JP2005147271A - Controller for continuously variable transmission - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a controller which reflects, on the control, the stability when controlling the clamping force of a continuously variable transmission, thereby enabling appropriate setting of the clamping force. <P>SOLUTION: The controller comprises a control stability detecting means (step S105) which detects the stability of control and a clamping force reduction amount determining/ calculating means (step S107) which determines or calculates the amount of the clamping force reduced according to the stability detected with the control stability means. The control stability detecting means comprises a time window integration processing means (step S202) which performs integration processing in accordance with the current time constant and a control stability reduction detecting means (step S205 and step S206) which compares the time window integration value obtained by the time window integration processing means and a predetermined value to detect the reduction in stability. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a control device for a continuously variable transmission whose torque capacity changes according to a clamping pressure.

  The belt-type continuously variable transmission and the traction-type continuously variable transmission transmit torque using the frictional force between the belt and the pulley and the shearing force of the traction oil between the disk and the roller. Therefore, the torque capacity of these continuously variable transmissions is set according to the pressure acting on the location where the torque is transmitted.

  The above-mentioned pressure in a continuously variable transmission is referred to as pinching pressure. Increasing the pinching pressure can increase the torque capacity and avoid slipping, but on the other hand, more power than necessary to generate high pressure. There are inconveniences such as consumption or reduction in power transmission efficiency. Therefore, in general, the clamping pressure is set as low as possible within a range in which unintended slip does not occur.

  For example, in a vehicle equipped with a continuously variable transmission, the engine speed can be controlled by the continuously variable transmission to improve fuel efficiency. In order to improve the transmission efficiency as much as possible, the clamping pressure is controlled to be set as low as possible within a range where no slip occurs. For that purpose, it is necessary to detect the pressure at which slipping starts (that is, the slip limit pressure). Conventionally, slip is detected by various methods, and the slip limit pressure is detected.

As an example, Patent Document 1 discloses a control device for setting a belt clamping pressure in a belt-type continuously variable transmission, which is in a state where the vehicle is running at a constant speed at a high speed on a flat road. A control device configured to allow a decrease in belt clamping pressure and to cancel the decrease in belt clamping pressure during unsteady running, where acceleration / deceleration is frequently performed or traveling on a rough road, is described. Has been. Japanese Patent Application Laid-Open No. H10-228561 describes an apparatus configured to control a hydraulic pressure (line pressure) related to the belt tension in consideration of a delay in a change in the line pressure.
JP 2003-120804 A JP-A-6-207657

  In the invention of the above-mentioned Patent Document 1, it is conceivable that the holding pressure is reduced when a predetermined condition is satisfied, and the reduction control is performed by, for example, feedback control. However, depending on how the control values such as the feedback control gain are set, there is a possibility that the responsiveness deteriorates, or conversely, hunting occurs and belt slippage occurs.

  The present invention has been made paying attention to the above technical problem, and provides a control device that sets an appropriate clamping pressure after confirming the stability of the control when controlling the clamping pressure of a continuously variable transmission. It is intended to do.

  The present invention is characterized in that the content of the reduction control of the clamping pressure is determined based on the stable state of the control. More specifically, the invention of claim 1 relates to a control stability detecting means for detecting control stability in a control device for a continuously variable transmission for controlling a clamping pressure for setting a torque capacity of the continuously variable transmission. A control apparatus comprising: a pinching pressure reduction unit that controls the reduction of the pinching pressure in accordance with the stability detected by the control stability detection unit.

  Further, the invention of claim 2 is the time window integration processing means according to claim 1, wherein the control stability detection means performs integration processing of the actual clamping pressure filtered by a filter frequency corresponding to a time constant, and the time A control apparatus comprising: a control stability decrease detecting unit that detects a decrease in control stability by comparing an integrated value obtained by the window integration processing unit with a predetermined value. is there.

  According to a third aspect of the present invention, in the first or second aspect, the control stability reduction detecting means compares the time window integrated value obtained by the time window integrating means with at least one threshold value. The control device further comprises a clamping pressure reduction amount determining means for determining a reduction amount of the clamping pressure based on the result of comparison by the comparison means.

  Further, the invention according to claim 4 is the method according to claim 1, wherein the control stability detecting means is obtained by a physical quantity learning means for obtaining a physical quantity of at least one of a dead time during control and a time constant, and the physical quantity learning means. And a clamping pressure reduction amount calculating means for calculating a reduction amount of the clamping pressure based on at least one of the dead time and the time constant.

  According to a fifth aspect of the present invention, there is provided a control device for a continuously variable transmission that feedback-controls a clamping pressure that sets a torque capacity of the continuously variable transmission, and a control state learning detection unit that detects a control state by learning control The control state evaluation means for evaluating the stability / instability of the control state when the feedback control is performed based on the state detected by the control state learning detection means, and the evaluation by the control state evaluation means It is a control device characterized by comprising clamping pressure reduction control means for controlling the reduction.

  Further, the control content is set based on the physical quantity learned by the control state learning detection means, and the clamping pressure is feedback-controlled. The learning control is configured to learn at least one of a time constant and a dead time. Further, the control state evaluation means is configured to evaluate the stability / instability of the control state based on the detected fluctuation state of the clamping pressure. The fluctuation state is a time window integrated value of a component of a predetermined frequency band of the detected clamping pressure. The frequency band is a band determined based on a physical quantity obtained by the learning control.

  Therefore, according to the invention of claim 1, the stability of the control is detected, and the clamping pressure is reduced based on this. Therefore, the current control state can be more accurately reflected in the amount of decrease in the pinching pressure, and the amount of decrease in the pinch pressure can be optimized.

  According to the invention of claim 2, the process of integrating the amount of change of the actual clamping pressure per unit time is performed. If the fluctuation of the actual clamping pressure becomes severe, the integrated value increases. Therefore, it is possible to accurately grasp the decrease in stability during control, and it is possible to accurately detect the decrease in stability.

  According to the invention of claim 3, at least one threshold value is set, and the stability is evaluated by comparing the threshold value with the time window integrated value. That is, every time the threshold value is exceeded, the amount of decrease in the clamping pressure becomes smaller. Therefore, the clamping pressure can be reduced stepwise as the stability decreases, so that the reduction control of the clamping pressure can be performed stably.

  Furthermore, according to the invention of claim 4, the dead time or the time constant is obtained by learning control. Then, the clamping pressure is reduced based on the dead time or the time constant obtained by the learning control. Therefore, the current control state can be more accurately reflected in the amount of decrease in the pinching pressure, and the amount of decrease in the pinch pressure can be optimized.

  According to the invention of claim 5, the current control state is detected by learning control prior to lowering the clamping pressure. Then, the control state detected by the learning control is evaluated to determine whether it is stable or unstable. The clamping pressure is controlled to be lowered depending on the result of the evaluation. Accordingly, since the current control state is reflected in the reduction control of the clamping pressure, the reduction control of the clamping pressure can be performed more appropriately.

  Next, the present invention will be described based on specific examples. First, an example of a drive system including a power source and a continuously variable transmission targeted in the present invention will be described. FIG. 4 schematically shows an example of a drive system including a belt-type continuously variable transmission 1. The continuously variable transmission 1 is connected to a power source 5 via a forward / reverse switching mechanism 2 and a fluid transmission mechanism 4 with a lock-up clutch 3.

  The power source 5 is composed of an internal combustion engine, or an internal combustion engine and an electric motor, or an electric motor. In the following description, the power source 5 is referred to as the engine 5. The fluid transmission mechanism 4 has a configuration similar to that of, for example, a conventional torque converter, and includes a pump impeller rotated by the engine 5, a turbine runner disposed to face the pump impeller, and a stator disposed therebetween. The turbine runner is rotated by supplying a spiral flow of fluid generated by a pump impeller to the turbine runner, and torque is transmitted.

  In such torque transmission via the fluid, inevitable slip occurs between the pump impeller and the turbine runner, and this causes a reduction in power transmission efficiency. Therefore, the input member such as the pump impeller and the turbine runner A lock-up clutch 3 that directly connects an output side member such as the above is provided. The lock-up clutch 3 is configured to be controlled by hydraulic pressure, and is controlled to a fully engaged state, a fully released state, and a slip state that is an intermediate state between them, and the slip rotation speed can be appropriately controlled. It has become.

  The forward / reverse switching mechanism 2 is a mechanism that is employed when the rotational direction of the engine 5 is limited to one direction, and outputs the input torque as it is or reversely outputs it. It is configured. In the example shown in FIG. 4, a double pinion type planetary gear mechanism is employed as the forward / reverse switching mechanism 2. That is, a ring gear 7 is arranged concentrically with the sun gear 6, and a pinion gear 8 meshed with the sun gear 6 and the pinion gear 8 and another pinion gear 9 meshed with the ring gear 7 are arranged between the sun gear 6 and the ring gear 7. The pinion gears 8 and 9 are held by the carrier 10 so as to rotate and revolve freely. A forward clutch 11 that integrally couples the two rotating elements (specifically, the sun gear 6 and the carrier 10) is provided, and the direction of the torque that is output by selectively fixing the ring gear 7 There is provided a reverse brake 12 that reverses.

  The continuously variable transmission 1 has the same configuration as a conventionally known belt-type continuously variable transmission, and each of a driving pulley 13 and a driven pulley 14 arranged in parallel to each other includes a fixed sheave, a hydraulic type The movable sheave is moved back and forth in the axial direction by the actuators 15 and 16. Therefore, the groove width of each pulley 13 and 14 is changed by moving the movable sheave in the axial direction, and accordingly, the winding radius of the belt 17 wound around each pulley 13 and 14 (the effective diameter of the pulleys 13 and 14). ) Changes continuously, and the gear ratio changes steplessly. The drive pulley 13 is connected to a carrier 10 that is an output element in the forward / reverse switching mechanism 2.

  The hydraulic actuator 16 in the driven pulley 14 is supplied with hydraulic pressure (line pressure or its correction pressure) according to the torque input to the continuously variable transmission 1 via a hydraulic pump and a hydraulic control device (not shown). Yes. Therefore, each sheave in the driven pulley 14 holds the belt 17 so that tension is applied to the belt 17, and a holding pressure (contact pressure) between the pulleys 13, 14 and the belt 17 is ensured. . On the other hand, the hydraulic actuator 15 in the drive pulley 13 is supplied with pressure oil corresponding to the speed ratio to be set, and is set to a groove width (effective diameter) corresponding to the target speed ratio. .

  The driven pulley 14 is connected to a differential 19 through a gear pair 18, and torque is output from the differential 19 to driving wheels 20. Therefore, in the above drive mechanism, the lockup clutch 3 and the continuously variable transmission 1 are arranged in series between the engine 5 and the drive wheels 20.

  Various sensors are provided in order to detect the operation state (running state) of the vehicle on which the continuously variable transmission 1 and the engine 5 are mounted. That is, the turbine speed sensor 21 that detects the input speed (the speed of the turbine runner) to the continuously variable transmission 1 and outputs a signal, and the input speed that detects the speed of the drive pulley 13 and outputs the signal. A sensor 22, an output rotation speed sensor 23 that detects the rotation speed of the driven pulley 14 and outputs a signal, and a hydraulic pressure sensor 24 that detects the pressure of the hydraulic actuator 16 on the driven pulley 14 side for setting the belt clamping pressure are provided. ing. Although not specifically shown, an accelerator opening sensor that detects a depression amount of the accelerator pedal and outputs a signal, a throttle opening sensor that detects a throttle valve opening and outputs a signal, and a brake pedal are depressed. A brake sensor or the like that outputs a signal in case is provided.

  A transmission is used to control the engagement / release of the forward clutch 11 and the reverse brake 12, the control of the clamping force of the belt 17, the control of the transmission ratio, and the control of the lockup clutch 3. An electronic control device (CVT-ECU) 25 is provided. The transmission electronic control unit 25 is configured mainly by a microcomputer as an example, and performs calculations according to a predetermined program based on input data and data stored in advance, and performs various operations such as forward, reverse, or neutral. And the required clamping pressure setting, the gear ratio setting, the engagement / release of the lock-up clutch 3, the slip rotation speed, and the like are executed.

  Here, an example of data (signal) input to the transmission electronic control unit 25 is as follows: a signal of the input rotation speed (input rotation speed) Nin of the continuously variable transmission 1 and an output of the continuously variable transmission 1. A signal of the rotation speed (output rotation speed) No is input from the corresponding sensor. An engine electronic control unit (E / G-ECU) 26 for controlling the engine 5 receives a signal of an engine speed Ne, an engine (E / G) load signal, a throttle opening signal, an accelerator pedal (not shown). )), The accelerator opening signal is input.

  According to the continuously variable transmission 1, the engine speed, which is the input speed, can be controlled steplessly (in other words, continuously), so that the fuel efficiency of a vehicle equipped with the engine speed can be improved. For example, the target driving force is obtained based on the required driving amount represented by the accelerator opening and the vehicle speed, and the target output necessary to obtain the target driving force is obtained based on the target driving force and the vehicle speed. The engine speed for obtaining the target output with the optimum fuel consumption is obtained based on a map prepared in advance, and the gear ratio is controlled so as to be the engine speed.

  In order not to impair such an improvement in fuel consumption, the power transmission efficiency in the continuously variable transmission 1 is controlled to a good state. Specifically, the torque capacity of the continuously variable transmission 1, that is, the belt clamping pressure, can transmit the target torque determined based on the engine torque, and the belt clamping pressure is as low as possible without causing the belt 17 to slip. It is controlled to become. For example, the belt clamping pressure controls the continuously variable transmission 1 in a so-called unsteady traveling state such as when acceleration / deceleration is performed relatively frequently or when traveling on a rough road with uneven or uneven road surfaces. It is set to a relatively high pressure such as the line pressure that is the total source pressure in the hydraulic system or its correction pressure.

  On the other hand, in steady running conditions such as running on a flat road at a certain speed or above, or in a quasi-steady running condition equivalent to this, the lowest pressure that can transmit input torque without slipping, that is, the limit clamping pressure In order to detect this, the belt clamping pressure is gradually reduced. The belt clamping pressure is set to a belt clamping pressure obtained by adding a predetermined safety factor or a pressure for setting a margin transmission torque for slipping to the detected limit clamping pressure. The belt clamping pressure in the continuously variable transmission is preferably as low as possible within a range where torque can be transmitted without causing slippage.

  However, if a change occurs in the operating state such as an increase in input torque due to an accelerator depressing operation or the like during a decrease control of the belt clamping pressure, if the stability of the clamping pressure control is low, the belt clamping force corresponding to the increase in the input torque is reduced. The increase in pressure may be delayed, causing belt slip. Similarly, even when an abnormality occurs in the hydraulic circuit, the stability of the clamping pressure control is reduced, and belt slippage may occur. In addition, in order to prevent belt slipping, considering the variation in responsiveness due to the influence of the hydraulic system, etc., setting the belt clamping pressure at normal times assuming a low stability state will make the clamping pressure as small as possible. There may be a case where the fuel efficiency improvement effect due to the reduction to a low pressure cannot be sufficiently obtained. Therefore, the control device of the present invention is configured so that an appropriate clamping pressure can be set by reflecting in the control the stability when the clamping pressure of the continuously variable transmission is controlled. A specific example of the control will be described below.

  1 and 2 are flowcharts showing an example thereof, and FIG. 3 is a part of a time chart when the control shown in the flowchart is performed.

  First, it is determined whether or not the dead time and time constant have been learned (step S101). If a negative determination is made in step S101, learning control (step S102) is performed, and the same determination is performed at the next routine execution. If the determination in step S101 is affirmative, that is, if learning control has been completed, it is determined whether or not the clamping pressure feedback control is possible (step S103). Here, the criteria for judgment are that the learned dead time and time constant are within a predetermined range, the shift position is in the drive range, the pressure sensor and the hydraulic circuit are normal, and the oil temperature is within the normal range. This is the case when all of the above is true.

  If the determination in step S103 is affirmative, clamping pressure feedback control is performed (step S104). The gain such as a coefficient in the clamping pressure feedback control is determined from the physical quantity calculated by the learning control such as the dead time and the time constant learned in step S102 and the oil temperature. When the feedback control is started, it is determined by the stability determination control whether the feedback control is in a stable state (step S105). Details of the stability determination control will be described later.

  Next, the result of the stability determination control is evaluated (step S106). If the determination in step S106 is affirmative, that is, if it is evaluated as stable, the clamping pressure reduction control is performed according to the learning value obtained in step S102 (step S107). If negative determination is made in step S103, that is, if feedback control is not possible, or if negative determination is made in step S106, that is, if feedback control is unstable, the clamping pressure reduction control is performed. The clamping pressure control is performed without performing the above.

  That is, initially, learning control is performed in step S102, and physical quantities such as dead time and time constant are obtained. If the determination in step S103 is affirmative, that is, waiting for the feedback control to be possible, the feedback control is performed. While the feedback control is performed and the stability is maintained, the clamping pressure reduction control is performed, and when it becomes unstable, a negative determination is made in step S106, the clamping pressure reduction control is stopped, and the clamping pressure is not reduced. Pressure control is performed (step S108).

  Next, stability determination control will be described. The stability determination control is control for determining the stability of the feedback control. First, the frequency of the band pass filter is determined based on the time constant obtained by the learning control in step S102 (step S201). Then, the actual clamping pressure is integrated over the time window (step S202). That is, the components in the predetermined frequency band of the fluctuation component of the clamping pressure are integrated in the time window.

  Then, it is determined whether a determination condition is satisfied (step S203). This judgment condition is that the clamping pressure command value within a predetermined time is within a predetermined range, that the time that can be integrated in the time window has elapsed, that the rate of change of the clamping pressure command value is within a predetermined range, This is established when all conditions such as no up command are established. That is, the change in the clamping pressure command value is small, and it is only necessary that the average value of the actual clamping pressure and the clamping pressure command value substantially match. If the determination condition is not satisfied, that is, if a negative determination is made in step S203, the following routine is omitted and the routine is exited.

  On the other hand, if the determination in step S203 is affirmative, that is, if the determination condition is satisfied, it is determined whether or not the time window integrated value is greater than a predetermined value (step S204).

  If the determination in step S204 is affirmative, it is determined that the feedback control is unstable (step S206). If a negative determination is made in step S204, the feedback control is determined to be stable (step S205).

  Next, the passage of time during control will be described with reference to FIG. FIG. 3 (a) shows a process when no control abnormality such as hunting has occurred. FIG. 3B shows the progress when hunting occurs.

  In the normal case (FIG. 3A), when feedback control is started (time t1, corresponding to step S104), the actual clamping pressure changes so as to match the command value (time t1 to time t2. In the figure, the actual clamping pressure has decreased). Then, a predetermined time (from time t1 to time t2) elapses, and stability determination control is started (time t2 corresponding to step S105). Since the actual clamping pressure does not fluctuate in the normal state, the clamping pressure bandpass integrated value, that is, the time window integrated value does not increase (time t2 to t3, corresponding to steps S201 to S203).

  When the determination condition is satisfied after a lapse of time (corresponding to steps S203 and S204 at time t3), the time window integrated value does not exceed the predetermined value, so a negative determination is made in step S204, and the feedback control is determined to be stable. (Corresponding to step S205). Then, a clamping pressure reduction command is output and the clamping pressure reduction control is performed (after time t3).

  On the other hand, when hunting occurs (FIG. 3B), when feedback control is started, the actual clamping pressure changes so as to match the command value as in the normal case (from t1). At time t2, the actual clamping pressure is decreasing in the figure). Then, stability determination control is started (time t2, corresponding to step S105).

  If the actual clamping pressure causes hunting during the period from the start of the stability determination control until the determination condition is satisfied (time t2 to t3), the cumulative rate of change of the actual clamping pressure per unit time is the clamping pressure bandpass integrated value. That is, it is output as a time window integrated value. Therefore, when this time window integrated value is compared with a predetermined value and exceeds the predetermined value, it is determined that the feedback control is unstable (at time t3, corresponding to step S206).

  In other words, since the control state stability is detected and the clamping pressure is reduced based on this, the current control state can be more accurately reflected in the clamping pressure reduction amount, and the clamping pressure reduction amount can be optimized. can do.

  In addition, since the process of accumulating the amount of change of the actual clamping pressure per unit time is performed, if the fluctuation of the actual clamping pressure becomes intense, the amount of change per unit time also becomes intense. Therefore, it is possible to accurately grasp the decrease in stability during control, and it is possible to accurately detect the decrease in stability.

  At least one threshold value is set, and the stability is evaluated by comparing the threshold value with the time window integrated value. Therefore, since the amount of decrease in the clamping pressure can be suppressed as the stability decreases, the clamping pressure reduction control can be performed stably.

  Although only one predetermined value is set in step S204 of the above embodiment, a plurality of values may be set. FIG. 4 is a time chart showing an example when two predetermined values are set. Feedback control is started (time t1), and stability determination control is started (time t2). When the predetermined time elapses (time point t2 to t3) and the determination condition is satisfied and the time window integrated value exceeds the second predetermined value, the feedback control is determined to be a quasi-unstable state (time point t3). And clamping pressure is controlled in the direction which suppresses the fall amount. When the time further elapses and the time window integrated value exceeds the first clamping pressure, it is determined that the feedback control is in an unstable state (at time t4). Then, the reduction control of the clamping pressure is stopped.

  In this case, at least two threshold values are set, and the stability is evaluated by comparing the threshold values with the time window integrated value. That is, every time the threshold value is exceeded, the amount of decrease in the clamping pressure becomes smaller. Therefore, since the reduction of the clamping pressure can be suppressed in a stepwise manner as the stability decreases, stable clamping pressure reduction control can be performed.

  In the embodiment described above, the stability determination is performed based on the time window integrated value, but this may be performed based on the time constant and the dead time. FIG. 5 is a flowchart showing an example in which stability determination is performed based on a time constant and dead time.

  First, it is determined whether or not dead time and time constant have been learned (step S301). If a negative determination is made in step S301, learning control (step S302) is performed, and the same determination is performed again at the next routine execution. Details of the learning control will be described later. If the determination in step S301 is affirmative, that is, if the learning control has been completed, it is determined whether or not the feedback control of the clamping pressure is possible (step S303). Here, the criteria for judgment are that the learned dead time and time constant are within a predetermined range, the shift position is in the drive range, the pressure sensor and the hydraulic circuit are normal, and the oil temperature is within the normal range. This is the case when all of the above is true.

  If the determination in step S303 is affirmative, clamping pressure feedback control and clamping pressure reduction control are performed (step S304). The proportional term in this clamping pressure feedback control is determined from the dead time and time constant learned in step S302 and the oil temperature. If a negative determination is made in step S303, that is, if feedback control is impossible, the clamping pressure control is performed without performing the clamping pressure reduction control (step S305).

  That is, initially, learning control is performed in step S302, and a dead time and a time constant are obtained. If the determination in step S303 is affirmative, that is, waiting for the feedback control to be possible, the feedback control is performed. While the feedback control is performed and the stability is maintained, the clamping pressure reduction control is performed. When the feedback control cannot be performed due to instability, a negative determination is made in step S303, and the clamping pressure reduction control is stopped. To do.

  Next, learning control will be described. Learning control is control that is performed when a dead time and a time constant indicating the stability of control are obtained. FIG. 6 is a flowchart showing an example of the learning control.

  First, it is determined whether or not a precondition for performing learning control is established (step S401). The preconditions are that the vehicle is in a quasi-steady running state, is not a rough road, the vehicle speed is within a predetermined range, and that the clamping pressure up command is not output, etc. That is, the precondition is satisfied when the change in the command value of the clamping pressure is very small.

  If a negative determination is made in step S401, that is, if the precondition is not satisfied, it is determined whether or not the current clamping pressure is stepping up (step S409). If the determination in step S409 is affirmative, step-up of the clamping pressure is stopped so that the precondition is satisfied (step S410). Further, when a negative determination is made in step S409, that is, when the step-up is not in progress, the precondition is not satisfied, and other causes can be considered. Therefore, the learning control is executed until the other causes are resolved. Not done.

  On the other hand, if the determination in step S401 is affirmative, it is determined whether or not the current step-up is in progress (step S402). If a negative determination is made in step S402, the clamping pressure is stepped up in order to perform step input to the control system in order to obtain a learning value (dead time, time constant) (step S403). Thereafter, the process proceeds to step S404. On the other hand, if the determination in step S402 is affirmative, step S403 is in progress, so step S403 is skipped and the process proceeds to step S404.

  Next, it is determined whether a predetermined time has elapsed from when the clamping pressure is stepped up until the clamping pressure is stabilized (step S404). If a negative determination is made in step S404, that is, if the predetermined time has not elapsed, the process waits until it elapses. If the predetermined time has elapsed and the determination in step S404 is affirmative, the dead time and time constant at that time are detected, and the clamping pressure is lowered to restore the clamping pressure (step S404). S405).

  Then, it is determined whether learning is successful (step S406). Here, the conditions for success of learning are when the variation range of the clamping pressure command value within a predetermined time is within a predetermined value, the dead time and the time constant are within a predetermined value, etc. is there.

  If the determination in step S406 is affirmative, that is, if learning is successful, learning is completed (step S408). On the other hand, if a negative determination is made in step S406, that is, if learning fails, the learning value is cleared (step S408), and learning is performed again.

  FIG. 7 is a time chart showing an example of control in the above-described embodiment. When learning control is started (time t1, equivalent to step S401) and the clamping pressure command value is stepped up (time t1, step S403), the time constant (t2 to t3) is delayed by a dead time (t1 to t2). ) The actual clamping pressure increases with a slope corresponding to. During this time, the change in the actual clamping pressure is recorded, and when a predetermined time elapses (from time t3 to time t4), the time constant and dead time are detected. Then, learning is completed (time t4, corresponding to step S407), and the clamping pressure is reduced (time t4 to t5, corresponding to step S304).

  Note that the pinching pressure that finally becomes constant after the pinching pressure drop can be determined according to the learned dead time and time constant. Therefore, since the current control state is reflected in determining the reduction amount of the clamping pressure, the reduction amount of the clamping pressure can be determined more appropriately.

  Furthermore, a dead time or a time constant is obtained by learning control. Then, the clamping pressure is reduced based on the dead time or the time constant obtained by the learning control. Therefore, the current control state can be more accurately reflected in the amount of decrease in the pinching pressure, and the amount of decrease in the pinch pressure can be optimized.

  Prior to lowering the clamping pressure, the current control state quantity is detected by learning control. Then, the control state quantity detected by the learning control is evaluated to determine whether it is stable or unstable. The reduction amount of the clamping pressure is determined and controlled according to the result of the evaluation. Accordingly, since the current control state is reflected in determining the reduction amount of the clamping pressure, the reduction amount of the clamping pressure can be determined more appropriately.

  Here, the relationship between each of the above specific examples and the present invention will be briefly described. The functional means in step S105 described above corresponds to the “control stability detecting means”, and the functional means in step S107 is referred to as “clamping pressure drop”. It corresponds to “means”. The functional means in step S202 corresponds to “time window integration processing means”, and the functional means in steps S205 and S206 correspond to “control stability decrease detection means”. Further, the functional means in step S204 corresponds to “comparison means”, and the functional means in step S107 corresponds to “clamping pressure drop amount determination means”. The functional means in step S302 corresponds to “physical quantity learning means”, and the functional means in step S304 corresponds to “clamping pressure decrease amount calculating means”. In addition, each functional unit in steps S102 and S302 corresponds to a “control state learning detection unit”, and each functional unit in steps S103, S105, S106, and S303 corresponds to a “control state evaluation unit”. The functional means in steps S107 and S304 correspond to “clamping pressure drop control means”.

  In addition, the process in step S201 can use an appropriate filter process besides the process by a band pass filter. In short, it is only necessary to be able to capture hunting during feedback control. That is, any process that can capture the instability of the control system may be used. For example, instability of the control system may be determined from the time constant and dead time instead of the filter processing.

It is a flowchart for demonstrating an example of control by the control apparatus of this invention. 5 is a flowchart for explaining an example of stability determination control in the control by the control device of the present invention. It is a time chart for demonstrating an example of control by the control apparatus of this invention. It is a time chart for demonstrating an example when the predetermined value of control by the control apparatus of this invention is taken in two steps. It is a flowchart for demonstrating the other example of control by the control apparatus of this invention. 5 is a flowchart for explaining an example of learning control in the control by the control device of the present invention. It is a time chart for demonstrating the other example of control by the control apparatus of this invention. It is a figure which shows typically an example of the drive system containing the continuously variable transmission made into object by this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Continuously variable transmission, 3 ... Lock-up clutch, 5 ... Engine (power source), 13 ... Drive pulley, 14 ... Driven pulley, 15, 16 ... Hydraulic actuator, 17 ... Belt, 20 ... Drive wheel, 25 ... Shift Electronic control unit for machine (CVT-ECU).

Claims (5)

In the control device for a continuously variable transmission that controls the clamping force that sets the torque capacity of the continuously variable transmission,
Control stability detecting means for detecting control stability;
A control device for a continuously variable transmission, comprising: a clamping pressure reduction means for performing a clamping pressure reduction control according to the stability detected by the control stability detection means. A time window integration processing means for performing an integration process of the actual clamping pressure filtered by the filter frequency corresponding to the current time constant;
And a control stability decrease detecting unit configured to detect a decrease in control stability by comparing the integrated value obtained by the time window integration processing unit with a predetermined value. The control device for a continuously variable transmission according to Item 1. Comparing means for comparing the time window integrated value obtained by the control stability decrease detecting means with the time window integrating means with at least one threshold value;
The control of a continuously variable transmission according to claim 1 or 2, further comprising a clamping pressure reduction amount determining means for determining a reduction amount of the clamping pressure based on a result of comparison by the comparison means. apparatus. A physical quantity learning means for obtaining a physical quantity of at least one of a dead time at the time of control and a time constant;
2. The nipping pressure reduction amount calculating unit that calculates a nipping pressure decrease amount based on at least one of a dead time and a time constant obtained by the physical quantity learning unit. Control device for step transmission. In a control device for a continuously variable transmission that feedback-controls the clamping force that sets the torque capacity of the continuously variable transmission,
Control state learning detecting means for detecting the state of control by learning control;
Control state evaluation means for evaluating the stability and instability of the control state when performing the feedback control based on the state detected by the control state learning detection means;
A control device for a continuously variable transmission, comprising: a clamping pressure reduction control unit that controls the reduction of the clamping pressure in accordance with an evaluation result by the control state evaluation unit.

Images