JP6753754B2 - Engine system controller - Google Patents

Description

The present invention relates to a diesel engine, a diesel throttle, an EGR device, and an engine system control device applied to an engine system including a variable displacement turbocharger.

Conventionally, in an engine system using a diesel engine, various technologies have been used in order to reduce NOx and improve fuel efficiency. For example, in the engine system described in Patent Document 1, the flow velocity of the exhaust gas flowing into the exhaust gas recirculation (EGR: Exhaust Gas Recirculation) device or turbocharger that recirculates a part of the exhaust gas from the exhaust side to the intake side of the diesel engine is determined. It is equipped with a variable capacity turbocharger that can be changed.

Japanese Unexamined Patent Publication No. 2003-193875

In recent years, in order to further reduce NOx and improve fuel efficiency, it is required to control the engine system with higher accuracy.
An object of the present invention is to provide a control device for an engine system capable of controlling an engine system including a diesel engine with high accuracy.

The control device of the engine system that solves the above problems is a control device of the engine system that controls the controlled object of the engine system, and the engine system is a diesel engine, a diesel throttle, an EGR device, and a variable displacement turbocharger. The controlled object includes the diesel throttle, the EGR valve of the EGR device, and the variable nozzle of the variable displacement turbocharger, and includes a plurality of parameters relating to the state quantity of the diesel engine. A target value calculation unit that calculates a target value based on the acquisition value of the acquisition unit for each of the acquisition unit that acquires the value and the plurality of control parameters that are a part of the plurality of parameters, and the plurality of control parameters. For each, the target deviation calculation unit that calculates the target deviation based on the deviation between the target value and the acquisition value of the acquisition unit, and the deviation vector including the target deviation as a component and the gain matrix are multiplied to control the control target. It includes an instruction value calculation unit that calculates a control instruction value, and the control parameters are an intake EGR rate that is a ratio of EGR gas to a working gas that is sucked by the diesel engine and a suction supercharged by the variable displacement turbocharger. Includes boost pressure, which is the pressure in the passage through which air flows.

According to the above configuration, the control instruction value to be controlled is calculated by multiplying the deviation vector including the target deviations of the intake EGR rate and the boost pressure in the components and the gain matrix. A diesel throttle having an influence on both the intake EGR rate and the boost pressure, an EGR valve of the EGR device, and a variable nozzle of the variable displacement turbocharger are set as control targets. That is, according to the above configuration, the control instruction value for the control target is determined in consideration of the target deviations of the intake EGR rate and the boost pressure, and the influence of each of the intake EGR rate and the boost pressure on the control target. It is calculated. As a result, the control instruction value of the control target is calculated so that each of the control parameters becomes a target value while considering the influence of other control parameters. As a result, the engine system can be controlled with high accuracy.

The control device of the engine system puts the diesel engine into a steady state under the acquired value of the acquisition unit for the state parameter which is a parameter related to the state quantity of the diesel engine and whose value changes depending on the control of the controlled object. It further includes an ideal value calculation unit that calculates an ideal value that is a value at a certain time, and a state deviation calculation unit that calculates a state deviation that is a deviation between the ideal value and the acquisition value of the acquisition unit for the state parameter. It is preferable that the deviation vector includes the target deviation and the state deviation as components.

According to the above configuration, the control instruction value of the control target is calculated in consideration of the state deviation, which is the deviation of the state parameter, in addition to the target deviation of the control parameter. As a result, the control instruction value is set to a value such that the control parameter becomes a target value and the diesel engine approaches a steady state. As a result, the engine system can be controlled to a state closer to the steady state with the best exhaust gas performance.

In the control device of the engine system, it is preferable that the target deviation calculation unit calculates the integrated value of deviations for each of the control parameters as the target deviation.
According to the above configuration, since the integrated value of the deviation of each control parameter is set to the target deviation, it is possible to configure a servo system that causes each control parameter to reach the target value without deviation.

The control device of the engine system further includes a state quantity estimation unit that calculates estimated values of the plurality of parameters using a model, and the acquisition unit acquires the estimated values calculated by the state quantity estimation unit. Is preferable.

The acquisition unit acquires the estimated value calculated by the state quantity estimation unit using the model as the value of the parameter related to the engine state quantity. Therefore, for example, the engine system can be controlled with higher accuracy than when the acquisition unit acquires the observed value of the sensor as the value of the parameter related to the state quantity of the engine.

The control device of the engine system includes an observation unit for observing parameters related to the state quantity of the diesel engine, and the state quantity estimation unit inputs a control instruction value to the control target as a component of an input vector of the observation unit. It is preferable that the estimated value is calculated by the Kalman filter theory using the observed value as a component of the observation vector. The parameters observed by the observation unit are the boost pressure, the intake air amount which is the amount of air taken in by the diesel engine, the intake manifold temperature which is the temperature in the intake manifold, and the state amount on the intake side with respect to these diesel engines. It is also preferable to include the turbine rotation speed, which is the rotation speed of the turbine in the variable displacement turbocharger, that is, the state amount indicating the balance between the state on the intake side and the state on the exhaust side with respect to the diesel engine.

According to the above configuration, the estimated value calculated by the state quantity estimation unit can be made closer to the value indicating the current state quantity of the engine. Then, the engine system can be controlled with higher accuracy by calculating the control instruction value of each control target using the estimated value.

The figure which shows the schematic structure of the engine system which carries out one Embodiment of the control device of an engine system. A block diagram showing an example of Kalman filter theory. The functional block diagram which shows an example of the control device of an engine system. The functional block diagram which shows an example of the control calculation part. The functional block diagram which shows an example of the state quantity estimation part.

An embodiment of the control device of the engine system will be described with reference to FIGS. 1 to 5. First, the overall configuration of the engine system will be described with reference to FIG.
As shown in FIG. 1, the engine system includes a diesel engine 10 (hereinafter, simply referred to as an engine 10) that uses light oil as fuel. Six cylinders 12 are formed in the cylinder block 11 of the engine 10. In each cylinder 12, fuel is injected from the injector 13 with respect to the sucked working gas, and the mixture of the working gas and the fuel is burned. The crankshaft 10a of the engine 10 is driven by the combustion of the air-fuel mixture in each cylinder 12 in a predetermined order.

An intake manifold 14 for supplying working gas to each cylinder 12 and an exhaust manifold 15 into which exhaust gas from each cylinder 12 flows are connected to the cylinder block 11.

An air cleaner (not shown), a compressor 18 of a turbocharger 17, and an intercooler 19 are attached to the intake passage 16 connected to the intake manifold 14 in this order from the upstream side. Further, the intake passage 16 has a diesel throttle 20 (a diesel throttle 20) whose flow path cross-sectional area of the intake passage 16 can be changed on the downstream side of the intercooler 19 and on the upstream side of the connection portion with the EGR passage 25 described later. Hereinafter, the throttle 20) is simply attached.

An exhaust passage 21 is connected to the exhaust manifold 15. A turbine 23 connected to the compressor 18 via a connecting shaft 22 is attached to the exhaust passage 21. Further, the exhaust manifold 15 is connected to the EGR passage 25 of the EGR device 24 which is connected to the intake passage 16 and introduces a part of the exhaust gas into the intake passage 16 as EGR (Exhaust Gas Recirculation) gas. An EGR cooler 26 is attached to the EGR passage 25. The portion of the EGR passage 25 that connects the EGR cooler 26 and the exhaust manifold 15 is composed of a first EGR passage 25A and a second EGR passage 25B. An EGR valve 27 capable of changing the flow path cross-sectional area of the EGR passage 25 is attached to the EGR passage 25 on the downstream side of the EGR cooler 26. A mixed gas of exhaust gas and intake air is supplied to the cylinder 12 as a working gas when the EGR valve 27 is in the open state, and intake air is supplied as a working gas when the EGR valve 27 is in the closed state. ..

The turbocharger 17 is a variable capacity turbocharger (VNT: Variable Nozzle Turbo) in which a variable nozzle 28 is arranged on a turbine 23. The opening degree of the variable nozzle 28 is changed by driving an actuator 29 provided with a stepping motor, and the cross-sectional area of the flow path of the exhaust gas flowing into the turbine 23 is changed.

The engine system includes an intake air amount sensor 31, a boost pressure sensor 32, an in-mani temperature sensor 33, an engine speed sensor 34, a turbine speed sensor 35, and an accelerator opening sensor 36.

The intake air amount sensor 31 observes the intake air amount Ga, which is the mass flow rate of the intake air, upstream of the compressor 18. The boost pressure sensor 32 is a boost pressure Pb which is the pressure of the intake air supercharged by the turbocharger 17 downstream of the throttle 20 and upstream of the connection portion between the intake passage 16 and the EGR passage 25. To observe. The intake manifold temperature sensor 33 observes the intake manifold temperature Tim, which is the temperature inside the intake manifold 14, which is the temperature of the working gas sucked by the engine 10. The engine speed sensor 34 observes the engine speed Ne, which is the speed of the crankshaft 10a. The turbine rotation speed sensor 35 observes the turbine rotation speed Nt, which is the rotation speed of the connecting shaft 22 of the turbocharger 17. The accelerator opening sensor 36 observes the accelerator opening ACC, which is the amount of depression of the accelerator pedal operated by the driver. The various sensors 31 to 36 can function as an observation unit for observing parameters related to the state quantity (operating state) of the engine 10, and constitute a sensor group 40 (see FIG. 3). The signals output from the various sensors 31 to 36 are input to the ECU 50, which is a control device that controls the engine system in an integrated manner.

The ECU 50 is mainly composed of a microcomputer having a processor, a memory, an input / output interface, and the like. The ECU 50 acquires the observed values of various sensors 31 to 36, and the control target such as the throttle 20, the EGR valve 27, and the variable nozzle 28 is based on the estimated value of the state quantity obtained by the Kalman filter theory using the acquired observed values. 60 (see FIG. 3) is controlled.

Before explaining the configuration of the ECU 50, the Kalman filter theory will be described with reference to FIG.
As shown in FIG. 2, in the Kalman filter theory, the state vector whose component is the state quantity related to the controlled object is x (k), and the input vector whose component is the control instruction value for the controlled object is u (k). In this case, the equation of state is set as in Eq. (1). The input vector u (k) is, for example, a state vector x (k), an ideal vector xss (k) having an ideal state quantity value of the engine under various operating conditions as a component, and a state related to a controlled object. It is calculated by feedback control based on the target vector r (k) having the target value of the quantity as a component. Further, when the observation vector whose component is the observation value of the preset sensor is y (k), the observation equation is set as in the equation (2).

Note that k is an integer of 0 or more and is a discrete time. Further, A is a system matrix, B is an input matrix, and C is an observation matrix, and these matrices are preset based on the results of first-principles modeling or simulation such as an equation of motion. Further, w (k) is system noise and v (k) is observation noise. These system noises w (k) and observed noises v (k) are treated as white noises that are independent of each other according to a normal distribution. Therefore, the average values of the system noise w (k) and the observed noise v (k) are set to 0. The dispersibility sigma _v ² of variance sigma _w ² and the observation noise v (k) of the system noise w (k) is set based on the results of simulations that were performed previously.

In the Kalman filter theory, a prediction step for calculating a predicted estimated value x ^ ⁻ (k), which is an estimated value of the state vector x (k) at time k, based on information up to time k-1, and information up to time k. The filtering step of calculating the filtering estimated value x ^ (k), which is the estimated value of the state vector x (k) at the time k, is performed based on the above. Since the system noise w (k) and the observed noise v (k) are white noises that follow a normal distribution independent of each other, the predicted estimated value x ^ ⁻ (k) is expressed by Eq. (3), and the filtered estimated value x. ^ (K) is represented by the equation (4). The initial value x (0) of the state vector x (k) is set based on the result of a simulation performed in advance. In equation (4), G (k) is Kalman gain. The Kalman gain G (k) may be a constant value set based on the result of a simulation or the like performed in advance, or may be updated every time k.

When the Kalman gain G (k) is updated, in the prediction step, as shown in equation (5), the time is based on the posterior error covariance matrix P (k-1) at time k-1 and the system matrix A. The prior error covariance matrix P ⁻ (k) at k is calculated. On the other hand, in the filtering step, the Kalman gain G (k) is updated every time k as shown in the equation (6). Further, in the filtering step, as shown in the equation (7), the posterior error covariance matrix P (k) based on the pre-error covariance matrix P ⁻ (k) at time k, the updated Kalman gain G (k), and the like. k) is calculated.

The initial value P ⁻ (0) of the prior error covariance matrix P ⁻ (k) is set based on the result of a simulation or the like performed in advance. In Formula (5) ~ (7), I is a unit matrix, ^{A T} is the transposed matrix of the system matrix A, a ^{C T} is a transposed matrix of the observation matrix C. Q is a covariance matrix for the observed noise v (k) and is set based on the result of a simulation performed in advance.

That is, in the Kalman filter theory, first, the predicted estimated value x ^ ⁻ (k), which is the estimated value of the state vector x (k) at the time k, is calculated based on the information up to the time k-1. Then, when the observation vector y (k) is input at time k, the observation prediction value (= Cx ^ ⁻ (k)), which is the prediction value of the observation vector y (k), is subtracted from the observation vector y (k). Then, the observation prediction error (= y (k) -Cx ^ ^- (k)) is calculated.

Next, the Kalman gain G (k) is multiplied by the observed prediction error, and the correction term (= G (k) × (y (k) −Cx ^ ⁻ (k) for the predicted estimated value x ^ ⁻ (k). ))) Is calculated. Then, the correction term is added to the predicted estimated value x ^ ⁻ (k), and the filtering estimated value x ^ (k), which is the estimated value of the state vector x (k) based on the information up to the time k, is calculated.

Next, the filtering estimate x ^ (k) multiplied by the system matrix A (= Ax ^ (k)) and the input vector u (k) multiplied by the input matrix B (=). By adding Bu (k)), the predicted estimated value x ^ ⁻ (k + 1) at time k + 1 is calculated. The predicted estimated value x ^ ^- (k + 1) is set as the predicted estimated value x ^ ^- (k) at the next time k by delaying the time by one, and the operation of the input vector u (k) or the observed predicted value (= Cx) It is used for the calculation of ^ ^- (k)).

The configuration of the ECU 50 will be described in detail with reference to FIGS. 3 to 5.
As shown in FIG. 3, the ECU 50 includes a control calculation unit 51 that calculates an input vector u (k) having a control instruction value for a control target 60 such as a throttle 20, an EGR valve 27, and a variable nozzle 28 as a component. Further, the ECU 50 applies the Kalman filter theory using the observed values of various sensors to obtain a predicted estimated value x ^ ⁻ (k) of the state vector x (k) having a parameter indicating the state quantity of the engine 10 as a component. A state quantity estimation unit 52 for calculation is provided.

As an acquisition unit, the control calculation unit 51 receives a predicted estimated value x ^ ⁻ (k) and an observation vector y (k) whose components are observation values of a plurality of sensors preset from the sensor group 40. get. Based on the predicted estimated value x ^ ⁻ (k) and the observation vector y (k), the control calculation unit 51 comprises an ideal vector xss (k) having an ideal value of a state parameter as a component and a target value of a control parameter as a component. The target vector r (k) to be calculated is calculated. The control calculation unit 51 calculates the input vector u (k) by state feedback control based on the ideal vector xss (k), the target vector r (k), and the predicted estimated value x ^ ⁻ (k). The control calculation unit 51 outputs the input vector u (k) to the control target 60 and the state quantity estimation unit 52.

The state quantity estimation unit 52 uses the observation vector y (k), the input vector u (k) input from the control calculation unit 51, and the Kalman filter theory, and constitutes a plurality of parameters indicating the state quantity of the engine 10. The predicted estimated value x ^ ⁻ (k) is calculated. The state quantity estimation unit 52 outputs the calculated predicted estimated value x ^ ⁻ (k) to the control calculation unit 51.

In the ECU 50, the observation parameters that are components of the observation vector y (k) include the intake air amount Ga, the boost pressure Pb, the intake manifold temperature Tim, the turbine speed Nt, the engine speed Ne, and the accelerator opening ACC. It has been.

The components of the state vector x (k) include an intake EGR rate ηi and a boost pressure Pb as control parameters that are components of the target vector r (k). Further, the components of the state vector x (k) include the intake air amount Ga, the intake manifold temperature Tim, and the turbine rotation speed Nt. The intake EGR rate ηi indicates the ratio (= Gr / Gwg) of the EGR gas amount Gr (weight) to the working gas amount Gwg (weight). Further, the components of the state vector x (k) include an intake density ρim, which is the density of the working gas in the intake manifold 14, an exhaust pressure Pem, which is the pressure of the exhaust gas in the exhaust manifold 15, and the density of the exhaust gas in the exhaust manifold 15. The exhaust density ρem, which is, and the exhaust EGR ratio ηe, which is the ratio (= Gr / Gex) of the EGR gas amount Gr (weight) to the exhaust gas amount Gex (weight) discharged by the engine 10, and the like are included.

The components of the input vector u (k) include a throttle opening Ath which is a control instruction value of the throttle 20, an EGR valve opening Ar which is a control instruction value of the EGR valve 27, and a nozzle which is a control instruction value of the variable nozzle 28. The opening degree Anz is included.

The components of the state vector x (k) described above include state parameters. In the present embodiment, the state parameters are boost pressure Pb, intake density ρim, intake EGR rate ηi, exhaust pressure Pem, exhaust density ρem, exhaust EGR rate ηe, and the like. As described above, the state parameter is a parameter whose value changes when the operating state of the engine 10 changes through the control of the controlled object 60 such as the throttle 20, the EGR valve 27, and the variable nozzle 28. Further, the control parameter is a parameter for guiding the engine 10 to a target operating state by controlling the control target 60.

As shown in FIG. 4, the control calculation unit 51 includes a state quantity setting unit 51a, a state deviation calculation unit 51b, a target deviation calculation unit 51c, and an indicated value calculation unit 51f.
The state quantity setting unit 51a sets an ideal vector xss (k) having an ideal value of each state parameter as a component as an ideal value calculation unit. The ideal value is a value that each state parameter should take when the engine 10 is in a steady state with the best exhaust gas performance under each operating condition.

For example, the state quantity setting unit 51a calculates the ideal vector xss (k) based on the engine speed Ne and the fuel injection amount Gf which is a control instruction value to the injector 13. The fuel injection amount Gf is calculated based on the required torque Tr based on the engine speed Ne and the accelerator opening ACC, and may be input to the injector 13 separately from the input vector u (k), or the input vector u. It may be input to the injector 13 as one of (k). When the input vector u (k) includes the fuel injection amount Gf, the fuel injection amount Gf used by the state amount setting unit 51a at the time of calculation is the previous control instruction value. The state quantity setting unit 51a holds an ideal value map for each state parameter in which ideal values corresponding to the engine speed Ne and the fuel injection amount Gf are defined based on the results of experiments and simulations performed in advance. Then, the state amount setting unit 51a calculates the ideal vector xss (k) by selecting the ideal value corresponding to the engine speed Ne and the fuel injection amount Gf from the ideal value map of each state parameter.

The state quantity setting unit 51a sets a target vector r (k) having a target value of each control parameter as a component as a target value calculation unit. The target value is a value that each control parameter should take in order to shift the engine 10 to the target operating state.

For example, the state quantity setting unit 51a calculates the target vector r (k) based on the engine speed Ne and the fuel injection amount Gf. The state quantity setting unit 51a calculates the target vector r (k) by calculating the target value for each control parameter. The state quantity setting unit 51a holds a target value map for each control parameter in which target values corresponding to the engine speed Ne and the fuel injection amount Gf are defined based on the results of experiments and simulations performed in advance. Then, the state amount setting unit 51a calculates the target vector r (k) by selecting the target value corresponding to the engine speed Ne and the fuel injection amount Gf from the target value map of each control parameter.

The state deviation calculation unit 51b is a subtractor, and the state deviation vector xe (k) having the deviation of the state parameter as a component by subtracting the predicted estimated value x ^ ⁻ (k) from the ideal vector xss (k). Is calculated.

The target deviation calculation unit 51c has a subtractor 51d and an integration circuit 51e. The subtractor 51d calculates the control deviation vector re (k) having the deviation of each control parameter as a component by subtracting the predicted estimated value x ^ ⁻ (k) from the target vector r (k). By integrating the control deviation vector re (k), the integration circuit 51e calculates the target deviation vector xi (k) having the integrated value of the deviation of the control parameters as a component.

The indicated value calculation unit 51f has a deviation vector e (k) (= [xe (k) xi (k)] composed of a gain matrix K, a state deviation vector xi (k), and a target deviation vector xi (k). By multiplying with ^T ), the throttle opening Ath, the EGR valve opening Ar, and the nozzle opening Anz constituting the input vector u (k) are calculated. This gain matrix K is set based on the results of experiments performed in advance and the results of simulations, and is set based on, for example, the least squares estimation method.

That is, in the ECU 50, the control calculation unit 51 calculates the input vector u (k) based on the deviation of the state parameter and the deviation of the control parameter, and the calculated input vector u (k) is the control target 60 and Output to the state quantity estimation unit 52.

As shown in FIG. 5, the state quantity estimation unit 52 calculates the engine model 52a that handles the average value of one cycle of the engine 10, the error calculator 52b that calculates the observation prediction error, and the correction term by the Kalman filter theory. A correction term calculation unit 52c is provided.

The engine model 52a is based on the input vector u (k), the system matrix A, the input matrix B, the predicted estimated value x ^ ⁻ (k), and the correction term calculated by the correction term calculation unit 52c, and the predicted estimated value x ^. ^- (K + 1) is calculated. Engine model 52a is predicted estimate ^x ^ - by delaying one time (k + 1), the predicted estimates ^x ^ - handled as (k) - a (k + 1) predicted estimate ^{x ^} at the next time k .. Further, the engine model 52a calculates the observed predicted value (= Cx ^ ⁻ (k)) based on the predicted estimated value x ^ ⁻ (k) and the observation matrix C.

The error calculator 52b calculates the observation prediction error (= y (k) -Cx ^ ^- (k)) by subtracting the observation prediction value from the observation vector y (k). The correction term calculation unit 52c calculates the correction term for the predicted estimated value x ^ ⁻ (k) by multiplying the observation prediction error by the Kalman gain G (k), and the calculated correction term is applied to the engine model 52a. Output.

The operation of the ECU 50 having the above configuration will be described.
The ECU 50 is a state in which the state parameter is composed of a predicted estimated value x ^ ⁻ (k), an ideal vector xss (k), and a deviation between the ideal vector xss (k) and the predicted estimated value x ^ ⁻ (k). Deviation vector xe (k), these are calculated. Further, the ECU 50 uses the integrated value of the predicted estimated value x ^ ⁻ (k), the target vector r (k), and the deviation between the target vector r (k) and the predicted estimated value x ^ ⁻ (k) for the control parameters. The target deviation vector xi (k) to be constructed is calculated. Then, the ECU 50 calculates the input vector u (k) by multiplying the gain matrix K by the deviation vector e (k) whose constituent elements are the state deviation vector xe (k) and the target deviation vector xi (k). To do.

As described above, in the ECU 50, the target vector r (k) is configured by the intake EGR rate ηi and the boost pressure Pb, which are control parameters closely related to the throttle opening Ath, the EGR valve opening Ar, and the nozzle opening Anz. .. Then, the input vector u (k) is calculated by multiplying the gain matrix K with the deviation vector e (k) including the integrated deviation which is the integrated value of the deviations of each control parameter. Therefore, each of the arithmetic expressions for calculating the throttle opening Ath, the EGR valve opening Ar, and the nozzle opening Anz constituting the input vector u (k) includes the term including the integrated deviation of the intake EGR rate ηi and the boost pressure Pb. It has a term including an integrated deviation. That is, each of the throttle opening Ath, the EGR valve opening Ar, and the nozzle opening Anz takes into consideration the integrated deviation of the intake EGR rate ηi and the integrated deviation of the boost pressure Pb, and these intake EGR rate ηi and boost pressure. Both Pb are set to values close to the target value.

As a result, for example, the feedback control of the throttle opening Ath based on the boost pressure Pb and the feedback control of the EGR valve opening Ar and the nozzle opening Anz based on the intake EGR rate ηi are performed with higher accuracy than when they are performed independently of each other. The engine system can be controlled under.

Further, the deviation vector e (k) includes the state deviation vector xe (k) having the deviation of each state parameter as a component. Therefore, the arithmetic expression for obtaining each of the opening degree Ath, Ar, and Anz has a term including the deviation of the state parameter for each state parameter. That is, each of the opening degree Ath, Ar, and Anz is set to a value that brings the value of each state parameter closer to the ideal value in consideration of the deviation of each state parameter. That is, each opening degree Ath, Ar, and Anz is set to a value in which both the intake EGR rate ηi and the boost pressure Pb, which are control parameters, approach the target value, and each state parameter approaches the ideal value. ..

According to the ECU 50 of the above embodiment, the effects listed below can be obtained.
(1) Each of the opening degree Ath, Ar, and Anz is set to a value that brings both the intake EGR rate ηi and the boost pressure Pb, which are control parameters, close to the target value. This makes it possible to control the engine system with high accuracy.

(2) Further, each of the opening degree Ath, Ar, and Anz is set to a value that brings both the intake EGR rate ηi and the boost pressure Pb close to the target value and each of the state parameters close to the ideal value. As a result, the engine 10 can be operated in a state closer to the steady state. As a result, the performance of the exhaust gas can be improved.

(3) The target deviation calculation unit 51c calculates the target deviation vector xi (k) by integrating the control deviation vector re (k). That is, the target deviation calculation unit 51c sets the integrated value of the deviation for each of the intake EGR rate ηi and the boost pressure Pb as the target deviation. This makes it possible to configure a servo system that allows each control parameter to reach the target value without deviation.

(4) The state quantity estimation unit 52 calculates a correction term based on the Kalman filter theory based on the predicted estimated value x ^ ⁻ (k) of the time k calculated by the engine model 52a, and uses the correction term at the time k + 1. Calculate the predicted estimated value x ^ ⁻ (k + 1). By calculating the predicted estimated value x ^ ^- (k) using the Kalman filter theory in this way, the accuracy of the predicted estimated value x ^ ^- (k) is improved, so that the engine system is controlled with higher accuracy. can do.

(5) The turbine speed Nt is one index showing the balance between the state on the intake side and the state on the exhaust side. Therefore, by applying the turbine rotation speed Nt as one of the observed values to the Kalman filter theory in addition to the intake air amount Ga, boost pressure Pb, and intake manifold temperature Tim, which indicate the state of the intake side of the engine 10, the predicted estimated value The accuracy of x ^ ^- (k) is further improved.

The above embodiment can be modified as appropriate and implemented as follows.
-The method by which the state quantity estimation unit 52 calculates the estimated value of each parameter is not limited to the method using the Kalman filter theory. For example, a method in which an engine model 52a is constructed based on the results of simulations and experiments performed in advance, and each parameter is calculated by applying an observation vector y (k) and an input vector u (k) to the engine model 52a. It may be. Further, for example, it is a method in which each parameter is calculated by applying the observation vector y (k) and the input vector u (k) to a statistical model constructed based on the results of simulations and experiments performed in advance. You may.

The ECU 50 may have a configuration in which the state quantity estimation unit 52 is omitted. In this case, the state quantity setting unit 51a calculates the current value and the target value of each of the control parameter and the state parameter based on the components of the observation vector y (k). Even with such a configuration, an effect similar to the above (1) to (3) can be obtained.

For example, the intake EGR rate ηi, which is a control parameter, can be configured as follows. That is, the state amount setting unit 51a is based on a state equation using the intake manifold pressure Pim, the intake manifold temperature Tim, the engine rotation speed Ne, the exhaust amount D of the engine 10, and the like, which are the pressures of the working gas in the intake manifold 14. Gwg is calculated, and the EGR gas amount Gr is calculated by subtracting the intake air amount Ga from the working gas amount Gwg. Then, the state amount setting unit 51a calculates the calculated value of the intake EGR rate ηi based on the working gas amount Gwg and the EGR gas amount Gr. Further, the state amount setting unit 51a calculates a target value of the intake EGR rate ηi based on the engine speed Ne and the accelerator opening degree ACC. Then, the subtractor 51d calculates the deviation between the target value and the current value of the intake EGR rate ηi. The EGR gas amount Gr may be observed by the sensor, and the current value of the intake EGR rate ηi (= Gr / (Ga + Gr)) may be obtained based on the observed value itself.

In the ECU 50, the target deviation calculation unit 51c may have a configuration in which the integrator circuit 51e is omitted. Even with such a configuration, the control deviation vector re (k) can be calculated as the target deviation vector xi (k). Further, the target deviation calculation unit 51c may have a differentiating circuit for differentiating the control deviation vector re (k). In such a case, the target deviation vector xi (k) has a differential value of the deviation as a component.

-In the ECU 50, the method of obtaining the ideal value of the state parameter is not limited to the method using the engine speed Ne, the fuel injection amount Gf, and the ideal value map. For example, the ideal value may be calculated by substituting the value of the component selected from the observation vector y (k) or the predicted estimated value x ^ ⁻ (k) into a predetermined calculation formula.

-In the ECU 50, the method of obtaining the target value of the control parameter is not limited to the method using the engine speed Ne, the fuel injection amount Gf, and the target value map. For example, the target value may be calculated by substituting the value of the component selected from the observation vector y (k) or the predicted estimated value x ^ ⁻ (k) into a predetermined calculation formula.

In the ECU 50, the control calculation unit 51 may calculate the control instruction value by the state feedback control based on the gain matrix K and the target deviation vector xi (k). That is, the control calculation unit 51 may perform state feedback control using only the control parameters without using the state parameters.

The boost pressure Pb may be any pressure in the passage through which the intake air supercharged by the turbocharger 17 flows, and the intake manifold pressure Pim may be used as the boost pressure Pb.
The control parameters are not limited to the intake EGR rate ηi and the boost pressure Pb, but include the parameters described as the above-mentioned state parameters such as the intake air amount Ga in addition to the intake EGR rate ηi and the boost pressure Pb. It may be.

-The state parameter may be any parameter whose value changes depending on the control of the control target 60. Therefore, it may be a part of the parameters described as the state parameters in the above embodiment, or may include other parameters in addition to the state parameters described in the above embodiment.

10 ... Diesel engine, 10a ... Crank shaft, 11 ... Cylinder block, 12 ... Cylinder, 13 ... Injector, 14 ... Intake manifold, 15 ... Exhaust manifold, 16 ... Intake passage, 17 ... Turbocharger, 18 ... Compressor, 19 ... Intercooler , 20 ... diesel throttle, 21 ... exhaust passage, 22 ... connecting shaft, 23 ... turbine, 24 ... EGR device, 25 ... EGR passage, 26 ... EGR cooler, 27 ... EGR valve, 28 ... variable nozzle, 29 ... actuator, 31 ... intake air volume sensor, 32 ... boost pressure sensor, 33 ... intake manifold temperature sensor, 34 ... engine rotation speed sensor, 35 ... turbine rotation speed sensor, 36 ... accelerator opening sensor, 40 ... sensor group, 50 ... ECU, 51 ... Control calculation unit, 51a ... state quantity setting unit, 51b ... state deviation calculation unit, 51c ... target deviation calculation unit, 51d ... subtractor, 51e ... integration circuit, 51f ... instruction value calculation unit, 52 ... state quantity estimation unit, 52a ... Engine model, 52b ... Error calculator, 52c ... Correction term calculation unit, 60 ... Control target.

Claims (4)

An engine system control device that controls the control target of the engine system.
The engine system includes a diesel engine, a diesel throttle, an EGR device, and a variable displacement turbocharger.
The controlled object includes the diesel throttle, the EGR valve of the EGR device, and the variable nozzle of the variable capacity turbocharger.
An acquisition unit that acquires the values of a plurality of parameters related to the state quantity of the diesel engine, and
For each of the plurality of control parameters that are a part of the plurality of parameters, a target value calculation unit that calculates a target value based on the acquisition value of the acquisition unit, and a target value calculation unit.
A target deviation calculation unit that calculates a target deviation based on a deviation between the target value and the acquisition value of the acquisition unit for each of the plurality of control parameters.
For the state parameter that is a parameter related to the state amount of the diesel engine and whose value changes depending on the control of the controlled object, an ideal value that is a value when the diesel engine is in a steady state under the acquired value of the acquisition unit is set. The ideal value calculation unit to calculate and
A state deviation calculation unit that calculates a state deviation that is a deviation between the ideal value and the acquisition value of the acquisition unit for the state parameter.
And a command value calculator for calculating a control command value for the controlled object by multiplying the deviation vector and the gain matrix including the components of the target deviation and the state deviation,
The control parameters include an intake EGR rate, which is the ratio of EGR gas to the working gas sucked by the diesel engine, and a boost pressure, which is the pressure in the passage through which the intake air supercharged by the variable displacement turbocharger flows. Engine system controller. The control device for an engine system according to claim 1 , wherein the target deviation calculation unit calculates an integrated value of deviations for each of the control parameters as the target deviation. The engine according to claim 1 or 2 , further comprising a state quantity estimation unit that calculates estimated values of the plurality of parameters using a model, and the acquisition unit acquires the estimated values calculated by the state quantity estimation unit. System control device. It is equipped with an observation unit that observes parameters related to the state quantity of the diesel engine.
The state quantity estimation unit calculates the estimated value by the Kalman filter theory using the control instruction value for the controlled object as a component of the input vector and the observed value of the observation unit as a component of the observation vector.
The parameters observed by the observation unit are the boost pressure, the intake air amount which is the amount of air taken in by the diesel engine, the intake manifold temperature which is the temperature in the intake manifold, and the rotation of the turbine in the variable displacement turbocharger. The control device for an engine system according to claim 3 , which includes a turbine rotation speed which is a number.