JP3301351B2 - Control device and method of designing the control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device and a method for designing the control device, and more particularly, to a general control target, which outputs a manipulated variable command value so as to match its state to a target state, and controls the control target. The present invention relates to a control device for controlling a state and a method for designing such a control device.

[0002]

2. Description of the Related Art With the development of control theory, various control devices have been designed and manufactured. One of them will be described with reference to a clutch slip control device as an example. Various types of clutch slip control devices that control slippage of a lock-up clutch of a torque converter, for example, are known. In these slip control devices, when the input and output of the torque converter are directly connected, the vibration of the engine is directly transmitted to the transmission side in a low engine rotation range to degrade the riding comfort, while the input and output are wide in a wide range of rotation speed. Releasing the combination of the two solves the contradictory problem that the fuel consumption reduction effect is not effectively utilized.

Conventionally, such a slip control device has been continuously improved with an emphasis on how to achieve both high responsiveness and control stability. For example, an operation amount command value at the current time is set to a target slip rotational speed and an actual slip speed. A control deviation amount, which is a difference from the slip rotation speed, a differential amount and an integral amount of the deviation amount, or a differential amount of the deviation and a second-order differential amount thereof are calculated. 586
issue). Alternatively, a method has been proposed in which these quantities are developed into time-series quantities to calculate the increment of the manipulated variable (Japanese Patent Laid-Open Publication No. Sho.
No. 4-30966). If these controls are sufficiently adjusted, they perform the specified operations to stably maintain and control the slip amount to the target value and realize high follow-up to the target value without deteriorating stability. can do.

[0004]

However, these control devices have a problem that sufficient control cannot be realized for a control object whose control characteristics fluctuate, for example, a system for controlling a slip state of a clutch. . When the difference between the lock-up clutch and the hydraulic control system for controlling the amount of slip is large, or when the control system is designed in an initial state, the friction material and the hydraulic oil deteriorate to reduce the friction characteristics of the lock-up clutch, that is, the μ of the clutch. If the -v characteristic changes to impair the stability of the slip speed, the slip amount cannot be controlled to the target value in a stable and high-speed manner. This point will be briefly described with reference to the drawings.

Changes in the characteristics of the slip control system are illustrated by the gain and phase of the transfer function from the manipulated variable command value to the slip speed. FIG. 28 is a graph showing the variation between the solids in the characteristics at the time of design. The characteristics of the clutch vary between solids, particularly in a high frequency range, due to the instability of the friction material and the way in which the friction material is pressed. Further, as shown in FIG. 29, when the friction material wears out or the hydraulic oil deteriorates with time and thermally, the characteristics of both the gain and the phase characteristics decrease from the characteristics at the time of design. In a conventional control device, it was not possible to achieve both high responsiveness and stability in response to such a large change in the characteristics of the controlled object.

Further, in addition to the problem of the change in the control characteristics, there is a problem that such a control target has a large dead time in the control. For example, when considering the control of a clutch of an automatic transmission, if the direct control of a high hydraulic pressure that actually drives the clutch is performed, the control valve becomes large. In general, a method of controlling the hydraulic pressure by a solenoid valve and controlling the operating hydraulic pressure of the clutch by changing the pilot hydraulic pressure is adopted. In this case, there is a considerable delay from when the solenoid valve is controlled until the clutch actually operates and its output-side rotational speed changes. This delay is a dead time in control, and has been a problem in improving controllability.

[0007] Such a problem of characteristic change always exists in a real device, and therefore, in a control device actually used,
In many cases, the stability of the control is prioritized so that the response of the control system is moderate. As a result, in a situation where a steady operating state continues for a long time, the target slip rotation speed is stably maintained, but the torque fluctuation of the engine on the input side is cut off, or the transient operation in which the target slip rotation speed fluctuates. The torque transmission efficiency could not be increased following the state.

[0008] The control device of the present invention is intended to solve such a problem and to realize high responsiveness in control without losing stability even when a characteristic change occurs in a controlled object. . A preferred application example of such a control device method is a clutch slip control device. The present invention has the following features to attain the object.

[0009]

The control device of the present invention obtains a manipulated variable command value of a controlled object having a large dead time in terms of control characteristics and outputs the manipulated variable command value to the controlled object. A control state detection means for detecting a control state of the control target, and a control state detection means for detecting the control state of the control target, and A state prediction means for predicting a plurality of state quantities of the control target, and a model obtained by identifying the control target by a low-order linear model, and a sliding mode control using the predicted plurality of state quantities as an input. And a sliding mode control means for outputting the manipulated variable command value.

According to such a control device, a controlled object having a large dead time due to control characteristics is not directly controlled, but a plurality of state quantities of the controlled object after a dead time from the control state. Is predicted, and the sliding mode control is performed using this. Therefore, the sliding mode control can be performed using a model obtained by identifying a low-order linear model, and even for a control object having a large dead time in the control characteristics, the characteristics of the sliding mode control can be sufficiently improved. Can be withdrawn. Therefore, even when the characteristics of the control target change significantly, when there is a large disturbance, or when the dead time fluctuates,
It is possible to sufficiently enjoy control stability and responsiveness by the sliding mode control. These advantages cannot be obtained if the control target is simply identified by a higher-order model.

As the large dead time in the control characteristics, it is possible to consider that the dead time is the same as or longer than the response time of the controlled object. Of course, the dead time may be shorter than the response time as a control target. Such a control device can be, for example, a device that controls a clutch of an automatic transmission.
Although it is conceivable to directly control the slip rotation speed, in such controlled objects, the response time of the controlled object is on the order of several tens of milliseconds, and the same as dead time in many cases. The control device of the present invention is particularly suitable when a system having such a long dead time is to be controlled.

In such a configuration, the prediction of a plurality of state quantities of the control target after the dead time can be performed using the internal state quantity of the control target and the past operation amount command value. A configuration in which a plurality of state quantities after a dead time are predicted by the Smith method is also suitable.

According to the design method of the present invention for designing such a control device, an operation amount command value of a controlled object having a large dead time in terms of control characteristics is obtained, and the operation amount command value is output to the controlled object. A method of designing a control device that matches a state of a control target to a target state, comprising: (a) identifying a model of the control target from a relationship between an operation amount command value of the control target and a control state; (b) ) Separating the model into a dead time part and a dynamics part, (c) approximating the dynamics part by a low-order linear model, and (d) a state based on the linear model and the dead time. Designing a predictive controller; and (e) setting design parameters for sliding mode control using the output of the state predictive controller as a control input based on the linear model. It is summarized in that with.

According to this design method, after a model to be controlled is identified, the model is separated into a dead time part and a dynamics part, and the dynamics part is approximated by a low-order linear model. Thereafter, a state prediction controller is designed based on the linear model and the dead time, and design parameters for the sliding mode control using the output of the state prediction controller as a control input are set.

As the low-order linear model in the step (c), a second-order linear model can be used. Also,
After the step (c), the linear model is provided with a step of expanding the model by adding an integrator for integrating a deviation between a state of a control target and a target value thereof, and the steps (d) and (e).
Then, it is also possible to use this expanded linear model. Such a method of designing a control device can be applied to a method of designing a control device that directly controls a slip rotation speed by controlling a clutch of an automatic transmission as a control target.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to further clarify the configuration and operation of the present invention described above, a preferred embodiment of the present invention will be described below. FIG. 1 is a schematic configuration diagram showing an overall configuration of a clutch control device in an automatic transmission 10 as one embodiment of the present invention. Since the control of the automatic transmission 10 is related to the control of the engine EG, the engine control computer 50 is shown together with the automatic transmission control computer 30 in FIG. The automatic transmission 10 is integrally formed with an engine EG whose intake air amount is adjusted by a throttle valve THV whose opening is adjusted by an accelerator pedal AP, and has a hydraulic control for controlling a hydraulic pressure described later inside. The device 20 is incorporated. Further, a sensor group such as a sensor for detecting an engine rotational speed is provided in the engine EG, the automatic transmission 10, and its vicinity, and signals from these sensor groups are transmitted to the automatic transmission control computer 30 and the engine control computer 50. It is connected to the. As will be described later, the clutch control device of the embodiment performs control so that the slip speed Nfuki matches the target rotation speed N * at the shift point of the automatic transmission 10 while inputting signals from various sensor groups 40. .

The slip control device is provided with various sensor groups 40. When the signals from the sensor group 40 are listed, the signals input to the engine control computer 50 include an engine speed, an intake air flow rate,
There are intake air temperature and the like. Further, signals input to the engine control computer 50 and the automatic transmission control computer 30 include a throttle opening, a vehicle speed, an engine water temperature, a brake switch, and the like. Further, signals input to the automatic transmission control computer 30 include a shift position, a pattern select switch, an overdrive switch, outputs of a C0 sensor and a C2 sensor described later in the hydraulic control device 20, a transmission oil temperature, and a manual shift switch. and so on.

The engine control computer 50 is connected to a throttle motor M for controlling the opening of a second throttle valve provided in the intake air passage, a fuel injection valve or an igniter provided for the engine EG, and the like. Controls the amount and ignition timing. The automatic transmission control computer 30 controls the solenoid valve provided in the hydraulic control device 20 to control the automatic transmission 10 through control of each hydraulic pressure controlled by the solenoid valve. The automatic transmission control computer 3
The control signal from 0 to the solenoid valve is also input to the engine control computer 50, and the engine control computer 50 can know the operation state of the automatic transmission 10. The engine control computer 50 and the automatic transmission control computer 30 also output information on the current gear position and a signal (Q /
N) etc. are exchanged. Therefore, the engine control computer 50 can control the fuel injection amount and the ignition timing of the engine EG by knowing the operation state of the automatic transmission 10, and the automatic transmission control computer 30 can control the current operation state of the engine EG. , It is possible to perform the control for switching the gear position.

The clutch control device of the embodiment is a device realized inside the automatic transmission control computer 30. This clutch control device controls the opening degree of a valve called a B3 control valve of an automatic transmission to control the clutch-release-side rotation speed Nfuki to a target rotation speed N *. Before describing a design method and an internal configuration of this clutch control device, the automatic transmission 10
Will be briefly described. FIG. 2 is a schematic diagram showing the structure of the automatic transmission 10. As shown in FIG. FIG. 3 is an explanatory diagram showing how each control valve and switch are in a state at each shift bed.

As shown in FIG. 2, the mechanical parts of the automatic transmission 10 are mainly composed of a sub-transmission mechanism D comprising a front type overdrive planetary gear unit and a forward 4-speed and reverse 1-speed consisting of a simple connection of three planetary gear trains. It has a five-speed configuration in which a speed change mechanism M is combined, and this mechanism is connected to a torque converter T with a lock-up clutch L.

The auxiliary transmission mechanism D includes a one-way clutch F-0 associated with a sun gear S0, a carrier C0, and a ring gear R0 constituting a front type overdrive planetary gear unit, and a multi-plate clutch C-0 connected in parallel to the one-way clutch F-0. And a multi-disc brake B-0 connected in series with the brake. The main transmission mechanism M includes gear units P1 to P3 each having one sun gear S1 to S3, one carrier C1 to C3, and one ring gear R1 to R3, and these gear units P1 to P3 are directly connected as transmission elements as appropriate. Thus, a simple connection three planetary gear train is configured. In relation to the shift elements of each gear unit P1 to P3,
Multiple disc clutches C1 to C3, band brake B-1, multiple disc brakes B-2 to B-4, one-way clutch F-1,
F-2 is provided. The connection configuration of these gear units and the arrangement of the clutch and brake are well known. The automatic transmission 10 has a clutch C-0.
C0 sensor SN1 for detecting the rotation of the drum of the clutch C2, a C2 sensor SN2 for detecting the rotation of the drum of the clutch C-2, and a slip speed sensor SN3 for detecting the slip speed Nfuki which is the number of rotations of the multiple disc clutches C1 to C3 on the releasing side. And so on. Although not shown in FIG. 2,
Each clutch and brake is provided with a hydraulic servo mechanism composed of a piston / cylinder mechanism for engaging and disengaging the friction members.

In this automatic transmission 10, the rotation of the output shaft of the engine EG shown in FIG. 1 is transmitted to the input shaft IN of the auxiliary transmission mechanism D via the torque converter T, and the input shaft IN
Is constantly rotated. The rotation of the carrier C0 is interrupted to the output shaft OU by releasing each frictional engagement element under the control of the hydraulic control device 20 or by engaging only the brake B-0. And a so-called neutral state is established.

To set the automatic transmission 10 to the first speed shift state, the clutch C-0 is engaged to directly connect the subtransmission mechanism D, and the clutch C-1 of the main transmission mechanism M is engaged. Except for the one-way clutch F-2 being engaged, all other frictional engagement elements are released. At this time, the rotation of the input shaft IN enters the sun gear S3 of the gear unit P3 via the ring gear R0 and the clutch C-1. Since the reverse rotation of the ring gear R3 of the gear unit P3 is prevented by the one-way clutch F-2, the rotation of the input shaft IN is finally output as the first speed rotation from the carrier C3 to the output shaft OU. You.

FIG. 3 shows the state of each frictional engagement element when the automatic transmission 10 is set to the second speed, the third speed, and so on. In FIG. 3, a circle indicates that the clutch and the brake are in the engaged state, and that the one-way clutch is in the locked state. Further, the mark ● indicates engagement only during engine braking, and the mark Δ indicates engagement or release. Further, the mark ◎ indicates engagement that is not involved in power transmission. As shown in FIG. 3, by controlling the state of each engagement element, the output shaft OU of the automatic transmission 10 is set in the neutral state, the reverse rotation state (the vehicle moves backward), the first, second, third, and third states. The shift state is such that the fourth and fifth speed rotations are output. The operation states of the sub transmission mechanism D and the main transmission mechanism M when outputting each speed rotation can be considered in the same manner as in the case of the first speed rotation described above, and therefore, detailed description is omitted.

Here, the objects directly controlled by the clutch control device of the present embodiment will be described in detail. FIG. 4 shows a B-3 control valve 25 to be controlled by the clutch control device of the embodiment, and a solenoid valve SL3 for adjusting the pilot pressure of the B-3 control valve 25.
FIG. 4 is a schematic diagram showing a brake B-3 driven by operating hydraulic pressure output from a B-3 control valve 25. Here, the brake B-3 changes the number of revolutions output via the gear unit P1 by changing its frictional state, so that the brake B-3 functions as a clutch that controls the number of slip revolutions as a whole. It can be said that. By controlling the solenoid valve SL3, the rotation speed of the output shaft OU in the second speed rotation controlled by the B-3 control valve 25, and thus the clutch, can be adjusted.

The solenoid valve SL3 is a solenoid valve that is duty-controlled, and the spool valve opening is adjusted by controlling the duty of the pulse of the voltage signal applied to the coil, and the pilot pressure determined by the spool valve opening is adjusted. Is supplied to the B-3 control valve 25.
The B-3 control valve 25 includes a spool 24 urged in one direction by a spring 22. The pilot pressure supplied to the first port 25A via the solenoid valve SL3 and the urging force of the spool 24 by the spring 22 , The spool 24 is operated. Since the spool 24 stops at a position where the biasing force of the spring 22 and the pilot pressure balance, the automatic transmission control computer 30 adjusts the pilot pressure by controlling the duty of the operating voltage applied to the solenoid valve SL3. , The position of the spool 24 can be controlled freely. When a force (control force) FV for driving the spool 24 upward by the pilot pressure acts, the spool 24 generates a frictional force FRD due to friction between the spool 24 and the cylinder 25D of the B-3 control valve 25. This friction is a frictional force due to the static friction coefficient until the spool 24 starts to move, and becomes a frictional force due to the kinetic friction coefficient once the spool 24 starts to move. However, in order to simplify the design of the system, FIG. As shown, the frictional force FRD is smaller than the driving force FV with respect to the control force acting on the spool 24 due to the differential pressure between the pilot pressure and the urging force of the spring 22.
Is proportional to this in a predetermined range, and becomes constant when it is larger than the predetermined range.

An original pressure for operating the automatic transmission 10 is supplied to a second port 25B of the B-3 control valve 25. When the pilot pressure supplied via the solenoid valve SL3 rises and the spool 24 is pushed up against the urging force of the spring 22, a part of the original pressure is output via the third port 25C, and the main transmission mechanism M brake B-
3 is supplied. As a result, the brake B3 generates a braking force according to the supplied original pressure, and changes the operation state of the planetary gear unit P1. Therefore, there is a large delay between the time when the duty du of the solenoid valve SL3 is changed and the time when the planetary gear unit P1 actually operates and the rotational speed of the shaft is converted. In this embodiment, the delay (dead time) is about 50 to 70 [msec].

Next, a control configuration in this embodiment will be described. FIG. 6 is a block diagram of a clutch control system according to the present embodiment. As shown in the figure, the direct target of control is B-3
The B-3 oil pressure supplied from the B-3 control valve 25 to the main transmission mechanism M is adjusted, the gear train and the clutch pack are controlled by the oil pressure, and the slip speed Nfu in the automatic transmission 10 is controlled.
ki will change. Most slip speed Nf
The uki is not determined only by the B3 hydraulic pressure, but is also affected by the output rotation speed of the torque converter T that is rotated by the rotation of the output shaft of the engine EG operated based on the throttle opening θ. In this embodiment, the slip rotation speed Nf
A controller 100 that receives uki as a control input and controls the B-3 control valve 25 by the duty of the voltage applied to the solenoid valve SL3 is configured in the automatic transmission control computer 30. Output u of controller 100 is B-3 control valve 2
5 corresponds to the manipulated variable command value for the solenoid valve S
It is output as a duty signal of L3.

Next, the design method of the controller 100 and the internal configuration of the designed controller will be described. First, the internal configuration of the controller 100 will be roughly described. This controller 100 has the configuration shown in FIG.
As shown in the figure, the state prediction controller 110, the sliding mode controller 120, the feedforward controller 13
0, an integrator 140, and a differentiator 150. Slip speed N which is a control input to the controller 100
fuki is a sensor SN provided inside the main transmission mechanism M.
3 is input. The target value N * of the slip speed Nfuki is determined in the automatic transmission control computer 30 from the current operating state of the engine EG and the automatic transmission 10, the vehicle speed, and the like.

The integrator 140 accumulates the deviation between the actual slip speed Nfuki and the target value N *. The output x3 (k) of the integrator 140 is one of the state variables handled by the controller 100, and is used for state prediction control. Output to the container 110. On the other hand, the differentiator 150 obtains a deviation of the slip speed Nfuki between a time before the predetermined sampling time T and the current time. That is, the differentiator 150 is means for calculating a differential value of the slip speed Nfuki, and the output x2 (k) of the differentiator 150 is also output to the state prediction controller 110 as one of the state quantities. State prediction controller 11
0 together with these state quantities x3 (k) and x2 (k)
The slip speed Nfuki itself is also input as the state quantity x1 (k). These state quantities x1 (k) and the like are shown as quantities at the k-th time of the control repeated at the sampling time ΔT, that is, as discretized ones.

The state prediction controller 110 calculates the three state quantities x1 (k), x2 (k) and x3 (k)
The internal state quantities xx1 (k), xx2 (k) after the dead time,
xx3 (k). Where xx (k)
Indicates the predicted value of the state quantity at the time point n times after the current control timing (k) at the sampling time ΔT. That is, the internal state quantities xx1 (k) and the like output by the state prediction controller 110 are prediction values that preempt the internal state of the control target after the dead time has elapsed. The design method of the state prediction controller 110 and its contents will be described later in detail.

Next, the sliding mode controller 120 will be briefly described. Sliding mode controller 1
Reference numeral 20 denotes a controller that controls the duty of the solenoid valve SL3 that controls the B-3 control valve 25 based on the so-called sliding mode control theory.
The theory of sliding mode control is described in detail in "Sliding Mode Control" (Corona, Kenzo Nonami, Hiroki Tahiro). However, when the phase space is divided into two regions using the switching function prepared for the controlled object, the system Is a theory of a variable structure control system based on the finding that the behavior in the topological space is composed of two completely different behaviors. One of the two different behaviors is the arrival mode (non-sliding mode), starting from any location on the phase plane, the trajectory moves towards the switching line and reaches the switching line in finite time. Another behavior is
In sliding mode, this trajectory asymptotically approaches the origin on the phase plane.

In order to realize the sliding mode control, the following method is used. First, it is assumed that the control target is represented by the following equation (1).

[0034]

(Equation 1)

Here, x is an n-dimensional vector indicating an internal state quantity, and u is an m-dimensional vector indicating a control input of the system. At this time, m is represented by a vector expression such as σ (x).
Switching functions, and the variable control structure is given by the following equation (2).
Is represented by

[0036]

(Equation 2)

As a result, if the switching surface σ (x) and the variable structure control input are designed so as to satisfy the following conditions, the sliding mode control is realized. Design a switching surface σ (x) that represents a lower order system dynamics than a given controlled object order. A variable structure such that any state x outside the switching surface reaches the switching surface within a finite time. Control input u (x, t)
To design. It should be noted that a low-order system designed by
It is desirable to have −m dimensions. With this design, a sliding mode is generated on the switching surface so as to follow the desired system dynamics. In this case, the whole system becomes asymptotically stable on a global basis.
The sliding mode controller 1 used in the present embodiment
Reference numeral 20 denotes a servo system in which the slip speed target value Nfuki changes. The design of the servo control system for sliding mode control is detailed from page 141, line 1 to page 144, line 2 of the same book.

Next, a description will be given of a method of designing a clutch control device according to the present embodiment and the contents of the control device obtained as a result of the above-described configuration. FIG. 8 is a process chart illustrating a design method of the clutch control device of the embodiment. The design method will be described according to this process chart. (1) Step S10—Identify a Model of a Hydraulic Control System Among Control Targets First, an operation of identifying a model of a control target is performed. The control object is a solenoid valve SL3 as shown in FIG.
Is received as an operation amount command value, the B-3 control valve 25 in the hydraulic control device 20 is operated by the duty u, and the control hydraulic pressure PB3 supplied from the B-3 control valve 25 causes the automatic transmission 10 to operate. This is a system in which the gear train section operates and finally the slip speed Nfuki of the clutch on the release side changes. For this system, a number of relationships between the operation command value u and the control pressure PB3 of the B-3 control valve 25 are obtained, and an operation of identifying a model of the clutch hydraulic control system is performed.

(2) Step S20—Identify the Model of the Gear Train Part of the Control Target Next, the model of the gear train part, which is the latter stage of the control target, is inputted with the control oil pressure PB3 and the slip speed Nfuk.
With i as an output, a large number of data are taken, and a work of identifying a model of the gear train portion from these input / output data is performed.

(3) Step S30-Combining the Models Obtained in Steps S10 and S20 The models obtained in the above two steps are connected in series to create an identification model representing the entire control target.

(4) Step S40—Separation of Model The model obtained in step S40 is separated into a dead time part and a dynamics part, and the dynamics part is subjected to a process of approximating it with a second-order linear model. The dynamics part separated in this step is expressed by the following equation (3) in a continuous time system.

[0042]

(Equation 3)

(5) Step S50—Expansion of the System Next, this system is controlled by following the target value to determine the slip speed Nfu.
In order to use a servo system in which ki changes, the deviation is expanded using x3 which is a time integral of the deviation between the target slip speed N * and the detected slip speed Nfuki. That is, the output of the integrator 140 shown in FIG. 7 is added. The expanded system is represented by the following equation (4).

[0044]

(Equation 4)

(6) Step S60-Determination of Control Period In order to convert this control system into a discrete time model in the next step S70, a process of determining a control period (sampling time) ΔT is performed. The control cycle is determined based on the time required for the automatic transmission control computer 30 to perform the required matrix operation. In this embodiment, as described later, the control period ΔT is determined to be about 5 milliseconds.

(7) Step S70—Conversion to Discrete Time Model Next, using the control cycle ΔT determined in step S60,
Equation (4) is converted to a discrete-time model to obtain a state equation of the following equation (5). Equation (5) shows a case where control is repeatedly performed at a control cycle ΔT, where k is a variable indicating the current control timing, and k + 1 indicates the next control timing. X (k) is x1 (k), x2 (k), x3
(K). Further, Ad, Bd, C
d is a coefficient vector when the continuous-time model represented by the equation (4) is converted to a third-order discrete-time model.

[0047]

(Equation 5)

The state quantity handled in the state equation of equation (5) is an amount x1 corresponding to the slip speed Nfuki.
(K), differential value x2 (k) of slip speed Nfuki,
And the integral value x3 (k) of the deviation between the actual slip speed Nfuki and the target value N *. The derivative value x2 (k) and the integral value x3 (k) are defined as follows. x2 (k) = {x1 (k) -x1 (k-1)} / {T x3 (k) = x3 (k-1) + {N * -x1 (k)}.
ΔT At this point, the design of the dynamics part was completed.

(8) Step S80—State prediction controller 110
The state prediction controller 110 is designed based on the tertiary model obtained by the above steps and the dead time of the controlled object. The state prediction controller 110 predicts the internal state quantity of the control target after a time corresponding to the dead time L, and based on the state equation of Expression (5), the past n times corresponding to the dead time Command value u (kn), u (k- (n)
-1)),... U (k-1) and current state quantity X (k)
From this, the estimated value xx1 of the internal state quantity L time delay ahead
(K), xx2 (k), and xx3 (k). Here, the number n of the past manipulated variable command values used for the calculation
Can be obtained as n = L / △ T (where n
Is a positive integer). That is, the state prediction controller 110 has a model of the control system of the clutch speed to be controlled inside, and according to this model, the internal state quantity X (k) at the time of performing prediction and the past n times Is estimated from the manipulated variable command value u. State prediction controller 11 designed based on equation (5)
0 can be represented by the following equation (6).

[0050]

(Equation 6)

Through the above steps, the state prediction controller 110
The design has been completed. Next, the sliding mode controller 1
Design 20. (9) Step S90-Sliding Mode Controller 120
The parameters of the sliding mode controller 120 are
It is set based on the cubic model of the above equation (5). First,
In the design of the sliding mode controller 120, the switching plane is defined as in the following equation (7). In the following description, the suffix (k) indicating the control timing is omitted altogether to avoid complicating the description, but the sliding mode controller 120 is also a discrete time system like the state prediction controller 110. It is still described and designed.

[0052]

(Equation 7)

Here, if control is performed so that σ · (dσ / dt) <0 is satisfied, a sliding mode occurs, and the state quantity is constrained to the switching hyperplane σ = 0. The manipulated variable command value u (k) satisfying Expression (7) is configured as the sum of two control inputs Ueq and Unl shown in Expression (8).

[0054]

(Equation 8)

As shown in the above-described equations (7) and (8), the parameters in the design of the sliding mode controller 120 have three coefficients of S1, κ, and ξ. Among these coefficients, S1 is a parameter that advances the phase of the closed loop system, and κ is a parameter that directly affects the response of the slip speed Nfuki (that is, the state quantity x1). Therefore, by appropriately increasing or decreasing these parameters, the sliding mode controller 120 performs the sliding mode control with desired responsiveness and stability.
Can be set.

(10) Step S100—Connection of System The state prediction controller 110 and the sliding mode controller 120 described above are connected to each other to form the controller 10 as a whole.
0. As a result, the controller 1 shown in FIG.
00 is configured. Although FIG. 7 illustrates the feedforward controller 130, the feedforward controller 130 has a target value N *.
When a large change occurs in the control amount, a predetermined control amount is added to the operation amount command value u which is the output of the controller 100 without waiting for the detection of the deviation of the slip speed Nfuki from the target value N *. It will improve. Since the above-described sliding mode controller 120 has excellent stability, hunting and the like can be performed even when the operation amount command value is adjusted by a predetermined amount by the feedforward controller 130 when the target value N * greatly changes. Stable control is realized without causing any trouble.

The controller 1 obtained by the above steps
00, the slip speed Nfuki was actually controlled by using the B-3 control valve 25 of the automatic transmission 10. Hereinafter, this control example will be described in comparison with control characteristics of a conventional control device. In the present embodiment, a1 = 269.0 a2 = 38.5 b = 5.16 was obtained as a coefficient in the continuous-time model of the control object obtained in step S50 and expressed as equation (4).

When this is converted into a discrete-time model in step 70 by setting the sampling time ΔT to 4 [msec], each coefficient Ad, Bd, Cd of the state equation (5) is obtained.
Was as follows.

[0059]

(Equation 9)

The coefficients Ad ⁿ and D of the state prediction controller 110 were obtained from the coefficients of the third-order discrete time model, and the following values were obtained. Since a delay time of the control system for the slip speed Nfuki is expected to be 64 [msec] as an initial value, the control cycle is 4 [msec] for the n-times data handled by the state prediction controller 110. Therefore, 64/4 =
16 times.

[0061]

(Equation 10)

Further, the coefficients of the sliding mode controller 120 represented by the equations (7) and (8) were set to the following values by repeating the experiment. S1 = 35, κ = 8.75, α1 = 0, α2 = 0, α3
= 30

Next, the processing executed by the automatic transmission control computer 30 of this embodiment will be described. The automatic transmission control computer 30 has a well-known CPU, RAM, R
It has OM and the like, and controls each solenoid valve of the hydraulic control device 20 of the automatic transmission 10 according to a program stored in the ROM in advance. Automatic transmission control computer 3
In the case of 0, control is performed to switch the shift point of the automatic transmission 10 based on the vehicle speed, the accelerator opening, and the like. As one of the shift point controls, the slip speed Nfuki is adjusted to a desired value. This control is performed by the controller 10 described above.
The control is performed by the automatic transmission control computer 30. That is, the controller 1
00, and furthermore, the internal state prediction controller 110 and the sliding mode controller 120 are realized by the automatic transmission control computer 30 executing programs corresponding to these controllers. FIG.
It is a flow chart which shows an example of control of this slip speed Nfuki.

When the slip speed Nfuki control routine shown in FIG. 10 is started, first, the slip speed Nfuk
i, a process of inputting the throttle opening θth and the like is performed (step S200). Next, based on the throttle opening degree θth and the like, a process of calculating a target rotation speed N * is performed (step S201), and a process of calculating a deviation e between the actual slip speed Nfuki and the target rotation speed N * is performed (step S202). ). Further, a process of integrating the deviation e (step S203) and a process of obtaining a differential value of the slip speed Nfuki (step S204) are performed. The integral value of the deviation e corresponds to x3 in the above equation, the differential value corresponds to x2, and the slip speed Nfuki corresponds to x1. Therefore, these state quantities are then included in the above equation (6),
A process of obtaining predicted values xx1, xx2, and xx3 is performed (step S205), and further, using these predicted values,
Based on the sliding mode control formula (8), a process for obtaining the duty of the solenoid valve SL3, which is the operation amount command value u (k), is performed (step S206). This operation amount command value u (k) is actually
Output to L3 (step S207). As a result, the opening degree of the solenoid valve SL3 is controlled, and the slip speed Nfuki is adjusted to the target rotation speed N * via the operation amount of the B-3 control valve 25.

By repeatedly executing this processing, the slip speed Nfuki of the automatic transmission 10 is constantly adjusted to the target rotation speed N *. However, when the shift point is switched, the target rotation speed N * is changed. Is changed. A control example of the slip speed Nfuki in such a case will be described below in comparison with the conventional PID control. FIG.
4 is a graph showing how the actual slip speed Nfuki is controlled when the target value of the clutch release rotational speed of the automatic transmission 10, that is, the slip speed Nfuki increases by 26 rpm. At this time, the oil temperature in the hydraulic control device 20 of the automatic transmission 10 was 80 [° C.], and the friction coefficient of the B-3 control valve 25 was 0.13. It can be seen that the controller 100 of the present embodiment achieves good responsiveness and stability with respect to the change in the target rotation speed N *. On the other hand, FIG. 12 shows a control example by the conventional PID control. From the comparison between FIG. 11 and FIG. 12, the excellent control characteristics of the controller 100 including the state prediction controller 110 and the sliding mode controller 120 of the present embodiment are understood, but the superior characteristics of the controller 100 of the present embodiment are understood. This characteristic becomes more apparent when the characteristics of the control target such as the B-3 control valve 25 and the gear train change.

FIG. 13 shows the control characteristics of the controller 100 of the embodiment when the dead time of the entire system including the B-3 control valve 25 to be controlled and the gear train increases by about 30 [msec]. . On the other hand, under the same conditions, the control characteristics of the conventional PID control device are shown in FIG.
The results are shown in FIG. As shown in the figure, according to the controller 100 of the present embodiment, even if the dead time fluctuates, the control characteristics are such that the undershoot and the overshoot slightly increase, but in the conventional PID control, the slip speed Nfu is reduced.
ki fluctuates greatly, does not converge, and causes hunting. This is because the PID control optimally adjusts the various coefficients on the assumption that the characteristics of the control target do not significantly change, and cannot cope with a change in the dead time of the control target. Similarly, when the temperature of the hydraulic oil in the hydraulic control device 20 changes to 40 [° C.], as shown in FIG. 15, in the control by the controller 100 of the present embodiment, even if the target rotation speed N * fluctuates. Slip speed Nfu
ki is controlled to follow this, whereas in the conventional PID control, the slip speed Nfu is controlled as shown in FIG.
ki fluctuates greatly. 17 and FIG.
8 shows a control example between the controller 100 and the conventional PID control when the friction coefficient of the clutch changes to 0.11. Also in this case, the controller 100 of the present embodiment
Shows excellent control characteristics.

Further, according to the present embodiment, the control object is designed to be separated into a dead time portion and a quadratic linear model. As a result, in the dynamics part, it is possible to obtain extremely excellent control characteristics while using a relatively simple model of a second-order linear model (the final model is a third-order model for a servo system). Was. In order to express a system including dead time as it is, a higher-order model must be used. However, sliding mode control is less effective for higher order models. In addition, when a higher-order model is used, the solution becomes complicated, the amount of calculation during actual control increases, and it is difficult to shorten the control cycle. On the other hand, in this embodiment, since the dead time part and the dynamics part are separated, and the sliding mode control is applied to the dynamics part, a simple model can be used, and the configuration can be simplified and the control can be simplified. The contribution to shortening the cycle is significant. In addition, since the excellent characteristics of the sliding mode control can be sufficiently extracted, when the characteristics of the controlled object fluctuate greatly (for example, a large fluctuation of the coefficient of friction, a large fluctuation of the oil temperature)
In addition, control satisfying responsiveness and stability can be realized. Furthermore, since the state quantity after the dead time is predicted using the state prediction controller 110 for the dead time portion, the responsiveness can be improved for an object using the phenomenon of oil pressure and friction with a large dead time. Control with excellent stability can be performed while securing the same.

The prediction of the dead time may be performed by a configuration other than the state prediction controller 110. For example, dead time compensation by the Smith method shown in FIG. 19 can be adopted. In FIG. 19, the state quantities xx1, xx2, xx3 in which the dead time is compensated by the Smith method compensator 110 are shown.
And the compensation components xM1, xM2, xM3 for the disturbance, and the state quantity x of the sliding mode controller 120 is added.
h1, xh2, xh3. The dead time compensation by the Smith method is designed using a transfer function, and is a well-known method designed on the assumption that the dead time is constant.

Next, a second embodiment of the present invention will be described. In the second embodiment, the control device of the invention is applied to a slip control device for a lock-up clutch. FIG.
FIG. 2 is a diagram showing a vehicle power transmission device to which the slip control device is applied. First, a hardware configuration of a clutch slip control device according to an embodiment will be described. As shown in FIG. 20, the power of engine 210 is supplied to torque converters 212 and 3 having lock-up clutches.
The transmission is transmitted to driving wheels via a stepped automatic transmission 214 composed of a set of planetary gear units and the like, and further through a differential gear device (not shown). Torque converter 212 is fixed to pump impeller 218 connected to crankshaft 216 of engine 210 and to input shaft 220 of automatic transmission 214. The torque converter 212 includes a turbine wheel 222 that rotates by receiving oil from a pump wheel 218, a stator wheel 228 fixed to a housing 226 that is a non-rotating member via a one-way clutch 224, and a damper 230. And a lock-up clutch 232 connected to the input shaft 220 through the lock shaft. The lock-up clutch 232 is an input / output member of the torque converter 212, that is, the crankshaft 216.
And the input shaft 220 is directly connected. When the oil pressure in the engagement-side oil chamber 235 of the torque converter 212 is higher than the release-side oil chamber 233, the lockup clutch 232 is engaged, and the rotation of the crankshaft 216 is transmitted to the input shaft 220 as it is. . On the other hand, when the oil pressure in the release-side oil chamber 233 in the torque converter 212 is higher than that in the engagement-side oil chamber 235, the lock-up clutch 2
32 is disengaged, and the torque converter 212
Its original function, that is, torque is converted at an amplification factor corresponding to the input / output rotation speed ratio, and rotation of the crankshaft 216 is
20.

The automatic transmission 214 includes the input shaft 220 and the output shaft 234, and one of a plurality of forward gears and a reverse gear is selected by a combination of operations of a plurality of hydraulic friction engagement devices. It is configured as a stepped planetary gear device that is in a state of mechanical engagement. A shift control hydraulic control circuit 244 for controlling the gear position of the automatic transmission 214 and an engagement control hydraulic control circuit 246 for controlling engagement of the lock-up clutch 232 are provided. The shift control hydraulic control circuit 244 includes a solenoid no. 1 and solenoid N
o. 2 is provided with a first solenoid valve 248 and a second solenoid valve 250 which are respectively turned on and off by the second solenoid valve 250. The clutch and the brake are selectively operated by a combination of the operations of the first solenoid valve 248 and the second solenoid valve 250. Thus, any one of the first to fourth speeds is achieved.

The engagement control hydraulic control circuit 246 includes a linear solenoid valve 252, a switching valve 254, and a slip control valve 256. This linear solenoid valve 252 is
The constant modulator pressure Pmodu generated in the shift control hydraulic control circuit 244 is used as the base pressure, and the solenoid no. It operates linearly in response to the current flowing through 3. That is, the linear solenoid valve 252
Is the drive current Iso from the electronic control unit (ECT) 242.
The output pressure Plin of the magnitude corresponding to the magnitude of 1 is continuously generated. This output pressure Plin is supplied to the switching valve 254 and the slip control valve 256. The switching valve 254 has a release side position where the lockup clutch 232 is released and an engagement side position where the lockup clutch 232 is engaged. Further, the slip control valve 256 operates using the regulator pressure Pcl generated according to the throttle valve opening by a clutch pressure regulating valve (not shown) in the shift control hydraulic control circuit 244 as a base pressure.

The switching valve 254 is provided with a spring 258 for urging a spool valve (not shown) toward a release side position.
And a first port 2 to which the regulator pressure Pcl is supplied.
60, a second port 262 to which the output pressure of the slip control valve 256 is supplied, and a third port connected to the release-side oil chamber 233.
It has a port 264, a fourth port 266 connected to the engagement side oil chamber 235, and a fifth port 268 connected to the drain. When the output pressure Plin of the linear solenoid valve 252 supplied to the switching valve 254 falls below a predetermined value, the switching valve 254 switches its spool valve to the spring 2.
The release side position according to the urging force of 58 (the state of FIG. 20)
The second port 262 is closed, and the first port 260, the third port 264, and the fourth port 266 are closed.
And the fifth port 268. Therefore, when the output pressure Plin of the linear solenoid valve 252 acting on the spool valve of the switching valve 254 falls below a predetermined value, the spool valve of the switching valve 254 moves to the release side in accordance with the biasing force of the spring 258. At the same time, the hydraulic pressure Poff in the release-side oil chamber 233 is set to the regulator pressure Pcl, and at the same time, the hydraulic pressure Pon in the engagement-side oil chamber 235 is set to the atmospheric pressure, and the lock-up clutch 232 is released. Therefore, at this time, the torque converter 212 performs the original operation of the torque converter for converting and transmitting the torque.

On the other hand, when the output pressure Plin of the linear solenoid valve 252 acting on the spool valve element of the switching valve 254 exceeds a predetermined value, the spool valve element of the switching valve 254 receives the urging force of the spring 258. The position is switched to the engagement side position to close the fifth port 268, and the first port 260, the fourth port 266, and the second
The port 262 and the third port 264 communicate with each other. Therefore, the hydraulic pressure Pon in the engagement-side oil chamber 235 is set to the regulator pressure Pcl, and at the same time, the hydraulic pressure Poff in the release-side oil chamber 233 is pressure-controlled by the slip control valve 256, and the lock-up clutch 232 is slip-controlled. Alternatively, they are engaged.

The slip control valve 256 includes a spring 270 for urging a spool valve (not shown) to increase the output pressure. This spool valve element has an engagement side oil chamber 23 for generating a thrust toward the output pressure increasing side.
5 is operated, and the release-side oil chamber 23 is used to generate a thrust toward the output pressure decreasing side.
3, the hydraulic pressure Poff and the output pressure Plin of the linear solenoid valve 252 are applied. For this reason,
The slip control valve 256 operates so that the differential pressure ΔP (= Pon−Poff) corresponding to the slip amount becomes a value corresponding to the output pressure Plin of the linear solenoid valve 252, as shown in Expression (11). In the equation (11), F is the biasing force of the spring 270, A1 is the pressure receiving area of the hydraulic pressure Pon at the spool valve, A2 (where A1 = A2) is the pressure receiving area of the hydraulic pressure Poff, and A3 is the pressure receiving area of the output pressure Plin. Area.

[0075]

[Equation 11]

Therefore, in the engagement control hydraulic control circuit 246 configured as described above, the engagement side oil chamber 235 is provided.
And Poff in the release side oil chamber 233
Is changed according to the output pressure Plin of the linear solenoid valve 252, as shown in FIG. 21, so that the switching control of the switching valve 254 is performed by the output pressure Plin of the linear solenoid valve 252, and the switching valve 254 is engaged with the engagement position. Then, the slip control of the lock-up clutch 232 after being switched to can be performed.

Next, the configuration and the design of the electronic control unit 242 for controlling the slip control will be described in detail. The electronic control unit 242 includes a well-known CPU 282, an RO
This is a so-called microcomputer including an M284, a RAM 286, an interface circuit (not shown), and the like. In the present embodiment, the interface circuit of the electronic control unit 242 includes a throttle sensor 28 for detecting a throttle valve opening provided in an intake pipe of the engine 210.
8. An engine rotation speed sensor 290 for detecting the rotation speed of the engine 210, an input shaft rotation sensor 292 for detecting the rotation speed of the input shaft 220 of the automatic transmission 214, and a rotation speed of the output shaft 234 of the automatic transmission 214 An output shaft rotation sensor 294 and an operation position sensor 298 for detecting the operation position of the shift lever 296, that is, any one of the L, S, D, N, R, and P ranges are connected. From these sensors, the electronic control unit 242 receives a throttle valve opening degree θth, an engine rotation speed Ne (pump impeller rotation speed NP), an input shaft rotation speed Nin (turbine impeller rotation speed Nt), The output shaft rotation speed Nout and the operation position Ps of the shift lever 296 are input.

The CPU 282 of the electronic control unit 242
While using AM286 as a work area, RO
The first solenoid valve 248, the second solenoid valve 250, and the linear solenoid valve 2 are used to process the input signal according to the program stored in the M284 and execute the shift control of the automatic transmission 214 and the engagement control of the lock-up clutch 232.
52 is appropriately controlled. In the shift control, the ROM 2
A shift diagram corresponding to an actual shift speed is selected from a plurality of types of shift diagrams stored in 84, and from the shift diagram, a traveling state of the vehicle, for example, a throttle valve opening θth and an output shaft rotational speed Nout are determined. The shift speed is determined based on the calculated vehicle speed SPD, and the first solenoid valve 248 and the second solenoid valve 250 are driven so as to obtain the shift speed. In this way, the operations of the clutch and the brake of the automatic transmission 214 are controlled, and any one of the four forward speeds is engaged, so that a desired shift is realized.

Assuming the hardware configuration described above, the slip rotation speed NSLP of the lock-up clutch
Is controlled by the dead time + sliding mode control. This controller 300
Is actually realized by the processing executed by the electronic control unit 242, but is illustrated as a control block in FIG. 22 for convenience of understanding. This controller 300
Corresponds to the controller 100 of the first embodiment. In the second embodiment, the torque converter 212 is controlled by directly controlling the linear solenoid valve 252.
Is controlled. In the system for varying the slip rotation speed NSLP to be controlled, when the duty of the voltage applied to the linear solenoid 252 changes, the output pressure Plin from the solenoid 252 changes, and the switching valve 254 is switched by this oil pressure. 254
Is a system in which the slip control of the lock-up clutch 232 is performed after being switched to the engagement position, and has an extremely long dead time (about 300 [msec] in the present embodiment).

Therefore, the controller 300 includes the first
As in the embodiment, the state prediction controller 310, the sliding mode controller 320, and their associated integrators 340,
A differentiator 350 and the like are provided. The design method of the state prediction controller 310 and the sliding mode controller 320 is the same as that of the first embodiment. As shown in FIG. 8, the control method is based on the control amount command value u and the output pressure Plin of the linear solenoid 252. Then, a model of the linear solenoid 252 is identified (step S10), and then the lock-up clutch 2 is determined from the output pressure Plin and the slip rotation speed NSLP.
32 models are identified (step S20), and the two are combined to obtain a model representing the entire system (step S30).

Thereafter, the model is approximated to a dead time part and a dynamics part by a numerator and a dynamics part is approximated by a quadratic linear model (step S40), and an integrator for converting the model into a servo system as a continuous time model is added. The model is enlarged to the next model (step S50). After obtaining the tertiary model, on the one hand, the control cycle is determined (step S60), the conversion to the discrete time model (step S70), and the state prediction controller 310 is designed (step 80) to perform the state prediction control. The vessel 310 is obtained. On the other hand, each parameter of the sliding mode controller 320 is set based on the third order model (step 9).
0). Then, the state prediction controller 310 and the sliding mode controller 320 are combined to complete the controller 300 (step S100).

The state prediction controller 310 thus obtained,
An example of each parameter of the sliding mode controller 320 is shown below. As in the first embodiment, the state prediction controller 310 repeatedly executes the operation of the following expression (12) at a sampling time (control cycle) of 32 [msec]. Since the dead time of the control target is about 300 [msec], the state prediction controller 310 predicts the state quantity nine times later in the control cycle. That is, in the following equation, n = 9.

[0083]

(Equation 12)

In this embodiment, the values of the coefficients are as follows.

(Equation 13)

As in the first embodiment, the manipulated variable command value u (k) output from the sliding mode controller 320 is configured as the sum of two control inputs Ueq and Unl shown in the following equations.

[0086]

[Equation 14]

The coefficients in this embodiment are as follows: S1 = 1.186, κ = 0.75, α1 = 0, α2 =
0, α3 = 2.

The processing shown in FIG. 23 is executed inside the electronic control unit 242. FIG. 23 shows the electronic control unit 242.
4 is a flowchart illustrating a slip control processing routine executed by the control unit. When the electronic control unit 242 determines from the driving situation of the vehicle that the vehicle is in the area where the slip control is to be performed, the electronic control unit 242 repeatedly executes the slip control processing routine at intervals of about several milliseconds. Whether or not the condition for performing the slip control is determined from the output shaft rotation speed Nout and the throttle valve opening θth, an example of this condition is shown in FIG.

When it is determined that the vehicle is in the slip control region shown in FIG. 24, in the embodiment, the slip control processing routine shown in FIG. 23 is started, and the engine speed Ne and the input shaft speed are first transmitted via the interface circuit. A process for inputting Nin and the throttle valve opening θth is performed (step S400). Subsequently, the target slip rotation speed NSL is calculated from the input shaft rotation speed Nin and the throttle valve opening θth.
A process for obtaining P * is performed (step S410). The target slip rotation speed NSLP * is calculated as a three-dimensional map in advance.
The relationship between the three can be stored and determined by referring to this map.

FIG. 25 is a map for determining the target slip rotation speed NSLP * from the input shaft rotation speed Nin and the throttle valve opening θth. In this example, the target slip rotation speed NSLP * is 50 rpm or 150 rpm from the throttle valve opening θth and the input shaft rotation speed Nin.
After the target slip rotation speed NSLP * is determined in this manner (step S410), a process for determining the actual slip rotation speed NSLP of the torque converter 212 is performed (step S410).
420). The actual slip rotation speed NSLP can be obtained as a deviation between the engine rotation speed Ne and the input shaft rotation speed Nin. After that, the target slip rotation speed NSLP *
And the actual slip rotation speed NSLP as the control deviation amount e
Is performed (step S430).

Next, the deviation e is integrated to obtain an integrated value x3, and a process is performed to obtain a differential value x2 of the slip rotation speed NSLP (step S440). Thus, assuming that the slip rotation speed NSLP is the state quantity x1, three state quantities are obtained as in the first embodiment. Then, using these state quantities x1, x2, and x3, the state quantities xx1, x9 that are nine times ahead in the control cycle in accordance with the above equation (12).
x2 and xx3 are predicted (step S4).
50).

In the above description, the number of times this process routine is repeatedly executed (the number of the process) is not specifically described. However, the actual process is executed at intervals of 32 milliseconds. This is a discrete process that can distinguish the number of processes. The electronic control unit 242 stores the values required for the calculation in the processes up to n times before the currently activated process in the RAM.
286. The electronic control unit 242 then uses the predicted values xx1, xx2, and xx3 to set the duty of the voltage applied to the linear solenoid valve 252 as the control amount command value u (k) according to the above-described equation (14) of the sliding mode control. Perform the required processing (step S
460), the manipulated variable command value u (k) obtained in this way is transmitted to the linear solenoid valve 252 via the interface circuit.
After that, the process exits to “NEXT” and ends the processing routine.

FIG. 26 shows an example of control by the slip control device for the lock-up clutch according to the second embodiment described above. In the figure, the upper column shows the target slip rotation speed NSLP *
Shows the state of control of the slip rotation speed NSLP when the speed changes from 50 rpm to 150 rpm, and the lower column shows the manipulated variable command value u (k) output to the linear solenoid valve 252 in this case. In the upper part of FIG. 26, a solid line JJ indicates a control characteristic of the control device of the present embodiment,
A broken line BB shows control characteristics when the conventional PID control and the state prediction controller are combined. For comparison, FIG. 27 shows a control example in which the state prediction controller 310 is not provided. In FIG. 27, the control characteristic in the case of the sliding mode control is indicated by a solid line JN, and the control characteristic in the case of the conventional PID control is indicated by a broken line BN.

As is clear from these comparisons, according to the lock-up clutch slip control device of the second embodiment, the control target having a very long dead time (in this example, the dead time is several hundred [msec]). Even so, good stability and responsiveness can be obtained.

Further, in this embodiment, similarly to the first embodiment, the control object is designed to be separated into a dead time portion and a second-order linear model. As a result, the dynamics part is basically 2
While using a relatively simple model of the following linear model (the final model is a third order in order to make a servo system), extremely excellent control characteristics could be obtained. In the embodiment, since the dead time part and the dynamics part are separated, and the sliding mode control is applied to the dynamics part, a simple model can be used, and the configuration is simplified and the control cycle is shortened. are doing. In addition, since the state quantity after the dead time is predicted using the state prediction controller 310 for the dead time portion, the responsiveness can be improved for an object using the phenomenon of oil pressure and friction with a large dead time. It is possible to easily design a control device that performs control with excellent stability while ensuring the control.

Further, according to the control device designing method of these two embodiments, such control performance can be achieved without repeating the adjustment and the cut-and-error, and the characteristics such as the disturbance and the fluctuation of the friction coefficient can be achieved. An advantage is obtained in that a control device that can stably control a change or the like can be designed. With conventional controllers,
Matching in consideration of control characteristics at a plurality of operating points having different operating conditions and disturbances could not be dealt with by intuition of an experienced person or trial and error, and could hardly be dealt with. The control device designing method of the present invention can solve such a problem in a formal manner, so that the design man-hour and the adjustment man-hour can be reduced, and the labor for development can be remarkably reduced.

Further, in this embodiment, the present invention is applied to the slip control of the clutch for transmitting the power of the engine 210 to the automatic transmission 214, thereby cutting off the torque fluctuation from the engine 210 and increasing the torque transmission efficiency. The effect can be achieved not only in a steady state but also in a transient operation state, and further in a state in which characteristics have changed due to aging. As a result, an effect of improving fuel efficiency can be obtained.
The balance between such high responsiveness and stability is achieved by reducing the responsiveness of the control system in order to cope with characteristic changes while maintaining the stability of the conventional slip control device. Only under long-lasting conditions, the slip condition is outstanding compared to just achieving the target condition. Further, in the clutch slip control in which various operating conditions are considered, an effect that the response under these operating conditions can be further improved can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a clutch slip control device as a first embodiment.

FIG. 2 is an explanatory diagram schematically showing a configuration of the automatic transmission 10.

FIG. 3 is an explanatory diagram showing an operation state of each element of the hydraulic control device 20.

4 is an explanatory diagram illustrating a hydraulic system that controls a slip speed Nfuki via a B-3 control valve 25. FIG.

FIG. 5 is a graph showing a relationship between a driving force FV of a spool 24 and a frictional force in a B-3 control valve 25;

FIG. 6 is a block diagram showing a system for controlling a slip speed Nfuki.

FIG. 7 is a block diagram illustrating an internal configuration of a controller 100;

FIG. 8 is a process chart showing a process of designing the controller 100 of the embodiment.

FIG. 9 is a block diagram for identifying a model to be controlled.

FIG. 10 is a flowchart showing a control processing routine of the automatic transmission control computer 30 in the first embodiment.

FIG. 11 is a graph showing a control example by the controller 100 of the embodiment.

FIG. 12 is a graph showing a control example of PID control under the same conditions as the control example shown in FIG. 11;

FIG. 13 is a graph showing a control example when the dead time fluctuates in the embodiment.

FIG. 14 is a graph showing a control example of PID control under the same conditions as the control example shown in FIG.

FIG. 15 is a graph showing a control example when the oil temperature of the hydraulic system of the hydraulic control device 20 fluctuates in the embodiment.

16 is a graph showing a control example of PID control under the same conditions as the control example shown in FIG.

FIG. 17 shows a B-3 control valve 2 in an embodiment.
5 is a graph illustrating a control example when the friction coefficient of the spool No. 5 changes.

18 is a graph showing a control example of PID control under the same conditions as the control example shown in FIG.

19 is a block diagram showing a configuration example of a predictor based on the Smith method instead of the state prediction controller 110. FIG.

FIG. 20 is a schematic configuration diagram showing an overall configuration of a control device for a slip rotation speed of a lock-up clutch as a second embodiment.

FIG. 21 is a graph illustrating a relationship between a linear solenoid valve 252, an output pressure Plin, and oil pressures Poff and Pon.

FIG. 22 is a block diagram illustrating a configuration of a controller 300 according to the second embodiment.

FIG. 23 is a flowchart illustrating a slip control processing routine executed by an electronic control unit 242 according to the second embodiment.

FIG. 24 is an explanatory diagram showing a map for determining whether or not the vehicle is in a slip control area.

FIG. 25 is a graph for obtaining a target slip rotation speed NSLP *.

FIG. 26 is a graph showing a control example by the controller 300 of the second embodiment.

FIG. 27 is a graph showing a control example of PID control under the same conditions as the control example shown in FIG. 26.

FIG. 28 is a graph showing the variation between solids in characteristics at the time of design.

FIG. 29 is a graph showing changes over time in gain / phase characteristics of a clutch.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Automatic transmission 20 ... Hydraulic control device 22 ... Spring 24 ... Spool 25A ... 1st port 25B ... 2nd port 25C ... 3rd port 25D ... Cylinder 30 ... Automatic transmission control computer 40 ... Sensor group 50 ... Engine control computer 100 ... Controller 110 ... Smith method compensator 110 ... State prediction controller 120 ... Sliding mode controller 130 ... Feed forward controller 140 ... Integrator 150 ... Differentiator 210 ... Engine 212 ... Torque converter with lock-up clutch 214 ... Stepped type Automatic transmission 216 ... Crank shaft 218 ... Pump impeller 220 ... Input shaft 222 ... Turbine impeller 224 ... One-way clutch 226 ... Housing 228 ... Stator impeller 230 ... Damper 232 ... Lock-up clutch 233 ... Release side oil chamber 234 ... Out Force shaft 235 ... engagement side oil chamber 242 ... electronic control device 244 ... shift control hydraulic control circuit 246 ... engagement control hydraulic control circuit 248 ... first solenoid valve 250 ... second solenoid valve 252 ... linear solenoid valve 254 ... Switching valve 256 Slip control valve 258 Spring 260 First port 262 Second port 264 Third port 266 Fourth port 270 Spring 282 CPU 284 ROM 286 RAM 288 Throttle sensor 290 Engine rotation Speed sensor 292 Input shaft rotation sensor 294 Output shaft rotation sensor 296 Shift lever 298 Operating position sensor 300 Controller 310 State predictor 320 Sliding mode controller 340 Integrator 350 Difference machine

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Katsumi Kono 1 Toyota Town, Toyota City, Aichi Prefecture Inside Toyota Motor Corporation (72) Inventor Yasuya Nakamura 1 Toyota Town, Toyota City, Aichi Prefecture Inside Toyota Motor Corporation (56) References JP-A-4-370804 (JP, A) JP-A-3-288902 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G05B 11/00-13 / 04

Claims (2)

(57) [Claims] 1. A control for obtaining an operation amount command value of a control object having a large dead time in terms of control characteristics and outputting the operation amount command value to the control object so that the state of the control object coincides with a target state. A control state detection unit that detects a control state of the control target; a state prediction unit that predicts a plurality of state quantities of the control target after the dead time from the detected control state; Sliding mode control means for performing a sliding mode control using a model obtained by identifying a controlled object by a low-order linear model, and inputting the plurality of predicted state quantities, and outputting the manipulated variable command value And a control device comprising: 2. A control for obtaining an operation amount command value of a control object having a large dead time in terms of control characteristics and outputting the operation amount command value to the control object so that the state of the control object coincides with a target state. A method of designing an apparatus, comprising: (a) a step of identifying a model of a controlled object from a relationship between an operation amount command value of the controlled object and a control state; (C) approximating the dynamics part with a lower-order linear model; (d) designing a state prediction controller based on the linear model and the dead time; (e) A design method of a control device, comprising: a step of setting design parameters for sliding mode control using an output of the state prediction controller as a control input based on the linear model.