JP2655145B2 - Control device for internal combustion engine - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an internal combustion engine such as an automobile engine, and more particularly to a control device suitable for improving drivability, riding comfort, and the like.

[Conventional technology]

2. Description of the Related Art In the field of the automobile industry, there has been a demand for improvements in fuel economy, exhaust purification, and the like, as social interest in environmental conservation and energy resource saving by preventing air pollution has increased. Further, in recent years, improvement in performance as a man-machine system such as drivability and ride comfort has been strongly demanded. Therefore, the development of a control device that comprehensively controls the operation state of a gasoline engine for a vehicle and realizes required performance improvement has been widely promoted.

For example, using a microcomputer,
Cooling water temperature sensor, O ₂ to detect the presence or absence of oxygen in exhaust gas
It takes in signals from sensors to extract various data representing the operating state of the engine, such as sensors, and performs various controls such as air supply, fuel supply, ignition timing, idle speed, and exhaust gas recirculation to always optimize An electronic engine control device (hereinafter referred to as EEC) capable of obtaining an operating state of an engine has been used.

An example of a system in which such an EEC is applied to a fuel injection type internal combustion engine is proposed in Japanese Patent Application Laid-Open No. 55-134721, and this conventional example will be described with reference to FIGS.

FIG. 3 is a partial cross-sectional view schematically showing the entire control system of the engine. In the drawing, intake air is supplied to a cylinder 98 through an air cleaner 92, a throttle chamber 94, and an intake pipe 96. The gas burned in the cylinder 98 is discharged from the cylinder 98 through the exhaust pipe 70 to the atmosphere.

The throttle chamber 94 is provided with an injector 72 for injecting fuel, and the fuel ejected from the injector 72 is atomized in an air passage of the throttle chamber 94 and mixed with intake air to form an air-fuel mixture. Then, the air-fuel mixture passes through the intake pipe 96 and is supplied to the combustion chamber of the cylinder 98 by opening the intake valve 20.

A throttle valve 74 is provided near the outlet of the injector 72. The throttle valve 74 is configured to be mechanically linked with the accelerator pedal, and is driven by the driver.

An air passage 22 is provided upstream of the throttle valve 74 of the throttle chamber 94, and a hot-wire type air flow meter comprising an electric heating element, that is, a flow sensor 24 is provided in the air passage 22.
An electric signal AF that changes according to the air flow velocity is extracted.
Since the flow rate sensor 24 composed of the heating element (hot wire) is provided in the bypass air passage 22, the cylinder 8
In addition to being protected from high-temperature gas generated at the time of backfire from the air, it is also protected from being contaminated by dust and the like in the intake air. The outlet of the bypass air passage 22 is opened near the narrowest portion of the venturi, and the inlet is opened on the upstream side of the venturi.

Pressurized fuel is always supplied to the injector 72 from the fuel tank 30 via the fuel pump 32, and when an injection signal is given to the injector 72 from the control circuit 60,
Fuel is injected from the injector 72 into the suction pipe 6.

The air-fuel mixture sucked from the intake valve 20 is compressed by the piston 50 and burns by sparks from a spark plug (not shown), and this combustion is converted into operating energy.
The cylinder 98 is cooled by the cooling water 54. The temperature of the cooling water is measured by a water temperature sensor 56, and the measured value TW is used as the engine temperature.

An O ₂ sensor 142 is provided in the exhaust pipe 70, and O ₂
The presence or absence of ₂ is measured and a measurement value λ is output.

Further, a crankshaft (not shown) is provided at every reference crank angle and at a constant angle (for example, 0.5
A crank angle sensor that outputs a reference angle signal and a position signal for each degree is provided.

The output of the crank angle sensor, the output signal TW, an electrical signal AT from the output signal λ and the heating element 24 of the O ₂ sensor 142 is composed of a microcomputer control circuit of the water temperature sensor 56
The injector 72 and the ignition coil are driven by the output of the control circuit 60.

Further, the throttle chamber 54 is provided with a bypass 26 that straddles the throttle valve 74 and communicates with the intake pipe 96. The bypass 26 is provided with a bypass valve 61 that is opened and closed.

The bypass valve 61 is exposed to the bypass 26 provided around the throttle valve 74, is opened and closed by a pulse current, and changes the cross-sectional area of the bypass 26 according to the lift amount. The driving unit is driven and controlled by the output of 60. That is, an open / close cycle signal is generated by the control circuit 60 for controlling the drive unit, and the drive unit adjusts the lift amount of the bypass valve 61 based on the open / close cycle signal.

The exhaust gas recirculation (hereinafter referred to as EGR) control valve 90 is connected to the exhaust pipe 70
Control the passage between the exhaust pipe 70 and the suction pipe 96.
The EGR amount to 96 is controlled.

Therefore, the injector 72 shown in FIG. 3 is controlled to control the air-fuel ratio (A / F) and the fuel increase control, and the bypass valve 61 and the injector 72 control the engine speed at idle (ISC). And the EGR amount can be controlled.

FIG. 4 is an overall configuration diagram of a control circuit 60 using a microcomputer, and includes a central processing unit 102 (hereinafter C).
It is composed of a read only memory 105 (hereinafter, referred to as ROM), a random access memory 106 (hereinafter, referred to as RAM), and an input / output circuit 108. CPU10 above
2 calculates the input data from the input / output circuit 108 according to various programs stored in the ROM 104, and returns the calculation result to the input / output circuit 108 again. Intermediate storage required for these operations uses the RAM 106. CPU102, ROM104, RAM1
06, exchange of various data between the input / output circuits 108 is performed by a bus line 110 including a data bus, a control bus, and an address bus.

The I / O circuit 108 inputs and outputs 1-bit information to and from a first analog / digital converter 122 (hereinafter referred to as ADC1), a second analog / digital converter 124 (hereinafter referred to as ADC2), an angle signal processing circuit 126, and the like. And a discrete input / output circuit 128 (hereinafter referred to as DIO) for inputting data.

ADC1 includes a battery voltage detection sensor 132 (hereinafter referred to as VBS), a cooling water temperature sensor 56 (hereinafter referred to as TWS), an ambient temperature sensor 136 (hereinafter referred to as TAS), and an adjustment voltage generator 138 (hereinafter referred to as V
RS) and throttle sensor 140 (hereinafter OTHS) and O
The output from the _two sensors 142 (hereinafter, referred to as O ₂ S) is applied to a multiplexer 162 (hereinafter, referred to as MPX).
One of them is selected and the analog / digital conversion circuit 16 is selected.
4 (hereinafter referred to as ADC). The digital value output from the ADC 164 is held in a register 166 (hereinafter referred to as REG).

The output of the flow sensor 24 (hereinafter referred to as AFS) is ADC2.1
24, and is converted into a digital signal through an analog / digital conversion circuit 172 (hereinafter referred to as an ADC).
(Hereinafter referred to as REG).

From the angle sensor 146 (hereinafter, referred to as ANGLS), a signal indicating a reference crank angle, for example, a 180-degree crank angle (hereinafter, referred to as REF) and a signal indicating a minute angle, for example, a 1-degree crank angle (hereinafter, referred to)
POS) is applied to the angle signal processing circuit 126, where the waveform is shaped.

The DIO (128) includes an idle switch 148 (hereinafter referred to as IDLE-SW), a top gear switch 150 (hereinafter referred to as TOP-SW), and a starter which operate when the throttle valve 74 returns to the fully closed position.・ Switch 152 (hereinafter referred to as START-SW)
Is input.

Next, a pulse output circuit and a control target based on a calculation result of the CPU will be described. An injector control circuit 1134 (hereinafter referred to as INJC) is a circuit that converts a digital value of a calculation result into a pulse output. Accordingly, a pulse INJ having a pulse width corresponding to the fuel injection amount is generated by the INJC 1134 and applied to the injector 12 via the AND gate 1136.

The ignition pulse generation circuit 1138 (hereinafter referred to as IGNC) has a register for setting the ignition timing (hereinafter referred to as ADV) and a register for setting the starting time of the primary current supply of the ignition coil (hereinafter referred to as IGNC).
DWL), and these data are set by the CPU. A pulse IGN is generated based on the set data, and is supplied to an amplifier 62 for supplying a primary current to the ignition coil.
This pulse IGN is applied via an AND gate 1140.

The valve opening rate of the bypass valve 61 is determined by a pulse I applied from a control circuit (hereinafter referred to as ISCC) 1142 through an AND gate 1144.
Controlled by SC. ISCC1142 is a register that sets the pulse width ISCD and a register that sets the pulse period ISCP
And have

EGR amount control pulse generation circuit 1178 that controls EGR control valve 90
(Hereinafter referred to as EGRC) are registers that set the value that indicates the time interval ratio (duty) of the high / low state of the pulse and registers that set the value that indicates the pulse period
With EGRP. The output pulse EGR of the EGRC is applied to the transistor 90 via the AND gate 1156.

The 1-bit input / output signal is controlled by the circuit DIO (128). Input signals include IDLE-SW signal, START
-SW signal and TOP-SW signal. The output signal includes a pulse output signal for driving the fuel pump.
This DIO is provided with a register DDR192 for determining whether a terminal is used as an input terminal and a register DOUT194 for latching output data.

The mode register 1160 is a register (hereinafter referred to as MOD) for holding commands for instructing various states inside the input / output circuit 108. For example, by setting the mode register 1160 to a name, the AND gates 1136, 1140, 1144, All the 1156s are activated or deactivated.
By setting an instruction in the MOD register 1160 in this manner, it is possible to control the stop and start of the outputs of INJC, IGNC, ISCC, and EGRC.

A signal DIO1 for controlling the fuel pump 32 is output to DIO (128).

Therefore, if EEC is applied in this way, almost all controls related to the internal combustion engine, such as A / F control, can be performed appropriately, and strict exhaust gas regulations for automobiles can be sufficiently cleared, and excellent fuel efficiency can be achieved. Can get the engine.

By the way, in such A / F control in EEC, for example, control data of the injector 72 is obtained from the data AF24 representing the intake air amount and the engine speed N, and the result is obtained.
O ₂ is corrected by the feedback control by the data of the sensor 142, that as a predetermined A / F is obtained is well known, according to this type of control technique, the variation of mechanical parts, sensors and actuators, with time Even when the injection pulse deviates from the value for obtaining the optimal air-fuel ratio state due to changes in the environment and the like, the concentration of the specific component in the exhaust gas can be reduced by the O ₂ sensor
And the feedback correction is performed according to the detected value, so that the injection pulse is always controlled to the optimum value.

[Problems to be solved by the invention]

However, such control works effectively when the operating state of the engine is in a steady state or a state where it is slowly changing, but in a transient operating state where the operating state changes suddenly, the air-fuel ratio Since the feedback correction cannot follow the operating state, the air-fuel ratio state of the engine greatly deviates from the optimum value. For this reason, the purification efficiency of the catalytic converter for reducing harmful components in the exhaust gas is significantly deteriorated.

On the other hand, learning control has been proposed as a method of optimizing the injection pulse when the air-fuel ratio is greatly shifted as described above. An example of the learning control is disclosed in Japanese Patent Application Laid-Open No. Sho 57-26229. According to this method, the average value of the air-fuel ratio correction coefficient determined by the concentration of the specific component in the exhaust gas when the engine is in the idle operation state is obtained, and the error correction amount is learned and controlled so that the average value falls within a predetermined range. The average value of the air-fuel ratio correction coefficient determined by the concentration of the specific component in the exhaust gas when the engine is operating at a predetermined rotation speed different from idle is determined, and the error correction is performed so that the average value falls within a predetermined range. The amount is obtained by learning control, and a component that varies according to the rotation speed is obtained as a function of the rotation speed from each error correction amount at the time of idling and other than the idling to determine the entire error correction amount.

However, this type of method has a problem in that the amount of program is increased due to the difference between processes at idle and the process other than idle, and the convergence by the error correction amount and the unambiguous correction amount by the rotation speed due to learning control. However, no consideration has been given to correction in a transient state in which the load state of the engine changes suddenly as described above.

As described above, in the conventional control methods, no consideration is given to sufficiently suppressing the deterioration of the exhaust gas in the transient operation state in which the engine operation state changes abruptly. There is a problem that harmful components are generated in a spike form, and the exhaust gas is likely to deteriorate.

That is, for example, there is a so-called dead time (a control response delay) from when the driver steps on the accelerator to when the vehicle is actually accelerated. Similarly, there is a control response delay with respect to a sudden change in the target value of the operating state given to the control device. Due to the response delay, the deviation (control deviation) between the target value of the operation state and the current value of the operation state fed back from the internal combustion engine becomes larger by the response delay. Then, since the control device operates to reduce the large deviation, excessive control or the like occurs, and the above-described problem occurs.

Also, in the conventional engine control, no aggressive measures have been taken with respect to the target value control of the rotational speed in the sense of improving the drivability in terms of acceleration, and there is a sense of acceleration in terms of ergonomics. The ride was not considered.

For this reason, mechanical conditions such as up and down, curving, and loading load on the running road surface, environmental conditions such as temperature, clear rain, snowfall, line of sight, and running conditions such as highways, urban areas, congested roads, etc. The appropriate driving style, especially the responsiveness and mode of acceleration / deceleration, is entirely up to the driver's knowledge, know-how and driving skills, so sometimes unpleasant longitudinal acceleration, sudden acceleration, uneven driving, etc. As a result, the improvement has been strongly demanded from the viewpoint of ride comfort.

An object of the present invention is to reduce adverse effects due to excessive control in a transient operation state in which the engine operation state changes abruptly, and furthermore, appropriate driving skills and characteristics of each driver, as well as appropriate according to running conditions and the like. An object of the present invention is to provide a control device for an internal combustion engine that can obtain an acceleration pattern.

[Means for solving the problem]

In order to achieve the above object, a control device for an internal combustion engine according to the present invention includes a control unit, a linear model, a state observation unit, a delay unit, a standard acceleration pattern storage unit, and a target value generation unit. The detailed configuration of each part is as follows.

The linear model simulates a control characteristic of the internal combustion engine except for a response delay, receives an operation input that is input to the internal combustion engine, and obtains an estimated value of an operating state of the internal combustion engine corresponding to the operation input. The state observation unit receives an operation input input to the internal combustion engine and an estimated value of the operating state of the internal combustion engine obtained by the linear model, and obtains a state variable based on the input. The delay unit delays the estimated value of the operating state of the internal combustion engine obtained by the linear model by a time corresponding to the response delay. The standard acceleration pattern storage unit stores a plurality of standard acceleration patterns obtained by patterning a temporal change of an acceleration target value according to an operation state of the internal combustion engine in advance. The target value generator is
A detected value of the operating state of the internal combustion engine and the estimated value of the operating state delayed by the delay unit are input to determine a deviation therebetween, and an acceleration pattern corresponding to the operating state of the internal combustion engine is stored in the standard acceleration pattern storage unit. Is extracted from among a plurality of standard acceleration patterns stored in the memory, the extracted acceleration pattern is corrected based on the acceleration component of the deviation, and an acceleration target value is obtained based on the corrected acceleration pattern, and the deviation is calculated. The target values of the other operating states are obtained based on the above, and the target values of the operating state including the acceleration target values are generated by inputting the accelerator pedal input. The control unit receives the target value of the operating state including the target acceleration value generated by the target value generation unit and the estimated value obtained by the linear model as inputs, and is obtained by the state observation unit so as to reduce their deviation. It is assumed that the operation input to be input to the internal combustion engine is adjusted based on the state variable.

[Action]

According to the present invention having the above-described configuration, the estimated value of the operating state of the internal combustion engine excluding the response delay is obtained by the linear model,
Since the control unit determines the operation input according to the deviation from the target value using the estimated value as the feedback value, it is possible to reduce excessive control due to a response delay. As a result, in a transient operation state in which the engine operation state changes abruptly, adverse effects caused by excessive control, for example, deterioration of the air-fuel ratio can be reduced.

In addition, an acceleration pattern storage unit capable of outputting a plurality of acceleration patterns is provided, and an acceleration pattern according to the control state is determined by the target value generation unit, and the acceleration pattern is corrected based on the control deviation. Since the target value of the acceleration is generated based on the acceleration, it is possible to obtain an acceleration pattern suitable for the driving skill, characteristics, or running conditions of the driver of the vehicle, thereby improving acceleration, driving performance, riding comfort, and the like. Is done.

〔Example〕

 Next, examples according to the present invention will be described.

Birdware Configuration Regarding the birdware configuration of the embodiment of the present invention, dedicated hardware may be configured, but it is the same as the conventional EEC described in FIGS. 3 and 4, or as shown in FIG. In addition, an acceleration sensor 25 (hereinafter referred to as GS) is added to the control circuit of FIG. 4 to measure the longitudinal acceleration, and the control operation of the control circuit 60 including the microcomputer is adapted to achieve the object of the present invention. Therefore, it can be realized by changing a part of the program stored in the ROM 104.

Overview In order to achieve the object of the present invention, it is necessary to be able to sufficiently control fluctuations in the driving state in a short period of time, such as acceleration and deceleration from starting and cruising at a constant speed. However, it is considered theoretically difficult to control the transient state as in the present case, and in the case of a car, a certain time interval from when the driver depresses the accelerator pedal to when the car actually accelerates, so-called wasteful time. The extra time makes it even more difficult. The detrimental effect of this dead time on driving a car is what many drivers experience. Causes of occurrence include delay in the arrival of the air-fuel mixture, rising of a gradual torque for sequentially exploding for each cylinder, inertia of the engine structure, and the like.

In order to achieve the above-mentioned objects, the present invention has been made to solve the above-mentioned technical difficulties by (1) means for estimating a response for the purpose of compensating for a response delay due to a dead time, and (2) reliable means for a transient state. An acceleration control means for controlling is adopted.

The prediction means is a linear model as described below. That is, this linear model is a linear approximation of the behavior characteristics of an actual engine in the vicinity of a characteristic finite number of operating states in a continuously changing operating state in transient operation. As the linear model used for the prediction means, a linear model approximated in the characteristic operating state closest to the operating state of the actual engine is selected from the plurality of linear models. It should be noted that these linear models do not have a dead time that actually exists, and the response of the engine to the control input one sample period after the current control timing in response to the operation input of the current control timing, for example, acceleration, rotation speed, A predicted value such as an air-fuel ratio is given.

The acceleration control means has the following effects. That is, in the control in the transient operation state, it goes without saying that the response of the engine changes every moment, and the control target value itself may change. Under these circumstances, in order to sufficiently grasp the operating state of the engine, the rotational speed and the air-fuel ratio alone are not sufficient, and it is necessary to monitor the respective time changes. In particular, grasping the longitudinal acceleration given as the time differential value of the rotational speed or the time differential value of the vehicle speed is a factor that directly affects the drivability, acceleration, and riding comfort, and therefore, the control of the excessive driving state. Is extremely important. In other words, this operation can be performed by grasping the longitudinal acceleration, and the transient driving state can be controlled in more detail, so that the drivability, acceleration, and riding comfort can be improved.

First Embodiment FIGS. 1 and 2 show a first embodiment of the present invention. First, contents disclosed in this embodiment are roughly classified into a first system relating to acceleration control including a target value generator 3 and a standard acceleration pattern storage 4, and an exhaust gas including a linear system model 1 and a state observer 2. It is divided into the second system related to countermeasures. Hereinafter, each element will be described individually.

The linear model 1 has the same inputs as the engine that is the control target 6, that is, the throttle opening θth, the fuel injection time T _I , the exhaust gas recirculation amount δP _R , and the ignition timing δIT (where δ is a deviation from a standard set value. And the corresponding outputs are the longitudinal acceleration G of the vehicle body, the rotational speed N of the engine, and the air-fuel ratio.
Estimated values of A / F, give. The linear model 1 is provided one for each of a finite number of operating states, that is, for each combination of characteristic values in G, N, and A / F, and these are one for each of the values of G, N, and A / F. The input θth of the control target 6 in the vicinity,
The response to each of T _I , δP _R , and δIT is determined by linear approximation. Depending on the combination of G, N, and A / F values, that is, depending on the operating state, only one of these finite number of linear models is used. One is set, or a linear model is created by averaging or combining a plurality of linear models having similar characteristics, and the estimated value, Is used to calculate What should be noted about the linear model 1 is that it does not have a response delay time so-called dead time existing in the controlled object 6 while explaining the response to the input of the controlled object 6. From this feature, when an arbitrary input is applied to the control target 6 in an arbitrary operating state, an estimated value of an output appearing after a dead time can be immediately obtained, and the first aspect of the present invention. The point is attributable to this point.

The selection of the linear model 1 is based on the outputs G, N, A /
F and the estimate obtained from linear model 1, The dead time element 7 corrects the deviation from a value which is delayed by the dead time L seconds existing in the control target 6. This deviation is due to the error of the linear model 1 or the control object 6
Means the load fluctuation at the time. Further, learning control can be realized by accumulating the correction process as experience and reproducing the correction process when a similar situation is encountered again.

Next, the state observer 2 will be described. As in the present embodiment, there is a well-known method of using a state observer as one of the construction methods of a control law when considering a plurality of inputs and outputs for the control target 6. The function of the state accelerator 2 will be described. First, the inputs θth, T _I , δ in the linear model 1
P _R and δ _I T are collectively described as a vector U, and an estimated value as an output, Are collectively described as a vector, and the internal state that supports the behavior of the linear model 1 is described as a vector. If the state equation of the linear model 1 is n at the current time and n-1 is the immediately preceding sample time, (n) = A (n-1) -BU ( n-1) ...... (1) (n-1) = C (n-1) ...... (2) _{However, U = [θth T I δP} R δ I T] t (T is transposed) (3) Becomes Here, A, B, and C are all constant matrices, which are determined by the transfer function matrix T (Z) of the control target 6 shown in the formula (5) by the formula (6). Where Z indicates the Z-transform of the sample value, For example, T ₁₁ (Z) is a transfer function for θth, T ₂₂ (Z) is a transfer function for T _I , and T ₃₄ (Z) is δ _I T
Against And the like.

If a new matrix G is selected such that all the eigenvalues of the matrix (A-GC) are within the unit circle, an estimated value of the internal state variable can be obtained by equation (7).

(N) = (A-GC) X (n-1) + BU (n-1) + G (n-1) (7) Each of the estimated state variables and longitudinal acceleration, rotation speed, and air-fuel ratio Target values G *, N *, (A / F) *
(Hereinafter referred to as Y *) and the estimated value by the linear model 1,
, LQI optimal regulator control is performed based on the deviation of The evaluation function J is expressed by equation (8). Here, Q and R are weight matrices.

Where δ (n) = (n) − (n−1) (9) δU (n) = U (n) −U (n−1) (10) Optimal to minimize the evaluation function J The control input U * (n) Becomes In equation (11), the optimal gain K is Is determined. At this time, the matrix P is a solution of the Riccati equation of the equation (13).

However, As described above, the estimated amount of the state variable is obtained from the state observer 2 based on the equation (7), and the optimal control U * is generated from the controller based on the equation (11).

Next, the target value generator 3 will be described. The target value generator 3 is based on the accelerator pedal input AP and the estimated value of the linear model 1. E- ^LS and the output G, N, A / F (hereinafter Y
(Hereinafter referred to as ΔY) and outputs a target value Y * of the control amount. The target value Y * is generated as a time variable when controlling the transient operation state, and is set in advance in a manner that changes with time or calculated according to the sampling time.

In the present invention, the longitudinal acceleration G is described as requiring special control in the transient operation state.

However, when the longitudinal acceleration is not directly controlled and the engine speed N and the air-fuel ratio A / F are controlled only, that is, there is no particular indication that the present invention is applicable to the control system shown in FIG. Is evident. A particular problem with the longitudinal acceleration G is, for example, FIG.
When the accelerator pedal input AP is applied in a step-like manner as shown in FIG. 2, the so-called jerky vibration usually appears in the longitudinal acceleration as shown in FIG. 3 (b), which has been a major cause of discomfort to passengers. Was. On the other hand, in the present invention, as shown in FIG. 9C, the time-dependent manner G * of the target longitudinal acceleration G is determined in advance, and the actual longitudinal acceleration G
By controlling to follow G *, it is characterized that front and rear discomfort is eliminated and driving performance, acceleration and riding comfort are simultaneously improved. FIG. 3D shows the rotation speed corresponding to the state of G *.

However, the uniqueness of the target longitudinal acceleration G * is easily lost due to a variety of factors such as the mechanical characteristics of the vehicle body, road surface, environment, and load conditions, or drivability and acceleration desired by the driver. Responding to each of these complicated factors is no longer possible with a method of calculating Y * by a simple operation. That is, in this case, as shown in FIG. 1, for example, a standard acceleration pattern storage unit 4 is provided, and the target G * of the longitudinal acceleration predetermined for all of the finite number of possible combinations of the various factors is set as a time variable. It is stored as data, and only one is selected from these finite number of standard acceleration patterns in accordance with the various factors, that is, the deviation Δ or the signal signal8 from the driver or other measuring instrument or controller mounted on the vehicle. Or when a plurality of the above factors occur simultaneously, a standard acceleration pattern is created by averaging or synthesizing the standard acceleration patterns corresponding to each factor. Also, N *,
(A / F) * may be calculated according to G * data,
Alternatively, each of them may be stored in the standard acceleration pattern storage 4 as a standard pattern similarly to G *.

The linear model 1 is stored in the standard acceleration pattern storage 4.
Because the standard acceleration pattern based on is stored,
By inputting the aforementioned ΔY to the target value generator 3, if there is an error between the actual engine and the linear model 1, the target value is generated while correcting the error in the target value generator 3. be able to.

In addition, when the target value changes only at a constant value or only slowly, there is no problem as a control system for setting the target value to the constant value target G *, N *, (A / F) *. The operation process in the present invention will be described with reference to FIG. 7 based on the configuration diagram of FIG. First,
Calculation starts as soon as accelerator pedal input AP is detected (80)
I do. Next, the initial value Y ₀ * of the target value of the control amount is read (8
1) At this time, the other initial values are Y ₀ = O, Y ₁ = O, ₀
= O, = O. A first operation amount U in the order n of the control cycle is created (82). The deviation ΔYn between the output of the control target 6, that is, the control amount Yn and the estimated value _n−1 of the linear model 1 is calculated (83) according to the equation (16). L is synchronized with Yn by ^applying e- ^LS , which is a ^dead time (second), to _n-1 .

Δn ( _n-1 · e- ^LS- Yn) (16) Next, the linear model 1 may be used if necessary due to the small amount of Δn.
Is selected again and corrected (84). The estimated value n of the current control amount is estimated (85) by the subsequent linear model 1.
From the operation amount Um and the estimated value n, n is derived (86) according to the equation (7). Further, the time series data G * as a standard acceleration pattern is selected or corrected as needed from the memory 4 from the deviation Δn and the signal Signal (87). Based on the obtained time-series data G *, the control synchronization n +
The acceleration target value G _{n + 1} * with respect to ₁ is extracted, and other control amounts N _{n + 1} *,
_{(A / F) n + 1} * to calculate the determining the control amount of the target value Y _{n + 1} *.

Finally, the target value Y _{n + 1} * in the ( _{n + 1} ) _th cycle and the (n +) th
From the integral value of the deviation from n, which is the estimated value of the output of the controlled object 6 in one cycle, and the estimated value n of the internal state in the nth cycle, the manipulated variable _{Un + 1,} which is the optimal control input, according to (11).
* Is calculated, and the process proceeds to the next control synchronization to repeat the arithmetic processing.

With the above configuration, it is possible to avoid various adverse effects from the dead time existing in the controlled object, and at the same time, to accurately solve the optimal control problem in the multivariable input / output system. Enables the control of transient operating conditions such as starting and acceleration / deceleration that correspond to the sensitivity of the wearer.

Second Embodiment FIG. 8 shows a configuration diagram of a second embodiment of the present invention. First
The basic difference from the first embodiment is that both the control amount and the operation amount are scalar amounts, and the rotational speed N and the throttle opening θth
That is the point. The linear model 1 takes an input as θth and obtains an estimated value of the longitudinal acceleration. Further, the longitudinal model G obtained as the output of the controlled object 6 is delayed from the longitudinal acceleration G by the time delay L (second) to synchronize with G. The linear model 1 is re-selected or re-selected as necessary due to the deviation from e- ^LS. The configuration is modified. Here, the linear model 1 is a transfer function having θth as input and output as θth. The controller 5 inputs the deviation between the target value G * of the longitudinal acceleration and the estimated value, and both the target value N * of the rotational speed and the rotational speed N and the deviation of the controlled object 6. The controller 5 performs P1 control or PID control. In particular, the deviation N *-is large, that is, in the transient operation state, the emphasis is placed on minimizing the deviation G *-, and later, the deviation G *-
Is small, that is, in a constant-speed operation state, a weighted control law is provided to emphasize the minimization of the deviation N * -N. Thereby, the control of the longitudinal acceleration G and the rotation speed N can be performed complementarily, and the follow-up control to the target value series of the longitudinal acceleration during the transient operation and the constant value control of the rotation speed in the constant speed operation are performed in a time-division manner. Can be realized.

According to the above embodiment, the behavior of an engine having a dead time can be estimated in real time, and this estimating means is corrected to give the best estimation according to the operating state and load condition of the engine. Therefore, regardless of the constant speed operation and the acceleration / deceleration, it is possible to control the control amount, that is, one of the longitudinal acceleration, the rotation speed, or the air-fuel ratio, or if there are a plurality of them, simultaneously, thereby achieving fuel economy. Performance, exhaust purification, acceleration, drivability or riding comfort can be greatly improved.

〔The invention's effect〕

As described above, according to the present invention, the estimated value of the operating state of the internal combustion engine from which the response delay has been removed by the linear model is obtained, and the control unit uses the estimated value as a feedback value, and calculates the estimated value of the internal combustion engine according to the deviation from the target value. Because we are seeking operation input,
The adverse effects caused by excessive control in the transient operation state can be reduced.

In addition, according to the one provided with the acceleration pattern storage unit capable of outputting a plurality of acceleration patterns and the target value generation unit, it is possible to obtain an acceleration suitable for the driving skill, characteristics or running conditions of the driver of the vehicle. .

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is the same block diagram, FIG. 3 is a schematic diagram showing an example of electronic control of an engine, FIG. 4 is a block diagram of a conventional control circuit, FIG. 5 is a block diagram of a control circuit of the present invention, FIG. 6 is a characteristic diagram showing a time-dependent characteristic of an accelerator pedal input and a longitudinal acceleration of the vehicle, and FIG. 7 is a flow chart showing a calculation process of FIG.
FIG. 8 is a block diagram showing another embodiment. 1 ... linear model, 2 ... state observer, 3 ... target value generator, 4 ... standard acceleration pattern storage, 5 ... controller, 6 ... controlled object, 7 ... dead time element.

Claims (1)

(57) [Claims] A control unit, a linear model, a state observation unit,
A delay unit, a standard acceleration pattern storage unit, and a target value generation unit, wherein the linear model simulates a control characteristic of the internal combustion engine except for a response delay, and an operation input to be input to the internal combustion engine. And an estimated value of the operating state of the internal combustion engine corresponding to the operation input is obtained.The state observation unit is configured to calculate the estimated value of the internal combustion engine by the operation input and the linear model input to the internal combustion engine. The estimated value of the operating state is input, and a state variable is determined based on the input.The delay unit corresponds to the estimated value of the operating state of the internal combustion engine determined by the linear model, which corresponds to the response delay. The standard acceleration pattern storage unit stores a plurality of standard acceleration patterns obtained by patterning in advance a time change of an acceleration target value according to an operation state of the internal combustion engine. The target value generation unit inputs the detected value of the operation state of the internal combustion engine and the estimated value of the operation state delayed by the delay unit and obtains their deviation, An acceleration pattern corresponding to the operating state of the internal combustion engine is extracted from a plurality of standard acceleration patterns stored in the standard acceleration pattern storage unit, and the extracted acceleration pattern is corrected based on the acceleration of the deviation. Obtaining an acceleration target value based on the corrected acceleration pattern, and obtaining a target value of another operation state based on the deviation to generate a target value of an operation state including the acceleration target value. The input unit receives a target value of the operating state including the target acceleration value generated by the target value generation unit and an estimated value obtained by the linear model, and inputs a bias value thereof. Control apparatus for an internal combustion engine and adjusts the operation input to be input to the internal combustion engine based on the state variable determined by the state observing unit to reduce.