JPH0122456B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a boost pressure control device for a turbocharger.

(Prior art) A turbocharger uses the high temperature and high pressure energy of exhaust gas to rotate an exhaust turbine at high speed and drive a compressor on the same axis.As the rotation speed of the compressor increases, the intake air The intake pressure of the manifold can be increased to above atmospheric pressure. By supercharging a large amount of air into the engine in this way, the engine generates high torque and high output, and also improves fuel efficiency.

However, in the case of automobile engines that have a wide operating speed range, due to the characteristics of the turbocharger, it is possible to secure sufficient boost pressure in the medium-to-high speed operating range, but it is difficult to obtain sufficient exhaust pressure in the low-speed operating range. Therefore, the rotational speed of the exhaust turbine does not increase, the boost pressure decreases, and the engine tends to lack low-speed torque.

The driving efficiency of an exhaust turbine is determined according to the ratio A/R of the cross-sectional area A of the turbine scroll part, which is the introduction part of the exhaust gas into the turbine, and the radius R from its center. Even in a low-speed operating range, if this A can be reduced to increase the exhaust flow velocity, the turbine rotational speed can be increased and the supercharging pressure can be increased.

Therefore, a turbocharger equipped with a variable capacity means for varying the A/R of the exhaust turbine has been filed by the present applicant as Japanese Patent Application No. 162918/1982. However, sufficient boost pressure can be obtained.

The variable capacity means installed in the exhaust gas introduction part of the exhaust turbine to increase and decrease the opening degree is driven by an actuator whose operating pressure is the supercharging pressure generated downstream of the compressor, and a solenoid valve is used to adjust this operating pressure. The solenoid valve adjusts the pressure by releasing a portion of the operating pressure to the atmosphere.For example, the operation can be controlled by a control device such as a microcomputer depending on the operating state of the engine. It's summery.

A solenoid valve is an on-off type solenoid valve that opens and closes at a predetermined frequency, and the opening time ratio is controlled according to its control duty. For example, when the duty representing the opening time is 100%, the solenoid valve is fully open. In this case, the variable capacity means controls A to the minimum via the actuator to quickly increase the turbine rotation speed, and conversely, when the duty is 0%, the solenoid valve is fully closed and A becomes the maximum. As a result, the turbine rotation speed is suppressed.

In actual control, in order to further improve control accuracy, the target boost pressure set according to the operating state is compared with the actual boost pressure detected by the boost pressure sensor, and the By correcting the control duty, the boost pressure is feedback-controlled to match the target boost pressure.

On the other hand, although such a capacity variable means secures supercharging pressure in low-speed operating ranges where the exhaust gas flow rate is small, in high-speed, high-load operating ranges where the exhaust gas flow rate increases, even if the cross-sectional area A is maximized, the In this case, the speed of flow introduced into the turbine does not decrease, and the rotational speed increases rapidly, causing the boost pressure to exceed the allowable upper limit.

Therefore, when the boost pressure approaches the upper limit value, an exhaust bypass valve is provided that bypasses a portion of the exhaust gas from upstream to downstream of the exhaust turbine to suppress the boost pressure. When the exhaust bypass valve opens, the amount of exhaust gas flowing into the exhaust turbine is reduced, which prevents the turbine rotational speed from decreasing and boost pressure from exceeding the upper limit.

Regarding the control of this exhaust bypass valve, the supercharging pressure is detected and feedback control is performed in the same manner as the capacity variable means to improve the control accuracy of the supercharging pressure.

Note that if such a variable capacity means and exhaust bypass valve are both subjected to feedback control, there will be a region where the controls interfere with each other, so that the variable capacity means and exhaust bypass valve are not originally used to control boost pressure. It is also expected that the position will deviate from the optimum position.

For example, when the boost pressure approaches the upper limit in a high-speed operating range, the capacity variable means would normally open fully and not throttle the exhaust, but if the exhaust bypass valve adjusts its opening according to the boost pressure, the exhaust would be reduced. It is possible to optimally control the boost pressure while maintaining good engine output efficiency without increasing the pressure, but in this case, if the capacity variable means narrows the opening and increases the exhaust flow velocity, the exhaust bypass valve will increase accordingly. Even if it is opened extra, it is possible to control the boost pressure to the same level, but in this case, the capacity variable means will unnecessarily throttle the exhaust flow path, and the exhaust pressure will increase accordingly. This reduces the exhaust efficiency of the engine and ultimately reduces the output efficiency.

Therefore, when performing feedback control using these capacity variable means and the exhaust bypass valve, it is necessary to set the range in which feedback control is to be performed depending on the operating state of the engine (feedback control includes proportional control, integral control, etc.). However, it is not necessarily necessary to set separate control regions for all of these; for example, it is sufficient to set such a control region only for integral control.) When one is performing feedback control, the other is fixed. By doing so, problems caused by control interference can be avoided.

(Problems to be Solved by the Invention) Incidentally, the variable capacity means is a means that operates frequently starting from a low speed rotation range and has good responsiveness, but it may become difficult to move as time passes, resulting in so-called stiffness. , even though the control amount of this capacity variable means is calculated according to changes in the operating condition,
The operation of the variable capacity means cannot follow this control amount, resulting in a delay in the operation response. In such a case, the boost pressure can be raised to the target value by reluctantly reducing the opening degree over time in accordance with the control amount corresponding to the change in the operating state.

On the other hand, when the boost pressure exceeds the target value, feedback control, which performs control based on the actual boost pressure, increases the opening of the capacity variable means to lower the boost pressure. Even if the control amount is calculated, the displacement variable means is difficult to open due to this reluctance, and this delay in response causes a phenomenon in which the boost pressure continues to greatly exceed the target value. The more serious the degree of such operational response delay is, the longer the time period during which excessive boost pressure is applied to the engine continues.

Normally, when performing feedback control on both the variable capacity means and the exhaust bypass valve, if you do not provide independent control areas for both and perform feedback control on both, the exhaust bypass valve will help with the delay in control response of the variable capacity means. However, as described above, mutual interference in control occurs in such a device. Therefore, in order to avoid mutual interference in control, if a region is provided in which feedback control (for example, integral control) is performed in each region, the operation of the exhaust bypass valve in the control region in which the displacement variable means performs boost pressure control is generally fixed and the boost pressure is controlled primarily by the variable capacity means, so if a control response delay occurs in the variable capacity means, the boost pressure will continue to greatly exceed the target value.

In this way, if the boost pressure exceeds the target value and exceeds the allowable boost pressure limit, excess air (fuel) will be supplied to the engine, causing an excessive load on the engine. This will have a negative impact on durability. This tendency occurs during overboost control, which temporarily increases the target boost pressure during acceleration and increases the boost pressure to follow this, in order to improve the acceleration performance of the engine. This becomes noticeable because the target boost pressure is set higher than usual.

In addition, as a countermeasure in the event that the boost pressure exceeds the allowable limit value of the boost pressure, a fuel cut is performed uniformly when the boost pressure exceeds the limit value.
A phenyl safe function is provided to prevent excessive load from being applied to the engine. This fuel cut is performed as an emergency measure when the exhaust flow rate becomes large and it is not possible to prevent the boost pressure from rising above the limit value even if the exhaust bypass valve is fully opened. Although it can certainly protect the engine, If such a fuel cut exists, the engine torque will fluctuate rapidly, which will worsen drivability.

The present invention has focused on this point, and even if a control response delay occurs due to the reluctance of the displacement variable means in the feedback control region on the displacement variable means side, the boost pressure will not reach the target due to this control response delay. To provide a supercharging pressure control device that avoids a state in which the value is greatly exceeded and prevents the fuel from reaching a fuel cut in order to improve driveability even in a state where the capacity variable means becomes increasingly reluctant. With the goal.

(Means for Solving the Problems) FIG. 1 is an overall configuration diagram of the present invention. It is controlled according to the operating state by an exhaust bypass valve 10 that bypasses a part of the exhaust gas passing through the turbine from upstream to downstream.

Reference numeral 1 denotes an operating state detection means for detecting parameters representative of the operating state of the engine 8, such as the amount of intake air and engine speed, and 2 a supercharging means for detecting the actual supercharging pressure supercharged by the compressor. The pressure detecting means 3 is a deviation calculation means for calculating the deviation between the detected supercharging pressure and the target supercharging pressure.

Reference numeral 4 denotes a feedback control region determining means which determines the operating region in which feedback control is to be performed and also selects which of the variable capacity means or the exhaust bypass valve is to be subjected to feedback control.

Reference numeral 5 denotes a first control amount calculation means for calculating a control amount of the exhaust turbine capacity variable means 9 in accordance with a parameter representing the operating state including at least the integrated value of the deviation; 6 also includes at least the integrated value of the deviation. This is a second control amount calculation means that calculates the control amount of the exhaust bypass valve according to a parameter representing the operating state.

11 is a supercharging pressure abnormality detection means for detecting an abnormality when the supercharging pressure becomes an excessive supercharging pressure higher than a target value;
1 and 2, when an abnormality in the boost pressure is detected in the feedback control area on the side of the displacement variable means, the exhaust bypass valve 10 is opened in the direction corresponding to the boost pressure at that time, regardless of the determination result of the feedback control area. This is a control amount correction means for correcting the control amount of the exhaust bypass valve 10.

7 is a first control means for driving and controlling the capacity variable means 9 according to the control amount calculated by the control amount calculation means 5; The control amount correction means 1
Exhaust bypass valve 1 according to the control amount corrected in step 2.
This is second control means for driving and controlling the 0.

(Function) The feedback control region determining means 4 provides a feedback control region (for example, a region for integral control) for the variable capacity means 9 and the exhaust bypass valve 10, and these regions are independently controlled in each region. Therefore, in the feedback control region on the displacement variable means side, the boost pressure is mainly controlled by the displacement variable means 9 (for control other than integral control, for example proportional control, an independent control region is not provided). Sometimes, this proportional control also controls the exhaust bypass valve 10 at the same time, but even in that case, overall,
It is mainly controlled by the capacity variable means 9. ),
This variable capacity means 9 may cause a delay in control response due to reluctance due to changes over time, etc., and when the boost pressure exceeds the target value, the variable capacity means 9 is opened to lower the boost pressure. Even if the amount is calculated, the opening of the variable capacity means 9 is delayed due to this delay in operational response, causing a phenomenon in which the boost pressure continues to greatly exceed the target value. In contrast, in the present invention, the boost pressure abnormality detection means 11
When detecting an abnormality in the supercharging pressure in which the supercharging pressure exceeds the target value, and when the control amount correction means 12 detects this abnormality in the supercharging pressure in the feedback control region on the side of the displacement variable means, it The control amount of the exhaust bypass valve 10 is corrected in the direction of opening the exhaust bypass valve 10 according to the boost pressure at that time. Therefore, even when an abnormality is detected, the boost pressure is normally controlled by the capacity variable means 9, but when such an abnormality occurs, the exhaust bypass valve 10 is auxiliarily driven to open, reducing the exhaust flow rate to the exhaust turbine. let As a result, the boost pressure quickly converges to the target value, thereby preventing the boost pressure from continuing to exceed the target value, which would pose a problem in terms of engine durability.

Further, by avoiding an abnormality in supercharging pressure by auxiliary driving of the exhaust bypass valve 10, there is no need to perform a fuel cut, and drivability is improved.

(Embodiment) FIG. 2A is a flowchart in which essential parts of the present invention are extracted from a flowchart showing control of the capacity variable means and exhaust bypass valve, which will be described later as an embodiment of the present invention, and mainly shown in FIG. 5A. Corresponds to process number 212. I will explain this first.

In addition, the numbers in the figure represent processing numbers, the capacity variable means is described by the symbol VN, the exhaust bypass valve is described by the symbol WG, and signals representing operating conditions such as engine rotation speed and intake air flow rate are stored in memory. shall be taken as a thing.

The control shown in this flowchart is executed once every control cycle (50 ms), and is performed within a control device such as a microcomputer that controls the actuators for driving the VN and WG.

First, in the 500, among the parameters representing the operating state of the engine, for example, VN
Basic control amount (basic control duty) on side and WG side
is obtained by calculation. In addition, the basic control variables on the VN side and WG side are calculated in advance according to the operating conditions.
Two tables as shown in FIG. B may be stored in memory and determined by table lookup. In step 501, the deviation ERROR between the actual boost pressure and the target boost pressure, which is used for calculating the control amount during feedback control, is determined.

502, VN side, WG based on intake air flow rate
Determine which side should be subjected to feedback control. That is, if the intake air flow rate is smaller than the area determination air flow rate (QSVNTWG, QSVNTWGX), it is determined that the intake air flow rate is in the VN side control area, and the process proceeds to 503 to find the VN side feedback control amount (feedback correction amount). In addition, if the intake air flow rate is larger than the area judgment air flow rate,
Determine that it is a control area on the WG side and proceed to 507
Find the feedback control amount on the working group side.

At 504, it is detected whether the boost pressure is abnormal based on the actual boost pressure. In this case, the abnormality is detected when the boost pressure exceeds the target value, and if an abnormality is detected, the feedback on the WG side is adjusted in the direction in which the WG opens according to the boost pressure at that time. Correct the control amount.

The feedback control amount on the WG side corrected in this way is output to the control means on the VN side and the WG side at the same time as the feedback control amount on the VN side at 506, regardless of the determination result of the feedback control area. That is, the WG is auxiliary opened by the corrected feedback control amount, regardless of the result of determining the feedback control region, and the rotation of the exhaust turbine is lowered to lower the supercharging pressure to the target value.

With feedback control, the boost pressure should be controlled to the target value, but if there is a delay in control response due to reluctance due to VN changes over time, the VN Although the control amount is calculated to lower the boost pressure, the VN operation does not follow this control amount and there is a delay in reducing the boost pressure to the target value. For this reason, the engine is exposed to a high load condition during which excessive boost pressure acts, which is undesirable. However, when such an abnormality in boost pressure is detected in the feedback control area on the VN side, the WG opens auxiliarily and controls both the VN and WG at the same time.
It supplements the feedback control on the VN side and quickly lowers the boost pressure to the target value.

Next, the correction performed in 505 is, for example, the second
As shown in Figure C, boost pressure P1, P2, P3 (P0<P1
<P2<P3), the feedback control amount on the WG side is reduced by 25 to 75% in three steps. Here, in the same figure, the point lower than P0 is the target value of boost pressure, and the higher the boost pressure is, the greater the WG control amount is reduced, and the greater the reduction rate of this control amount, the greater the WG valve opening. . The reason why the correction is performed in stages in this way is that the degree of increase in boost pressure and the maximum boost pressure differ depending on the degree of VN astringency, that is, the degree of control response delay.

For example, if you want to make a correction to open the WG when the boost pressure rises sharply in order to make it safer, in this case, the boost pressure will increase unless the WG control amount is corrected to the side where it opens considerably. Since it would become expensive, the control amount was already reduced by 75% at P2 and WG
All you have to do is set it to open rapidly. However, with this setting, even if the increase in boost pressure is not severe, the WG will open suddenly when P2 is reached, so the decrease in boost pressure will be greater than the increase in boost pressure, so the boost pressure will increase. There is a sudden drop in torque, which causes large fluctuations in torque. In order to prevent this situation, the control amount of the WG is corrected to be reduced in stages according to the boost pressure.

In addition, as an emergency measure when the WG side control cannot reduce the boost pressure to the target value, the boost pressure can be further reduced by cutting the fuel at the allowable boost pressure limit value. It is set beyond the limit. In other words, in this invention, the fuel cut is not performed immediately when the boost pressure exceeds the target value, but the WG is operated auxiliary even in the boost pressure range up to the fuel cut to reach the target boost pressure. This improves drivability by controlling the value.

In addition, FIG. 2A and FIG. 5A of the embodiment described later,
Describing the correspondence of process numbers with Figure 5B, Figure 6, etc., 504 and 505 directly correspond to 212,
Other 500 are 201,206, deviation of 501
ERROR is 113,114,134,135,1
This is used to calculate 36 P, I, and D components. 503 becomes 113, 134, 136, 5
07 is 114, 135, 136, 506 is 11
3,114,134,135. Also,
Figure 2B is Figure 10A, Figure 10B, Figure 2C
corresponds to FIG.

Note that, instead of performing the correction immediately when these supercharging pressures P1 to P3 are reached, the correction may be performed via a delay means in order to smooth the response.

Next, FIG. 3 is a schematic diagram of the mechanical configuration of an embodiment of the present invention. In the figure, air is supplied to an engine 21 through an intake pipe 22 and an intake manifold 23, and is exhausted through an exhaust manifold 24 and an exhaust pipe 25. Intake pipe 2
The end bent to the left in figure 2 shows the amount of intake air.
An air flow meter 31 for measuring Qa is provided,
A compressor 35 , which constitutes a part of a turbocharger, is disposed at a bent portion of the intake pipe 22 , and pressurizes intake air supplied via an air flow meter 31 and supplies it to the engine 21 . A throttle valve 32 is disposed at the base end of the intake pipe 22 close to the intake manifold 23, and a relief valve 2 is provided in the intake pipe 22 between the throttle valve 32 and the compressor 35.
9 is provided.

A portion of the exhaust pipe 25 bent to the right in the figure forms a turbine chamber 38, and a turbine 37 is disposed within the turbine chamber 38. The turbine 37 is connected to the compressor 35 via a connecting shaft 36. ing. The turbine chamber 38 has a scroll 39 formed to surround the turbine 37, as shown in FIG. Therefore, it is formed gradually smaller. The introduction passage 40 to this scroll 39 and the scroll 39
A movable tongue portion 45 as a capacity variable means constituted by a flap valve is provided at the confluence portion of the terminal end portion 41 of the introduction passage 40. The base end portion is rotatably supported by a shaft 46.

This movable tongue portion 45 is connected to the exhaust pipe 2 near the upstream side, which is the introduction passage 40 to the turbine 37 in FIG.
It is located within 5. A shaft 46 rotatably supporting the movable tongue portion 45 is connected to the upper end of a rod 48 via an arm 47, and the lower end of the rod 48 is connected to a diaphragm 52 constituting an actuator 50 for driving the movable tongue portion. connected. The case 51 housing the diaphragm 52 is divided into an atmospheric chamber 53 and a positive pressure chamber 54 by the diaphragm 52, and the atmospheric chamber 53 has a spring biased to push the diaphragm 52 toward the positive pressure chamber 54. 55 are arranged. The positive pressure chamber 54 is connected to the intake pipe 22 on the downstream side of the compressor 35 via a connecting pipe 56, and supercharging pressure generated by the compressor 35 is supplied to the positive pressure chamber 54, and the diaphragm 52 is
It is pushed toward the atmospheric pressure 53 side against the pressure. Further, a solenoid valve 57 is provided in the middle of the connecting pipe 56, and when this solenoid valve 57 is driven and released by the control unit 80, the connecting pipe 56 is communicated with the compressor inlet via this solenoid valve 57. The pressure within the positive pressure chamber 54 decreases. More specifically, the solenoid valve 57 is duty-controlled by the control unit 80, and as the duty value increases, the degree of opening of the solenoid valve 57 increases and the pressure in the positive pressure chamber 54 decreases. Therefore, due to the action of the spring 55 in the atmospheric chamber 53, the diaphragm 52
moves downward, and this movement is transmitted to the movable tongue portion 45 via the rod 48, arm 47, and shaft 46, and the movable tongue portion 45 moves in the direction of reducing the exhaust gas introduction passage 40 to the turbine 37, that is, Rotate in the closing direction. As a result, the flow rate supplied to the turbine 37 increases, and the boost pressure applied to the engine 21 by the compressor 35 increases. Also, conversely,
As the duty value decreases, the degree of opening of the solenoid valve 57 decreases and the pressure in the positive pressure chamber 54 increases, so the diaphragm 52 moves upward against the spring 55, and the movable tongue 45 It rotates in the direction to open the passage 40. As a result, the flow rate supplied to the turbine 37 becomes slower, and the boost pressure applied to the engine 21 by the compressor 35 decreases.

An exhaust bypass valve (wastegate valve) 60 is provided at a connection portion between the exhaust bypass passage 26 that bypasses the turbine 37 and the exhaust manifold 24 . This exhaust bypass valve 60 is connected to one end of a rod 63 via an arm 61 and a connecting member 62, and the other end of the rod 63 is connected to a diaphragm 72 of an actuator 70 for driving the exhaust bypass valve. A case 71 housing the diaphragm 72 is divided into an atmospheric chamber 73 and a positive pressure chamber 74 by the diaphragm 72, and the atmospheric chamber 73 has a spring biased to push the diaphragm 72 toward the positive pressure chamber 74. 75 are provided. The positive pressure chamber 74 is connected to the compressor 35 via a connecting pipe 76.
The compressor 35 is connected to the intake pipe 22 on the downstream side, and supercharging pressure generated by the compressor 35 is supplied to the positive pressure chamber 74.

Further, in the middle of the connecting pipe 76, a solenoid valve 77 is provided.
When the solenoid valve 77 is driven and opened by the control unit 80, the connecting pipe 76 is communicated with the compressor inlet via the solenoid valve 77, and the pressure in the positive pressure chamber 74 is reduced.
More specifically, the solenoid valve 77 is duty-controlled by the control unit 80, and as the duty value increases, the degree of opening of the solenoid valve 77 increases, and the pressure in the positive pressure chamber 74 decreases. The diaphragm 72 moves downward under the action of the spring 75, and this movement is transmitted to the exhaust bypass valve 60 via the rod 63, the connecting member 62, and the arm 61, and the valve 60 moves in the direction of closing the bypass passage 26. Furthermore, as the duty value decreases, the degree of opening of the solenoid valve 77 decreases and the pressure in the positive pressure chamber 74 increases, so the diaphragm 72 moves upward against the spring 75, thereby opening the exhaust bypass valve 60. move in the direction. The exhaust bypass valve 60 is provided in order to prevent damage to the engine 21 due to excessively high supercharging pressure of intake air supplied to the engine 21 by the turbocharger when the engine 21 is in a high-speed, high-load state. , part of the exhaust gas from the engine 21 is discharged to the outside to reduce the amount of exhaust gas supplied to the turbine 37 so that an appropriate boost pressure can be introduced into the engine 21.

The control unit 80 is composed of a microcomputer consisting of a microprocessor, memory, and an input/output interface including an A/D converter, and the amount of intake air is supplied from the air flow meter 31 to the control unit 80 via the interface. At the same time, the rotational speed of the engine 21 is supplied from a crank angle sensor 30 provided on the left side of the engine 21, and the boost pressure is also supplied from a boost pressure sensor 33. The control unit 80 controls the solenoid valves 57, 77 according to this information.
Appropriately control the duty value of the signal that drives the
By varying the cross-sectional area of the exhaust gas introduction passage 40 to the turbine 37 via the movable tongue portion 45 and by varying the amount of exhaust gas to the turbine 37 via the exhaust bypass valve 60, the engine 21
The boost pressure of the intake air supplied to the engine is appropriately controlled according to the amount of intake air Qa to increase torque from low-speed to high-speed operating ranges.

Next, FIGS. 5A to 9 show flowcharts in which control of the variable capacity means (movable tongue) and the waste gate valve (exhaust bypass valve) is realized using a microcomputer. In addition, the numbers in the diagram represent processing numbers, the variable capacity means is described by the symbol VN, the waste gate valve is described by the symbol WG,
It is assumed that signals representing operating conditions such as engine rotational speed and intake air flow rate are stored in memory.

To explain from Fig. 5A, boost pressure control
VNWGCONTROL calculation processing is performed every predetermined cycle (control cycle), and the control target value (target boost pressure) of boost pressure obtained from various operating conditions P2ADAPT
The control amount for controlling VN and WG is calculated so that the actual supercharging pressure P2 matches the actual supercharging pressure P2. To explain step by step, first, at step 200, the air flow rate index QS is determined from the input intake air amount QA. Note that in actual control, this QS is used as calculation data, but in the following explanation, for convenience, QS will be explained as the intake air amount. In 201, the basic control duty of WG is set to BASEDUTY1 from the QS value.
to ask. At step 202, 35% of the control duty is added to BASEDUTY1. This is due to a setting error on the WG side and component variations.
This is the amount of correction to prevent the opening of the opening. For example, in the VN side control area, if the WG opens due to a setting error or component variation on the WG side, the VN will try to increase the boost pressure and shift to the closed side, and the VN side will no longer be able to perform normal learning. It becomes impossible. Therefore, the control amount on the WG side is increased in advance to ensure the reliability of learning control on the VN side. In step 203, in order to eliminate steady-state deviation from the control target value, a learning amount LEARNWG determined by learning control is added to BASEDUTY1.
At 204, it is checked whether overboost control is being performed, which temporarily increases boost pressure during sudden acceleration to improve acceleration performance.If it is determined that overboost control is being performed, the process advances to 205, where overboost control is performed. Acceleration correction amount for
Add to BASEDUTY1. Here,
LEARNWG and acceleration correction amount give the feedforward control amount on the WG side. In addition, LEARNWG
The calculation and overboost control will be described later.

On the other hand, in steps 206 to 210, the feed forward control amount on the VN side is obtained from BASEDUTY0.
The only difference between the processes 201-205 and 206-210 is that the basic control duty determined for BASEDUTY0 is subtracted by 5% in 207. This is a correction amount to prevent the WG from opening when the VN basic control duty table deviates to the closing side due to setting deviations on the VN side or component variations. In 208, the amount of learning is the same as on the WG side.
Add LEARNVN. Furthermore, this
The calculation of LEARNVN will also be described later. Here, since the basic control duty of VN and WG is given as the characteristics shown in FIG. 10A and FIG. 10B, for example, the basic control duty is obtained by storing the table shown in FIG. Then, it can be found by one-dimensional lookup. However, FIG. 10C is for VN, and H represents hexadecimal representation.

In 211, a feedback correction amount is calculated for the difference between the actual boost pressure P2 and the target boost pressure P2ADAPT, and it is further added to the feedback control amount obtained earlier to obtain the final control signal amount.
Find BASEDUTY0 and BASEDUTY1 respectively. The feedback control P2FBCONT performed in step 211 will be described later.

212 is the main part of the present invention. That is, in step 212, overshoot prevention at the initial stage of rapid acceleration and fail-safe processing in the event of component failure are performed. To explain the process for preventing overshoot, the boost pressure rises rapidly during sudden acceleration, but in the case of a turbocharger with a VN, the boost pressure rises faster than a normal turbocharger, so as shown in Figure 16. This will cause an overshoot. In particular, in the example shown in the figure, the supercharging pressure exceeds 500 mmHg during overboost control, which may impair the durability of the engine. In order to prevent this, the control duty (control signal) on the WG side is temporarily reduced at the beginning of rapid acceleration to increase the flow rate of exhaust gas that bypasses the turbine 37 and is released, thereby lowering the supercharging pressure. More specifically, control duty correction on the WG side is performed using the supercharging pressure as shown in FIG. That is, when the boost pressure increases due to sudden acceleration, the control duty on the WG side is reduced by 50% when the predetermined boost pressure P0 is exceeded. However, if the level of P0 is set low (e.g. 375mmHg) to prevent overshoot, the boost pressure will decrease thereafter, so the reduction due to the slice level of P0 should only be made for 0.3 seconds after the boost pressure reaches P0. . After 0.3 seconds have elapsed, normal fail-safe processing is performed to gradually reduce the control duty on the WG side using P1 to P3 (>P0) as slice levels. That is,
When the VN has a delay in control response due to changes over time, and it becomes difficult to control the boost pressure normally using the boost pressure control on the VN side, it is necessary to protect the engine and before performing a fuel cut as a last resort. ,
By opening the WG auxiliary, the supercharging pressure is controlled to the target value, and the deterioration in drivability that occurs when fuel cut is performed is prevented. Also, in consideration of the case where the WG does not open, if the situation continues to exceed P4, a fuel cut request flag is set to perform a fuel cut using the engine control system as a last resort. This is how it was finally requested
The control duty on WG and VN sides will be explained later.
After being transferred to ONDUTY1 and ONDUTY0, they are output to the aforementioned solenoid valves 77 and 57, respectively, via the output interface. In addition, overshoot countermeasures and failsafe are ONDUTY1,
This is done by correcting ONDUTY0.

Next, feedback control is performed at 211.
P2FBCONT will be explained using the flowchart in FIG. Here, we will determine the start conditions for integral control of VN feedback control, determine the conditions for switching feedback control from the VN side to the WG side, initialize the feedback control variables, and determine the conditions for switching feedback control from the WG side to the VN side. In addition to making the judgment, the feedback correction amount and learning amount are calculated, and the final control amount is determined.
Store to ONDUTY0 and ONDUTY1.

To explain step by step, at 100, processing is performed to reduce the control target value P2ADAPT in order to avoid abnormal combustion when the intake air flow rate increases.
For example, a one-dimensional table as shown in Fig. 11A is stored in the memory ROM, and the intake air flow rate is
When QS exceeds the predetermined intake air flow rate QSPDOWN0, the control target value is gradually decreased. In 101
A flag FP2FBVNWG indicating whether the control area is on the VN or WG side is checked, and if it is '1', it is determined that feedback control is being performed on the WG side, and the process proceeds to step 111. 10 if '0'
Steps 2 and subsequent steps determine the operating range in which feedback control is to be performed and calculate the amount of learning.

First, in step 102, it is checked whether the actual supercharging pressure P2 is smaller than the range judgment supercharging pressure P2JUDGE (230 mmHg) for determining the operating range in which feedback control is performed, and if it is smaller, the operating range is not determined and the process proceeds to step 111. This is to prevent the feedback control from switching to the WG side before entering the overboost control after a sudden acceleration determination, which will be described later, is performed. In other words, in overboost control, sudden acceleration is determined by comparing the acceleration time from 100 mmHg to 200 mmHg with the determination standard TJUDGE, and the acceleration time TJUDGE
If it is smaller than , it is determined that there is sudden acceleration, but
If the above operating range is to be determined even when boost pressure P2 is smaller than P2JUDGE, in addition to determining sudden acceleration, in order to determine which side of VN or WG to perform feedback control, we will explain later. A comparison is made between the intake air flow rate QS and the feedback control region determination air flow rate QSVNTWG. Therefore, when sudden acceleration is determined and overboost control is performed to improve responsiveness during sudden acceleration, if the intake air flow rate becomes larger than QSVNTWG, it is determined that it is in the WG side control region, so sudden acceleration is detected. Feedback control will switch from the VN side to the WG side regardless of the judgment of
is smaller than P2JUDGE, the above judgment is not performed.

At step 103, the feedback control area determination air flow rate QSVNTWG during normal operation is stored in the register ACC. This area judgment air flow rate QSVNTWG
The condition for switching the feedback control from the VN side to the WG side is when the value exceeds 1, which is the air amount line A in FIG. That is, in the same figure, line A
The area on the left side is the control area on the VN side, and the area on the right side is the control area on the WG side. At 104, it is checked whether or not there is a sudden acceleration. If the flag FACCEL is '1', it is determined that there is a sudden acceleration and the process proceeds to 105; if it is not determined that there is a sudden acceleration, the process proceeds to 107. this flag
FACCEL is a sudden acceleration determination flag that is set to '1' when sudden acceleration is determined, and will be explained in the acceleration determination process described later. At 105, it is checked whether the overboost control has ended or not. If it is determined that the overboost control is in progress, the process proceeds to 106, and if it is determined that the overboost control has ended, the process proceeds to 107.
At 106, the area determination air flow rate QSVNTWGX (>QSVNTWG) during overboost control is stored in ACC. This area judgment air amount QSVNTWGX
is the air amount line B shown in FIG. That is, since the control region on the VN side expands during overboost control, the region determination air flow rate is increased from line A to line B accordingly.

At step 107, the area determination air flow rate stored in the ACC is compared with the actual intake air flow rate QS. QS
If it is larger than ACC, it is determined that the control area is not on the VN side, and in step 108, a flag FP2FBVNWG indicating whether the control area is on the VN side or the WG side is set to '1'. As a result, FP2FBVNWG = 1 means that the control area on the VN side has been switched to the control area on the WG side, and at 109
Start the timer to start learning control on the WG side and calculate the learning amount on the VN side at 110
Perform ACCLEARNVN. This learning amount calculation will be described later. 111 flag FP2FBVNWG
If it is '0', go to 113 and set '1'.
In this case, the learning amount calculation ACCLEARNWG on the WG side is performed in step 112. This learning amount calculation will also be described later. In this way, the operation range for performing feedback control as shown in Fig. 15 is determined, and VN, WG
The learning amount on the side will be calculated.

Next, from step 113 onwards, the feedback correction amount is calculated on each of the VN and WG sides. Here, we will discuss proportional-integral-derivative control, and the proportional, integral, and differential components calculated from the deviation will be expressed as P
Abbreviated as minutes, I minutes, and D minutes. First of all, in 113
I calculated the P portion on the VN side and found it earlier.
Add to BASEDUTY0 and store the added result in memory M2. This P component is calculated by the following calculation, taking into consideration the stability of the control and the case where the basic control duty BASEDUTY0 is deviated. That is,
The P portion on the VN side is KPVN x (ERROR) ² . Here, ERROR is the deviation between the target boost pressure and the actual boost pressure (ERROR=P2ADAPT−P2), and KPVN is the calculation gain. Similarly in 114
P on the WG side is calculated and added to BASEDUTY1 mentioned above, and the added result is stored in memory M2+2. However, the P portion on the WG side is KPWG×ERROR, where KPWG is the operational gain. Note that the first
In Figure 4, the P minute on the VN side is shown by a broken line, and the P minute on the WG side is shown by a solid line.

The P component of the feedback control obtained in this way is always added, but since the integral-differential control is performed at a predetermined boost pressure or higher, it is determined in steps 115 to 118 whether or not the integral-differential control is to be performed. First, at 115, the actual boost pressure P2 is stored in the register ACC, and at 116, the target boost pressure P2ADAPT is stored.
Check if is 375mmHg.
If P2ADAPT is 375 mmHg, the process proceeds to 118, but if P2ADAPT is smaller than 375 mmHg, P2DOWNVALUE is added to ACC in 117. Normally, when the boost pressure reaches P2MIN (320 mmHg), it is determined that integral-derivative control is possible at the next step 118, but if the air flow rate becomes high at step 100 and the control target value is lowered, This is to allow it to be determined that control is possible from a lower boost pressure. In other words, in the low-medium air flow rate region where the control target value is 375 mmHg, the control region in which integral-derivative control is performed is determined by comparing the actual boost pressure and the determined boost pressure P2MIN (320 mmHg), but if the control target value is
In a high air flow rate region where the air flow rate decreases to 375 mmHg or less, it is preferable to also reduce the determined supercharging pressure P2MIN to secure a control region for performing integral-differential control. For this reason,
All you have to do is to compare P2 stored in register ACC with the value obtained by subtracting a predetermined value from P2MIN, but this can be done by adding the above predetermined value to P2 in advance and comparing this added value with P2MIN. The same result will be obtained if you compare. The predetermined value in this case is the P2DOWNVALUE,
P2DOWNVALUE may be a constant value or may be a value that changes depending on QS.

At step 118, it is checked whether ACC is greater than or equal to P2MIN, and if it is greater than or equal to P2MIN, it is determined that integral-differential control is possible, and the process proceeds to step 127. In other words, P2MIN is set as a predetermined boost pressure lower than the target boost pressure, and when P2MIN is exceeded,
This is given as a condition for starting integral control of VN feedback control. If it is not determined that control is possible, the QS returns the specified air flow rate at 119.
Check whether it is less than QSWGAREA, and if it is smaller, reset the various control sequence flags and initialize various feedback variables at 120, and at the same time rewrite the learning amount on both the VN and WG sides at 121. . In other words, the boost pressure
Smaller than P2MIN and air flow rate
When it is smaller than QSWGAREA, it is given as a condition for initializing the feedback control variable and switching control from the WG side to the VN side. Therefore,
If it is larger than QSWGAREA, the process proceeds to step 122 in order to prevent the flag from being reset or the variables from being initialized if the boost pressure drops instantaneously in a high air flow area. In other words, if the accelerator pedal is temporarily released during full-throttle acceleration in the high air flow range, the boost pressure may decrease faster than the intake air flow rate, and in this case, the air flow rate remains high and the WG As a result, the supercharging pressure falls below the above-mentioned P2MIN even though it is within the side control region. Therefore, if we decide to reset the above flags and initialize the feedback variables in this case, the accumulated value of the deviation ERROR up to the previous time on the WG side will be
If ERRORIWG disappears, the amount of control on the WG side becomes small, and if there are variations in parts, this will lead to control deviations, so in order to prevent this, the above-mentioned flags are not reset. be.

At 122, the values stored in the memories M2 and M2+2 (results of adding various correction amounts to the basic control duty) are transferred to ONDUTY0 and ONPITY1.
Note that for this transfer, an upper limit value and a lower limit value are set, and each value on the VN and WG side is limited between the upper limit value and the lower limit value. At 123, it is checked whether QS is smaller than the judgment air flow rate QSDUTYCUT, and if it is, the control duty is set at 124.
Set ONDUTY0 and ONDUTY1 to the minimum value. This process is performed to increase durability by not moving the control solenoid valves 57, 77 at low air flow rates such as during idle. At step 125, error prevention processing for acceleration determination is performed.
The contents of this process will be explained in the acceleration determination process described later. After this, proceed to 212 of VNWGCONTROL.

Next, a case where it is determined in step 118 that the integral-derivative control is in the controllable region and the process proceeds to step 127 will be described. From 127 onwards, the judgment result (101 to 1
06 processing) based on VN,
The WG side performs each calculation. First, at 127, check whether the flag FP2FBVNWG is '1', and if it is '1', at 128, check the deviation from the previous time on the WG side.
This deviation in the accumulated value of ERROR ERRORIWG
Add ERROR. If '0', proceed to 129
Accumulated value of deviation ERROR up to the previous time on the VN side
Add the current deviation ERROR to ERRORIVN.
At step 130, it is checked whether or not overboost control has started. If it is determined that it has started, at step 131, the correction amount VNCOEFI during overboost control is added to ERRORIVN. This is to add the feedforward control amount corresponding to the increase in the control target value during overboost control. If overboost control has not started, proceed to 132.
Check FP2FBVNWG, if '1' then WG
Since this is the side control area, the process proceeds to step 133 and a predetermined value is subtracted from the integrated value ERRORIVN of the deviation ERROR on the VN side. That is, after the feedback control is switched from the VN side to the WG side, the control amount on the VN side is gradually decreased from the control amount immediately before switching. If the control amount on the VN side is maintained at the control amount immediately before switching even after the feedback control is switched to the WG side, the flow rate in the introduction passage 40 will increase as the exhaust gas flow rate increases, and the pressure will decrease. Due to the pressure drop, the movable tongue portion 45 is rotated in a direction to close the introduction passage 40, thereby reducing the capacity of the turbocharger. In contrast, the integrated value of the deviation ERROR on the VN side
When the predetermined value is subtracted from ERRORIVN, the movable tongue 4
5 is rotated in the direction to open the introduction passage 40 and is fully opened, so that sufficient exhaust capacity can be secured even when the WG side controls it, and the performance of the turbocharger can be maximized. on the other hand,
If FP2FBVNWG is '0', proceed to 134 and VN
The I component on the side is calculated by KIVN×ERRORIVN and added to the above-mentioned M2. Here, KIVN is the operational integral gain. At the same time, this I minute is used as the learning amount on the VN side for learning control.
Store in VNLEARN. In 135, I on the WG side
The amount is calculated by calculating KIWG×ERRORIWG and this is added to M2+2 mentioned above. Here, KIWG is the computational integral gain. At the same time, this I portion is used as the learning amount on the WG side for learning control.
Save to WGLEARN. At step 136, D is calculated and the result is stored in memory M1. The calculation for this D is obtained by KD×(ERROR1−ERROR). Here, KD is the operational differential gain.
Specifically, FP2FBVNWG checks which side is being controlled, VN or WG, and if it is in the VN side control area, the VN side gain KDVN is controlled, and the WG side is controlled.
If it is in the control region of the WG side, the gain KDWG of the WG side is selected and calculated. In addition, ERROR1 is the previous
It is ERROR. In 137, FP2FBVNWG'
When it is 1', add D on the WG side to the WG side at 138, store the added result in M2+2, and set it to '0'.
In this case, the value D on the VN side is added to the VN side in step 139, and the added result is stored in M2. In step 140, the current deviation ERROR (=P2ADAPT−P2) is calculated for the D component to be performed in the next calculation process.
Store in ERROR1. 141 and 142, the values stored in the memories M2 and M2+2 (results of adding various correction amounts to the basic control duty) are set as the final control duty, ONDUTY0,
Move to ONDUTY1. Note that for this transfer, an upper limit value and a lower limit value are set, and each value on the VN and WG side is limited between the upper limit value and the lower limit value. After this, proceed to 212 of VNWGCONTROL.

Next, overboost control, which temporarily increases boost pressure during sudden acceleration to improve acceleration performance, will be further explained. Basically, overboost control is realized by correcting the aforementioned feedforward control amount and increasing the control target value. Figure 5B shows the process of setting and resetting various overboost flags.
The flowchart of BOOSTCNTR and FIG. 7A show the flowchart of sudden acceleration determination processing ACCELJUDGE.

First, the determination of sudden acceleration will be explained step by step based on FIG. 7A. This process is executed once every 10 ms, in addition to the control calculation process described earlier. At 300, the boost pressure is stored in P2. 30
In step 1, it is checked whether P2 has exceeded 100 mmHg, and if it has not, in step 302, various control sequence flags are reset and various variables are initialized. If it is 100 mmHg or more, the process advances to 303, and if it is exceeded for the first time, the process advances to 304 to start the acceleration time measurement timer. 305 calculates the time that is the criterion based on the engine rotation speed at 100 mmHg, gear position, etc., and stores this in TJUDGE. This judgment standard is the judgment line shown in Figure 13B, that is, the judgment line given by the engine rotation speed (rpm) (×10ms) at 156250/100mmHg, and the acceleration time τ, which will be described later, is below this judgment line. If it is in the area of
It is determined that the acceleration is sudden. Note that the numbers in the figure represent the gear positions from 1st to 4th gear, and up to 3rd gear there is no problem as it falls below the judgment line, but in 4th gear, the acceleration time τ (supercharging The time it takes for the pressure to go from 100mmHg to 200mmHg) crosses the judgment line and is distributed in the area surrounded by the broken line in the figure. Therefore, in the 4th speed low rotation range, it is necessary to move the judgment line beyond this area, and the value obtained by adding a predetermined value to the value of the judgment line is used as the judgment standard. For these reasons, the gear position is also taken into consideration as a criterion.

If 303 exceeds 100mmHg for the second time or later, 3
Proceed to step 07 and check whether P2 exceeds 200mmHg, and if it is less than 200mmHg, no judgment of sudden acceleration is made. If it exceeds 200mmHg, the timer value reset at 304, that is, the acceleration time τ shown in Fig. 13A (as mentioned above, the boost pressure changes from 100mmHg to
Judgment criteria determined by 305 (time to reach mmHg)
Checks whether it is smaller than TJUDGE, and if it is smaller, proceeds to 309 and determines that it is a sudden acceleration.FACCEL
Set to '1'. After that, the process returns to engine control. In this way, it is possible to input the boost pressure necessary for control and judge the sudden acceleration state, and this information is used in the VNWGCONTROL shown in Figure 5A and the VNWGCONTROL in Figure 6.
P2FBCONT, as shown in Figure 5B, which will be explained next.
Use with BOOSTCNTR.

Next, the process BOOSTCNTR for optimally performing overboost control will be explained using FIG. 5B. This BOOSTCNTR is executed once before executing VNWGCONTROL, and various information necessary for overboost control is transferred. To explain the order in detail below, first, in step 214, it is checked whether the overboost control has ended. This is done by checking the results of the processing for terminating the overboost control performed in steps 236 to 239 described below. If it is determined that the control has been completed, the process proceeds to 241 and a process is performed to gradually lower the control target value (at the same time, the feed forward control amount during overboost control is also gradually reduced). If the control has not been completed, proceed to 215 and check the sudden acceleration determination flag FACCEL, which is set or reset in the sudden acceleration determination process described above. If it is '0', the process is terminated, and if it is '1', sudden acceleration is detected. has been determined, the program proceeds to 216 to check whether overboost control is permitted. This is because whether or not overboost control is performed differs depending on the engine and car model, so by storing this information in the memory ROM, for example, the same program can handle different specifications with or without overboost control. If overboost control is permitted, proceed to 217 and check whether the engine water temperature is below 100°C. 100
If the temperature is above ℃, overboost control is not performed because abnormal combustion is likely to occur. If the temperature is less than 100°C, the process proceeds to 218 and the feed forward correction start flag FP2WG on the WG side is set to '1'. 21
At step 9, it is checked whether the supercharging pressure exceeds 250 mmHg, and if it does not, the process is terminated. If it exceeds, proceed to 221. If it exceeds 250 mmHg for the first time, the process proceeds to 222, where the sudden acceleration judgment error prevention timer is started, and the feed forward correction start flag FP2VN on the VN side is set to '1'. The measurement time of this erroneous judgment prevention timer is checked in the sudden acceleration judgment error prevention process at 125 in Fig. 6 mentioned above, and if 3 seconds or more elapses from exceeding 250 mmHg to 320 mmHg, it is judged that sudden acceleration has occurred. Instead, the sudden acceleration determination flag FACCEL and the feed forward correction start flag FP2VN on the VN side are set to '1'. This is because when accelerating from 1/4 of the throttle valve opening in 2nd gear as shown in Fig. 17, the time τ used for sudden acceleration judgment is short and it is judged that there is a sudden acceleration, but the acceleration time is short and the acceleration is This is to prevent the supercharging pressure from entering overboost control at the end of the cycle and causing a sudden change in drivability. i.e. from 250mmHg
If the time T0 measured by the misjudgment prevention timer until the temperature reaches 320 mmHg is T0≧3, it is not considered to be a sudden acceleration. Next, use 221 for the second time onwards.
If it exceeds 250mmHg, proceed to 223 and boost pressure P2
Check whether it exceeds 345mmHg. If it exceeds 345mmHg, proceed to 225 and check whether it exceeds 345mmHg for the first time. 22 if you cross it for the first time
Step 6 starts a timing timer that measures the timing to increase the control target value, and ends the process. If the temperature exceeds 345 mmHg for the second time or later at 225, the process proceeds to 228, where it is checked whether the timing timer started at 226 has elapsed for a predetermined time (0.3 seconds), and if the time has elapsed, the process proceeds to 229. If the predetermined time has elapsed for the first time, the process proceeds to 230, where the overboost control amount corresponding to the engine water temperature is determined and the control target value is increased. That is, the optimum overboost control amount is provided by increasing a predetermined amount that decreases as the water temperature increases from the control target value of 425 mmHg during overboost control using a one-dimensional table according to the water temperature as shown in FIG. 11B.

Next, in the case of proceeding from 228 and 229 to 232 and onwards, from 232 onwards, the overboost control termination condition is checked. i.e. 232,2
In order to measure the time elapsed since the boost pressure exceeded 375mmHg in 234, it was 375mmHg for the first time in 234.
If the overboost control time is exceeded, the process proceeds to 235 and a timer for measuring overboost control time is activated. 234 in 2
If it exceeds 375 mmHg after the first time, the process proceeds to 236 and checks whether the control time measurement timer activated in 235 has exceeded a predetermined time. If it exceeds, the process advances to 239 and overboost control is ended. If the knocking level is not exceeded, the process proceeds to 237, where the knocking level is checked, and if it is greater, the overboost control is terminated to prevent knocking. If it is smaller, proceed to 238 and QS determines the air amount to cut off overboost control.
Check whether it is larger than QSBOOSTCUT, and if it is larger than 23 to prevent abnormal combustion.
Proceed to step 9 to end overboost control. In this way, the BOOSTCNTR process processes various information for overboost control.

Next, learning control for correcting deviations in feedforward control amounts on the VN and WG sides will be explained. First, on the VN side, the timing to calculate the learning amount is 110 in Figure 6, and from the VN side
This is the time to switch to feedback control on the working group side. The amount of learning is assumed to be I stored in VNLEARN at 134 in the same figure. This means that the steady-state deviation when the boost pressure is controlled on the VN side is added to the feedforward control amount in advance from the next control. The actual learning amount calculation will be explained based on FIG. 8. First, at 400, it is checked by FACCEL whether overboost control is being performed. Learning amount calculation is of course possible even in engines that do not perform overboost control, but
In an engine that performs overboost control, the control area on the VN side expands in the overboost control area, so the value of I is large. Since control accuracy can be improved by performing control using this large value, in this embodiment, the learning amount calculation is performed immediately after overboost control is performed. That is, in an acceleration situation where overboost control is not performed, the learning amount is not calculated. If overboost control is performed, proceed to 401 and step 1 in Figure 6.
Subtract the correction amount during overboost control (15% in terms of control duty) from the steady-state deviation VNLEARN found in 34, and check the subtracted result again.
Let it be VNLEARN. This is because the basic control duty is given to be optimal when not in overboost control. In 402 this
The sum of VNLEARN and LEARNVN is
Store in VNLEARNVALUE. Here
LEARNVN is the previous learning result, and this value
This is to ensure that the learning amount converges to the optimal value by adding LEARNVN and the current learning result VNLEARN. The latest learning results memorized in this way
VNLEARNVALUE is 121 in Figure 6, that is, the boost pressure is smaller than 320mmHg and QS is
By updating when the feedback control reset condition, which becomes smaller than QSWGAREA, is satisfied, it will be reflected in the next control. In this example, the update timing is set to be when two conditions are satisfied, but it may be at least as long as the condition that the pressure is lower than the predetermined boost pressure is satisfied.

Next, learning control on the WG side will be explained. The timing for calculating the learning amount is 11 in Figure 6.
2, and the feedback control is switched to the WG side after 1.2 seconds have elapsed. The amount of learning is 135.
Let it be the I minute stored in WGLEARN. This means that the steady-state deviation when the boost pressure is controlled on the WG side is added to the feedforward control amount in advance from the next control. The actual learning amount calculation will be explained based on FIG. 9. First, in 404, the measured value of the learning control start timer WGLEARNTIMER on the WG side, which was started at 109 in FIG. 6, that is, when the feedback control was switched to the WG side. Check whether the time is 1.2 seconds or more. If the time is less than 1.2 seconds, no learning amount calculation is performed. If more than 1.2 seconds have passed, proceed to 405.
Similarly to the VN side, the current steady deviation WGLEARN and the previous learning amount LEARNWG obtained at 135 in Figure 6
Add and store in WGLEARNVALUE. This latest learned value is updated at the same timing as on the VN side, that is, at 121 in FIG. thus
The feedforward control amount is corrected by calculating and updating the learning amount at the optimal timing on each side of the VN and WG.

Further, the various time measurement timers used in the above explanation are realized by a configuration in which the timers are incremented by one every fixed period of time (10 ms) by TIMER processing, as shown in FIG. 7B.

Figure 18 is due to the reluctance of VN due to changes over time.
FIG. 7 is a diagram showing the effects when this embodiment is applied when the VN causes a delay in control response. For comparison, the solid line indicates when the VN is operating normally.
The case where the VN causes a delay in control response is shown by a broken line, and the case of this embodiment is shown by a dashed-dotted line.

As shown in the figure, when acceleration is started from time T0, the supercharging pressure increases quickly and continues to increase even after reaching the target value. In this case, if the VN becomes difficult to open due to reluctance due to changes over time, even though the VN control duty is calculated in the direction of opening the VN due to feedback control based on the detection of the actual boost pressure. , VN is unable to follow the control duty, and as shown by V1, there is a control response delay with respect to V0 (indicating a normal case), and the boost pressure continues to rise even after reaching the target value. It deviates even further from the target value, and the excessive boost pressure continues for a long time until the VN fully opens. In such a case, the engine load becomes significantly large, and there is a possibility that engine durability may be impaired.

Conventionally, if this situation continues, a countermeasure is taken, for example, to perform a fuel cut at T1, but if a fuel cut is performed, the engine suddenly stops producing output at this point, resulting in sudden torque fluctuations. This causes discomfort to the driver.

On the other hand, in this embodiment, when the boost pressure reaches P1 (P1 > target value) due to detection of an abnormality in the boost pressure, the control duty on the WG side is increased to 25% as in S1.
If the boost pressure is still increased even after the reduction correction is made, the control duty is corrected to be reduced by 50% as shown in S2 when the boost pressure reaches P2 (P2>P1). In this case, the feedback control area is on the VN side, so
Normally, the operation of the WG is fixed to avoid inconveniences caused by overlapping control areas, but when there is an abnormality in boost pressure, the WG is auxiliary, regardless of the result of determining the feedback control area. It opens. In other words, the WG was kept in the fully closed position, but from the time T2 it opens wide according to the control duty on the WG side, and as a result, the boost pressure increases.
It quickly converged to the target value without even reaching P3 (P3>P2). That is, in this example, P1,
Since the increase in supercharging pressure was suppressed by the second stage correction at P2, no correction was made at the third stage P3.

In this way, in this embodiment, even if the VN becomes difficult to open due to stiffness due to changes over time, the boost pressure is adjusted so that the engine is not exposed to excessive boost pressure exceeding the target value. This improves engine durability, and at the same time, even if there is a delay in the VN's control response, the WG assists the VN and activates it before performing a fuel cut as a last resort. This prevents deterioration in drivability due to fuel cut.

(Effects of the Invention) As described above, according to the present invention, in a boost pressure control device that provides feedback control areas for the capacity variable means and the exhaust bypass valve and controls them independently in order to avoid redundant control, When an abnormality in boost pressure is detected and an abnormality in boost pressure is detected in the feedback control area on the capacity variable means side, the exhaust bypass valve is opened in the direction corresponding to the boost pressure at that time, regardless of the determination of the feedback control area. Since the control amount of the bypass valve is corrected, even if the displacement variable means has a delay in control response due to agitation due to aging, etc., and the boost pressure greatly exceeds the target value, such an overload can be avoided. In the event of an abnormality in boost pressure, the exhaust bypass valve opens to assist the capacity variable means and reduces the rotation of the exhaust turbine, suppressing an excessive rise in boost pressure, and as a result, preventing excessive load from being placed on the engine. Engine durability is improved. Also,
Since the supercharging pressure is controlled to the target value without immediately performing a fuel cut in response to an abnormality in the supercharging pressure, it is possible to prevent deterioration of drivability due to the fuel cut.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram for clearly showing the configuration of the present invention. Figure 2A is a flowchart that extracts the main parts from the flowchart of this embodiment, Figure 2B is a VN,
A characteristic diagram showing the basic control amount of the WG, FIG. 2C, is a diagram illustrating correction of the control amount on the WG side according to this embodiment. FIG. 3 is a schematic diagram of the mechanical structure of an embodiment of the present invention, and FIG. 4 is a sectional view of the scroll portion of the turbocharger. 5A, 5B, 6, 7A, 7B, 8, and 9 are flowcharts showing the operation of this embodiment. 10th
Figures A and 10B are characteristic diagrams showing the basic control duty of VN and WG, respectively, and Figure 10C is a table showing the relationship between QS and basic control duty. FIG. 11A is a characteristic diagram showing the control target value for QS,
FIG. 11B is a characteristic diagram showing the decrease in the overboost control amount with respect to the engine water temperature. 12th
The figure is a characteristic diagram showing the reduction rate of control duty with respect to boost pressure to explain overshoot countermeasures and failsafe. FIG. 13A is a diagram for explaining τ, and FIG. 13B is a characteristic diagram for explaining the relationship between engine rotational speed and τ at 100 mmHg. 1st
FIG. 4 is a characteristic diagram illustrating the relationship between ERROR and P (correction amount), and FIG. 15 is a characteristic diagram illustrating the respective control regions on the VN and WG sides with respect to QS.
FIG. 16 is a supercharging pressure characteristic diagram illustrating overshoot during sudden acceleration, and FIG. 17 is a supercharging pressure characteristic diagram illustrating prevention of errors in sudden acceleration determination. FIG. 18 is an explanatory diagram showing the effects of this embodiment. DESCRIPTION OF SYMBOLS 1... Operating state detection means, 2... Boost pressure detection means, 3... Deviation calculation means, 4... Feedback region discrimination means, 5... First control amount calculation means, 6
...Second control amount calculation means, 7...First control means, 8...Second control means, 9...Capacity variable means, 10...Exhaust bypass valve, 11...Supercharging pressure abnormality detection Means, 12... Controlled amount correction means, 13...
Engine, 21...Engine, 22...Intake pipe,
23... Intake manifold, 24... Exhaust manifold, 25... Exhaust pipe, 26... Bypass passage, 30... Crank angle sensor, 31... Air flow meter, 32... Throttle valve, 33... Supercharging pressure sensor , 35... Compressor, 36... Connection shaft,
37...Turbine, 38...Turbine chamber, 39...
...scroll, 40...introduction passage, 41...terminus, 45...movable tongue, 46...shaft, 47...arm, 48...rod, 50...actuator, 52...diaphragm, 54... Positive pressure chamber, 5
6... Connecting pipe, 57... Solenoid valve, 60... Exhaust bypass valve, 61... Arm, 62... Connecting member,
63... Rod, 70... Actuator, 72
... diaphragm, 74 ... positive pressure chamber, 76 ... connecting pipe, 77 ... solenoid valve, 80 ... control unit.

Claims (1)

[Claims] 1. A turbocharger that controls the supercharging pressure of an engine according to operating conditions using a capacity variable means that varies the capacity of an exhaust turbine of the turbocharger and an exhaust bypass valve that bypasses a part of the exhaust gas passing through the exhaust turbine from upstream to downstream. The boost pressure control device includes an operating state detection means for detecting the engine operating state, a supercharging pressure detection means for detecting the supercharging pressure, and a deviation calculation for calculating the deviation between the detected supercharging pressure and the target supercharging pressure. means, feedback control region determining means for determining an operating region in which feedback control is to be performed and selecting which of the variable capacity means or the exhaust bypass valve is to be subjected to feedback control; and an operating state including at least the integrated value of the deviation. first and second control amount calculation means for calculating the control amount of the exhaust turbine capacity variable means and the exhaust bypass valve according to the parameters represented;
First and second control means that drive and control the capacity variable means and the exhaust bypass valve according to the control amount calculated by the second control amount calculation means, and supercharging pressure abnormality detection that detects an abnormality in the supercharging pressure. and when an abnormality in the boost pressure is detected in the feedback control area on the side of the variable capacity means, the exhaust bypass valve is opened in a direction in which the exhaust bypass valve is opened in accordance with the boost pressure at that time, regardless of the determination result of the feedback control area. 1. A supercharging pressure control device for a turbocharger, comprising: control amount correction means for correcting a control amount of a valve.