JP4638827B2 - Control apparatus and control method using sliding mode control - Google Patents

Description

  The present invention relates to a control device that controls a controlled system, such as a continuously variable transmission, that changes a state of a control system toward a state corresponding to the adjustment amount according to an input adjustment amount.

As a continuously variable transmission for a vehicle, a belt type continuously variable transmission is known. A control device for controlling the belt type continuously variable transmission includes a pulley position and a primary rotational speed (the rotational speed of a driving pulley). ) To be controlled, the gear ratio of the continuously variable transmission is controlled so as to be a gear ratio suitable for the driving state of the vehicle.
In order to adjust this speed ratio, in general, in a belt type continuously variable transmission, a secondary hydraulic control system that sets a secondary hydraulic pressure (hydraulic pressure for adjusting the belt clamping force of the driven pulley) according to the input torque, and an input A gear ratio control system for setting a primary oil pressure (hydraulic pressure for adjusting the pulley position of the drive pulley) required to obtain a predetermined gear ratio according to the torque and the secondary oil pressure is provided.

Here, in the gear ratio control system, in order to realize the target gear ratio, the corresponding target pulley position or target primary rotational speed is calculated, and feedback control based on a deviation from the actual value of the pulley position or primary rotational speed. Thus, the target value of the primary oil pressure is calculated, and the actual oil pressure is adjusted to be the target value of the primary oil pressure.
In performing such feedback control, hunting of the gear ratio is performed by adjusting the control gain or correcting the target value according to various driving modes (start, steady, kickdown, manual shift, brake, etc.). Thus, the responsiveness and convergence of the shift control have been optimized.

In this type of continuously variable transmission, as a technique for adjusting the control gain, for example, Patent Document 1 discloses that the control gain is increased compared to the normal travel mode to increase the speed at which the gear ratio moves to the larger side during sudden braking. In other cases, a technique for keeping the control gain small is disclosed.
Further, as a technique for correcting the target value, for example, Patent Document 2 discloses that the PID controller performs excessive integration control when the target primary rotational speed decreases suddenly, such as when the throttle is suddenly closed. In order to prevent gear ratio hunting from occurring due to undershooting of the primary rotational speed, a technique for limiting the reduction speed of the target primary rotational speed is disclosed.

  However, in the above prior art, since it is basically PID control, the robustness is low and there are the following problems. That is, (1) the control software becomes complicated in order to adjust the control gain or correct the target value for each driving mode. (2) It takes time to develop control software because it is necessary to set an optimal control gain and target value correction amount for each driving mode.

As such feedback control that compensates for the drawbacks of PID control, sliding mode control is known that has high robustness and few problems (1) and (2) (for example, Patent Document 3 and Patent Document 4). , Patent Document 5, Patent Document 6, etc.).
In these techniques, basically, a sliding mode controller performs feedback control on the servo mechanism based on a target value obtained by feedforward calculation.

However, in the sliding mode control, hunting of the control is likely to occur due to its nature, and although it does not become a big problem in the field such as the robot where the operation condition is limited, the control of various devices in many other fields, for example, When applied to the control of a continuously variable transmission for a vehicle, the hunting cannot be ignored, and there is a risk that the ride comfort of the vehicle may be reduced.
Thus, for example, a technique has been proposed in which such hunting is suppressed when the continuously variable transmission is controlled in the sliding mode (see Patent Document 7).

  In this technique, the sliding mode control amount calculated by the sliding mode control unit and the feedforward control amount calculated by the primary hydraulic feedforward term calculation unit are added, and the primary hydraulic pressure is determined by the added control amount (target primary hydraulic pressure). The control actuator is adjusted. As a result, the target primary hydraulic pressure is not all based on the control amount by the sliding mode control, but includes the control amount by the feedforward control, and the amount of control by the feedforward control is included. However, it will be smaller than if it were not. Therefore, the nonlinear feedback term gain in the sliding mode control can be reduced, and hunting can be prevented while satisfying the responsiveness.

More specifically, the following techniques are described.
In the calculation of the sliding mode control amount, the nonlinear feedback term in the control amount is calculated at least in the boundary layer provided on the switching surface on the sliding mode control phase space. Nonlinear function f (s) that gives a smaller absolute value than the value obtained by proportionally calculating the distance between the state point s of the continuously variable transmission control system) and the switching surface (surface represented by the switching function). The value of the nonlinear feedback term is obtained on the basis of the value obtained by mapping the state point s in (4). Further, the state point s in this case is defined as in Expression f based on the deviation (claim 5).
s = k1 · err ′ + k2 · err + k3 · ierr [Formula f]
When the distance between the state point s of the controlled system on the phase space and the switching surface is larger than a preset value, the integral calculation of the deviation in the equation (f) is stopped, and The amount of integration until the integration operation is stopped is held (claim 6).

Therefore, for example, when the primary hydraulic pressure is controlled in a continuously variable transmission, if the state point (PID control amount) s related to the hydraulic control is more than a certain distance from the switching surface (reference state), the integral operation of the deviation (PID control) The calculation of the integral control amount term of the amount) is stopped, and the integral amount until the stop of the integral operation is held for the integral control amount.
Japanese Patent No. 2505420 JP-A-7-167234 JP-A-8-249067 Japanese Patent Laid-Open No. 61-271509 JP-A 61-271510 JP-A-2-297602 JP-A-11-194801

  In the technique of Patent Document 7 described above, an increase in the integral control term of the nonlinear feedback term in the sliding mode control is prevented, which is effective in preventing control hunting. However, when the state point is not more than a certain distance from the switching surface. Since no control hunting countermeasures are taken against the nonlinear feedback term in sliding mode control, there is a possibility that control hunting cannot be sufficiently prevented depending on the setting.

  The present invention has been devised in view of such problems, and can be applied to the control of, for example, a continuously variable transmission for a vehicle, so as to more reliably prevent control hunting in sliding mode control. For example, an object of the present invention is to provide a control device and a control method using sliding mode control that can improve the ride comfort and driving feeling of a vehicle by preventing control hunting.

  In order to achieve the above target, the control device using the sliding mode control of the present invention includes an actual value acquisition unit that acquires an actual value of a specific parameter related to a control target, and a target value that sets a target value of the specific parameter A switching function design unit that designs a switching function σ for sliding mode control based on a deviation between a setting unit, a real value acquired by the actual value acquisition unit, and a target value set by the target value setting unit; Nonlinear input computing means for computing a nonlinear input of sliding mode control based on the switching function σ designed by the switching function designing means, and the sliding input control by the sliding mode control using the control input including the nonlinear input computed by the nonlinear input computing means Control means for controlling an object to be controlled, and the target value setting means includes the control A first target value setting means for setting a first target value of the specific parameter according to a target state of a target; and a deviation between the actual value and the first target value with respect to the first target value. And a second target value generating means for generating a second target value as the target value by using a delay filter that outputs a delay corresponding to the magnitude of the first deviation. Item 1).

It is preferable that the time constant of the delay filter is set to be larger as the first deviation is larger and smaller as the first deviation is smaller.
The time constant of the delay filter is a constant value when the first deviation is within a first predetermined range, and the time deviation is greater than the first predetermined range. Is preferably set so as to increase in proportion to an increase in the magnitude of the first deviation.

Furthermore, it is preferable that an integral gain changing unit is provided for changing the integral gain of the switching function σ according to the first deviation so that the integral gain becomes smaller as the first deviation becomes larger.
A control apparatus using sliding mode control related to the present invention includes an actual value acquisition unit that acquires an actual value of a specific parameter related to a control target, a target value setting unit that sets a target value of the specific parameter, A switching function design unit that designs a switching function σ of sliding mode control based on a deviation between an actual value acquired by the actual value acquisition unit and a target value set by the target value setting unit, and the switching function design unit Based on the designed switching function σ, non-linear input calculation means for calculating a non-linear input for sliding mode control, and the control target is controlled by sliding mode control using a control input including the non-linear input calculated by the non-linear input calculation means. Control means and an integral gain of the switching function σ You are characterized in that it includes an integral gain changing means for changing in response to the first difference as the first deviation is larger reduced.

The integral gain changing means sets the integral gain of the switching function σ to a maximum value when the first deviation is within a second predetermined range, and the magnitude of the first deviation is the first In the case where it is larger than the predetermined range of 2 and within the third predetermined range, the smaller the first deviation, the larger the first deviation, and the smaller the first deviation, the smaller the first deviation. Saga it is preferable to set the minimum value is greater than the third predetermined range (claim 5).

Further, it further comprises reset command means for resetting the integral term of the switching function σ designed by the switching function design means to 0, and the reset command means is in control prior to completion of control for the controlled object. when the first state may determine does not perform the reset command, said when the control target is a second state which may be determined that a control start from the steady state before the control is performed for the reset command is preferably (claim 6 ). Whether the control is being performed or the control is started from the steady state depends on the deviation between the actual value and the first target value or the time constant of the delay filter (claims 2 and 3). Judgment can be made.

For example, when the control target is a movable member (movable pulley) for adjusting a transmission ratio of a continuously variable transmission for a vehicle, and the specific parameter is controlled as a position of the movable member (movable pulley) (claim) 8 ) In shifting, when there is a shift command from the steady state, the first target displacement of the movable member changes suddenly, and the first deviation between the actual displacement and the first target displacement increases. Therefore, if the deviation between the actual displacement and the first target displacement is smaller than a predetermined value (when the time constant of the delay filter is smaller than the predetermined time constant), If the deviation between the actual displacement and the first target displacement is large (when the time constant of the delay filter is larger than the predetermined time constant), it can be determined that the shift is started from the steady state.

Further, linear input calculation means for calculating an equivalent control input that is a linear input of sliding mode control based on the mathematical model of the control target is further provided, and the control means includes a nonlinear input calculated by the nonlinear input calculation means, it is preferable to control the controlled object the sum of the calculated equivalent control input by the linear input calculating means as a control input (claim 7).

Furthermore, it is preferable that the object to be controlled is a movable member (movable pulley) for adjusting a gear ratio of the continuously variable transmission for a vehicle, and the specific parameter is a position of the movable member (movable pulley) ( Claim 8 ).
The control method using the sliding mode control of the present invention includes an actual value acquisition step of acquiring an actual value of a specific parameter related to a control target, a target value setting step of setting a target value of the specific parameter, A switching function design step for designing a switching function σ for sliding mode control based on a deviation between the actual value acquired in the actual value acquisition step and the target value set in the target value setting step, and the switching function design step Based on the designed switching function σ, the nonlinear input calculation step for calculating the nonlinear input of the sliding mode control, and the control target is controlled by the sliding mode control by the control input including the nonlinear input calculated by the nonlinear input calculation step. The target value setting The determining step includes a first target value setting step of setting a first target value of the specific parameter according to the target state of the control target; and the actual value and the first target value with respect to the first target value. A second target value generating step for generating a second target value as the target value by using a delay filter that outputs a delay according to the magnitude of the first deviation that is a deviation from the value. (Claim 9 ).

Wherein the switching function design step, the integral gain of the switching function sigma, it is preferable to change in response to the first difference (claim 10).
A control method using sliding mode control related to the present invention includes an actual value acquisition step of acquiring an actual value of a specific parameter related to a controlled object, a target value setting step of setting a target value of the specific parameter, A switching function design step for designing a switching function σ for sliding mode control based on a deviation between the actual value acquired in the actual value acquisition step and the target value set in the target value setting step, and the switching function design step Based on the designed switching function σ, the nonlinear input calculation step for calculating the nonlinear input of the sliding mode control, and the control target is controlled by the sliding mode control by the control input including the nonlinear input calculated by the nonlinear input calculation step. Control step and the switching The example function design step, the integral gain of the switching function sigma, that features a changing in response to the first difference above which is a deviation of the actual value and the first target value.

  Each of the control methods further includes a reset command step for resetting the integral term of the switching function σ designed by the switching function design step to 0. In the reset command step, the control command before the completion of the control on the control target is provided. The reset command is not issued when the control state is in the first state where it can be determined that the control is being performed, and the reset command is performed when the control target is in the second state where it can be determined that control has started from the steady state before control. Preferred (claim 11).

According to the control device using the sliding mode control of the present invention (Claim 1) and the control method (Claim 9 ), the first value and the first value corresponding to the target state of the controlled object are compared with the first value. Since the second target value obtained by giving a delay corresponding to the magnitude of the first deviation, which is a deviation from the target value, is used as the target value, chattering of nonlinear input in sliding mode control and overshoot of the control target are suppressed. be able to. Thereby, for example, when the position of the movable member (movable pulley) is controlled as the specific parameter, the movable member (movable pulley) for adjusting the gear ratio of the continuously variable transmission for the vehicle is controlled. According to the eighth aspect of the present invention, it is possible to improve the shift stability of the vehicle equipped with such a continuously variable transmission to the target gear ratio, thereby improving the fuel consumption of the vehicle and obtaining a stable ride feeling.

  In other words, when the target state of the controlled object is changed greatly, the first target value is changed greatly, but there is a limit to the speed of change to the target state of the controlled object, and the state of the controlled object (actual value) ) And the first target value continue to be large. For this reason, when the switching function is designed based on the first deviation, the integral term of the deviation included in the switching function is cumulatively added to the large first deviation, resulting in an excessive value. Chattering and overshoot of the control target will be caused. On the other hand, if the switching function is designed based on the second target value obtained by giving a delay corresponding to the magnitude of the first deviation with respect to the first target value, the first target value changes greatly. However, since the magnitude of the second deviation is suppressed, it is possible to suppress an excessive increase in the integral term of the deviation included in the switching function, and to prevent or suppress the occurrence of chattering of the nonlinear input and overshoot of the controlled object. be able to.

  In particular, the time constant of the delay filter for generating the second target value is set to be larger as the first deviation is larger and smaller as the first deviation is smaller (Claim 2), or By setting so as to increase in proportion to the increase in the size of one deviation (Claim 3), the follow-up to the original target value (first target value) for the controlled object is ensured while the nonlinear input An effect of preventing or suppressing chattering and occurrence of overshoot of the controlled object can be obtained. When the first deviation is smaller than a certain value (within the first predetermined range), the time constant of the delay filter is set to a constant value (Claim 3), so that the second target value is set to the first value. It is possible to smoothly converge to one target value, and control can be performed smoothly.

According to the control device (claim 4 ) and the control method (claim 10 ) using the sliding mode control of the present invention, the integral gain of the switching function σ is changed according to the first deviation. Since the integral gain is changed according to the first deviation so as to be smaller as the first deviation is larger, chattering of nonlinear input in sliding mode control and overshoot of the controlled object can be suppressed. Thereby, for example, when the position of the movable member (movable pulley) is controlled as the specific parameter, the movable member (movable pulley) for adjusting the gear ratio of the continuously variable transmission for the vehicle is controlled. According to the eighth aspect of the present invention, it is possible to obtain the effects of improving the fuel efficiency of the vehicle, improving the feeling of riding, and improving the driving feeling.

  That is, if the deviation (first deviation) between the state of control object (actual value) and the target value (first target value) is large, the integral term of the deviation included in the switching function becomes an excessive value, and the nonlinear input Although chattering and overshoot of the controlled object are caused, the integral gain of the switching function σ is changed so as to be smaller as the first deviation is larger. It is possible to prevent or suppress the occurrence of input chattering and overshoot of the control target.

Further, according to the control device using the sliding mode control of the present invention (Claim 6 ) and the control method (Claim 11 ), the switching is performed when it can be determined that the control is being performed before the control of the control target is completed. The reset command for setting the integral term of the function σ to 0 is not performed. When the control target is in a state where it can be determined that the control has started from the steady state before the control, the reset command is performed, so that the integral term of the switching function σ is obtained. An appropriate value can be given and the control accuracy can be increased. Thereby, for example, when the position of the movable member (movable pulley) is controlled as the specific parameter, the movable member (movable pulley) for adjusting the gear ratio of the continuously variable transmission for the vehicle is controlled. ( 8 ) The response delay of the gear ratio can be suppressed to improve the speed change performance and improve the fuel efficiency of the vehicle.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
1 to 4 illustrate one embodiment of the present invention, which will be described with reference to these drawings.
FIG. 1 is a block diagram showing the configuration of the control device, FIG. 2 is a system configuration diagram of a vehicle equipped with a CVT hydraulic control device related to the control, FIG. 3 is a graph showing the control characteristics, and FIG. It is a flowchart explaining these.

[System configuration of a vehicle equipped with a CVT hydraulic control device according to this control]
FIG. 2 is a system configuration diagram of a vehicle (automobile) equipped with a CVT hydraulic control apparatus according to this control. As shown in FIG. 2, the power of the engine 10 is transmitted to the CVT 30 via the torque converter 20 and the forward / reverse clutch 25. The CVT (belt type continuously variable transmission) 30 includes a primary pulley 31 on the driving side and a secondary pulley 32 on the driven side, and transmits power by a belt 33 interposed therebetween.

  In such a CVT 30, the oil pressure applied to both the drive side and driven side pulleys 31, 32 is controlled independently, thereby performing a shift, and a primary pulley (drive side pulley, movable pulley) 31 and a secondary pulley (driven side pulley). ) 32 includes a primary slide pulley 31a and a secondary slide pulley 32a. The primary slide pulley 31a and the secondary slide pulley 32a are slid by hydraulic pressure, thereby changing the belt rotation radii in the primary pulley 31a and the secondary pulley 32a independently to perform a shift.

  The oil pump 40 supplies oil to the primary pressure regulating valve 71 and the secondary pressure regulating valve 72 via the first pressure regulating valve 51 and supplies oil to the primary solenoid 61 and the secondary solenoid 62 via the second pressure regulating valve 52. It is. Further, the primary solenoid 61 and the secondary solenoid 62 are solenoid valves controlled by the CVT control unit (control device) 100, and are connected to the primary pressure regulating valve 71 and the secondary pressure regulating valve 72, respectively, to perform control by sending a signal pressure. .

(Hydraulic control of CVT pulley)
The hydraulic pressure generated by the oil pump 40 is adjusted to the line pressure by the first pressure regulating valve 51 and supplied to the primary pressure regulating valve 71 and the secondary pressure regulating valve 72. Further, the pilot pressure is set by the second pressure regulating valve 52 and supplied to the primary solenoid 61 and the secondary solenoid 62. The CVT control unit 100 controls the primary solenoid 61 and the secondary solenoid 62, adjusts the supplied pilot pressure to a desired signal pressure, and supplies it to the primary pressure regulating valve 71 and the secondary pressure regulating valve 72.

  The primary pressure adjusting valve 71 and the secondary pressure adjusting valve 72 adjust the line pressure based on the supplied signal pressure, and supply the oil pressure to the primary slide pulley 31a and the secondary slide pulley 32a, respectively, to slide them. As described above, the CVT control unit 100 controls the primary solenoid 61 and the secondary solenoid 62 to achieve shifting of the CVT 30.

[Control configuration of CVT control unit]
FIG. 1 is a control block diagram of the CVT control unit 100. As shown in FIG. 1, the CVT control unit 100 includes an external input acquisition unit 101, a thrust / difference thrust calculation unit 110, a sliding mode control unit (sliding mode controller) 120, addition units 102 and 103, and hydraulic pressure. A feedback control unit (control means) 130 and an actual gear ratio calculation unit 150 are provided to control the hydraulic valve solenoid 140 (primary solenoid 61 and secondary solenoid 62) and to control the state of the transmission unit (pulley & belt) of the CVT 30. To do.

  Among these, the external input acquisition unit 101 includes information such as the engine speed Ne of the engine to which the CVT 30 is connected and the input torque (engine load) Te from the engine to the transmission, and the vehicle speed VSP from a vehicle speed sensor (not shown). , The primary pulley rotation speed Npri from the primary rotation sensor (not shown), the throttle opening from the throttle opening sensor (not shown) [information related to the input torque (engine load) Te] TV0 and the like are collected. The shift command (target transmission ratio) ip * is calculated. The target speed ratio ip * is calculated from the shift line based on the vehicle speed VSP, the primary pulley rotation speed Npri, and the throttle opening TV0.

  The thrust calculation unit 110 controls the balance thrust calculation unit 111 that calculates the balance thrust F * to be applied to the belt pulley so that the target speed ratio ip * can be maintained, and the speed change speed at the time of the shift determined in advance from the experimental results ( That is, a differential thrust calculation unit 112 for calculating a differential thrust ΔF (for obtaining a predetermined shift speed) is provided. The balance thrust F * and the differential thrust ΔF are added by the adder 102, and the sum Uff of the balance thrust F * and the differential thrust ΔF obtained thereby is used as a feedforward amount for pulley control of the continuously variable transmission. Output to add to pulley.

In addition, the actual speed ratio calculating unit 150 calculates the actual speed ratio rip (= Npri / N) from the primary speed Npri input from the primary speed sensor (not shown) and the secondary speed Nsec input from the secondary speed sensor (not shown). Nsec).
The sliding mode controller 120 calculates a control amount based on the target speed ratio ip * and the actual speed ratio rip so that the actual speed ratio converges stably without a steady error even if there is some disturbance to the target speed ratio. And output to add to the belt pulley.

  The sliding mode controller 120 includes a gear ratio-stroke converter (first target value setting means) 121a, a gear ratio-stroke converter 121b, a corrected target displacement generator (second target value generator) 122, Deviation calculation unit 123, switching function integral gain setting unit (integration gain changing unit) 124, switching function design unit (switching function design unit) 125, integration reset command unit (reset command unit) 126, linear input calculation Unit (linear input calculation means) 127, nonlinear input calculation section (nonlinear input calculation means) 128, and addition section 129. The gear ratio / stroke converter 121a and the corrected target displacement generator 122 constitute a target value setting unit.

The gear ratio-stroke conversion unit 121a obtains the pulley target displacement (target stroke amount, corresponding to the first target value) x ₁ ^* from the target gear ratio ip * input from the external input acquisition unit 101. Further, the gear ratio-stroke converting unit 121b obtains the actual displacement x (actual stroke amount, corresponding to the actual value) x of the pulley from the actual gear ratio rip inputted from the actual gear ratio calculating unit 150. Since there is a correlation between the gear ratio and the pulley stroke, these conversion units 121a and 121b, for example, map this correlation into a table or a table, and perform the conversion based on them.

In the deviation calculation unit 123, a deviation (first deviation) e ′ (e ′ = x−x ₁ ^* ) between the actual displacement x of the pulley and the target displacement x ₁ ^*, and a bar is added to e in the figure and the mathematical expression. Calculated). The deviation calculation unit 123 also calculates a deviation (second deviation) e (e = x−x ₂ ^* ) between the actual displacement x of the pulley and a correction target displacement x ₂ ^* described later.
The corrected target displacement generation unit 122 is provided with a delay filter and gives a delay to the target displacement x ₁ ^* converted by the conversion unit 121a to correspond to the corrected target displacement (corrected target stroke amount, second target value). Output) x ₂ ^* . In this delay filter, a time constant Tc related to the delay is given according to the magnitude of the first deviation e ′ calculated by the deviation calculator 123.

  That is, as indicated by a thick line in FIG. 3, the time constant Tc is set to increase as the magnitude of the first deviation e ′ increases. However, here, when the magnitude of the first deviation e ′ is within a predetermined range (first predetermined range), the time constant Tc is fixed to the minimum value, and the magnitude of the first deviation e ′ is within the predetermined range ( When it is larger than the first predetermined range), the time constant Tc increases linearly in proportion to the increase in the magnitude of the first deviation e ′.

Thus, by generating a corrected target displacement x ₂ ^* by giving a delay corresponding to the magnitude of the first deviation e ′ to the target displacement x ₁ ^* , the displacement x, x ₁ ^* , The time variation of x ₂ ^* is shown. At the start of shifting, the pulley target displacement x ₁ ^* rises at once, and the deviation e ′ between the actual displacement x and the target displacement x ₁ ^* increases rapidly. Since the corrected target displacement x ₂ ^* changes with a larger delay as the magnitude of the deviation e ′ increases in accordance with e ′, the deviation e between the actual displacement x and the target displacement x ₂ ^* does not increase excessively.

When the first deviation e ′ is smaller than a certain value (within a first predetermined range, that is, between e ₁ ′ to e ₄ ′ shown in FIG. 3), the time constant of the delay filter is constant. By setting the value (minimum value), the corrected target displacement x ₂ ^* is output with a certain delay from the target displacement x ₁ ^* , so that the corrected target displacement x ₂ ^* converges smoothly to the target displacement x ₁ ^*. It can be made to.
The switching function integral gain setting unit 124 sets the integral gain S ₁ of the switching function σ used for calculation of the nonlinear input of the sliding mode control based on the first deviation e ′ [S ₁ = f (e ′)]. That is, as shown in FIG. 3, when the magnitude of the first deviation e ′ is small within a certain range (within the second predetermined range, that is, between e ₂ ′ to e ₃ ′ shown in FIG. 3), The integral gain S ₁ is set to the maximum value, and the magnitude of the first deviation e ′ is larger than the second predetermined range, ie, a third predetermined range (ie, between e ₁ ′ to e ₄ ′ shown in FIG. 3). In the case where the integral gain S ₁ is set with such a characteristic that it decreases as the magnitude of the first deviation e ′ increases, and the magnitude of the first deviation e ′ is greater than the third predetermined range ( That is, when the first deviation e ′ is smaller than e ₁ ′ and larger than e ₄ ′ shown in FIG. 3, the integral gain S ₁ is set to a minimum constant value (minimum value).

In particular, when the first deviation e ′ is outside the second predetermined range and within the third predetermined range (ie, between e ₁ ′ to e ₂ ′ or e ₃ ′ to e ₄ ′), the corrected target displacement using the time constant Tc of the lag filter used in generating unit 122, by the following relationship (a), it sets the integral gain S _1.

切り換え関数設計部１２５では、偏差演算部１２３で求めたプーリの実変位ｘと補正目標変位ｘ₂ ^*との偏差（第２偏差）ｅ（ｅ＝ｘ−ｘ₂ ^*）と、切り換え関数用積分ゲイン設定部１２４で求めた積分ゲインＳ₁を用いて次式（Ｂ）によりスライディングモード制御の切り換え関数σを計算する。

In the switching function design unit 125, the deviation (second deviation) e (e = x−x ₂ ^* ) between the actual displacement x of the pulley and the corrected target displacement x ₂ ^* obtained by the deviation calculation unit 123, and the integration for the switching function The switching function σ of the sliding mode control is calculated by the following equation (B) using the integral gain S ₁ obtained by the gain setting unit 124.

積分リセット司令部１２６では、スライディングモード制御の切り換え関数σの積分器（上記のＳ₁∫ｅｄｔの項）のリセット（積分項を０にすること）の要否に応じて、切り換え関数設計部１２５に積分リセット司令を出力する。つまり、本実施形態の場合、制御中において積分器をリセットすべき状態は変速開始時であり、逆に、変速中（変速が開始されたが完了していない場合）には積分器を有効に効かせなければ制御を滑らかに行なえないのでリセットすべきではない。ただし、このような積分リセット司令については、変速開始時と判断した時のみに行なうこととする。

The integration reset command unit 126 switches the switching function design unit 125 according to the necessity of resetting the integrator (the term of S ₁ ∫edt) of the sliding mode control switching function σ (setting the integration term to 0). Output integral reset command. In other words, in the present embodiment, the state in which the integrator should be reset during control is at the start of shifting, and conversely, the integrator is enabled during shifting (when shifting is started but not completed). If it is not effective, the control cannot be performed smoothly and should not be reset. However, such an integral reset command is performed only when it is determined that shifting is started.

At the start of the shift, the first deviation e 1 is increased because the first target displacement x ₁ ^* is greatly separated from the actual pulley displacement x. However, during the subsequent shift, the first deviation e ′ is gradually decreased. It can be determined from 1 deviation e ′ whether the shift is started or during the shift. Further, here, in the case where the first deviation e ′ is larger than a certain extent (outside the first predetermined range), the first deviation e ′ and the time constant Tc of the delay filter are in a proportional relationship, so that the integral reset is performed. When the first deviation e ′ is small, that is, when the time constant Tc of the delay filter is smaller than the predetermined time constant Tc _min , the command unit 126 determines that the shift control is in the middle of shifting, and in this case Does not reset the integrator (S ₁ ∫edt term) of the switching function σ of the sliding mode control. As a result, unnecessary integral reset during shifting is prevented, and shifting is completed smoothly. In other words, by prohibiting the resetting of integration during the shift, the pulley thrust does not fluctuate suddenly and the shift is performed smoothly, so that the driving feeling is not lost.

When the first deviation e ′ is large, that is, when the time constant Tc of the delay filter is larger than the predetermined time constant Tc _min, it is determined that the shift is started from the steady state of the shift control, and the sliding mode control is performed. Reset the integrator of the switching function σ. As a result, it is possible to reduce the dead time and delay at the start of the shift and improve the shift response.
In addition, the linear input calculation unit 127 _{generates an} equivalent control input u _eq ′ closest to the mathematical model to be controlled under the sliding mode control logic (indicated by a hat mark on u in the mathematical expression). It calculates by following Formula (C).

非線形入力演算部１２８では、切り換え関数設計部１２５で求めたスライディングモード制御の切り換え関数σと線形入力演算部１２７で求めた等価制御入力ｕ_eq´（状態フィードバック制御入力）に基づいて、制御対象の数式モデルの不確かさや外乱などがあってもスライディング条件を満たすようなスライディングモード制御の非線形入力ｕ_nl´を次式（Ｄ）により計算する。

In the non-linear input calculation unit 128, based on the switching function σ of sliding mode control obtained by the switching function design unit 125 and the equivalent control input u _eq ′ (state feedback control input) obtained by the linear input calculation unit 127, A non-linear input u _nl ′ for sliding mode control that satisfies the sliding condition even if there is uncertainty or disturbance in the mathematical model is calculated by the following equation (D).

加算部１２９では、スライディングモード制御の非線形入力ｕ_nl´と等価制御入力ｕ_eq´とを加算（＝Ｕscm）する。

The adder 129 adds (= Uscm) the sliding mode control nonlinear input u _nl ′ and the equivalent control input u _eq ′.

Then, the adder 103 adds the feedforward control input Uff from the adder 102 and the sliding mode control input Uscm from the adder 129.
The hydraulic feedback control unit (control means) 130 converts the thrust of the final control input calculated by the adding unit 103 in this way into the hydraulic pressure required for each pulley in consideration of each operation state, and issues a hydraulic command. The feedback control is performed so that the actual hydraulic pressure follows the hydraulic pressure command. However, in the hydraulic feedback control unit 130, specifically, the hydraulic pressure is controlled through current control of a solenoid that is a drive source of the hydraulic valve, so the hydraulic pressure command is converted into current control of the solenoid 140 that is the drive source of the hydraulic valve. Then, the solenoid 140 of the hydraulic valve is controlled.

In this way, by controlling the solenoid 140 of the hydraulic valve, the continuously variable transmission 30 is shifted so that the pulley of the continuously variable transmission moves in the axial direction by the hydraulic pressure supplied from the hydraulic valve. It has become.
[Input design of sliding mode control from mathematical model]
Here, the belt type continuously variable transmission is represented by a mathematical model, and the derivation of the above-described nonlinear input u _nl ′ and equivalent control input u _eq ′ will be described.

The belt-type continuously variable transmission can be expressed by a mathematical model such as the following formula (1).
Mathematical model of continuously variable transmission

定常誤差を無くすため、サーボ系スライディングモード制御の切り換え関数σを次式（２）のように定義する［次式（２）は前式（Ｂ）と同じもの］。

In order to eliminate the steady-state error, the switching function σ of the servo system sliding mode control is defined as the following equation (2) [the following equation (2) is the same as the previous equation (B)].

スライディングモード制御をサーボ系に適用するため、式（２）に基づいて、次式（３）のように積分値を付加した拡大形を用いる。

In order to apply the sliding mode control to the servo system, based on the equation (2), an enlarged form to which an integral value is added as in the following equation (3) is used.

安定なスライディングモード制御システムを設計するためには、前記に定義された切り換え関数σがゼロに収束するような制御を考える。

In order to design a stable sliding mode control system, a control is considered in which the switching function σ defined above converges to zero.

Therefore, the Lyapunov function V = (1/2) σ ^T σ is introduced as in the following equation (4).

ここで、明らかにＶ≧０である。
したがって、Lyapunov関数が常に単調減少する（Ｖ＜０）ならば、σは有限の時間でゼロになる。つまり、σが有限の時間でゼロになるには、Ｖ＝σ・σ＜０となればよい。

Here, clearly V ≧ 0.
Therefore, if the Lyapunov function always decreases monotonically (V <0), σ becomes zero in a finite time. That is, in order for σ to become zero in a finite time, V = σ · σ <0.

From the Lyapunov stability condition, an equivalent control input (state feedback control) u _eq ′ and a nonlinear input u _nl ′, which are control inputs in the sliding mode, are obtained.
First, as shown in the following equation (5), an equivalent control input u _eq ′ closest to the mathematical model equation (3) is obtained from σ = 0.

また、外乱やシステムの不確かさに対してもスライディング条件を満たすために、次式（６）のように、切り換え平面（σ＝０）で不連続な項（−Ｋsgn（σ））の非線形入力ｕ_nl´を導入する。

Also, in order to satisfy the sliding condition against disturbance and system uncertainty, nonlinear input of a discontinuous term (−Ksgn (σ)) on the switching plane (σ = 0) as shown in the following equation (6). u _nl ′ is introduced.

ここで、前記の式（３）は、拡大系の無段変速機の数学モデルであるが、これを、不確かさをもつ実プラントモデルとすると次式（７）のように表すことができる。

Here, the above equation (3) is a mathematical model of an expansion continuously variable transmission, and can be expressed as the following equation (7) if it is an actual plant model with uncertainty.

ここで、安定条件Ｖ＝σ・σ＜０を満たす十分に大きなＫを次式（８）のように求める。

Here, a sufficiently large K that satisfies the stability condition V = σ · σ <0 is obtained as in the following equation (8).

ここで、上式のうちのα，β，ε，ηは、不確かさや外乱などに応じて決める設計パラメータである。

Here, α, β, ε, and η in the above equations are design parameters determined in accordance with uncertainty, disturbance, and the like.

Since the nonlinear input u _nl ′ is similar to the equivalent control input u _eq ′, the nonlinear input u _nl ′ can also be expressed as in the following equation (9).

−Ｋsgn（σ）の切り換え関数sgn（σ）についての不連続チャタリングを防ぐため、飽和（saturation）関数を用いて次式（１０）のように連続化する［式（１０）は前記の式（Ｄ）と同じもの］。

In order to prevent discontinuous chattering with respect to the switching function sgn (σ) of −Ksgn (σ), it is continuous using the saturation function as in the following expression (10). Same as D)].

ここで、上式のΦは、切り換え面の境界層の幅を決める設計パラメータ、即ち、不連続のsgn関数を連続化する区間を決める設計パラメータである。

Here, Φ in the above equation is a design parameter that determines the width of the boundary layer of the switching surface, that is, a design parameter that determines an interval in which the discontinuous sgn function is made continuous.

The final control input of the sliding mode control of the servo system is summarized as the following expression (11) as an addition value of the nonlinear input u _nl ′ and the equivalent control input (state feedback control) u _eq ′ [Expression (11) is A combination of the above formula (C) and formula (D)].

［ＣＶＴコントロールユニット（制御装置）による制御方法］
本発明の一実施形態にかかるＣＶＴコントロールユニット（制御装置）１００は、上述のように構成されているので、例えば図４に示すように変速制御が行なわれる。なお、図４のフローは所定の制御周期で繰り返し実行される。

[Control Method by CVT Control Unit (Control Device)]
Since the CVT control unit (control device) 100 according to one embodiment of the present invention is configured as described above, shift control is performed, for example, as shown in FIG. Note that the flow of FIG. 4 is repeatedly executed at a predetermined control cycle.

  That is, first, in the external input acquisition unit 101 and the actual gear ratio calculation unit 150, various data from each detection unit, for example, the engine speed Ne, the input engine torque Te, the vehicle speed VSP, the primary pulley speed Npri, and the secondary pulley speed. Several Nsec, throttle opening TV0, etc. are read (step S10: actual value acquisition step). Then, the external input acquisition unit 101 and the actual gear ratio calculation unit 150 calculate the target ip (target gear ratio) ip * and the actual ip (actual gear ratio) rip (step S20: target value acquisition step).

These target transmission ratio ip *, actual gear ratio rip the gear ratio of the sliding mode controller 120 - converted stroke conversion unit 121a, in each stroke (displacement), first target displacement x ₁ ^*, the actual displacement x is Calculated (step S30: first target value acquisition step, actual value acquisition step of target value acquisition step). Then, the deviation calculator 123 calculates a deviation (first deviation) e ′ (e ′ = x−x ₁ ^* ) between the actual displacement x and the first target displacement x ₁ ^* (step S40), and this deviation e ′. Are compared with the thresholds e ₁ ′ and e ₄ ′ (step S50).

Here, if the magnitude of the deviation e ′ is less than or equal to the threshold value e ₁ ′, e ₄ ′, that is, if the deviation e ′ is between the threshold values e ₁ ′ and e ₄ ′ (within the first predetermined range), the delay filter The time constant Tc is set to the minimum value (step S60), and the magnitude of the deviation e ′ is larger than the threshold value −e ₁ ′, e ₄ ′, that is, the deviation e ′ is outside the threshold value e ₁ ′, e ₄ ′ ( If it is outside the first predetermined range, the time constant Tc of the delay filter is increased in proportion to the deviation e '(step S70).

In this way, by generating a corrected target displacement x ₂ ^* with a delay corresponding to the magnitude of the first deviation e ′ with respect to the target displacement x ₁ ^* , the larger the magnitude of the deviation e ′, the larger the deviation e ′. Since the corrected target displacement x ₂ ^* changes with a delay, the deviation e between the actual displacement x and the target displacement x ₂ ^* does not increase excessively.
The corrected target displacement generator (second target value generator) 122 converts the first target displacement x ₁ ^* into the second target displacement x ₂ ^* by the delay filter having the time constant Tc set as described above (step S80). : Second target value acquisition step). Then, a deviation (second deviation) e (e = x−x ₂ ^* ) between the actual displacement x and the second target displacement x ₂ ^* is calculated in the deviation calculating unit 123 (step S90). This deviation e is compared with threshold values e ₂ ′ and e ₃ ′ (step S100).

In the switching function integral gain setting unit (integral gain changing means) 124, the magnitude of the deviation e is equal to or less than the threshold −e ₂ ′, e ₃ ′, that is, the deviation e is between the thresholds e ₂ ′ and e ₃ ′ (first if within a second predetermined range), the integral gain S ₁ of the switching function σ and the maximum value s ₁ (step S110), the magnitude threshold of the deviation e -e ₂ ', e _3' than atmospheric, that is, If the deviation e is outside the threshold values e ₂ ′ and e ₃ ′ (outside the second predetermined range), the integral gain S ₁ of the switching function σ is set to the first deviation e ′ by the relationship of the above equation (A). The size is set so as to decrease as the size increases (step S120: integral gain changing step).

The switching function design unit (switching function design means) 125 uses the integral gain S ₁ set in this way to design the switching function σ as shown in the above equation (B) (step S130: switching function design step). ).
Then, the time constant Tc is compared with the predetermined time constant Tc _min (step S140). If the time constant Tc is equal to or greater than the predetermined time constant Tc _min , the first deviation e ′ is greater than a certain value. The process proceeds from step S140 to step S144, and it is determined whether or not the integration reset flag Fr is zero. If the integral reset flag Fr is 0, it is determined that the shift is started from the steady state of the shift control, and in this case, the integrator of the switching function σ of the sliding mode control is reset (step S160). As a result, it is possible to reduce the dead time and delay at the start of the shift and improve the shift response.

Thereafter, the integration reset flag Fr is set to 1 (step S162).
The integration reset flag Fr is a flag for preventing the process of resetting the integrator of the switching function σ of the sliding mode control from being performed a plurality of times, and when the integrator is reset at the start of shifting. 1 (step S162). Then, when a predetermined condition (the following step S141) is satisfied in preparation for the next shift start, the value is returned to 0 again (the following step S142).

On the other hand, if the time constant Tc is less than a predetermined time constant Tc _min, a case where the size of the first deviation e'small within a certain, it is determined that the process of the shifting of the shift control (i.e., not the time shift start) In this case, it is further determined whether or not the deviation e is between the threshold value e ₂ ′ and the threshold value e ₃ ′ (e ₂ ′ ≦ e ≦ e ₃ ′) (step S141). Is between the threshold value e ₂ ′ and the threshold value e ₃ ′, the process proceeds to step S142 to clear the integral reset flag Fr to 0, and the process proceeds to step S150, where the deviation e is the threshold value e ₂ ′ and the threshold value e ₃ ′. If it does not decrease in between, the process proceeds to step S150 without passing through step S142. In step S150, the integrator of the switching function σ for sliding mode control (the term of S ₁ ∫edt described above) is not reset.

By the processing in step S150, unnecessary integral reset during the shift is prevented and the shift is completed smoothly. In other words, by prohibiting the resetting of integration during the shift, the pulley thrust does not fluctuate suddenly and the shift is performed smoothly, so that the driving feeling is not lost.
The integration reset flag Fr is set to 1 when the integrator is reset at the start of the shift (step S162), and needs to be returned to 0 in preparation for the next shift start. If the reset flag Fr is returned to 0, the integrator may be reset a plurality of times. The condition for returning the integral reset flag Fr to 0 in step S141 is to avoid such a situation. That is, assuming that the deviation e fluctuates before and after the threshold value e ₁ ′ or the threshold value e ₄ ′, a state where e <e ₁ ′ or e ₄ ′ <e (that is, the time constant Tc is equal to or greater than the predetermined time constant Tc _min State) e ₁ ′ ≦ e ≦ e ₄ ′ (that is, a state in which the time constant Tc is smaller than the predetermined time constant Tc _min ), immediately thereafter, e <e ₁ ′ or e ₄ ′. <state of e (i.e., the time constant Tc is more state for a predetermined time constant Tc _min) it is considered that reverts back to the time constant Tc is a predetermined time constant Tc _min direct integration reset flag Fr is smaller than It can be considered that the integrator is reset again when it is returned to zero. Therefore, the time constant Tc is the integration reset flag Fr immediately be changed to a smaller state than a predetermined time constant Tc _min from the above state for a predetermined time constant Tc _min without returning to zero, the deviation e is the threshold value e ₂ ' And the threshold e ₃ ′ are sufficiently small (e ₂ ′ ≦ e ≦ e ₃ ′), the integration reset flag Fr is cleared to 0 (step S 142).

As a result, even if the time constant Tc is equal to or greater than the predetermined time constant Tc _min , the process proceeds to step S150 by the determination in step S144, and therefore the integrator of the sliding function control switching function σ is not reset. Then, when the time constant Tc is surely maintained below the predetermined time constant Tc _min (e ₂ ′ ≦ e ≦ e ₃ ′) due to the progress of the speed change operation, as described above, step S140 is performed. , S141 through step S142, the integration reset flag Fr is reset to zero.

As described above, at the start of shifting from the steady state of the shift control, the integrator of the switching function σ of the sliding mode control is reset. Otherwise, the reset is not performed, and thereafter, the linear input calculation unit 127 The equivalent control input u _eq ′ (indicated by a hat mark on u in the equation) is calculated by the above equation (C) (step S170: linear input calculation step).

Further, the nonlinear input calculation unit 128 calculates the nonlinear input u _nl ′ of the sliding mode control by the above equation (D) based on the switching function σ and the equivalent control input u _eq ′ (state feedback control input) ( Step S180: Nonlinear input calculation step).
Then, the adding unit 129 adds (= Uscm) the nonlinear input u _nl ′ and the equivalent control input u _eq ′ of the sliding mode control (step S190).

On the other hand, the balance thrust F * and the differential thrust ΔF are calculated by the thrust calculator 110, and the balance thrust F * and the differential thrust ΔF are added by the adder 102 to obtain the feedforward control input Uff (step S200). ).
The adder 103 adds the feedforward control input Uff from the adder 102 and the sliding mode control input Uscm from the adder 129 (step S210).

The hydraulic feedback control unit (control means) 130 converts the calculated thrust of the final control input into a hydraulic pressure required for each pulley (step S220), and this hydraulic pressure command is converted into a solenoid 140 serving as a hydraulic valve drive source. The solenoid 140 of the hydraulic valve is controlled (step S230: control step).
In this way, the solenoid 140 of the hydraulic valve is controlled, and in the continuously variable transmission 30, the pulley of the continuously variable transmission moves in the axial direction by the hydraulic pressure supplied from the hydraulic valve, and the speed is changed.

According to this control, the sliding mode control is performed using the second target displacement x ₂ ^* which is delayed with respect to the original target displacement (first target displacement) x ₁ ^* of the pulley position to be controlled. Further, chattering of nonlinear input in sliding mode control and overshoot of the controlled object can be suppressed. As a result, the position control of the movable pulley for adjusting the gear ratio of the continuously variable transmission for the vehicle can be stably performed, the shift stability to the target gear ratio of the vehicle is improved, and the fuel efficiency of the vehicle is improved. A stable ride feeling can be obtained.

In particular, the time constant of the delay filter for generating the second target displacement x ₂ ^* is set to the minimum value when the magnitude of the first deviation e ′ is small within a certain range (within the first predetermined range). When the magnitude of the first deviation e ′ is large (outside the first predetermined range), the second target displacement is set so as to increase in proportion to the increase in the magnitude of the first deviation e ′. x ₂ ^* can be smoothly converged to the first target displacement x ₁ ^* , which is the original target value, and control can be performed smoothly.

  Further, since the integral gain of the switching function σ is changed so as to decrease as the first deviation e ′ increases, chattering of nonlinear input in sliding mode control and overshoot of the controlled object can be suppressed. Thereby, the position control of the movable pulley for adjusting the transmission ratio of the continuously variable transmission can be appropriately performed, and the effects of improving the fuel consumption of the vehicle, the feeling of riding, and the driving feeling can be obtained.

Further, based on the time constant, it is determined whether the control before the completion of the shift control is being performed, or whether the shift control is started from the steady state before the shift control. There is no reset command to set the integral term of the switching function σ to 0, and this reset command is issued at the start of shifting, so that unnecessary integral reset during shifting is prevented and smooth shifting is completed. At the start of the shift, the integrator of the sliding function control switching function σ can be reset to reduce the dead time and delay at the start of the shift, thereby improving the shift response.
(Other)
As described above, the best mode for carrying out the present invention has been described based on the embodiment. However, the specific configuration of the present invention is not limited to the above embodiment, and the scope of the invention is not deviated. Design changes and the like are included in the present invention.

For example, in the above-described embodiment, the shift control of the belt type continuously variable transmission has been described. However, the control device and the control method of the present invention are not limited to the shift control of other continuously variable transmissions, and other input adjustment amounts. Thus, the present invention can be widely applied to a control system that controls a controlled system that changes the state of the control system toward a state corresponding to the adjustment amount, for example, a continuously variable transmission.
Further, the setting of the time constant of the delay filter, the setting of the switching function σ integral gain, and the reset determination of the integrator are not limited to the above-described embodiment, and can be changed without departing from the spirit thereof. .

It is a block diagram which shows the structure of the control apparatus concerning one Embodiment of this invention. 1 is a system configuration diagram of a vehicle equipped with a CVT hydraulic control apparatus according to an embodiment of the present invention. It is a graph which shows the characteristic of control concerning one embodiment of the present invention. It is a flowchart explaining the control method concerning one Embodiment of this invention.

Explanation of symbols

10 Engine 20 Torque converter 25 Forward / reverse clutch 30 CVT (Belt type continuously variable transmission)
DESCRIPTION OF SYMBOLS 31 Primary pulley on driving side 31a Primary slide pulley 32 Secondary pulley on driven side 32a Secondary slide pulley 33 Belt 40 Oil pump 51 First pressure regulating valve 52 Second pressure regulating valve 61 Primary solenoid 62 Secondary solenoid 71 Primary pressure regulating valve 72 Secondary pressure regulating valve 100 CVT control unit (control device)
DESCRIPTION OF SYMBOLS 101 External input acquisition part 102,103 Adder part 110 Thrust, difference thrust calculating part 111 Balance thrust calculating part 112 Differential thrust calculating part 120 Sliding mode control part (sliding mode controller)
121a Gear ratio-stroke converter (first target value setting means)
121b Gear ratio-stroke conversion unit 122 Correction target displacement generation unit (second target value generation unit)
123 Deviation calculation unit 124 Integral gain setting unit for switching function (integral gain changing means)
125 Switching function design section (switching function design means)
126 Integral Reset Command (Reset Command)
127 linear input calculation unit (linear input calculation means)
128 Nonlinear Input Calculation Unit (Nonlinear Input Calculation Unit)
129 Addition unit 130 Hydraulic feedback control unit (control means)
140 Hydraulic valve solenoid (primary solenoid 61 and secondary solenoid 62)
150 Actual gear ratio calculation unit

Claims (11)

An actual value acquisition means for acquiring an actual value of a specific parameter related to the controlled object;
Target value setting means for setting a target value of the specific parameter;
A switching function design unit that designs a switching function for sliding mode control based on a deviation between the actual value acquired by the actual value acquisition unit and the target value set by the target value setting unit;
Nonlinear input computing means for computing a sliding mode control nonlinear input based on the switching function designed by the switching function design means;
Control means for controlling the object to be controlled by sliding mode control by a control input including a nonlinear input calculated by the nonlinear input calculating means;
The target value setting means is a first target value setting means for setting a first target value of the specific parameter in accordance with a target state of the control target, and the actual value and the first target value with respect to the first target value. And a second target value generating means for generating a second target value as the target value by using a delay filter that outputs a delay corresponding to the magnitude of the first deviation that is a deviation from the first target value. A control device using sliding mode control.   The sliding mode control according to claim 1, wherein the time constant of the delay filter is set to be larger as the magnitude of the first deviation is larger and smaller as the magnitude of the first deviation is smaller. Control device.   The time constant of the delay filter is a constant value when the magnitude of the first deviation is within a first predetermined range, and when the magnitude of the first deviation is greater than the first predetermined range. 3. The control device using sliding mode control according to claim 1, wherein the control device is set to increase in proportion to an increase in the magnitude of the first deviation. The integral gain changing means for changing the integral gain of the switching function according to the first deviation so as to become smaller as the first deviation is larger is provided. A control device using the sliding mode control according to item 1 . The integral gain changing means sets the integral gain of the switching function to a maximum value when the magnitude of the first deviation is within a second predetermined range, and the magnitude of the first deviation is the second value. If the first deviation is smaller, the larger the first deviation is, the smaller the first deviation is. The larger the first deviation is, the smaller the first deviation is. 5. The control device using sliding mode control according to claim 4 , wherein when the value is larger than a third predetermined range, the minimum value is set. A reset command means for resetting the integral term of the switching function designed by the switching function design means to 0;
The reset command means does not perform the reset command in the first state where it can be determined that the control target is in control before completion of the control, and the control target starts control from a steady state before control. The control device using sliding mode control according to any one of claims 1 to 5 , wherein the reset command is issued in a second state that can be determined. Linear input calculation means for calculating an equivalent control input which is a linear input of sliding mode control based on the mathematical model of the controlled object;
The control means controls the object to be controlled using a sum of a nonlinear input calculated by the nonlinear input calculating means and an equivalent control input calculated by the linear input calculating means as a control input. A control device using the sliding mode control according to any one of 1 to 6 . The control object is a movable member for adjusting the transmission ratio of the continuously variable transmission for a vehicle, wherein the specific parameter, characterized in that the position of the movable member, any one of claims 1-7 A control device using the sliding mode control according to item 1. An actual value acquisition step of acquiring an actual value of a specific parameter related to the controlled object;
A target value setting step for setting a target value of the specific parameter;
A switching function design step for designing a switching function for sliding mode control based on a deviation between the actual value acquired by the actual value acquisition step and the target value set by the target value setting step;
A non-linear input calculation step of calculating a non-linear input of sliding mode control based on the switching function designed by the switching function design step;
A control step of controlling the control object by sliding mode control with a control input including the nonlinear input calculated by the nonlinear input calculation step,
The target value setting step includes a first target value setting step of setting a first target value of the specific parameter according to a target state of the control target, and the actual value and the first target value with respect to the first target value. And a second target value generation step of generating a second target value as the target value by using a delay filter that outputs a delay according to the magnitude of the first deviation, which is a deviation from the first target value. And a control method using sliding mode control. An integral gain changing step of changing the integral gain of the switching function used in the switching function design step according to the first deviation so as to be smaller as the first deviation is larger, A control method using the sliding mode control according to claim 9 . A reset command step for resetting the integral term of the switching function designed by the switching function design step to zero;
In the reset command step, the reset command is not performed in the first state where it can be determined that the control target is being controlled before the control is completed, and the control target is controlled from the steady state before the control. The control method using sliding mode control according to claim 9 or 10 , wherein the reset command is issued in a second state that can be determined.

Images