valve 14 for adjusting the exhaust gas recirculation amount. ing. That is, the negative pressure pump 16 is connected to the EGR valve 14 via the negative pressure passage 15.
The negative pressure passage 15 is provided with a solenoid valve 17 for negative pressure control and a negative pressure sensor 18. The EGR valve 14 is provided with a lift sensor 19 for detecting the lift amount. An intake pressure sensor 21 and an intake air temperature sensor 22 are provided in a gathering portion of the intake passage 2, and a sensor 23 for detecting a crank angle is provided on a crankshaft of the engine.
The sensor 23 is also used for discriminating a cylinder and detecting an engine speed.

A fuel injection pump 24 is connected to the fuel injection valve 5 of each cylinder via a fuel passage. A sensor 26 for detecting a fuel supply pressure is provided on a common rail (common path) 25 of the fuel passage. 27 is an accelerator opening sensor for detecting the amount of depression of the accelerator pedal.

-EGR Valve and Supercharger- As shown in FIG. 2, a valve rod 14b is fixed to a diaphragm 14a that partitions a valve box of the EGR valve 14, and this valve rod 14b
A valve body 14c for linearly adjusting the degree of opening of the EGR passage 13 and a lift sensor 19 are provided at both ends. The valve body 14c is urged in the closing direction by a spring 14d. The negative pressure passage 15 is connected to the negative pressure chamber of the valve box. An electromagnetic valve 17 provided in the negative pressure passage 15 communicates and shuts off the negative pressure passage 15 with a control signal (current) from the control unit 5, so that the EG of the negative pressure chamber is reduced.
The R-valve driving negative pressure is adjusted, and thereby, the valve body 14c
, The degree of opening of the EGR passage 13 is linearly adjusted.

That is, as shown in FIG. 3, as the current increases, the EGR valve driving negative pressure increases (that is, the pressure decreases). As shown in FIG. 4, the EGR valve driving negative pressure is proportional to the EGR valve driving negative pressure. The lift amount of the valve body 14c changes. However, hysteresis is observed.

As shown in FIGS. 5 and 6, the supercharger 7
A VGT (Variable Geometry Turbo) in which a flap 7b for changing a cross-sectional area A of the inlet and changing a radial position R of the inlet at the same time is provided at the inlet of the turbine chamber 7a. . As shown in FIG. 5, the supercharging efficiency is increased by reducing the A / R by positioning the flap 7b such that its tip is close to the peripheral wall of the turbine chamber 7a. As shown in FIG. Is positioned so as to be closer to the center of the turbine chamber 7a.
Increasing / R lowers the supercharging efficiency.

<Overall Configuration of Control System for Direct Injection Engine> The objects to be controlled are an EGR valve (a control valve for an exhaust gas recirculation passage supplied from an exhaust passage to an intake passage), a VGT, and a fuel injection valve. These controls are executed by a control program electronically stored on the memory of the control unit 5.

-Exhaust gas recirculation control- As shown in FIG. 7, the control unit 5 is a two-dimensional system which records an experimentally determined optimum target torque trqsol in the case of changes in the accelerator opening accel and the engine speed Ne. Map 31, engine speed Ne, target torque trqsol
And fresh air volume (the amount of intake air and does not include fuel.
same as below. 3) A three-dimensional map 3 recording the experimentally determined optimum target fuel injection amount Fsol in the change of FAir.
2. The optimal target air-fuel ratio A / Fsol determined experimentally in the change of the engine speed Ne and the target torque trqsol
Are stored electronically in a memory.

The target air-fuel ratio A / Fsol is a reference for determining the exhaust gas recirculation amount for achieving both reduction of NOx and reduction of smoke. That is, FIG. 8 shows the relationship between the air-fuel ratio of a diesel engine and the amount of NOx in exhaust gas (an example).
As shown in the graph, when the air-fuel ratio increases, the NOx amount tends to increase. Therefore, it can be seen that if the exhaust gas recirculation amount is increased to lower the air-fuel ratio (to make it richer), the generation of NOx decreases.

However, as shown in FIG. 9, the relationship between the air-fuel ratio of the engine and the smoke value in the exhaust gas shows that the air-fuel ratio is rich, and when the air-fuel ratio falls below a certain air-fuel ratio, the amount of smoke increases rapidly. I do. For this reason, there is a limit in increasing the exhaust gas recirculation amount. To achieve the above balance, the target air-fuel ratio should be as rich as possible so as to reduce NOx, and before the smoke amount starts to increase rapidly. It can be said that it is necessary to control the exhaust gas recirculation amount with this value set as a target.

In this control system, the fresh air amount obtained from the air flow sensor 6 and the new air amount obtained from the O ₂ sensor 9 are switched by the new air amount switching unit 34 and used for controlling the exhaust gas recirculation amount. . This switching will be described later.

(When the new air volume obtained by the air flow sensor 6 is used) The target torque calculation unit 41 uses the accelerator opening degree accel detected by the sensors 1 and 2 and the engine speed Ne to store the memory. The target torque trqsol is determined with reference to the two-dimensional map 31 above. This target torque trqsol and the fresh air amount measured by the air flow sensor 6 and sent through the fresh air amount switching unit 34
Using the FAir and the engine speed Ne, the target injection amount calculation unit 42 determines the target injection amount Fsol with reference to the three-dimensional map 32 in the memory. On the other hand, the above target torque
Using the trqsol and the engine speed Ne, the target air-fuel ratio calculation unit 43 determines the target air-fuel ratio A / Fsol for achieving the above-mentioned compatibility by referring to the two-dimensional map 33 in the memory.

The target injection amount Fsol and the target air-fuel ratio A /
The target fresh air amount FAsol is calculated in the target fresh air amount calculation unit 44 using Fsol (FAsol = Fsol × A / Fsol). With the target fresh air amount FAsol as a target, feedback control of the new air amount FAir is performed in the fresh air amount feedback control unit 45. This control is the same as performing the air-fuel ratio feedback control, but instead of directly adjusting the fresh air supply amount itself, the fresh air amount is changed by adjusting the exhaust gas recirculation amount. That is,
The target EG is not determined for the fresh air correction amount.
The operation amount EGRsol of the R valve will be determined.

(In the case where the fresh air amount obtained by the O ₂ sensor is used) The O ₂ concentration in the exhaust gas detected by the O ₂ sensor 9 is converted to the air-fuel ratio A / F in the air-fuel ratio converter 46. This conversion is performed by referring to a table that is stored electronically on the memory of the control unit 5 and represents the relationship between the two. Using the measured air-fuel ratio A / F and the target injection amount Fsol, a new air amount FA
ir is calculated (FAir = A / F × Fsol). The fresh air amount FAir by the O ₂ sensor 9 is used for calculating the target injection amount Fsol via the fresh air amount switching unit 34, as in the case of the air flow sensor 6 described above. The target injection amount Fsol used for calculating the fresh air amount FAir is a previous value.

The air-fuel ratio feedback control section 48 performs feedback control of the air-fuel ratio A / F with the target air-fuel ratio A / Fsol obtained from the two-dimensional map 33 as a target. This control also changes the fresh air amount by adjusting the exhaust gas recirculation amount to bring the air-fuel ratio A / F closer to a target value.
Rsol will be determined.

(Control of Exhaust Gas Recirculation Amount Control Means) In the EGR valve drive amount switching unit 49, the target operation amount EGRs of the EGR valve obtained by using the fresh air amount by the air flow sensor 6.
ol and E obtained using the fresh air amount obtained by the O ₂ sensor 9.
One of the target operation amount EGRsol of the GR valve is selected,
Used for controlling the EGR valve. This selection switching is linked to the previous switching of the fresh air amount switching unit 34.

-VGT Control- The control unit 5 electronically stores a two-dimensional map 51 storing the experimentally determined optimum target turbo efficiency VGTsol in a change in the target torque trqsol and the engine speed Ne on a memory. Using the target torque trqsol and the engine speed Ne obtained by the two-dimensional map 31, the target turbo efficiency VGTsol is calculated in the target turbo efficiency calculation unit 52 with reference to the map 51. The VGT is controlled using this.

-Control of Fuel Injection- The control unit 5 has a two-dimensional map 53 that records the experimentally determined optimum common rail pressure CRPsol in the change of the target torque trqsol and the engine speed Ne.
Is stored electronically on a memory. Then, the target torque trqsol obtained from the two-dimensional map 31
The common rail pressure calculating unit 54 calculates the target common rail pressure CRPsol by referring to the map 53 using the engine speed Ne and the engine speed Ne, and controls the common rail pressure using the target common rail pressure CRPsol. This controlled common rail pressure CR
The excitation time of the electromagnetic fuel injection valve 4 is determined and controlled based on P and the target injection amount Fsol.

<Overall Flow of Exhaust Gas Recirculation / Fuel Injection Control>
FIG. 10 shows the entire flow of the control. That is, the intake air amount FAir is obtained for each cylinder based on the intake air amount detected by the air flow sensor 6 or the O ₂ sensor 9 and the crank angle detected by the crank angle sensor 23 (steps 1 to 3). Also, the engine speed N detected by the crank angle sensor 23
e, The target fuel injection amount Fsol is obtained based on the accelerator opening accel detected by the accelerator opening sensor 27 and the intake air amount FAir (steps 4 to 6).

A transient determination is made as to whether or not the engine is in an accelerating operation state based on the accelerator opening accel, the engine speed Ne, and the like (step 7). During a steady operation, a basic target air-fuel ratio is set, and the target intake air is set. The amount is determined and EG
The basic control of the R valve is performed, and the basic control is corrected by the EGR valve control for each cylinder based on the intake air amount FAir for each cylinder (steps 8 to 11). The correction control of the EGR valve for each cylinder corresponds to air-fuel ratio feedback control for achieving both reduction of NOx and reduction of smoke. During acceleration operation, a target air-fuel ratio during acceleration is set, and EGR valve control and injection amount control during acceleration are performed (step 1).
2-14).

<Detection of Intake Air Flow Rate and Calculation of Intake Air Volume for Each Cylinder> The air flow sensor 6 used for this detection is as follows.
It is a constant temperature type hot film type and includes a heater disposed in the intake passage 2 so as to be orthogonal to the direction of intake air flow, and hot films disposed upstream and downstream with the heater interposed therebetween. This is a backflow detection type that detects backflow based on the level of the temperature. FIG. 11 shows an example of the detected intake air flow rate. The hatched portion in the drawing is the backflow, and it can be seen that the integrated value from which the backflow has been subtracted, that is, the amount of air actually sucked into each cylinder fluctuates.

FIG. 12 shows the calculation of the intake air amount for each cylinder when the air flow sensor 6 is used (step 1 in FIG. 10).
3) shows a specific flow. As the intake air flow rate is integrated and the elapsed time is measured, each time the crank angle reaches 180 degrees, the integral value Q of the intake air flow rate for that 180 degrees is calculated as the intake air amount Qi of the cylinder (i). age,
The required time (crank timer time T) is determined by the cylinder (i)
And the average value of the obtained intake air amounts Qi of the four cylinders is obtained as the basic intake air amount Qav (steps A1 to A7). Each of the four cylinders is given a cylinder number "0, 1, 2, 3" for convenience.

Further, the cylinder (i
-1), the rate of change ΔQi = Qi / Qi-1 of the intake air amount of the cylinder (i) and the rate of change ΔTi = Ti / T of the crank interval.
i-1 is calculated, and a change index ΔQti = ΔQi / ΔTi of the intake air amount taking into account the time of the intake stroke is calculated (steps A8 to A8).
A10). Here, ΔTi is taken into consideration in order to eliminate disturbances due to torque fluctuations (angular velocity fluctuations of the crankshaft) as much as possible. This processing is particularly effective during idling operation in which torque fluctuations are large. Then, based on the change index ΔQti, the intake air amount characteristic ΔQt ′ of each cylinder
(i) is obtained by the following equation (step A11).

ΔQt ′ (i) = ΔQti × r + ΔQti ′ (1-r) where 0 <r ≦ 1, that is, ΔQti ′ is the previous value of the change index ΔQti,
The previous value is reflected at a predetermined ratio in the current change index ΔQti. Thereby, the individual difference between cylinders regarding the amount of intake air gradually becomes clear.

<Transient Determination> FIG. 13 shows a specific flow of the transient determination (steps 4 to 7 in FIG. 10). The transient determination includes a determination based on a change in the accelerator opening and a determination based on a change in the fuel injection amount. During the acceleration operation of the engine, it is necessary to increase the intake air amount in accordance with the increase in the fuel injection amount. To that end, it is necessary to rapidly reduce the exhaust gas recirculation amount. This is a transient determination for performing such exhaust gas recirculation amount reduction control.

That is, the fuel injection amount F is read from the three-dimensional map 32 in FIG. 7 using the accelerator opening Acc, the engine speed Ne, and the intake air amount Qav, and the present value Acc and the previous value of the accelerator opening are used. Based on Acc ′, a change amount ΔAcc = Acc−Acc ′ is obtained (steps B1 to B3). The acceleration determination reference αcc is read from the two-dimensional map using the fuel injection amount F and the engine speed Ne (step B4).

The αcc is the accelerator opening change amount ΔAc
The acceleration is determined on the basis of c. For example, the engine speed Ne increases as the engine speed Ne increases (acceleration determination is difficult), and the fuel injection amount F increases as the engine injection speed F decreases (acceleration determination is easily performed). The optimal values for the changes in the fuel injection amount F and the engine speed Ne are experimentally determined and electronically stored in the memory. Since the exhaust gas recirculation amount is originally large during low-load operation, when the accelerator change (fuel injection amount increase change) is large, the larger the fuel injection amount is, the more quickly the control can be shifted to the reduction control of the exhaust gas recirculation amount. The above αcc is reduced.

When the acceleration coefficient α = ΔAcc / αcc is larger than 1, it is determined that the engine is in an accelerating operation state. When the acceleration is determined, the EG during the transition is determined based on the acceleration coefficient α and the target air-fuel ratio TA / F separately obtained.
The R valve operation amount KTegr is read from the map (step B
5-B7). This is because when the change in the direction of enlargement of the accelerator opening is abrupt (when the accelerator pedal is suddenly depressed), the exhaust request is given priority over the reduction of NOx due to the exhaust gas recirculation. This is to reduce the amount of reflux quickly. Therefore, the map of the EGR valve operation amount KTegr is created by experimentally obtaining the operation amount so that the opening degree of the EGR valve decreases as the acceleration coefficient α increases, and is electronically stored in the memory. Is what it is.

When the acceleration is determined based on the accelerator opening, the operation amount of the EGR valve is determined on the basis of the determination. This is for performing a fuel injection control that checks based on the injection amount and matches the acceleration demand.

That is, a change rate ΔF = F / F ′ is obtained based on the current value F and the previous value F ′ of the fuel injection amount, and a two-dimensional map is obtained using the fuel injection amount F and the engine speed Ne. From the acceleration determination criterion Fk (step B
8, B9). This Fk is set similarly to the above αcc and is electronically stored in the memory. When the injection amount change coefficient β = ΔF / Fk is greater than 1, fuel injection control during acceleration is performed, and when it is small, steady-state exhaust gas recirculation control is performed (steps B10 and B11).

<Exhaust gas recirculation control at steady state>
The engine speed Ne and accelerator opening Acc are shown in
From the two-dimensional map 31 of FIG.
The target air-fuel ratio TA / F is read from the two-dimensional map 33 using the Ttrq and Ne, and the target intake air amount TQ = TA / F × F is obtained (steps C1 to C).
3). Then, an intake air amount deviation Qerr = TQ-Qav is obtained, and a basic EGR valve operation amount Tegr is obtained by IPD control so that the deviation Qerr becomes zero (step C).
4, C5).

The air-fuel ratio at which both the reduction of NOx and the reduction of smoke can be achieved slightly varies depending on the engine speed Ne and the engine torque Ttrq (in other words, the fuel injection amount F). And the case where it is not performed is relatively different. In other words, when supercharging is performed, the mixing of air and fuel in the combustion chamber is improved, and the unburned fuel is reduced (smoke is reduced). In the state (low engine speed), the former can set the target air-fuel ratio to a richer side, which advantageously works to reduce NOx. In an engine without a turbocharger, a fixed target air-fuel ratio may be determined regardless of the engine speed Ne and the magnitude of the engine load.

Therefore, the condition for steady determination that the state in which the absolute value of the accelerator opening change amount ΔAcc is smaller than the predetermined threshold value Thacc continues for a predetermined number of n cycles and fuel injection is being performed is checked ( Step C6).
This is because the control of this flow aims at improving the emission during the idle operation and during the steady operation thereafter. At the time of deceleration (F = 0), exhaust gas recirculation is not performed, so that the opening of the EGR valve is zero.

When the steady operation is confirmed, the EGR valve correction operation amount ΔTegr (i) for each cylinder is obtained from the previously obtained intake air amount characteristic ΔQt ′ (i) and the EGR correction gain E (i) ( Step C7). That is, ΔTegr (i) = ΔQt ′ (i)
× E (i) + ΔTegr ′ (i). ΔTegr ′ (i) is the previous value of the EGR valve correction operation amount of the cylinder i. In this integration, the value of ΔQt '(i) itself is emphasized, but E
This is to make the GR valve correction operation amount further reach an appropriate correction amount according to the individual difference between the cylinders.

When the EGR valve correction operation amounts of all four cylinders are obtained, the average value ΔE of the four cylinder EGR valve correction operation amounts is obtained.
Tegr-av is required. This average value should originally be zero, but by performing the processing in step C7,
The average value may be negative or positive due to various factors. This impairs the original purpose of correcting and controlling the EGR valve operation amount of each cylinder based on the basic EGR valve operation amount Tegr. Therefore, when the average value is negative, the absolute value is added to the ΔTegr (i) of each cylinder, and when the positive value is obtained, the absolute value is subtracted in the opposite direction, so that the process of always setting the average value to zero is performed every time (step). C8,
C9). The ΔEGr (i) thus obtained is added to the basic EGR valve operation amount Tegr to obtain an EGR valve operation amount Tegr (i) for each cylinder (step C10).

<Exhaust gas recirculation control at the time of acceleration determination based on acceleration coefficient α>-Single EGR valve-When acceleration is determined at step B6 in FIG.
The target EGR valve operation amount KTegr during the transition obtained in step B7 varies depending on the magnitudes of the acceleration coefficient α and TA / F. When the acceleration coefficient α is large, the opening of the EGR valve 14 becomes zero. Therefore, in this case, since the exhaust gas recirculation is not performed, the intake air amount of each cylinder increases, and even if the fuel injection amount increases, the engine output can be increased without increasing the smoke amount.

In this case, however, a control for giving a preset, which will be described later, to the EGR valve 14 is performed so that the control can be immediately shifted to the subsequent exhaust gas recirculation control.

-Preset control of EGR valve- During the exhaust gas recirculation control, the valve main body 14c allows the spring 14 to operate even when the EGR passage 13 is closed.
By applying a predetermined EGR valve drive negative pressure (preset negative pressure) to the negative pressure chamber so that the force pressed against the valve seat by d becomes small, and thus the pressing force becomes zero,
The pressing force in the closing direction by the spring 14d is balanced with the negative pressure for driving the EGR valve. That is, FIG.
As shown in (1), the preset negative pressure is the EGR valve driving negative pressure at the time when the EGR valve is controlled to close and the EGR valve lift reaches zero. A specific control flow for applying the preset negative pressure to the EGR valve 14 is shown in FIG.

That is, when the EGR valve operation amount Tegr is equal to EG
When the R valve lift amount is an operation amount that becomes zero, the value EGRVlift of the lift sensor 19 is read (steps D1, D
2). If this EGRVlift is larger than the EGR valve lift amount zero EGRV0, the EGR valve drive control is performed until the EGRVlift reaches EGRV0 (steps D3 and D4). That is, the negative pressure for driving the EGR valve is reduced until it reaches the preset negative pressure EGRV0.
When the EGR valve operation amount Tegr is an operation amount for which the preset value does not become zero due to exhaust gas recirculation, normal EGR valve drive control is performed (steps D1 to D4).

In the above embodiment, an EG with a lift sensor is used.
In the case of the R valve, in steps D2 and D3, instead of the lift sensor 19, an EGR valve driving negative pressure is detected to determine the preset state, or as shown in FIG. When there is a certain relationship, the drive amount may be detected to determine the preset state. here,
The drive amount may be either the drive negative pressure itself or the duty value of the negative pressure control solenoid valve for generating the drive negative pressure.

Therefore, when the engine shifts from the steady operation state to the acceleration operation state, the preset negative pressure acts on the EGR valve 14 even if the exhaust gas recirculation amount is set to zero in order to enhance the acceleration response. Thereafter, when the exhaust gas recirculation is restarted, the EGR valve 14 is quickly opened with almost no response delay according to the increase of Tegr, and
The opening will correspond to the size of r. Therefore, NO
This is advantageous for reducing x.

In the case where a plurality of EGR valves are provided in parallel, the example shown in FIG. 17 is configured such that the EGR passage 13 is branched in the middle and then joined again, and the branch EGR passage 13A,
13B is provided with EGR valves 14A and 14B. One branch EGR passage 13A has a small passage area, and an EGR valve 14A provided therein is a linear variable valve whose opening continuously changes according to the amount of energization.
The other branch EGR passage 13B has a passage area of the branch EG.
The EG provided therein is larger than the R passage 13A.
The R valve 14B is an on / off valve whose valve body changes to two positions, open and closed, depending on the on / off of energization.

FIG. 18 shows a control flow during transition when the on / off valve 14B is provided in addition to the linear variable valve 14A. That is, the current EGR valve 1
While reading the negative pressure value Pegr of 4A from the output of the negative pressure sensor 18, the target EGR valve drive negative pressure TPegr is stored in a table (Step B7 in FIG. 13) using the target EGR valve operation amount KTegr (step B7 in FIG. 13). It is read from a table that indicates the correspondence between KTegr and TPegr that is electronically stored in the memory in advance (steps E1 and E2). Then, the differential pressure between them (Pegr-TPegr) is equal to a predetermined value THopen.
If it is larger, the on / off valve 14B is closed, and if not, it is opened (steps E3 to E5).

Therefore, when the operating state of the engine shifts from the steady state to the accelerated state, as described above, the target exhaust gas recirculation amount is switched from large to small in order to increase the acceleration response (at steady state). Tegr → Transient KTegr), P
egr-TPegr> THopen, at which time the on / off valve 14B is immediately closed. Therefore, the amount of exhaust gas recirculation is rapidly reduced, and the amount of intake air can be rapidly increased to improve acceleration responsiveness.
At this time, the linear variable valve 14A
Since the opening is controlled based on Tegr, it is possible to prevent the NOx amount from excessively increasing.

On the other hand, during steady operation of the engine without the acceleration judgment, the on / off valve 14B is open, and the exhaust gas recirculation amount is controlled by adjusting the opening of the EGR passage 13 by the linear variable valve 14A. Therefore, the cross-sectional area of the EGR passage for recirculating a large amount of exhaust gas during normal operation of the engine is secured.

In the case where an EGR valve driving negative pressure path is provided in parallel In this example, as shown in FIG. 19, the negative pressure path 15 of the EGR valve 14 branches into a path 15a and a path 15b, and
5a is connected to a solenoid valve (linear on-off valve) 17,
An on / off valve 61 is provided at 5b. The passage diameter of the passage 15b is larger than that of the passage 15a.

That is, when the engine is in an accelerating state, the EGR valve 1 is used to quickly increase the intake air amount.
4 is required to be closed immediately.
The negative pressure chamber of the R valve 14 may be set to the atmospheric pressure. However, since the solenoid valve 17 is provided with a throttle in the passage for releasing air to obtain operation stability, it takes time to change from a state of a large negative pressure to a state of an atmospheric pressure. Therefore, in this example, the EGR valve 14 can be quickly brought to the atmospheric pressure by the on / off valve 61.

FIG. 20 shows a control flow of this embodiment. While the current negative pressure value Pegr of the EGR valve 14A is read from the output of the negative pressure sensor 18, the target EGR during the transition is read.
Using the valve operation amount KTegr (step B7 in FIG. 13),
The target EGR valve drive negative pressure T Pegr is stored in a table (K Tegr and T Pegr previously stored electronically in a memory).
Is read from the table (corresponding to the relationship between the two) (steps M1 and M2). Then, the differential pressure between the two (TPegr−Peg
When r) is larger than the predetermined value THVopen, the on / off valve 61 is opened, otherwise, it is closed (steps M3 to M5).

Therefore, when there is a request to close the EGR valve 14 and the negative pressure difference is large, the on / off valve 61 is opened, so that the EGR valve 14 is connected to the negative pressure chamber through the passage 15b at atmospheric pressure or overpressure. The valve closes promptly when air is supplied. Passage 1 is required for this quick valve closing.
The large diameter of 5b also contributes.

-A case where a plurality of EGR valves are provided in series- As shown in FIG. 21, two EGR valves 14A and 14B are provided in an EGR passage 13 in series, and one EGR valve 14A and 14B is provided.
The GR valve 14A is a linear variable valve whose opening continuously changes in accordance with the amount of energization, and the other EGR valve 14B is an on / off valve whose valve body changes to two positions of opening and closing by energizing ON / OFF. is there.

FIG. 22 shows a flow of the exhaust gas recirculation control at the time of the acceleration judgment. That is, while reading the deviation Qerr of the intake air amount, the deviation threshold THQerr is read from the map using the fuel injection amount F and the engine speed Ne (steps F1 and F2). The threshold value THQerr is related to the magnitude of the degree of acceleration demand. As the fuel injection amount F increases and the engine speed Ne decreases, THQerr increases.
err is set to be small and stored electronically in memory. When the deviation Qerr is smaller than the deviation threshold THQerr, the on / off valve 14B remains open, but the deviation Qerr becomes smaller than the deviation threshold THQerr.
When it exceeds Qerr, the on / off valve 14B is controlled to be closed (steps F3 to F5).

Therefore, when the acceleration demand is high, the EGR
Since the passage 13 is immediately closed by the on / off valve 14b, it is possible to increase the intake air amount quickly to suppress the smoke and to accelerate the fuel by increasing the amount of fuel.

<Control at Acceleration Judgment Based on Injection Amount Change Coefficient β> -Exhaust Gas Recirculation Control / Fuel Injection Amount Control- This is shown in FIG. 23. Injection amount change coefficient β, fuel injection amount F
Using the engine speed Ne and the three-dimensional map in which the optimum transient target air-fuel ratio KTA / F in these changes is recorded, KTA / F is read (step G1). The transient target air-fuel ratio KTA / F is leaner than the steady-state target air-fuel ratio TA / F so as to reduce the amount of exhaust gas recirculation and thereby quickly increase the engine output while suppressing the generation of smoke. Side is set. The KTA / F is set to be leaner in accordance with the fuel injection amount F as the load becomes lower, the injection amount change coefficient β becomes larger, and the engine speed Ne becomes lower. The optimum target air-fuel ratio for each change is experimentally determined and electronically stored in the memory.

Based on the obtained transient target air-fuel ratio KTA / F and fuel injection amount F, a transient target intake air amount TQ is calculated (step G2). Then, based on this TQ, the EGR valve operation amount is determined in the same manner as in the above-mentioned steady operation, and a rapid reduction control of the exhaust gas recirculation amount is performed.

As a result, the exhaust energy applied to the supercharger 7 is further increased, so that the intake air amount is rapidly increased, and a delay in acceleration response to depression of the accelerator pedal, that is, a so-called turbo lag is prevented.

On the other hand, when the amount of fuel is increased by depressing the accelerator pedal, the air-fuel ratio becomes rich, which is disadvantageous in reducing smoke. Therefore, a certain limit is imposed on the increase in fuel in order to temporarily suppress the increase. That is, the limit air-fuel ratio LimitA / F is read from the map of the fuel injection amount F and the engine speed Ne (step G).
3). This limit air-fuel ratio LimitA / F is for suppressing the generation of smoke, and the limit smoke amount is set to be larger than the limit smoke amount in a steady state. For example, 2BU
The amount of smoke is set to about this level, and this level does not hinder the increase in the output torque of the engine.

The relationship between the target air-fuel ratio TA / F in the steady state, the target air-fuel ratio KTA / F in the transient state, and the limit air-fuel ratio LimitA / F is as shown in FIG. 24. Target air-fuel ratio KTA / F during transition to the lean side from air-fuel ratio TA / F
Is set, and the limit air-fuel ratio LimitA / F is set to be richer than the target air-fuel ratio TA / F in the steady state. The limit air-fuel ratio LimitA / F can be basically set to the lean side as the fuel injection amount increases, and to the rich side as the engine speed increases. The optimum value experimentally obtained for the change is electronically recorded on a memory.

Based on the obtained limit air-fuel ratio LimitA / F and the current intake air amount Q (i), the fuel injection amount limit FLimi
t is calculated, and the smallest one of the basic injection amount F, the limit FLimit, and the maximum injection amount Fmax is determined as the target injection amount TF.
(Steps G4 and G5). The basic injection amount F is a fuel injection amount uniquely determined by the engine speed Ne and the accelerator opening Acc, and the maximum injection amount Fmax
Is the upper limit of the fuel injection amount that does not cause the engine to be destroyed.

Therefore, even if the exhaust gas recirculation amount is reduced during the transition, an excessive increase in the fuel injection amount is suppressed, so that the acceleration demand can be satisfied while suppressing an excessive increase in the smoke amount.

-Parallel control of exhaust gas recirculation amount and fuel injection amount- The previous control was to limit the fuel injection amount while performing exhaust gas recirculation control. The control is performed based on the fuel ratio, and is shown in FIG.

That is, when the acceleration state is determined in the previous transient determination, the optimal target for transient exhaust gas recirculation control in the transition is determined by using the injection amount change coefficient β, the fuel injection amount F, and the engine speed Ne. The KTA / F is read by referring to the three-dimensional map in which the air-fuel ratio KTA / F is recorded (step P
1). The target air-fuel ratio KTA / F for exhaust gas recirculation during transition is
As in the previous case, the air-fuel ratio is set leaner than the steady-state target air-fuel ratio TA / F.

When the fuel is increased by depressing the accelerator pedal, the air-fuel ratio becomes rich, which is disadvantageous in reducing smoke. Therefore, the target air-fuel ratio KFTA / for fuel injection amount control can be increased without increasing the smoke amount excessively.
F is read from a map of the fuel injection amount F and the engine speed Ne (step P2). The KFTA / F is an electronically stored optimum value experimentally obtained for the change in the fuel injection amount F and the engine speed Ne. That is, the KFTA / F for controlling the fuel injection amount is:
Similar to the critical air-fuel ratio LIMITA / F, the target air-fuel ratio T at steady state
The limit smoke amount is set to a richer side than the A / F, and the limit smoke amount on which the setting is based is larger than the limit smoke amount in a steady state, for example, about 2 BU.

As for the exhaust gas recirculation control, a transient target intake air amount TQ is calculated based on the obtained target air-fuel ratio KTA / F and the fuel injection amount F (step P3). Then, based on the TQ, the EGR valve operation amount KTegr in the transient state is determined in the same manner as in the previous steady operation, and a rapid reduction control of the exhaust gas recirculation amount is performed (step P4). As a result, the exhaust energy given to the turbocharger 7 further increases, so that the intake air amount increases quickly,
A delay in acceleration response to depression of the accelerator pedal, that is, a so-called turbo lag is prevented.

With respect to the fuel injection amount control, the transient fuel injection amount KF is calculated based on the obtained target air-fuel ratio KFTA / F and the current intake air amount Q (i) (step P).
5). Then, the smallest value of the transient fuel injection amount KF and the maximum injection amount Fmax is set as the target injection amount TF (step P6).

Therefore, during the transition (during acceleration), more fuel can be injected more aggressively than in the steady state within a range in which the smoke amount does not increase angularly, thereby increasing the engine output torque and The exhaust energy given to the supercharger 7 increases, which is advantageous for improving the acceleration.

-VGT Control- Next, the control of the VGT supercharger 7 performed at the time of the acceleration determination will be described. That is, when the acceleration is determined by the injection amount change coefficient β, the target turbo efficiency VGTsol is obtained from the map 51 using the target torque Trqsol and the engine speed Ne.
Read. Then, based on the obtained VGTsol, VGT
The rotation position of the flap 7b of the supercharger 7, that is, A / R
Adjust

Therefore, at the time of acceleration, even if the exhaust energy given to the turbocharger 7 is reduced due to the large amount of exhaust gas recirculation up to that time, the supercharging efficiency is increased by reducing the A / R, and the intake air is increased. The desired acceleration performance can be obtained by increasing the amount.

-Fuel injection timing advance control-Next, fuel injection timing advance control performed during acceleration will be described. That is, in this engine, the fuel injection timing in the steady state is set at a position that is considerably retarded from the MBT, and is controlled so as to gradually advance as the fuel injection amount increases. On the other hand, when the acceleration is determined based on the injection amount change coefficient β, the injection timing is further advanced from the corresponding injection timing in the steady state according to the magnitude of β.

The advance of the injection timing has an effect that the ignition is delayed, so that the mixing of the fuel and the air becomes good and the rapid combustion occurs. Thus, while NOx increases, smoke decreases. However, the air-fuel ratio is originally rich due to a large amount of exhaust gas recirculation, and even if the advance angle during acceleration is increased as described above, NOx does not excessively increase, and smoke is reduced by this advance angle control. The advantageous effect described above can be obtained.

-Relationship between control at the time of acceleration judgment by α and control at the time of acceleration judgment by β- The above-described EGR valve control at the time of acceleration judgment by acceleration coefficient α should be performed at the time of acceleration judgment by injection amount change coefficient β. The VGT control and the injection timing advance control at the time of acceleration determination based on the injection amount change coefficient β can also be performed at the time of acceleration determination based on the acceleration coefficient α.

[0097] <distinguish between air flow sensor 6 and the O ₂ sensor 9> - characteristics of the sensor - air flow sensor 6, the characteristics as shown in FIG. 26, but the detection error ΔQ according flow rate Q is increased becomes larger, The increase of ΔQ is smaller than the increase of flow rate Q. Therefore, as shown in FIG. 27, the flow rate error rate ΔQ /
Q is large in the low flow region but small in the high flow region. On the other hand, in the case of the pump current generation type linear O ₂ sensor 9, the detection error rate E increases as the air-fuel ratio increases as shown in FIG. Further, the intake air amount obtained by the O ₂ sensor 9 is not the one for the cylinder currently in the intake stroke, but the one for the cylinder whose intake stroke timing is several cylinders before.

Therefore, the two sensors 6 and 9 are switched so that the respective advantages can be fully utilized and used for measuring the intake air amount of each cylinder. That is, the use of the two sensors 6 and 9 is switched based on the operating region of the engine and the result of the accuracy comparison between the two sensors.

-Sensor Switching Flow- This flow is shown in FIG. When the transient (acceleration state) is determined by the transient determination described above, the intake air amount determined by the air flow sensor 6 is selected, and the EGR valve drive amount switching unit 49 in FIG. Then, transient exhaust gas recirculation control (A / F control) is performed (steps H1 to H5).
In the case of using the O ₂ sensor, a response delay occurs because the amount of intake air several cylinders before is detected. However, in the case of using the output of the air flow sensor 6, there is no such delay. Can be quickly reduced to increase the acceleration response.

If the engine is in a steady operating state, a determination is made to select one of the air flow sensor 6 and the O ₂ sensor 9 (step H6, this point will be described later). When the O ₂ sensor 9 is selected, the O ₂ concentration detected by the sensor 9 is used to electronically store the data on a memory representing the relationship between the O ₂ concentration and the air-fuel ratio as shown in FIG. The air-fuel ratio A / F is determined with reference to the map, and the intake air amount at that time is calculated based on the A / F and the corresponding fuel injection amount several cylinders earlier, and is used for control. should do the intake air amount is switched to the intake air amount by the O ₂ sensor 9 (step H7~H10). Then, the EGR valve drive amount switching unit 49 in FIG. 7 is switched to the EGR valve control by the O ₂ sensor 9, and the steady-state exhaust gas recirculation control (A / F control) is performed (step H1).
1, H12).

On the other hand, when the air flow sensor 6 is selected, the intake air amount to be used for control is switched to the intake air amount by the sensor 6, and the EGR
Valve control is switched to exhaust gas recirculation control (A / F
(Step H7 → H13 → H14 →
H12).

Therefore, in the steady state, the more accurate one of the air flow sensor 6 and O ₂ is used, which is advantageous for the intended control, and even when the O ₂ sensor 9 is selected. There is no problem because it is during steady operation.

-Sensor Selection Flow- This flow is shown in FIG. The operating state of the engine (engine speed, accelerator opening, etc.) is read, and the air flow sensor 6 is selected when the intake air flow rate is in a large operating range with reference to a map electronically stored in the memory. (Steps J1 to J3). As is apparent from FIG. 27, when the intake air flow rate is large, the detection error rate of the air flow sensor 6 is small, so that the intake air flow rate can be accurately detected, and control is performed in real time based on the detection result. Can be performed.

FIG. 32 shows a map used for determining the operation region. This indicates a region (shaded area) where the intake air flow rate is small in the change of the engine speed and the engine load, and is set experimentally. Basically, the air flow sensor 6 is selected in a region where the engine speed is high, and the O ₂ sensor 9 is selected in a region where the engine speed is low.

If the intake air flow rate is small and the engine is in the operating range, a detection error AFSerror in the intake air flow rate is read out based on the output of the air flow sensor 6 with reference to a map corresponding to FIG. 27 (step J4). Further, it reads the O ₂ concentration based on the output of the O ₂ sensor 9, the sensor 9
Reading the detection error O ₂ error by referring to a map corresponding to FIG. 28 at, but this O ₂ concentration when the predetermined amount or more to select the air flow sensor 6 for measuring the intake air amount (step J5, J6) .

When the O ₂ concentration is equal to or higher than the predetermined amount, the operating range is such that the air-fuel ratio exceeds λ = 1 and becomes lean to a predetermined level or higher (for example, when A / F ≧ 40).
At this time, since the detection error of the O ₂ sensor 9 becomes large as is clear from FIG. 13, the air flow sensor 6 is selected for measuring the intake air amount.

On the other hand, when the O ₂ concentration is less than the predetermined amount, the detection errors of the two sensors 6 and 9 are compared, and when the detection error of the air flow sensor 6 is smaller, the sensor 6 determines the amount of intake air. select the measurement, when towards the detection error of the O ₂ sensor 9 is small selects the sensor 9 for measuring the intake air amount (step J7~J9). Therefore, in the low flow rate region, the O ₂ sensor 9 with _high accuracy is used for measuring the intake air amount. Even in this case, when the air-fuel ratio is lean above a predetermined level, and the detection error of the sensor 9 increases. When it is large, optimization is achieved by using the airflow sensor 6.

However, instead of the above switching method, the air flow sensor 6 is used in an idling operation region or a low load operation region where the air-fuel ratio is lean such that A / F ≧ 40, as shown in FIG. In particular, the O ₂ sensor 9 may be used when the air-fuel ratio shifts to the rich side by increasing the exhaust gas recirculation amount in order to reduce NOx in the idle operation region.

<EGR Cylinder Variation Elimination Control> This control is to reduce the EGR rate variation between cylinders.

Regarding the variation in the EGR rate between the cylinders The recirculation of the exhaust gas is caused by the difference between the pressure in the intake passage 2 and the pressure in the exhaust passage 3. When only one EGR passage is provided as shown in FIG. 1, the in-pipe pressure at the EGR passage connection position of the intake passage 2 and the in-pipe pressure at the EGR passage connection position of the exhaust passage 3 are, for example, due to changes in the crank angle. It changes as shown in FIG. This pressure pulsation
In the case of an engine speed of 2000 rpm, the pressure in the pipe on the intake passage side is represented by a symbol “In”, and the pressure in the pipe on the exhaust passage side is represented by “Ex”. The change in the differential pressure between “In” and “Ex” is as shown in FIG. 35, and due to this differential pressure, the exhaust gas (burned gas) intermittently flows into the intake passage as shown by the broken line in FIG. I do. The solid line in FIG. 36 represents a change in the flow rate of the intake air (including the recirculated exhaust gas) flowing through the intake passage 2.

Engine speed 1500 rpm and 100
The changes of “In” and “Ex” at each 0 rpm are as shown in FIGS. 37 and 38, respectively. FIG.
As apparent from the comparison with the above, the manner of change of “In” and “Ex” differs depending on the engine speed. FIG. 39 shows the change in the differential pressure at each engine speed collectively, as shown in FIG. 39.
The peak height is different, and the peak height is reversed depending on the crank angle. For example, around 180 degrees, 1500 rpm is high and 1000 ppm is low,
In the vicinity of the temperature, 1000 ppm is high and 2000 ppm is low.

Therefore, a change in the EGR rate of each cylinder due to the engine speed is shown by a broken line in FIG. That is, a phenomenon occurs in which the EGR rates for the cylinders # 1 to # 4 are reversed depending on the engine speed. This is the cylinder-to-cylinder variation of the EGR rate in question here.

Eliminating Inter-Cylinder Variation by Correcting EGR Operation Amount Based on the relationship between the engine speed and the EGR rate of each cylinder, an EGR for reducing the cylinder-to-cylinder difference in the EGR rate between cylinders is considered. A correction coefficient for the valve operation amount Tegr (i) is experimentally determined according to the engine speed and electronically stored in a memory. Then, the correction coefficient is calculated from the memory using the engine speed Ne, and the EGR valve operation amount Tegr (i) is corrected based on the correction coefficient to obtain the EGR valve 1.
4 is used for control.

Thus, the air-fuel ratio of all cylinders is set to the target air-fuel ratio TA /
It can be adjusted to F 2, which is advantageous for reducing both NOx and smoke.

-Elimination of Inter-Cylinder Variation by Employing Alternative EGR Pipes-As shown in FIG. 41, the intake passage 2 and the exhaust passage 3
The GR pipe 13A (EGR1) and the EGR pipe 13B (EGR2) are connected by an EGR valve 14A (EGR1) and 14B (EGR2), respectively. The EGR pipes 13A (EGR1) and 13B (E
GR2) are independent from each other and connected to the intake passage 2 in the flow direction of the intake air, and connected to the exhaust passage 3 in the same manner. The EGR valves 14A (EGR1) and 14B (EGR2) are provided with separate opening adjusting means, respectively, and are configured to be controllable independently of each other.

The inter-cylinder variation of the EGR rate is caused by the difference between the exhaust pulsation and the intake pulsation depending on the engine speed. Such a pulsation is caused by the place where the EGR passage is connected to the intake passage 2 or the exhaust passage 3. Depending on the mode. Here, the two independent EGR pipes 13 are used as described above in order to eliminate the cylinder-to-cylinder variation using this fact.
A (EGR1) and 13B (EGR2) are provided, and are connected to positions in the intake passage 2 where the manner of intake pulsation differs, and to positions where the manner of exhaust pulsation in the exhaust passage 3 differs. That is, according to the engine speed, the EGR pipe 1
If 3A (EGR1) and 13B (EGR2) are selected and used (including a combination of both), the inter-cylinder variation can be eliminated.

-Specific Control Contents- As shown in FIG. 42, the operating range of the engine is set to a lower range N1 based on the predetermined engine speed Ne _- EXC.
And the high region N2, and the EGR tube 1 in the low region N1.
3A (EGR1), the EGR pipe 13B (E
GR2). Further, hysteresis THN1 and THN2 are provided for the switching based on the engine speed.

That is, as shown in FIG. 43, at the time of steady operation of the engine, the engine speed Ne is initially read with the current region as N1 (steps K1 and K2). It is determined whether the current region is any one. Since the current region is N1, the engine speed Ne-HYS1 for region switching is set to the upper limit of the hysteresis (steps K3 and K4). And
When the engine speed Ne is higher than the upper limit of the hysteresis, the region is set to N2, the EGR pipe 13A (EGR1) is used for main control, and the EGR pipe 13B (EGR2) is used for sub-control (steps K5 to K7). . The meaning of the main control and the slave control will be described later. When the engine speed Ne is equal to or lower than the upper limit of the hysteresis, the region remains N1 and the EGR pipe 13B (EGR
2) is subjected to main control, and the EGR pipe 13A (EGR1) is subjected to sub-control (step K8).

On the other hand, the area judgment in step K3 is N2
If so, the engine speed Ne-HYS2 for the region switching is set to the lower limit of the hysteresis (step K3 → K9). When the engine speed Ne is lower than the lower limit of the hysteresis, the region is set to N1, and the EGR pipe 13B (EGR
2) is subjected to main control, and the EGR pipe 13A (EGR1) is subjected to sub-control (steps K10 to K12). The engine speed
When Ne is equal to or greater than the lower limit of the hysteresis, the EGR pipe 13A (EGR1) is subjected to main control while the area remains N2, and the EGR pipe 13B (EGR2) is subjected to sub-control (step K10 → K1).
3).

Here, the main control is a feedback control of the EGR valve based on the intake air amount so that the target air-fuel ratio TA / F described above (an air-fuel ratio capable of achieving both reduction of NOx and reduction of smoke) is obtained. Means to do. For example, when the control shifts to the main control when the EGR pipe is closed, the opening of the EGR valve increases and converges to the target opening as shown in FIG. On the other hand, the sub-control means that the EGR valve is open-controlled so that its opening degree changes to a fully closed state at a predetermined change rate.

Therefore, the EG based on the engine speed is determined.
By switching the R pipes 13A (EGR1) and 13B (EGR2), the pulsation characteristics affecting the exhaust gas recirculation change, and as shown by the solid line in FIG. 40, the magnitude of the EGR rate of each of the cylinders # 1 to # 4 depends on the engine speed. Thus, reverse rotation can be avoided, and the difference in EGR rate between the cylinders can be made substantially uniform at all engine speeds.

In this embodiment, the case where the area is divided into two areas N1 and N2 has been described. However, the present invention is not limited to the two areas, and the same applies to the case where the area is divided into a plurality of areas N1, N2,.

Therefore, regarding such a difference in the EGR rate between the cylinders, this is treated as an individual difference between the cylinders, and weighting according to the individual difference is performed in the exhaust gas recirculation control (control of the EGR rate) for each cylinder. By giving each control amount,
It is possible to avoid a difference between the EGR rate and the air-fuel ratio of each cylinder. That is, it is possible to solve the problem that the amount of generated NOx and smoke varies among the cylinders and the amount of generated NOx and smoke in the entire engine increases.

The result shown by the solid line in FIG. 40 indicates that the EGR pipe 13A is divided into regions at an engine speed of 1750 rpm.
(EGR1) and 13B (EGR2) are used by switching.
Also, in the figure, the scale of the EGR rate is considerably enlarged in order to make the difference between the EGR rates of the cylinders clear.

If the length or volume of the EGR pipe changes, the relationship (phase relationship) between the intake pulsation and the exhaust pulsation that affects the EGR rate changes. The above-described inter-cylinder variation can be eliminated by the same control as in the above case.

FIG. 45 shows a passage configuration in the case where the length of the EGR pipe is changed according to the change in the engine speed. That is, one E having both ends connected to the intake passage and the exhaust passage.
The GR tube 13 is divided into an EGR tube 13A having a short length and an EGR tube 13B having a long length on the way, and each of the GR tubes 13 is provided with on-off valves 13a and 13b. When one of the two is opened, the other is closed so that the EGR pipe 13A is closed.
And the EGR pipe 13B are selectively used to eliminate inter-cylinder variations. The EGR pipe 13A and the EGR pipe 1
The switching between 3B and 3B may be performed by arranging one valve at the junction of both pipes to switch the communication direction, or by providing a valve in each pipe and operating each of them individually. Good.

FIG. 46 shows two branched EGR tubes 13A,
13B, an on-off valve 13a is provided in only one of them, and by opening and closing the on-off valve 13a, a short EGR pipe 13a is provided.
The case where only A is used and the case where both EGR pipes 13A and 13B are used are switched to eliminate inter-cylinder variation.

As means for switching the passage length of the EGR tube, a bellows or other expandable portion is provided in the middle of the tube.
The length of the elastic portion may be changed. Furthermore, when changing the volume of the EGR pipe,
A chamber may be connected in the middle of the EGR pipe, a piston for changing the volume may be arranged in the chamber, and the piston may be driven.

<Switching the Number of EGR Pipes Used> The required amount of exhaust gas recirculation differs depending on the operating state of the engine and other factors. Therefore, as shown in FIG.
A (EGR1) and 13B (EGR2) are provided, and the number of used engines is switched according to the operating state of the engine.

FIG. 47 shows the details of the switching control of the number of EGR pipes used. The EGR combination determination is based on whether the EGR pipes 13A (EGR1) and 13B (EGR2) are used together according to the operating state of the engine,
It is determined whether one of them is used (step L1). That is, using the fuel injection amount F, when it is decreasing, the two EGR pipes 13A (EGR1), 13B
(EGR2), the two EGR pipes 13A (EGR1), 13A during normal operation and during acceleration operation in which the fuel injection amount F has not decreased.
Select and use any one of B (EGR2). This alternative use is performed by switching control in the operating regions N1 and N2 for eliminating the above-described cylinder-to-cylinder variation in the EGR rate.

If it is determined that the two EGR tubes 13A (EGR1) and 13B (EGR2) are used together, a flag indicating that the EGR tube 13A (EGR2) is used is set. The EGR valve 14A (EGR1) of the pipe 13A (EGR1) is set fully open, and the EGR valve 1 of the EGR pipe 13B (EGR2) is opened.
4B (EGR2) is feedback-controlled based on the intake air amount (steps L2 to L6). If the currently closed EGR pipe is the EGR pipe 13B (EGR2), the EGR pipe 13B (EGR
GR2), the EGR valve 14B (EGR2) is fully opened, and the EG
The EGR valve 14A (EGR1) of the R pipe 13A (EGR1) is feedback-controlled for each cylinder based on the detected intake air amount so that the required opening degree (EGR amount at which the target air-fuel ratio TA / F is obtained) (step). L4 → L7, L8).

If it is determined that the operating state is not the state in which both the EGR pipes 13A (EGR1) and 13B (EGR2) are used together, the EGR valve 14 of the EGR pipe 13A (EGR1) is currently being used together.
When the A (EGR1) is feedback-controlled based on the intake air amount, the other EGR valve 14B which is fully open
Set the target opening of (EGR2) to fully closed (step L9-
L11). When the EGR valve 14B (EGR2) is fully closed, the in-use flag is reset (step L
12, L13). Currently used in combination with EGR tube 13B (EGR2)
When the EGR valve 14B (EGR2) is feedback-controlled based on the intake air amount, the target opening of the EGR valve 14A (EGR1) is set to fully closed, and the EGR valve 14A (EGR1) is set.
Are reset, the combined use flag is reset (steps L14 and L15).

Therefore, when the required amount of exhaust gas recirculation (EGR) increases rapidly, by using both the EGR passages 13A and 13B together, the actual EGR passage can be met to meet the requirement.
The amount can be increased rapidly.

That is, as shown in FIG.
That is, when the depression of the accelerator pedal is returned and the fuel injection amount decreases, the required amount of EGR rapidly increases.
This is to prevent the air-fuel ratio from becoming excessively lean and increasing the NOx amount. However, as shown in the middle part of FIG.
Due to the feedback control of the R valve, the actual EGR amount cannot be rapidly increased to meet the demand. On the other hand, by performing the above-described combined control, the actual EGR amount can be rapidly increased to meet the demand as shown in the lower part of the figure, and the air-fuel ratio temporarily becomes excessively lean. A sudden increase in NOx can be avoided.

In the above example, the combined use of the EGR pipe is determined based on the operating state of the engine. However, when the catalytic converter 12 is provided in the exhaust passage 3,
This may cause clogging due to particulate components in the exhaust gas and increase the exhaust pressure, and the exhaust pressure may change, such as removal of such clogging material resulting in a decrease in the exhaust pressure. . Further, even in the case of using together with a VGT supercharger, the exhaust pressure changes depending on the position of the VGT variable vane. This change in the exhaust pressure directly affects the EGR amount due to the EGR passage. Therefore, for example, when the exhaust pressure is high, 1
The exhaust gas recirculation control can be performed by the EGR pipes, and when the exhaust pressure is low, the exhaust gas recirculation control can be performed by the plurality of EGR pipes.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of an engine.

FIG. 2 is a configuration diagram of an EGR valve and a drive system thereof.

FIG. 3 is a graph showing a relationship between a drive current and a drive negative pressure of an EGR valve.

FIG. 4 is a graph showing the relationship between the drive negative pressure of an EGR valve and the lift amount thereof.

FIG. 5 is a front view showing a part of the VGT supercharger in a small A / R state.

FIG. 6 is a front view showing a part of the VGT supercharger in an A / R large state.

FIG. 7 is a configuration diagram of an engine control system.

FIG. 8 is a graph showing a relationship between an air-fuel ratio and a NOx emission amount.

FIG. 9 is a graph showing a relationship between an air-fuel ratio and a smoke value.

FIG. 10 is a diagram showing an overall flow of control.

FIG. 11 is a graph showing a time change of an intake air flow rate of the engine.

FIG. 12 is a flowchart for calculating an intake air amount.

FIG. 13 is a flowchart of a transient determination.

FIG. 14 is a flowchart of an EGR valve operation amount calculation.

FIG. 15 is a flowchart of control for giving a preset.

FIG. 16 is a graph showing a relationship between an EGR valve lift amount and a drive amount.

FIG. 17 is another engine configuration diagram.

FIG. 18 is a flowchart of an EGR valve on / off control in the case of an EGR valve parallel arrangement.

FIG. 19 is a configuration diagram of an EGR valve having a drive negative pressure path in parallel.

FIG. 20 is a control flow chart of an EGR valve having a drive negative pressure path in parallel.

FIG. 21 is a configuration diagram showing an EGR passage in which two EGR valves are arranged in series.

FIG. 22 is a flowchart of the EGR valve on / off control in the case of an EGR valve serial arrangement.

FIG. 23 is a flow chart of transient fuel injection amount control.

FIG. 24 is a graph showing a relationship between a target air-fuel ratio in a steady state, a target air-fuel ratio in a transient state, and a limit air-fuel ratio in a transient state.

FIG. 25 is a flowchart of parallel control of exhaust gas recirculation and fuel injection during transition.

FIG. 26 is a graph showing the relationship between the output of the air flow sensor and the detected flow rate / detection error.

FIG. 27 is a graph showing the relationship between the output of the air flow sensor and the detection error rate.

FIG. 28 is a graph showing the relationship between the output of the linear O ₂ sensor and the excess air ratio λ · detection error.

FIG. 29 is a flowchart of sensor switching control.

FIG. 30 is a graph showing the relationship between the O ₂ concentration of the exhaust gas and the air-fuel ratio.

FIG. 31 is a flowchart of sensor selection control.

FIG. 32 is a map diagram for determining an engine operation area for sensor selection.

FIG. 33 is a map diagram for determining another engine operation region for sensor selection.

FIG. 34 is a graph showing pressure fluctuations of intake air and exhaust gas at an engine speed of 2000 rpm.

FIG. 35 is a graph showing fluctuations in the differential pressure between intake air and exhaust gas at an engine speed of 2000 rpm.

FIG. 36 is a graph showing changes in the amount of intake air and the amount of EGR contained therein when the engine speed is 2000 rpm.

FIG. 37 is a graph showing pressure fluctuations of intake air and exhaust gas at an engine speed of 1500 rpm.

FIG. 38 is a graph showing pressure fluctuations of intake air and exhaust gas at an engine speed of 1000 rpm.

FIG. 39 is a graph showing a change in pressure difference between intake and exhaust at a plurality of engine speeds.

FIG. 40 shows the EGR of each cylinder in the related art and the present invention.
FIG. 4 is a graph showing a change in the rate according to the engine speed.

FIG. 41 is another engine configuration diagram.

FIG. 42 is a diagram showing a division of an engine operation region based on the engine speed.

FIG. 43 is a flowchart of EGR pipe selection control.

FIG. 44 is a graph showing a change in the EGR valve opening when the EGR pipe is switched.

FIG. 45: E with valves in two passages of different length
The block diagram of a GR tube.

FIG. 46 is a view showing another example of an EGR pipe provided with two passages having different lengths.

FIG. 47 is a flowchart of EGR tube combined control.

FIG. 48 is a graph showing the change over time of the EGR amount at the time of deceleration when using two EGR tubes together and when using a single EGR tube.

FIG. 49 is a graph showing the EGR rate and intake air amount deviation of each cylinder.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 4-cylinder diesel engine main body 2 Intake passage 3 Exhaust passage 4 Fuel injection valve 5 Control unit 6 Air flow sensor 7 VGT supercharger 9 Linear O ₂ sensor 13 EGR passage 13A EGR passage (or EGR tube) 13B EGR passage (or EGR tube) 14) EGR valve 14A EGR valve 14B EGR valve 15 Negative pressure passage 16 Negative pressure pump 17 Negative pressure control solenoid valve 18 Negative pressure sensor 19 EGR valve lift sensor 23 Crank angle sensor 24 Fuel injection pump 27 Accelerator opening sensor

Continued on the front page (51) Int.Cl. ⁶ Identification code FI F02D 41/14 310 F02D 41/14 310C 45/00 366 45/00 366E G01F 1/68 G01F 1/68

Claims (10)

[Claims] An exhaust gas recirculation passage for recirculating a part of exhaust gas to an intake system, means for adjusting the amount of exhaust gas recirculated, and means for measuring the amount of intake air for each cylinder, wherein the intake air for each cylinder is provided. An exhaust gas recirculation control device for a multi-cylinder direct injection engine in which the amount of exhaust gas recirculation is controlled based on the amount of fuel. Exhaust gas recirculation control means for controlling the operation of the exhaust gas recirculation amount adjusting means for each cylinder so as to achieve a target air-fuel ratio based on apparatus. 2. The exhaust gas recirculation control device for a direct injection engine according to claim 1, wherein the target air-fuel ratio changes as the amount of smoke in the exhaust gas increases as the air-fuel ratio becomes rich. The exhaust gas recirculation control device for a direct-injection engine, wherein the air-fuel ratio is obtained when the amount of smoke changes from a gradual increase to a rapid increase. 3. The exhaust gas recirculation control device for a direct injection engine according to claim 1, wherein the target air-fuel ratio of each of the cylinders is substantially the same. 4. The exhaust gas recirculation control device for a direct injection engine according to claim 1, wherein said intake air amount measuring means includes an air flow rate or an intake pipe provided in an intake passage. An exhaust gas recirculation control device for a direct injection engine, wherein an intake air amount is obtained for each cylinder based on an output from a sensor for detecting pressure. 5. The exhaust gas recirculation control device for a direct injection engine according to claim 4, wherein the intake air amount measuring means is a constant temperature type hot film type air flow sensor provided in an intake passage. An exhaust gas recirculation control device for a direct injection engine. 6. The exhaust gas recirculation control device for a direct injection engine according to claim 5, wherein the intake air amount measuring means measures only a forward air flow amount flowing through the cylinder. An exhaust gas recirculation control device for a direct injection engine, which is a hot film airflow sensor. 7. The exhaust gas recirculation control device for a direct injection engine according to claim 1, wherein the control means is configured to determine a difference between the measured intake air amounts between the cylinders. An exhaust gas recirculation control device for a direct injection engine, wherein a parameter representing an intake air amount characteristic of each cylinder is obtained by using the parameter, and the operation of the exhaust gas recirculation amount adjusting means is controlled for each cylinder based on the parameter. 8. The exhaust gas recirculation control device for a direct injection engine according to claim 7, further comprising: means for measuring a time of an intake stroke for each cylinder, wherein the control means measures the time for each cylinder. An exhaust gas recirculation control device for a direct injection engine, wherein an intake air amount is processed so as to reduce an intake air amount difference caused by a length of an intake stroke time of each cylinder to obtain the parameter. 9. The exhaust gas recirculation control device for a direct injection engine according to claim 8, wherein the control means changes the intake air amount of the cylinder with respect to a cylinder whose intake stroke timing is immediately before. ΔQti = ΔQi / ΔT obtained by dividing the rate ΔQi by the rate of change ΔTi of the intake stroke time of the cylinder with reference to the cylinder whose intake stroke is one time earlier.
i is calculated, and ΔQt ′ (i) obtained by reflecting the previous value of the cylinder at a predetermined ratio to the current value of ΔQti is calculated as a parameter representing the intake air amount characteristic of the cylinder. An exhaust gas recirculation control device for a direct injection engine, wherein operation of the exhaust gas recirculation amount adjusting means is controlled for each cylinder based on a parameter. 10. The exhaust gas recirculation control device for a direct injection type engine according to claim 1, wherein the engine is configured to generate an exhaust gas recirculation in an intake stroke of each cylinder. Means for detecting a predetermined parameter for specifying the operating state of the engine, wherein the differential pressure varies in a manner different from each other in accordance with the operating state of the engine; An exhaust gas recirculation control device for a direct injection engine, comprising: a correction means for correcting the control amount of the exhaust gas recirculation amount adjusting means so as to reduce a variation in the recirculation amount among the cylinders.