JP6791689B2 - Control device for continuously variable transmission - Google Patents

Description

The present invention relates to a control device for a continuously variable transmission.

In recent years, as an automatic transmission for a vehicle, a continuously variable transmission (CVT (Continuously Variable Transmission)) that can change the gear ratio steplessly without a shift shock has been widely put into practical use. The continuously variable transmission includes a variator including a primary pulley provided on the input shaft, a secondary pulley provided on the output shaft, and a power transmission member such as a chain that is hung on these pulleys. In the variator, the gear ratio is changed steplessly by changing the groove width of each pulley and changing the winding diameter of the power transmission member. In the control device of the continuously variable transmission, the speed change oil for changing the gear ratio of the variator and the clamp oil for preventing the power transmission member from slipping are controlled.

In addition, the continuously variable transmission is provided with a forward / backward switching mechanism for switching between forward and reverse (forward and reverse rotation of the drive wheels) of the vehicle. The forward / backward switching mechanism has a forward clutch that is engaged when the vehicle is moving forward and a reverse clutch that is engaged when the vehicle is moving backward. The control device for the continuously variable transmission controls the clutch oil pressure for engaging / releasing each clutch of the forward / backward switching mechanism.

As the hydraulic control for the clutch of the forward / backward switching mechanism, for example, in Patent Document 1, the clutch input torque input to the forward / backward switching mechanism is obtained, and the clutch input torque-clutch hydraulic map (for example, the clutch oil pressure map) is obtained according to the clutch input torque. It is disclosed that the clutch hydraulic pressure is obtained with reference to a linear map showing the relationship between the clutch input torque and the clutch hydraulic pressure, and the hydraulic pressure for engaging the clutch is controlled based on this clutch hydraulic pressure.

Japanese Unexamined Patent Publication No. 2013-24327

By the way, a torque other than the torque generated by the engine may be input to the continuously variable transmission. For example, when a tire hits a bollard while parking, an excessive torque is input to the continuously variable transmission from the wheel side. If the power transmission member slips on the variator due to this torque, the pulley and the power transmission member may be worn or damaged. As a method of avoiding slipping on the variator, for example, there is a method of sliding the clutch of the forward / backward switching mechanism before the variator. In this method, by supplying the minimum necessary hydraulic pressure to the clutch so that the clutch does not slip, the clutch slips before the variator when torque is input from the wheel side.

In the technique disclosed in Patent Document 1, the clutch oil pressure is acquired by using a linear map showing the relationship between the clutch input torque and the clutch oil pressure provided in advance. However, in the clutch, the friction coefficient of the friction material and the like vary from vehicle to vehicle. Therefore, in order to acquire the clutch oil pressure suitable for the clutch mounted on the vehicle (particularly, the minimum pressure required to prevent the clutch from slipping), the relationship between the clutch input torque and the clutch oil pressure is learned for each vehicle. There is a need.

Even if this learning is performed, if the learning accuracy is low, the clutch oil pressure obtained from the linear map after learning may be too high, for example, higher than the minimum required oil pressure to prevent the clutch from slipping. .. If the clutch oil pressure becomes too high above the minimum required oil pressure, the power loss of the oil pump that generates the oil pressure will increase, resulting in lower fuel consumption of the vehicle and more than the variator when torque is input from the wheel side. There is a risk that the clutch will not slip first. Therefore, it is desired to improve the accuracy of this learning.

The present invention has been made to solve the above problems, and it is possible to improve the accuracy of learning the relationship between the input torque in the forward / backward switching mechanism and the clutch oil pressure corresponding to the input torque. It is an object of the present invention to provide a control device for a transmission.

The continuously variable transmission control device according to the present invention engages a forward / backward switching mechanism having a forward clutch engaged when the vehicle moves forward and a reverse clutch engaged when the vehicle moves backward, and each clutch of the forward / backward switching mechanism. It is a control device of a continuously variable transmission provided with a hydraulic circuit for supplying hydraulic pressure for the operation, and is a torque acquisition means for acquiring input torque input to a forward / backward switching mechanism and hydraulic pressure supplied from the hydraulic circuit to each clutch. With respect to the hydraulic control means for controlling the clutch and at least one of the forward clutch and the reverse clutch, the input torque acquired by the torque acquisition means and the learning target clutch obtained by controlling the hydraulic control means slip. A learning value acquisition means for acquiring a learning value consisting of a minimum required hydraulic pressure and a learning value, a storage means for storing the learning value acquired by the learning value acquisition means, and a learning target for which the learning value is acquired by the learning acquisition means. For the clutch, using a predetermined number of learning values stored in the storage means, learning the relationship between the input torque and the clutch hydraulic pressure for engaging the learning target clutch according to the input torque, and learning to show the relationship. It is characterized by having a learning means for acquiring information.

According to the control device for the continuously variable transmission according to the present invention, the input torque and the clutch do not slip by controlling the minimum required oil pressure so that the clutch does not slip according to the input torque of the forward / backward switching mechanism. It is possible to accurately acquire the learning value consisting of the minimum required oil pressure. Further, according to the control device for the continuously variable transmission according to the present invention, by using a predetermined number of the learned values, the input torque and the clutch oil pressure corresponding to the input torque (particularly, the minimum necessary to prevent the clutch from slipping). It is possible to improve the accuracy of learning the relationship with the hydraulic pressure. As a result, the clutch oil supply to the clutch of the forward / backward switching mechanism and the clamp oil supply to the pulley of the variator can be reduced, so that the fuel consumption can be improved. Further, by supplying the clutch oil to each clutch based on the learning information (for example, a straight line) indicating the relationship between the input torque acquired in the learning and the clutch oil pressure, a torque other than the engine torque is input to the continuously variable transmission. In that case, the clutch of the forward / backward switching mechanism can be slid before the variator, and the variator can be protected.

In the continuously variable transmission control device according to the present invention, it is preferable that the learning means performs regression analysis using a predetermined number of learned values and calculates a regression line showing the relationship between the input torque and the clutch oil pressure. With this configuration, it is possible to obtain a highly accurate straight line or curve showing the relationship between the input torque and the clutch oil pressure corresponding to the input torque.

In the continuously variable transmission control device according to the present invention, the learning means preferably calculates a regression line indicating the relationship between the input torque and the clutch oil pressure. With this configuration, it is possible to obtain a highly accurate straight line showing the relationship between the input torque and the clutch oil pressure corresponding to the input torque.

In the control device for the continuously variable transmission according to the present invention, the learning value acquisition means inputs between a preset learning reflection upper limit straight line and a learning reflection lower limit straight line when a regression line has not yet been calculated by the learning means. When there is a coordinate point consisting of torque and the minimum hydraulic pressure required to prevent the learning target clutch from slipping, it is preferable to adopt the minimum required hydraulic pressure to prevent the learning target clutch from slipping as the learning value. .. With this configuration, the flood control with large variation can be excluded from the learning value, so that a learning value with high accuracy can be obtained and the learning accuracy can be improved.

In the control device for the continuously variable transmission according to the present invention, it is preferable that the learning reflection upper limit straight line and the learning reflection lower limit straight line are set in advance based on the friction coefficient of the friction material of the clutch to be learned. With this configuration, an appropriate learning reflection upper limit straight line and learning reflection lower limit straight line can be set in consideration of the variation in the friction coefficient of the friction material of the clutch.

In the control device for the stepless transmission according to the present invention, when the learning value acquisition means has already calculated the regression line, the learning reflection upper limit straight line and the minus side are separated by the first distance from the regression line to the plus side. A learning reflection lower limit straight line separated by a second distance is set in, and coordinates consisting of the input torque and the minimum hydraulic pressure required to prevent the learning target clutch from slipping between the set learning reflection upper limit straight line and the learning reflection lower limit straight line. It is preferable to adopt the minimum required hydraulic pressure as the learning value so that the input torque and the learning target clutch do not slip when a point exists. With this configuration, the learning reflection area can be narrowed down according to the regression line, and more accurate learning values can be obtained, and the learning accuracy can be improved.

In the control device for the continuously variable transmission according to the present invention, it is preferable that the second distance on the minus side is smaller than the first distance on the plus side. With such a configuration, it is possible to prevent the hydraulic pressure supplied to the clutch from becoming too low, so that the clutch can be prevented from slipping and the clutch can be protected.

The control device for the continuously variable transmission according to the present invention includes a slip determination means for determining whether or not the learning target clutch is slipping, and the learning value acquisition means slips when it is determined that the learning value acquisition means is slipping. By controlling the pressure increase with the hydraulic control means until it is determined that the clutch is not slipping by the determination means, the minimum required oil pressure is obtained so that the clutch to be learned does not slip, and when it is determined that the clutch is not slipping by the slip determination means, the clutch slips. It is preferable to acquire the minimum necessary oil pressure so that the clutch to be learned does not slip by controlling the pressure reduction with the hydraulic control means until it is determined that the clutch is slipping by the determination means. With this configuration, the oil pressure can be accurately controlled until the minimum required oil pressure is reached to prevent the clutch from slipping, and the minimum required oil pressure (learning value) to prevent the clutch from slipping according to the input torque can be accurately controlled. Can be obtained well.

According to the present invention, it is possible to improve the accuracy of learning the relationship between the input torque in the forward / backward switching mechanism and the clutch oil pressure corresponding to the input torque.

It is a block diagram which shows the structure of the control device of the continuously variable transmission which concerns on embodiment. It is a figure which shows the hydraulic circuit for the forward / backward switching mechanism of the valve body shown in FIG. It is a figure which shows an example of the learning reflection permission area before the regression line is calculated. It is a figure which shows an example of the learning reflection permission area after calculation of a regression line. It is a figure which shows an example of the straight line by one learning value. It is a figure which shows an example of the regression line by the predetermined number of learning values. It is a figure which shows an example of the excess oil pressure when the regression line is used. It is a flowchart which shows the main process of learning in the control device of the continuously variable transmission which concerns on embodiment. It is a flowchart which shows the flow of the learning value storage process in the control device of the continuously variable transmission which concerns on embodiment.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. In the figure, the same reference numerals are used for the same or corresponding parts. Further, in each figure, the same elements are designated by the same reference numerals, and duplicate description will be omitted.

In this embodiment, it is applied to the control device 1 of the chain type continuously variable transmission (CVT). The continuously variable transmission control device 1 according to the embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of a continuously variable transmission control device 1 according to an embodiment.

Before explaining the control device 1, the engine 2 and the continuously variable transmission 3 will be described. First, the engine 2 will be described. The engine 2 may be of any type, and is, for example, a horizontally opposed 4-cylinder gasoline engine. A continuously variable transmission 3 is connected to the crankshaft (output shaft) 2a of the engine 2. The engine 2 is controlled by an engine control unit (hereinafter referred to as "ECU (Engine Control Unit)") 20.

The ECU 20 is a control device that comprehensively controls the engine 2. The ECU 20 includes a microprocessor that performs calculations, a ROM that stores a program for causing the microprocessor to execute each process, a RAM that stores various data such as calculation results, a backup RAM whose stored contents are maintained by a 12V battery, and the like. It is configured to have input / output I / F and the like.

Various sensors such as an intake air amount sensor 21 and a crank angle sensor 22 are connected to the ECU 20 in order to acquire information necessary for control. The intake air amount sensor 21 detects the amount of air sucked from the air cleaner (not shown). The crank angle sensor 22 detects the rotation angle of the crankshaft 2a. The ECU 20 calculates (estimates) the engine torque (torque input to the continuously variable transmission 3) generated in the engine 2 by using, for example, the intake air amount or the like by a well-known estimation method. Further, in the ECU 20, for example, the engine rotation speed is calculated from the rotation angle of the crankshaft 2a. The ECU 20 controls the fuel injection amount, ignition timing, and the like based on the detected values acquired from these various sensors and the calculated various values. The detection values of various sensors acquired by the ECU 20 and the values of the engine torque, engine speed, etc. calculated by the ECU 20 are the transmission control unit (hereinafter, "TCU (Transmission Control Unit)" via CAN (Control Area Network) 60. (Called) 10 is transmitted. In this embodiment, the ECU 20, CAN60, etc. correspond to the torque acquisition means described in the claims.

Next, the continuously variable transmission 3 will be described. The continuously variable transmission 3 converts the driving force from the engine 2 and outputs it. The continuously variable transmission 3 includes a torque converter 30 and a forward / backward switching mechanism 31. Further, the continuously variable transmission 3 has a primary shaft 32 connected to the crankshaft 2a of the engine 2 via the torque converter 30 and a forward / backward switching mechanism 31, and a secondary shaft 33 arranged in parallel with the primary shaft 32. And have.

The torque converter 30 has a clutch function and a torque amplification function. The torque converter 30 mainly includes a pump impeller 30a, a turbine liner 30b, and a stator 30c. The pump impeller 30a is connected to the crankshaft 2a of the engine 2 to create an oil flow. The turbine liner 30b is arranged so as to face the pump impeller 30a, and receives the driving force of the engine 2 via oil to drive the output shaft (turbine shaft 30e). The stator 30c is arranged between the pump impeller 30a and the turbine liner 30b, rectifies the discharge flow (return) from the turbine liner 30b, and reduces it to the pump impeller 30a to generate a torque amplification action. Further, the torque converter 30 includes a lockup clutch 30d that directly connects the input (crankshaft 2a) and the output (turbine shaft 30e). The torque converter 30 amplifies and transmits the driving force of the engine 2 when the lockup clutch 30d is not engaged (when not locked up), and when the lockup clutch 30d is engaged (when locked up). The driving force of the engine 2 is directly transmitted to.

The forward / backward switching mechanism 31 has a function of switching between forward rotation and reverse rotation (forward and reverse of the vehicle) of the drive wheels. The forward / backward switching mechanism 31 mainly includes a double pinion type planetary gear train 31a, a forward clutch 31b, and a reverse clutch (reverse brake) 31c. The forward / backward switching mechanism 31 is configured to be able to switch the transmission path of the driving force of the engine 2 by controlling the states (engagement / release) of the forward clutch 31b and the reverse clutch 31c. An oil chamber is formed in the forward clutch 31b, and clutch oil pressure for engaging the clutch is supplied to this oil chamber. An oil chamber is formed in the reverse clutch 31c, and clutch oil pressure for engaging the clutch is supplied to this oil chamber.

When the drive range is selected by the driver's shift lever operation, the forward / backward switching mechanism 31 releases the reverse clutch 31c and then engages the forward clutch 31b, so that the rotation of the turbine shaft 30e is directly transferred to the primary shaft 32. introduce. In this case, the vehicle can be moved forward. On the other hand, when the reverse range is selected by the driver's shift lever operation, the forward / backward switching mechanism 31 operates the planetary gear train 31a by disengaging the forward clutch 31b and then engaging the reverse clutch 31c to operate the turbine. The rotation of the shaft 30e is reversed and transmitted to the primary shaft 32. In this case, the vehicle can be driven backward. When the neutral range or the parking range is selected by the driver's shift lever operation, the turbine shaft 30e and the primary shaft 32 are separated from each other by releasing the forward clutch 31b and the reverse clutch 31c (driving force of the engine 2). The forward / backward switching mechanism 31 is in a neutral state in which power is not transmitted to the primary shaft 32.

The primary shaft 32 is provided with a primary pulley 34. The primary pulley 34 has a fixed pulley 34a and a movable pulley 34b. The fixed pulley 34a is joined to the primary shaft 32. The movable pulley 34b faces the fixed pulley 34a and is mounted so as to be slidable in the axial direction of the primary shaft 32 and not to rotate relative to each other. The primary pulley 34 is configured so that the cone surface spacing (that is, the pulley groove width) between the fixed pulley 34a and the movable pulley 34b can be changed.

The secondary shaft 33 is provided with a secondary pulley 35. The secondary pulley 35 has a fixed pulley 35a and a movable pulley 35b. The fixed pulley 35a is joined to the secondary shaft 33. The movable pulley 35b faces the fixed pulley 35a and is mounted so as to be slidable in the axial direction of the secondary shaft 33 and not to rotate relative to each other. The secondary pulley 35 is configured so that the width of the pulley groove between the fixed pulley 35a and the movable pulley 35b can be changed.

A chain 36 for transmitting a driving force is hung between the primary pulley 34 and the secondary pulley 35. The variator 37 is composed of the primary pulley 34, the secondary pulley 35, and the chain 36. In the variator 37, the gear ratio is changed steplessly by changing the width of each pulley groove of the primary pulley 34 and the secondary pulley 35 and changing the ratio of the winding diameter of the chain 36 to each of the pulleys 34 and 35 (pulley ratio). To do. Assuming that the winding diameter of the chain 36 with respect to the primary pulley 34 is Rp and the winding diameter of the chain 36 with respect to the secondary pulley 35 is Rs, the gear ratio i is represented by i = Rs / Rp.

A primary drive oil chamber (hydraulic cylinder chamber) 34c is formed in the movable pulley 34b of the primary pulley 34. A secondary drive oil chamber (hydraulic cylinder chamber) 35c is formed in the movable pulley 35b of the secondary pulley 35. The primary drive oil chamber 34c is supplied with a shifting hydraulic pressure for changing the pulley ratio (gear ratio) and a clamp hydraulic pressure for preventing the chain 36 from slipping. Clamp oil is supplied to the secondary drive oil chamber 35c.

Each hydraulic pressure is supplied to the continuously variable transmission 3 by the valve body 40. A control valve mechanism is incorporated in the valve body 40. This control valve mechanism is an oil pump (not shown) by opening and closing an oil passage formed in a valve body 40 by using, for example, a plurality of spool valves provided in a hydraulic circuit and a solenoid valve for moving the spool valves. ) Is used to regulate the oil pressure (line pressure) discharged from the valve. In the valve body 40, the pressure-adjusted clamp hydraulic pressure is supplied to the primary drive oil chamber 34c and the secondary drive oil chamber 35c, and the pressure-adjusted variable speed hydraulic pressure is supplied to the primary drive oil chamber 34c.

In particular, the valve body 40 is provided with a hydraulic circuit 41 for adjusting (adjusting the pressure) the amount of oil supplied / discharged to each oil chamber of the forward clutch 31b and the reverse clutch 31b. The hydraulic circuit 41 will be described with reference to FIG. FIG. 2 is a diagram schematically showing a hydraulic circuit 41 for a forward / backward switching mechanism of the valve body 40.

The hydraulic circuit 41 includes a clutch linear solenoid valve 42, a clutch valve 43, and a manual valve 44. In the hydraulic circuit 41, the clutch hydraulic pressure supplied to the oil chamber of one of the forward clutch 31b and the reverse clutch 31c selected by the manual valve 44 is supplied by the clutch linear solenoid valve 42 and the clutch valve 43 controlled by the TCU 10. Adjust the pressure.

The clutch linear solenoid valve 42 is connected to an oil passage 46a and an oil passage 46b communicating with the clutch valve 43. Pilot pressure oil flows through the oil passage 46a. The clutch linear solenoid valve 42 is connected to the TCU 10. The clutch linear solenoid valve 42 is controlled to drive (position) the plunger in the axial direction according to the current value supplied from the TCU 10, and changes the clutch hydraulic control pressure. The clutch linear solenoid valve 42 supplies the clutch hydraulic control pressure to the clutch valve 43 via the oil passage 46b.

The clutch valve 43 is a spool valve and has a spool 43a that slides in the axial direction and a spring 43b that is arranged on one end side of the spool 43a. The clutch valve 43 is connected to an oil passage 46b communicating with the clutch linear solenoid valve 42, an oil passage 46c, and an oil passage 46d communicating with the manual valve 44. Line pressure oil flows through the oil passage 46c. The clutch valve 43 is controlled to drive (position) the spool 43a in the axial direction according to the clutch hydraulic control pressure supplied from the clutch linear solenoid valve 42. That is, the clutch valve 43 has a pushing force (clutch hydraulic control pressure x pressure receiving area) due to the clutch hydraulic control pressure supplied through the oil passage 46c, a spring force (urging force) of the spring 43b, and a pushing force due to the line pressure. The clutch oil pressure is adjusted by driving the spool 43a in the axial direction according to the balance with (line pressure × pressure receiving area). The clutch valve 43 supplies the clutch oil pressure to the manual valve 44 via the oil passage 46d.

The manual valve 44 is configured to move in conjunction with the shift lever 50. The manual valve 44 is connected to an oil passage 46d communicating with the clutch valve 43, an oil passage 46e communicating with the oil chamber of the forward clutch 31b, and an oil passage 46f communicating with the oil chamber of the reverse clutch 31c.

The shift lever (select lever) 50 is provided on the floor (center console) of the vehicle or the like. With the shift lever 50, for example, a drive range (D range), a manual range (M range), a parking range (P range), a reverse range (R range), and a neutral range (N range) can be selectively switched.

When the shift lever 50 is operated and the D range (forward travel range) is selected, the manual valve 44 moves in conjunction with the shift lever 50 so that the oil passages 46d and 46e communicate with each other and are adjusted by the clutch valve 43. The oil of the compressed clutch hydraulic pressure is supplied to the oil chamber of the forward clutch 31b via the oil passage 46e, and the oil is discharged from the oil chamber of the reverse clutch 31c. As a result, the forward clutch 31b is engaged by the clutch hydraulic pressure, the reverse clutch 31c is released, and the vehicle can move forward.

When the shift lever 50 is operated and the R range (reverse travel range) is selected, the manual valve 44 moves in conjunction with the shift lever 50 so that the oil passage 46d and the oil passage 46f communicate with each other, and the clutch valve 43 The pressure-adjusted clutch hydraulic oil is supplied to the oil chamber of the reverse clutch 31c via the oil passage 46f, and the oil is discharged from the oil chamber of the forward clutch 31b. As a result, the forward clutch 31b is released, the reverse clutch 31c is engaged by the clutch hydraulic pressure, and the vehicle can move backward.

When the shift lever 50 is operated to select the N range or the P range, the manual valve 44 moves in conjunction with the shift lever 50, so that the oil passage 46d, the oil passage 46e, and the oil passage 46f do not communicate with each other. , Oil is discharged from the oil chamber of the forward clutch 31b and the oil chamber of the reverse clutch 31c, respectively. As a result, the forward clutch 31b and the reverse clutch 31c are released, and the transmission of the engine driving force is cut off.

Then, the control device 1 of the continuously variable transmission will be described with reference to FIGS. 3 to 7. FIG. 3 is a diagram showing an example of a learning reflection permission region before calculating the regression line. FIG. 4 is a diagram showing an example of a learning reflection permission region after calculating the regression line. FIG. 5 is a diagram showing an example of a straight line based on one learning value. FIG. 6 is a diagram showing an example of a regression line based on a predetermined number of learning values. FIG. 7 is a diagram showing an example of excess flood pressure when a regression line is used.

The control device 1 is a control device that comprehensively controls the continuously variable transmission 3. In particular, the control device 1 according to the present embodiment has an input torque input to the forward / backward switching mechanism 31 and a clutch for engaging the clutch corresponding to the input torque for each of the clutches 31b and 31c of the forward / backward switching mechanism 31. The relationship with the oil pressure (corresponding to the friction coefficient of the friction material of each clutch 31b, 31c of each vehicle) is learned. In the present embodiment, the learning information indicating this relationship is represented by a straight line of the linear function, and consists of the slope and intercept of the linear function.

Each control of the control device 1 is carried out by the TCU 10. Like the ECU 20, the TCU 10 includes a microprocessor, a ROM, a RAM, a backup RAM, an input / output I / F, and the like.

Various sensors such as a turbine shaft rotation sensor 11, a primary pulley rotation sensor 12, an output shaft rotation sensor 13, and a range switch 14 are connected to the TCU 10 in order to acquire information necessary for control. Further, the TCU 10 receives various information such as the engine speed and the engine torque from the ECU 20 via the CAN 60.

The turbine shaft rotation sensor 11 detects the rotation speed of the turbine shaft 30e. The primary pulley rotation sensor 12 detects the rotation speed of the primary pulley 34 (primary shaft 32). The output shaft rotation sensor 13 detects the rotation speed of the secondary shaft 33 (output shaft). In the TCU 10, the vehicle speed is calculated from the rotation speed of the secondary shaft 33. The range switch 14 is connected so as to move in conjunction with the shift lever 50, and detects the selected position of the shift lever 50. The vehicle speed may be, for example, a vehicle speed (vehicle body speed) calculated from the wheel speeds detected by the wheel speed sensors provided on each wheel. In the TCU 10, for example, this vehicle speed is received from the ECU 20 or the like via the CAN 60.

The TCU 10 automatically shifts the gear ratio steplessly according to the driving state of the vehicle according to the shift map. In this control, for example, a target value of the primary rotation speed is set so as to have a predetermined gear ratio, and the actual primary rotation speed (rotation speed detected by the primary pulley rotation sensor 12) becomes the target value. By controlling each solenoid valve of the valve body 40, a speed change hydraulic pressure is generated to change the speed change ratio. The shift map is stored in the ROM in the TCU 10.

Further, the TCU 10 performs hydraulic control for engaging / releasing the clutches 31b and 31c of the forward / backward switching mechanism 31, and a straight line showing the relationship between the input torque for each clutch 31b and 31c and the clutch hydraulic pressure (particularly, return). Straight line) is learned. Therefore, the TCU 10 includes a hydraulic control unit 10a (corresponding to the hydraulic control means described in the claims), a slip determination unit 10b (corresponding to the slip determination means described in the claims), and a learning value acquisition unit. It has 10c (corresponding to the learning value acquisition means described in the claims) and a learning unit 10d (corresponding to the learning means described in the claims). In the TCU 10, each process of each of these parts 10a to 10d is realized by executing the program stored in the ROM by the microprocessor.

Further, the TCU 10 is configured with a storage unit 10e (corresponding to the storage means described in the claims) for storing data for learning. The storage unit 10e has a learning value storage area and a learning linear storage area for the forward clutch, and a learning value storage area and a learning linear storage area for the reverse clutch. Each learning value storage area is an area in which a predetermined number of learning values (input torque and the minimum oil pressure required to prevent the clutch from slipping) are stored. This predetermined number is determined by conformity. Each learning straight line storage area is an area in which the slope of a line and the intercept are stored.

Before explaining the specific processing of each part 10c-10d in the TCU 10, a straight line (clutch map) showing the relationship between the input torque and the clutch oil pressure will be described. This clutch map is provided for each clutch 31b and 31c. The clutch map is a map for obtaining clutch oil pressure for engaging the clutches 31b and 31c. The clutch map is a linear map showing the clutch oil pressure that changes linearly with respect to the input torque. This clutch oil pressure is the minimum oil pressure required to prevent the clutches 31b and 31c in the engaged state from slipping according to the input torque (torque transmitted by the clutches 31b and 31c). The input torque is the engine torque.

The straight line that becomes the clutch map is updated by learning. In particular, when a predetermined number of learning values are accumulated, it is a regression line obtained from the predetermined number of learning values. The initial map of the clutch map before learning is stored in advance in the ROM in the TCU 10. This initial map is, for example, a map shown by the learning reflection upper limit straight line UG1 shown in FIG. 3 and the like. Then, the specific processing of each part 10c-10d in TCU10 will be described.

First, the flood control unit 10a will be described. The hydraulic control unit 10a controls the clutch linear solenoid valve 42 in order to engage the clutches 31b and 31c. Specifically, when the D range is detected by the range switch 14 (when the D range is selected by the shift lever 50), the hydraulic control unit 10a refers to the clutch map (straight line) for the forward clutch 31b. Then, the clutch oil pressure corresponding to the input torque is acquired. When the R range is detected by the range switch 14 (when the R range is selected by the shift lever 50), the hydraulic control unit 10a refers to the clutch map for the reverse clutch and clutches corresponding to the input torque. Get the oil pressure.

The hydraulic control unit 10a sets a target value (instructed value) for hydraulic control based on the clutch hydraulic pressure acquired from the clutch map. For example, when the vehicle is in steady running, the clutch oil pressure acquired from the clutch map is set as the target value as it is. When the vehicle is accelerating or decelerating, the target value is set to the oil pressure obtained by adding a predetermined pressure to the clutch oil pressure acquired from the clutch map. Then, the hydraulic control unit 10a applies a current value required to reach the target value of the clutch hydraulic pressure to the clutch linear solenoid valve 42.

Further, the oil pressure control unit 10a controls the clutch linear solenoid valve 42 in order to increase or decrease the clutch oil pressure at a constant rate in response to an instruction from the learning value acquisition unit 10c. Specifically, when the learning value acquisition unit 10c gives an instruction to increase the pressure, the hydraulic control unit 10a clutches the oil pressure added by a constant pressure to the previously set target value of the clutch oil pressure every time a certain period of time elapses. Set as the target value of oil pressure. When the learning value acquisition unit 10c is instructed to reduce the pressure, the hydraulic control unit 10a sets as the target value of the clutch oil pressure by subtracting a constant pressure from the previously set target value of the clutch oil pressure every time a certain period of time elapses. To do. This constant time and constant pressure are determined by conformity. When the target value of the clutch oil pressure is set at regular time intervals, the oil pressure control unit 10a applies a current value required to reach the target value of the clutch oil pressure to the clutch linear solenoid valve 42.

Next, the slip determination unit 10b will be described. The slip determination unit 10b determines whether or not the clutches 31b and 31c are slipping. Specifically, the slip determination unit 10b compares the rotation speed of the turbine shaft 30e on the input side of the forward / backward switching mechanism 31 (clutch 31b, 31c) with the rotation speed of the primary shaft 32 on the output side to clutch the clutch. It is determined whether or not 31b and 31c are slipping. As this determination method, for example, the rotation speed of the primary shaft 32 is subtracted from the rotation speed of the turbine shaft 30e, and if the rotation speed difference is within a predetermined value, it is determined that the turbine is not slipping, and if it exceeds the predetermined value. Judges that it is slipping. At the time of this determination, the forward clutch 31b is slipping when the range switch 14 is detecting the D range, and the reverse clutch 31c is slipping when the range switch 14 is detecting the R range.

The learning value acquisition unit 10c will be described. The learning value acquisition unit 10c acquires a learning value consisting of an input torque and the minimum oil pressure required to prevent the clutch from slipping according to the input torque for each clutch 31b and 31c, and obtains the acquired learning value. It is stored (stored) in the storage unit 10e. As a learning implementation condition, it is implemented during steady running of the vehicle. This is because the flood pressure is stable during steady running. Further, as a learning implementation condition, it is implemented when the engine torque (input torque) is low. This is because the clutches 31b and 31c are temporarily slid when the learning value is acquired, and if the clutches 31b and 31c are slid when the engine torque is high, the clutches 31b and 31c may be worn or damaged. The learning value adoption condition (reflection condition) is the input clutch and hydraulic pressure when the two-dimensional coordinate point consisting of the acquired input torque and the minimum hydraulic pressure required to prevent the clutch from slipping is within the learning reflection permission area. Is adopted as a learning value (reflected in learning). This is because the oil pressure is unstable and has a large variation, and therefore, in order to improve the learning accuracy, the flood pressure having a large variation is excluded from the learning.

Specifically, the learning value acquisition unit 10c determines whether or not the engine speed is equal to or less than the predetermined rotation speed and the vehicle speed is within the predetermined vehicle speed range (the lower limit vehicle speed or more and the upper limit vehicle speed or less) in order to carry out the operation during steady driving. Judge whether or not. Further, the learning value acquisition unit 10c determines whether or not the engine torque is within a predetermined torque range (greater than or equal to the lower limit torque and less than or equal to the upper limit torque) for execution when the engine torque is low. The predetermined number of revolutions, the lower limit vehicle speed, the upper limit vehicle speed, the upper limit torque, and the lower limit torque are determined by conformity.

When the engine speed is equal to or less than the predetermined speed, the vehicle speed is within the predetermined vehicle speed range, and the engine torque is within the predetermined torque range, the learning value acquisition unit 10c determines that the clutches 31b and 31c are slipping by the slip determination unit 10b. If so, the hydraulic control unit 10a is instructed to increase the pressure. As a result, the oil pressure in the oil chamber of the forward clutch 31b gradually increases when the D range is detected by the range switch 14, and the oil in the reverse clutch 31c when the R range is detected by the range switch 14. The oil pressure in the chamber gradually increases. Further, the learning value acquisition unit 10c instructs the hydraulic control unit 10a to reduce the pressure when the slip determination unit 10b determines that the clutches 31b and 31c are not slipping. As a result, the oil pressure in the oil chamber of the forward clutch 31b gradually decreases when the D range is detected by the range switch 14, and the oil in the reverse clutch 31c when the R range is detected by the range switch 14. The oil pressure in the chamber gradually decreases.

When the pressure increase instruction is given, the learning value acquisition unit 10c stops the pressure increase instruction when the slip determination unit 10b determines that the clutches 31b and 31c are not slipping during the pressure increase, and does not slip with the input torque (engine torque). Acquires the indicated value of the clutch oil pressure (corresponding to the minimum oil pressure required to prevent the clutch from slipping) in the oil pressure control unit 10a when it is determined. As a result, the clutch oil pressure when the forward clutch 31b, which was slipping when the D range is detected by the range switch 14, is not slipped, is acquired, and when the R range is detected by the range switch 14, the clutch slips. The clutch oil pressure when the reverse clutch 31c that has been used does not slip is acquired.

When the decompression instruction is given, the learning value acquisition unit 10c stops the decompression instruction when the slip determination unit 10b determines that the clutches 31b and 31c are slipping during decompression, and determines that the clutch 31b and 31c are slipping with the input torque (engine torque). Acquires the indicated value of the clutch oil pressure (corresponding to the minimum oil pressure required to prevent the clutch from slipping) in the immediately preceding oil pressure control unit 10a. As a result, the clutch oil pressure immediately before the forward clutch 31b, which did not slip when the range switch 14 detects the D range, is acquired, and when the range switch 14 detects the R range, the clutch slips. The clutch oil pressure just before the reverse clutch 31c, which has not been used, starts to slip is acquired.

Whether or not the two-dimensional coordinate point composed of the input torque and the oil pressure is in the learning reflection permission region every time the learning value acquisition unit 10c acquires the minimum required oil pressure to prevent the input torque and the clutch from slipping. To judge. The learning reflection permission area is an area surrounded by the learning reflection upper limit straight line, the learning reflection lower limit straight line, and the upper limit torque and the lower limit torque. The learning reflection permission area (particularly, the learning reflection upper limit straight line and the learning reflection lower limit straight line) differs depending on whether the regression line has not been calculated yet or has already been calculated by the learning unit 10d.

First, with reference to FIG. 3, a learning reflection permission area when a regression line has not yet been calculated by the learning unit 10d (when a predetermined number of learning values have not been accumulated) will be described. In this case, in order to increase the learning reflection frequency, the learning reflection permission area is relatively widened. As shown in FIG. 3, the learning reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 are straight lines having a predetermined inclination passing through the origin in a two-dimensional coordinate system in which the horizontal axis is the input torque and the vertical axis is the clutch hydraulic pressure. .. The learning reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 are set for each clutch 31b and 31c based on the standard value of the friction coefficient of the friction material of the clutches 31b and 31c. The learning reflection upper limit straight line UG1 is set based on the lower limit value of the friction coefficient. The learning reflection lower limit straight line LG1 has a smaller slope than the learning reflection upper limit straight line UG1, and is set based on the upper limit value of the friction coefficient. The lower and upper limits of the coefficient of friction take into account the aging deterioration of the friction material. The upper limit torque UT and the lower limit torque LT are the upper limit torque and the lower limit torque of the torque range of the above-mentioned learning execution conditions, and are determined by conformity. As shown in FIG. 3, the learning reflection permission region PA1 defined by the learning reflection upper limit straight line UG1, the learning reflection lower limit straight line LG1, the upper limit torque UT, and the lower limit torque LT is a trapezoidal region.

Next, with reference to FIG. 4, the learning reflection permission area when the regression line has already been calculated by the learning unit 10d (when a predetermined number of learning values are accumulated) will be described. After the regression line is obtained by learning, it is considered that the learning value does not deviate significantly from the regression line. Therefore, the learning reflection permission area is made stricter (narrower) by using the regression line to suppress the variation in the learning value and improve the learning accuracy. As shown in FIG. 4, the learning reflection upper limit straight line UG2 and the learning reflection lower limit straight line LG2 are straight lines parallel to the regression line RL (a straight line having the same slope as the regression line RL). The learning reflection upper limit straight line UG2 and the learning reflection lower limit straight line LG2 are set for each of the clutches 31b and 31c based on the latest regression line RL. The learning reflection upper limit straight line UG2 is a straight line separated from the regression line RL by a certain first distance D1 on the plus side (a straight line whose intercept is D1 minutes larger than the regression line RL). The learning reflection lower limit straight line LG2 is a straight line separated from the regression line RL by a certain second distance D2 on the minus side (a straight line whose intercept is D2 smaller than the regression line RL). The second distance D2 on the minus side is smaller than the first distance D1 on the plus side. This is to prevent the clutches 31b and 31c from slipping by suppressing the oil pressure supplied to the clutches 31b and 31c from becoming too low. The first distance D1 on the plus side and the second distance D2 on the minus side are determined by conformity. As shown in FIG. 4, the learning reflection permission region PA2 defined by the learning reflection upper limit straight line UG2, the learning reflection lower limit straight line LG2, the upper limit torque UT, and the lower limit torque LT is a parallel quadrilateral region.

If the two-dimensional coordinate point consisting of the input torque and the minimum required hydraulic pressure to prevent the clutch from slipping is not within the learning reflection permission area, the learning value acquisition unit 10c determines the input torque and the minimum required hydraulic pressure. Not adopted as a learning value (not reflected in learning). On the other hand, when the two-dimensional coordinate point consisting of the input torque and the minimum required hydraulic pressure to prevent the clutch from slipping is within the learning reflection permission region, the learning value acquisition unit 10c is required to have the input torque and the minimum required. Adopt hydraulic pressure as a learning value (reflect in learning).

When adopted as a learning value, the learning value acquisition unit 10c stores the learning value in the learning value storage area for the forward clutch of the storage unit 10e when the D range is detected by the range switch 14, and the range switch When the R range is detected in 14, the learning value is stored in the learning value storage area for the reverse clutch of the storage unit 10e. Each time a predetermined number of learning values is acquired, the oldest learning value among the predetermined number of stored learning values is deleted in the learning value storage area, and the latest acquired learning value is deleted. Is remembered.

The learning value acquisition unit 10c determines whether or not the learning value accumulated in the learning value storage area for each clutch of the storage unit 10e has been acquired within a certain period of time. When there is a learning value that has not been acquired within a certain period of time, the learning value acquisition unit 10c deletes the learning value that has not been acquired within a certain period of time from the learning value storage area. At this time, all the learning values stored in the learning value storage area may be deleted (the learning value storage area is reset). A certain period of time is determined by conformity. The fixed period is, for example, six months or three months.

Since the friction materials provided on the clutches 31b and 31c deteriorate over time, the friction coefficient changes with the passage of time, and the minimum oil pressure required to prevent the clutch from slipping also gradually changes. Therefore, it is avoided to calculate the regression line using the learning values obtained before a certain period, and the regression line is calculated using the learning values acquired within the certain period.

The learning unit 10d will be described. The learning unit 10d calculates a straight line (clutch map) showing the relationship between the input torque and the clutch oil pressure for each clutch 31b and 31c by using the learning value in the learning value storage area of the storage unit 10e. In the calculation of this straight line, when a predetermined number of learning values are not accumulated in the learning value storage area, a straight line is calculated using the latest one learning value, and a predetermined number of learning values are accumulated in the learning value storage area. If so, the regression line is calculated using a predetermined number of learning values.

Specifically, when a predetermined number of learning values are not accumulated in the learning value storage area of the storage unit 10e, the learning unit 10d is updated every time the learning value is newly acquired by the learning unit acquisition unit 10d. A straight line (inclination and intercept of the linear function) showing the relationship between the input torque and the clutch oil pressure is calculated using one learning value. When the D range is detected by the range switch 14, the learning unit 10d stores the slope and intercept of the straight line in the learning linear storage area for the forward clutch of the storage unit 10e, and detects the R range by the range switch 14. If so, the slope and intercept of the straight line are stored in the learning straight line storage area for the reverse clutch of the storage unit 10e.

FIG. 5 shows an example of a straight line L obtained by one learning point LV.

When a predetermined number of learning values are stored in the learning value storage area of the storage unit 10e, the learning unit 10d uses the predetermined number of learning values each time the learning value is newly acquired by the learning unit acquisition unit 10d. Calculate the regression line (gradient and intercept of the linear function). When the D range is detected by the range switch 14, the learning unit 10d stores the slope and intercept of the regression line in the learning straight line storage area for the forward clutch of the storage unit 10e, and sets the R range by the range switch 14. If it is detected, the slope and intercept of the regression line are stored in the learning straight line storage area for the reverse clutch of the storage unit 10e. If the slope and intercept are already stored in the learning linear storage area, the newly obtained slope and intercept are overwritten.

As a method of calculating this regression line, for example, in a two-dimensional coordinate system in which the horizontal axis is the input torque and the vertical axis is the clutch hydraulic pressure, the input torque of a predetermined number of learning values and the minimum hydraulic pressure required to prevent the clutch from slipping. The slope and section of the regression line are obtained by performing regression analysis (for example, regression analysis by the least squares method) using the two-dimensional coordinate points consisting of. FIG. 6 shows an example of the regression line RL obtained by a predetermined number of learning points LV.

FIG. 7 shows the regression line RL shown in FIG. 6, and the straight line IL (ideal straight line) showing the relationship between the input torque and the minimum clutch oil pressure required to prevent the clutch from actually slipping according to the input torque. ) And is shown. When the engagement control of the clutches 31b and 31c is performed using the regression line RL, the oil pressure in the region SA1 where the regression line RL exceeds the straight line IL becomes an excessive oil pressure, but the excessive oil pressure amount is small.

Note that FIG. 5 shows a straight line L and a straight line IL (ideal straight line) obtained by one learning value, and when the clutches 31b and 31c are controlled to be engaged using this straight line L. , The oil pressure in the region SA2 where the straight line L exceeds the straight line IL becomes an excessive hydraulic pressure. Comparing the region SA2 shown in FIG. 5 and the region SA1 shown in FIG. 7, the region SA1 is smaller. Therefore, using the regression line RL reduces the excess oil pressure.

The flow of learning performed by the control device 1 (TCU10) will be described with reference to FIGS. 1 to 7 according to the flowcharts of FIGS. 8 to 9. FIG. 8 is a flowchart showing the main processing of learning. FIG. 9 is a flowchart showing the flow of the learning value storage process. In TCU10, the following processing is repeated at predetermined timings.

The TCU 10 determines whether or not the engine speed is equal to or lower than the predetermined speed (S10). If it is determined in the determination of S10 that the engine speed is higher than the predetermined speed, the TCU 10 does not perform learning.

When it is determined in the determination of S10 that the engine speed is equal to or less than the predetermined rotation speed, the TCU 10 determines whether or not the engine torque is within the predetermined torque range (S12). When it is determined in the determination of S12 that the engine torque is out of the predetermined torque range, the TCU 10 does not perform learning.

When it is determined in the determination of S12 that the engine torque is within the predetermined torque range, the TCU 10 determines whether or not the vehicle speed is within the predetermined vehicle speed range (S14). If it is determined in S14 that the vehicle speed is out of the predetermined vehicle speed range, the TCU 10 does not perform learning.

When it is determined in the determination of S14 that the vehicle speed is within the predetermined vehicle speed range, the TCU 10 performs learning. The learning described below is learning in which the learning target clutch is the forward clutch 31b when the D range is detected by the range switch 14 (when the D range is selected by the shift lever 50), and the range switch 14 When the R range is detected in (when the R range is selected by the shift lever 50), the learning is performed so that the learning target clutch is the reverse clutch 31c.

In the TCU 10, it is determined whether or not the clutches 31b and 31c are slipping (S16). When it is determined in the determination of S16 that the clutches 31b and 31c are slipping, the TCU 10 performs pressure increasing control for increasing the clutch oil pressure at a constant rate until the clutches 31b and 31c do not slip (S18). Then, the TCU 10 acquires the input torque (engine torque) and the clutch oil pressure (the minimum oil pressure required to prevent the clutch from slipping) when the clutch corresponding to the input torque does not slip (S18). On the other hand, when it is determined in S16 that the clutches 31b and 31c are not slipping, the TCU 10 performs decompression control for lowering the clutch oil pressure at a constant rate until the clutches 31b and 31c start slipping (S20). Then, in the TCU 10, the input torque and the clutch oil pressure immediately before the clutch corresponding to the input torque starts to slip (the minimum oil pressure required to prevent the clutch from slipping) are acquired (S20).

In the TCU 10, a learning value storage process including an input torque and a minimum required oil pressure to prevent the clutch from slipping is performed (S22). This learning value storage process shifts to the flowchart of FIG. 9, and first, the TCU 10 determines whether or not the regression line has not been calculated yet (S40).

When it is determined in the determination of S40 that the regression line has not been calculated yet, in the TCU 10, a learning in which a two-dimensional coordinate point consisting of the acquired input torque and the minimum hydraulic pressure required to prevent the clutch from slipping is preset. It is determined whether or not it exists in the learning reflection permission region PA1 between the reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 (S42). When it is determined in the determination of S42 that the area exists in the learning reflection permission area PA1, the TCU 10 adopts the acquired input torque and the minimum required oil pressure to prevent the clutch from slipping, and stores them in the storage unit 10e. (Accumulate) (S44). On the other hand, when it is determined in S42 that it does not exist in the learning reflection permission region PA1, the TCU 10 does not adopt the acquired input torque and the minimum required oil pressure to prevent the clutch from slipping as learning values (S46). ..

When it is determined in the determination of S40 that the regression line has already been calculated, the TCU 10 sets the learning reflection upper limit straight line UG2 and the learning reflection lower limit straight line LG2 based on the latest regression line, and obtains the input torque and the clutch. It is determined whether or not a two-dimensional coordinate point consisting of the minimum hydraulic pressure required to prevent slipping exists in the learning reflection permission region PA2 between the learning reflection upper limit straight line UG2 and the learning reflection lower limit straight line LG2 (S48). ). When it is determined in the determination of S48 that it exists in the learning reflection permission area PA2, the TCU 10 adopts the acquired input torque and the minimum required oil pressure to prevent the clutch from slipping, and stores them in the storage unit 10e. (Accumulate) (S50). On the other hand, when it is determined in S48 that it does not exist in the learning reflection permission region PA2, the TCU 10 does not adopt the acquired input torque and the minimum required oil pressure to prevent the clutch from slipping as learning values (S52). ..

Returning to the flowchart of FIG. 8, the TCU 10 determines whether or not the learning value stored in the storage unit 10e has been acquired within a certain period of time (S24). When it is determined in the determination of S24 that there is a learning value acquired outside the fixed period, the TCU 10 erases the learning value outside the fixed period from the storage unit 10e (S26). In this case, the regression line is not calculated.

When it is determined in the determination of S24 that all the learning values are acquired within a certain period, the TCU 10 determines whether or not a predetermined number of learning values are accumulated in the storage unit 10e (S28). If it is determined in S28 that a predetermined number of learning values have not yet been accumulated, the regression line is not calculated. In this case, the TCU 10 calculates the slope and intercept of a straight line indicating the relationship between the input torque and the clutch oil pressure using only the latest learning value, and stores the slope and intercept of this straight line in the storage unit 10e. To do.

When it is determined in the determination of S28 that a predetermined number of learning values are accumulated, the TCU 10 calculates the slope and intercept of the regression line indicating the relationship between the input torque and the clutch oil pressure by using the predetermined number of learning values. (S30). Then, the TCU 10 stores (updates) the slope and intercept of this regression line in the storage unit 10e (S32).

When the slope and intercept of the regression line are stored in the storage unit 10e, the TCU 10 acquires the clutch oil pressure according to the input torque with reference to this regression line (clutch map), and engages with this clutch oil pressure as the target value. Control (hydraulic control). As a result, the clutches 31b and 31c are engaged with substantially the minimum necessary hydraulic pressure with respect to the transmitted input torque. At this time, when an excessive torque is input to the continuously variable transmission 3 from the wheel side, the clutches 31b and 31c start to slip before the variator 37. This prevents the chain 36 from slipping on the variator 37.

According to the control device 1 of the continuously variable transmission according to the embodiment, the input torque and the clutch oil pressure (particularly, the clutch oil pressure) corresponding to the input torque are calculated by calculating the regression line (slope and intercept) using a predetermined number of learned values. , The minimum required oil pressure to prevent the clutch from slipping) can improve the accuracy of learning the relationship. As a result, the clutch hydraulic pressure (excessive hydraulic pressure) supplied to engage the clutches 31b and 31c of the forward / backward switching mechanism 31 can be reduced, and the clamp hydraulic pressure supplied to the pulleys 34 and 35 of the variator 37 ( The oil pressure for the margin) can be reduced. As a result, the power loss in the oil pump, which is the source of each of the flood controls, can be suppressed, and the fuel efficiency can be improved. Further, when the continuously variable transmission 3 is subjected to an excessive torque from the wheel side by engaging and controlling the clutches 31b and 31c using a regression line (clutch map) showing the relationship between the input torque and the clutch oil pressure. Since the clutches 31b and 31c can be slid before the variator 37, the variator 37 can be protected. Further, it is possible to prevent the clutch hydraulic pressure supplied for engaging the clutches 31b and 31c of the forward / backward switching mechanism 31 from becoming too low and the amount of slippage of the clutches 31b and 31c from becoming large (particularly, to prevent slippage). The clutches 31b and 31c can be protected.

According to the control device 1 of the continuously variable transmission according to the embodiment, when the regression line has not been calculated yet, it is between the preset learning reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 (learning reflection permission area PA1). By adopting it as a learning value when there is a two-dimensional coordinate point of the input torque and the minimum required hydraulic pressure to prevent the clutch from slipping (inside), the hydraulic pressure with large variation can be excluded from the learning value. Good learning values can be obtained and learning accuracy can be improved. By setting the learning reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 based on the friction coefficient of the friction material of the clutches 31b and 31c, it is appropriate in consideration of the variation in the friction coefficient of the friction material of the clutch 31b and 31c. The learning reflection upper limit straight line UG1 and the learning reflection lower limit straight line LG1 can be set.

According to the control device 1 of the stepless transmission according to the embodiment, when the regression line has already been calculated, the learning reflection upper limit straight line UG2 separated from the regression line by the first distance D1 and constant on the minus side is constant. The learning reflection lower limit straight line LG2 separated from the second distance D2 is set, and the input torque and the clutch do not slip between the learning reflection upper limit straight line UG2 and the learning reflection lower limit straight line LG2 (in the learning reflection permission area PA2). By adopting it as a learning value when there is a two-dimensional coordinate point with the required hydraulic pressure, it is possible to narrow down (make it stricter) the learning reflection permission area PA2 and obtain a more accurate learning value, and the learning accuracy. Can be further improved. By making the second distance D2 on the minus side smaller than the first distance D1 on the plus side, it is possible to prevent the clutch oil supply supplied to the clutches 31b and 31c from becoming too low, so that the amount of slippage of the clutches 31b and 31c can be reduced. It is possible to further suppress the increase (in particular, it is possible to prevent the clutches 31b and 31c from slipping), and it is possible to protect the clutches 31b and 31c.

According to the control device 1 of the continuously variable transmission according to the embodiment, when it is determined that the clutches 31b and 31c are slipping, the pressure is increased until it is determined that the clutches 31b and 31c are not slipping, so that the clutch does not slip. By acquiring the minimum required oil pressure and controlling the pressure reduction until it is determined that the clutches 31b and 31c are not slipping, the minimum required oil pressure to prevent the clutch from slipping is acquired. , The oil pressure can be accurately controlled until the minimum required oil pressure is reached so that the clutches 31b and 31c do not slip, and the minimum required oil pressure (learning value) can be accurately obtained so that the clutch does not slip with respect to the input torque. ..

Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments and can be modified in various ways. For example, in the above embodiment, it is applied to the chain type continuously variable transmission 3, but it can also be applied to other continuously variable transmissions such as a belt type.

In the above embodiment, the case where the forward / backward switching mechanism 31 is provided on the upstream side (engine 2 side) of the variator 37 is applied, but the case where the forward / backward switching mechanism 31 is provided on the downstream side of the variator 37. May be applied to.

In the above embodiment, a straight line (inclination, intercept) is acquired as learning information obtained by learning, but any information other than a straight line may be used as long as it shows the relationship between the input torque and the clutch hydraulic pressure. In particular, in the above embodiment, the regression line of the linear function is calculated by the regression analysis, but the regression curve of the nth function may be calculated by the regression analysis.

In the above embodiment, the minimum required oil pressure to prevent the clutch from slipping is set to use the indicated value in the hydraulic control unit 10a, but the detection value of the sensor that detects the actual oil pressure supplied to the clutches 31b and 31c is used. It may be configured.

In the above embodiment, the engine torque calculated (estimated) by the engine 2 is used as the input torque, but the detection value of the sensor that detects the actual torque input to the forward / backward switching mechanism 31 may be used. ..

In the above embodiment, until a predetermined number of learning values are accumulated, a straight line showing the relationship between the input torque and the clutch oil pressure is calculated using the latest learning value, but the predetermined number of learning values is calculated. The straight line of the initial map may be used without calculating the straight line until it is accumulated, or the straight line may be calculated using all the learning values less than a predetermined number.

In the above embodiment, both the forward clutch 31b and the reverse clutch 31c are used as the learning target clutchs, but either one of the forward clutch 31b and the reverse clutch 31c may be used as the learning target clutch.

1 Continuously variable transmission control device 3 Continuously variable transmission 10 TCU
10a Flood control unit 10b Slip determination unit 10c Learning value acquisition unit 10d Learning unit 10e Storage unit 20 ECU
31 Forward / backward switching mechanism 31b Forward clutch 31c Reverse clutch 34 Primary pulley 35 Secondary pulley 36 Chain 37 Variator 40 Valve body 41 Hydraulic circuit 42 Clutch Linear solenoid valve 43 Clutch valve 44 Manual valve 50 Shift lever 60 CAN

Claims (5)

A forward / backward switching mechanism having a forward clutch engaged when the vehicle moves forward and a reverse clutch engaged when the vehicle moves backward, and a hydraulic circuit for supplying hydraulic pressure for engaging the clutches of the forward / backward switching mechanism. It is a control device of a continuously variable transmission equipped with
Torque acquisition means for acquiring the input torque input to the forward / backward switching mechanism, and
A hydraulic control means for controlling the flood pressure supplied from the hydraulic circuit to each clutch, and
For the learning target clutch of at least one of the forward clutch and the reverse clutch, the input torque acquired by the torque acquisition means and the learning target clutch obtained by controlling by the hydraulic control means do not slip. A learning value acquisition means for acquiring a learning value consisting of the minimum required torque and
A storage means for storing the learning value acquired by the learning value acquisition means, and a storage means for storing the learning value.
When two or more predetermined numbers of the learning values are accumulated for the learning target clutch for which the learning values have been acquired by the learning acquisition means, the predetermined number of the learning values stored in the storage means is used. , A learning means for learning the relationship between the input torque and the minimum hydraulic pressure required to prevent the learning target clutch from slipping according to the input torque, and acquiring learning information indicating the relationship.
Equipped with a,
The learning means is a continuously variable transmission control device, characterized in that a regression analysis is performed using a predetermined number of the learned values and a regression line indicating the relationship between the input torque and the oil pressure is calculated . It said learning means, the control device for a continuously variable transmission according to claim 1, characterized in that calculating a regression line indicating the relationship between the input torque and the previous SL oil pressure. In the learning value acquisition means, when the regression line has not yet been calculated by the learning means, the input torque and the learning target clutch do not slip between the preset learning reflection upper limit straight line and the learning reflection lower limit straight line. 2. The method of claim 2 is characterized in that the minimum required hydraulic pressure is adopted as the learning value so that the input torque and the learning target clutch do not slip when there is a coordinate point consisting of the minimum required hydraulic pressure. The control device for the continuously variable transmission described in 1. When the regression line has already been calculated by the learning means, the learning value acquisition means has a learning reflection upper limit straight line that is a first distance away from the regression line on the plus side and a learning reflection lower limit that is a second distance away from the minus side. When a straight line is set and a coordinate point consisting of the input torque and the minimum hydraulic pressure required to prevent the learning target clutch from slipping exists between the set learning reflection upper limit straight line and the learning reflection lower limit straight line. The control device for a stepless transmission according to claim 2 or 3 , wherein the input torque and the minimum hydraulic pressure required to prevent the learning target clutch from slipping are adopted as the learning values. A slip determining means for determining whether or not the learning target clutch is slipping is provided.
When the learning value acquisition means determines that the clutch is slipping by the slip determination means, the learning target clutch does not slip by controlling the pressure increase by the hydraulic control means until it is determined that the slip determination means is not slipping. The clutch to be learned is obtained by acquiring the minimum necessary oil pressure for this purpose, and when it is determined that the clutch is not slipping by the slip determining means, the pressure is reduced by the hydraulic control means until it is determined that the slip is slipping. The control device for a continuously variable transmission according to any one of claims 1 to 4 , wherein the minimum necessary oil pressure is acquired so that the clutch does not slip.

Images