Hydraulic estimation device and storage medium

Abstract

The present invention provides a hydraulic pressure estimation device and a storage medium. The hydraulic pressure estimation device has a storage device and an execution device. The hydraulic pressure estimation device calculates an estimated pressure difference, which is an estimated value of the pressure difference between two oil chambers in a torque converter having two oil chambers and a lock-up clutch. The storage device stores mapping data that defines a mapping that has been learned by machine learning to output an estimated pressure difference variable as an output variable by inputting an input variable, where the estimated pressure difference variable is a variable representing the estimated pressure difference. The mapping includes an indicated pressure difference variable as one of the plurality of input variables, and the indicated pressure difference variable is a variable representing the indicated pressure difference. The execution device executes an acquisition process and a calculation process. The acquisition process acquires the value of the input variable, and the calculation process calculates the value of the output variable by inputting the value of the input variable into the mapping.

Description

Technical Field

[0001] The present invention relates to a hydraulic pressure estimation device and a storage medium. Background Art

[0002] The vehicle disclosed in Japanese Unexamined Patent Application Publication No. 2010-156359 includes an internal combustion engine, a torque converter, and an automatic transmission. The output shaft of the internal combustion engine is connected to the input shaft of the torque converter. The torque converter transmits torque between the input shaft and the output shaft via a fluid. The output shaft of the torque converter is connected to the input shaft of the automatic transmission.

[0003] The torque converter has a lock-up clutch. The lock-up clutch switches between an engaged state in which the input shaft and the output shaft of the torque converter are directly connected mechanically and a released state in which the connection is released. The lock-up clutch operates according to hydraulic pressure.

[0004] The vehicle has a control device. The control device switches the state of the lock-up clutch by controlling the hydraulic pressure. Summary of the Invention

[0005] In a technique such as Japanese Unexamined Patent Application Publication No. 2010-156359 that controls the hydraulic pressure related to the lock-up clutch, a deviation may occur between the indicated value of the hydraulic pressure and the actual hydraulic pressure. Therefore, in order to control the hydraulic pressure with good accuracy, it is necessary to accurately estimate the actual hydraulic pressure and control the hydraulic pressure in consideration of the deviation between the estimated value of the hydraulic pressure and the indicated value of the hydraulic pressure. In order to accurately estimate the actual hydraulic pressure, it is necessary to grasp the correlation between the indicated value of the hydraulic pressure and the actual hydraulic pressure and reflect the correlation in the estimation of the hydraulic pressure. For this purpose, it is necessary to derive a relational expression or the like that represents the correlation between the indicated value of the hydraulic pressure and the actual hydraulic pressure. However, such a relational expression is very complex and requires time and effort to derive.

[0006] The first technical solution of the present invention relates to a hydraulic estimation device that takes a torque converter having two oil chambers and a lock-up clutch that switches between an engaged state and a released state according to the pressure difference between the two oil chambers as a simulation object, calculates an estimated pressure difference based on an indicated pressure difference, where the indicated pressure difference is an indicated value of the pressure difference of the torque converter, and the estimated pressure difference is an estimated value of the pressure difference generated in the torque converter. This hydraulic estimation device includes a storage device and an execution device. The storage device is configured to store mapping data (Japanese: mapping data), which is data that defines a mapping (Japanese: mapping) that outputs an estimated pressure difference variable as an output variable by inputting an input variable and has been learned by machine learning. The estimated pressure difference variable is a variable representing the estimated pressure difference. Here, the mapping includes an indicated pressure difference variable as one of the plurality of input variables, and the indicated pressure difference variable is a variable representing the indicated pressure difference. The execution device is configured to execute an acquisition process and a calculation process. The acquisition process acquires the value of the input variable, and the calculation process inputs the value of the input variable acquired by the acquisition process into the mapping to calculate the value of the output variable.

[0007] In the hydraulic estimation device according to the above first technical solution, when the mapping is used to calculate the estimated pressure difference, it is only necessary to use appropriate data to learn the mapping so that the estimated pressure difference can be calculated appropriately. Therefore, in reflecting the correlation between the indicated pressure difference and the actual pressure difference in the calculation of the estimated pressure difference, there is no need for the effort as in the case of deriving a complex relational expression.

[0008] In the hydraulic estimation device according to the above first technical solution, the storage device and the execution device may also be provided in a vehicle equipped with the torque converter. The storage device may also be configured to store a plurality of mappings for each stage (Japanese: phase) obtained by dividing the period from when the lock-up clutch starts to switch to the engaged state until it starts to switch to the released state into a plurality of parts. The execution device may also be configured to execute the acquisition process, the selection process, and the calculation process. The acquisition process may acquire the value of the input variable when the indicated pressure difference is output to the torque converter. The selection process may select the mapping corresponding to the stage when the indicated pressure difference is output from the plurality of mappings for each stage. The calculation process may calculate the value of the output variable by inputting the value of the input variable acquired by the acquisition process into the mapping selected in the selection process.

[0009] In the hydraulic estimation device according to the above structure, since dedicated mappings are used for each stage, even if the correlation between the indicated pressure difference and the estimated pressure difference is different for each stage, the output variable can reflect the difference.

[0010] In the hydraulic pressure estimation device with the above structure, as one of the multiple input variables, an acceleration variable may also be included, where the acceleration variable is a variable representing the acceleration of the vehicle. According to the hydraulic pressure estimation device with the above structure, even when the correlation between the indicated pressure difference and the estimated pressure difference is different during vehicle acceleration and deceleration, an accurate output variable corresponding to the difference can be obtained.

[0011] In the hydraulic pressure estimation device with the above structure, as one of the multiple input variables, an accelerator variable may also be included, where the accelerator variable is a variable representing the operation amount of the accelerator pedal of the vehicle. According to the hydraulic pressure estimation device with the above structure, by including an accelerator variable as one of the input variables, an output variable corresponding to the torque applied to the torque converter can be obtained.

[0012] In the hydraulic pressure estimation device with the above structure, as one of the multiple input variables, a vehicle speed variable may also be included, where the vehicle speed variable is a variable representing the traveling speed of the vehicle. According to the hydraulic pressure estimation device with the above structure, even when there is a difference in the correlation between the indicated pressure difference and the estimated pressure difference corresponding to the traveling speed of the vehicle, an accurate output variable can be obtained.

[0013] In the hydraulic pressure estimation device with the above structure, as one of the multiple input variables, a gearshift variable may also be included, where the gearshift variable is a variable representing the gear ratio of the transmission of the vehicle. According to the hydraulic pressure estimation device with the above structure, even when there is a difference in the correlation between the indicated pressure difference and the estimated pressure difference corresponding to the gear ratio of the transmission, an accurate output variable can be obtained.

[0014] In the hydraulic pressure estimation device with the above structure, as one of the multiple input variables, an oil temperature variable may also be included, where the oil temperature variable is a variable representing the temperature of the working oil supplied to the torque converter. According to the hydraulic pressure estimation device with the above structure, an output variable corresponding to the temperature of the working oil that may affect the estimated pressure difference can be obtained.

[0015] A second aspect of the present invention is a non-transitory storage medium that stores commands that can be executed by one or more processors included in an execution device of a hydraulic estimation device and that cause the one or more processors to execute the following functions. The functions include: an acquisition process of acquiring a value of an input variable; and a calculation process of inputting the value of the input variable acquired by the acquisition process into a map to calculate a value of an output variable. Further, the hydraulic estimation device uses, as an analog object, a torque converter having two oil chambers and a lock-up clutch that switches between an engaged state and a released state according to a pressure difference between the two oil chambers, and includes a storage device and the execution device. The storage device stores map data, which is data defining a map that outputs a estimated pressure difference variable as the output variable by being input with the input variable and that has been learned by machine learning. The estimated pressure difference variable is a variable representing an estimated pressure difference, which is an estimated value of the pressure difference generated in the torque converter. Here, the map includes an indication of the pressure difference variable as one of the plurality of input variables. The indication pressure difference variable is a variable representing an indication pressure difference, which is an indication value of the pressure difference in the torque converter.

[0016] Regarding the storage medium according to the second aspect described above, when the map is used for calculating the estimated pressure difference, it is sufficient to use appropriate data to learn the map so that the estimated pressure difference can be calculated appropriately. Therefore, in reflecting the correlation between the indication pressure difference and the actual pressure difference in the calculation of the estimated pressure difference, there is no need for the effort as in the case of deriving a complex relational expression. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Hereinafter, features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described with reference to the drawings, in which like reference numerals denote like elements, and wherein:

[0018] Figure 1 is a schematic configuration diagram of a vehicle equipped with an embodiment of the present invention.

[0019] Figure 2 is Figure 1 a schematic configuration diagram of the torque converter shown in

[0020] Figure 3 is a diagram showing Figure 1 an example of a lock-up map shown in

[0021] Figure 4 is a diagram showing Figure 1 an example of the temporal change of the indication pressure difference in the engagement process executed by the control device shown in

[0022] Figure 5It is a flowchart showing the processing steps of the pressure difference estimation process executed by the control device.

[0023] Figure 6 It is a schematic diagram showing a modified example of the hydraulic pressure estimation device of the above-described embodiment. Detailed Embodiment

[0024] Hereinafter, an embodiment of a hydraulic pressure estimation device applied to a vehicle will be described with reference to the accompanying drawings.

[0025] First, the schematic structure of the vehicle will be described. As Figure 1 shown, the vehicle 500 has an internal combustion engine 10, a torque converter 20, an automatic transmission 80, a differential 71, and drive wheels 72. The internal combustion engine 10 is a drive source of the vehicle 500. A crankshaft, which is an output shaft of the internal combustion engine 10, is connected to an input shaft 21 of the torque converter 20. The torque converter 20 is a fluid coupling having a torque amplification function. An output shaft 22 of the torque converter 20 is connected to an input shaft of the automatic transmission 80. The automatic transmission 80 is a multi-stage transmission. An output shaft of the automatic transmission 80 is connected to the drive wheels 72 via the differential 71. The differential 71 allows a rotational speed difference to occur between the left and right drive wheels 72.

[0026] The torque converter 20 has an impeller 23 and a turbine runner 24. Specifically, the impeller 23 is connected to the input shaft 21 of the torque converter 20. The impeller 23 rotates integrally with the input shaft 21. The turbine 24 is located at a position facing the impeller 23. The turbine 24 is connected to the output shaft 22 of the torque converter 20. The turbine 24 rotates integrally with the output shaft 22. In the torque converter 20, torque is transmitted between the impeller 23 and the turbine 24 via a fluid.

[0027] As Figure 2 shown, the torque converter 20 has a lock-up clutch 30. The lock-up clutch 30 is a multi-plate clutch. That is, the lock-up clutch 30 has a plurality of first friction plates 32 and a plurality of second friction plates 34. In the present embodiment, the first friction plates 32 and the second friction plates 34 each have two pieces. In addition, only one first friction plate 32 is shown in Figure 2 . The first friction plates 32 and the second friction plates 34 are alternately arranged. The adjacent first friction plate 32 and second friction plate 34 face each other. The first friction plate 32 and the second friction plate 34 can move in their arrangement direction. The first friction plate 32 rotates integrally with the impeller 23. In addition, the second friction plate 34 rotates integrally with the turbine 24. Further, a set of friction plates composed of a plurality of first friction plates 32 and a plurality of second friction plates 34 is referred to as a friction plate group.

[0028] The torque converter 20 has a front cover 20a and a piston 37. The front cover 20a is located on one side of the friction plate group in the arrangement direction of the first friction plate 32 and the second friction plate 34. The front cover 20a forms a part of the housing of the torque converter 20. In addition, the piston 37 is located on the other side of the friction plate group in the arrangement direction of the first friction plate 32 and the second friction plate 34. The piston 37 can move in the arrangement direction of the first friction plate 32 and the second friction plate 34. When the piston 37 approaches the front cover 20a, the friction plate group is clamped between the piston 37 and the front cover 20a. In this case, the adjacent first friction plate 32 and second friction plate 34 are in contact with each other. When the piston 37 moves away from the front cover 20a, the state in which the friction plate group is clamped between the piston 37 and the front cover 20a is released. In this case, the adjacent first friction plate 32 and second friction plate 34 are arranged at positions separated from each other.

[0029] The torque converter 20 has a control oil chamber 25, a front oil chamber 26, and a rear oil chamber 27. Specifically, in the torque converter 20, the control oil chamber 25 is demarcated near the friction plate group. The control oil chamber 25 is connected to a control port 25a. The control oil chamber 25 receives the supply of working oil via the control port 25a. In addition, the working oil in the control oil chamber 25 is discharged to the outside via the control port 25a.

[0030] The front oil chamber 26 is demarcated near the control oil chamber 25. The front oil chamber 26 is connected to a supply port 26a. The front oil chamber 26 receives the supply of working oil via the supply port 26a. In addition, the rear oil chamber 27 is demarcated in the torque converter 20. The rear oil chamber 27 communicates with the front oil chamber 26. The rear oil chamber 27 is connected to a discharge port 27a. The working oil in the rear oil chamber 27 is discharged to the outside via the discharge port.

[0031] The front oil chamber 26 is adjacent to the control oil chamber 25 in the arrangement direction of the first friction plate 32 and the second friction plate 34. A part of the piston 37 extends between the front oil chamber 26 and the control oil chamber 25. The front oil chamber 26 is located on the side closer to the front cover 20a than the piston 37 in the arrangement direction of the first friction plate 32 and the second friction plate 34. In addition, although detailed illustrations are omitted, the piston 37 separates the front oil chamber 26 and the control oil chamber 25. Therefore, the piston 37 moves in the arrangement direction of the first friction plate 32 and the second friction plate 34 according to the pressure difference between the front oil chamber 26 and the control oil chamber 25.

[0032] The operation state of the lock-up clutch 30 is switched by the piston 37 operating according to the pressure difference between the front oil chamber 26 and the control oil chamber 25. Further, hereinafter, the value obtained by subtracting the hydraulic pressure PF of the front oil chamber 26 from the hydraulic pressure PC of the working oil in the control oil chamber 25 (hereinafter referred to as hydraulic pressure) will be referred to as the lock-up pressure difference ΔP. In the present embodiment, when the lock-up pressure difference ΔP is negative, that is, when the hydraulic pressure PC of the control oil chamber 25 is smaller than the hydraulic pressure PF of the front oil chamber 26, the piston 37 moves away from the front cover 20a. In this case, the adjacent first friction plate 32 and second friction plate 34 are arranged at positions separated from each other. That is, the operation state of the lock-up clutch 30 becomes the released state. Since the adjacent first friction plate 32 and second friction plate 34 are separated, the pump impeller 23 and the turbine 24 become a non-directly connected state.

[0033] On the other hand, when the lock-up pressure difference ΔP is zero or more, that is, when the hydraulic pressure PC of the control oil chamber 25 is equal to or higher than the hydraulic pressure PF of the front oil chamber 26, the piston 37 approaches the front cover 20a. And, the adjacent first friction plate 32 and second friction plate 34 come into contact with each other. That is, the operation state of the lock-up clutch 30 becomes the engaged state. Since the adjacent first friction plate 32 and second friction plate 34 are in contact, the pump impeller 23 and the turbine 24 become a mechanically directly connected state.

[0034] Further, the state when the operation state of the lock-up clutch 30 is the engaged state has two states, a semi-engaged state and a fully engaged state. When the lock-up pressure difference ΔP is correspondingly small, the operation state of the lock-up clutch 30 becomes the semi-engaged state. In this case, the adjacent first friction plate 32 and second friction plate 34 become a slipping state where they slide relative to each other. Further, when the lock-up pressure difference ΔP becomes a maximum value that can be tolerated and becomes quite large, the operation state of the lock-up clutch 30 becomes the fully engaged state. In this case, the adjacent first friction plate 32 and second friction plate 34 do not slide.

[0035] According to the respective operating states of the lock-up clutch 30 described above, the form of torque transmission through the torque converter 20 is switched. When the lock-up clutch 30 is in the released state, the torque transmission amount via the lock-up clutch 30 becomes zero. Further, the torque converter 20 performs torque transmission from the input shaft 21 to the output shaft 22 through torque transmission between the pump impeller 23 and the turbine 24 via the fluid. When the lock-up clutch 30 is in the semi-engaged state, the torque transmission efficiency varies according to the degree of slip between the first friction plate 32 and the second friction plate 34. The smaller the slip amount, the smaller the difference in rotational speed between the pump impeller 23 and the turbine 24, and the higher the torque transmission efficiency from the input shaft 21 to the output shaft 22. When the lock-up clutch 30 is in the fully engaged state, the pump impeller 23 and the turbine 24 rotate integrally. Thereby, torque is transmitted from the input shaft 21 to the output shaft 22 with substantially no loss.

[0036] The vehicle 500 is provided with a hydraulic circuit 40. The hydraulic circuit 40 is a flow path through which the working oil flows. The hydraulic circuit 40 is connected to the control port 25a, the supply port 26a, and the discharge port 27a. A plurality of solenoid valves 42 are provided in the middle of the hydraulic circuit 40. By switching these solenoid valves 42, the supply and discharge of the working oil to and from the control oil chamber 25, the front oil chamber 26, and the rear oil chamber 27 are performed.

[0037] As Figure 1 shown, the vehicle 500 is equipped with a vehicle speed sensor 58, an accelerator sensor 59, an acceleration sensor 64, and an oil temperature sensor 56. In addition, an accelerator pedal 60 is provided in the vehicle 500. The vehicle speed sensor 58 detects the vehicle speed SP which is the traveling speed of the vehicle 500. The accelerator sensor 59 detects the accelerator operation amount ACCP which is the operation amount of the accelerator pedal 60 operated by the driver. The acceleration sensor 64 detects the longitudinal acceleration W which is the acceleration in the longitudinal direction of the vehicle 500. The oil temperature sensor 56 detects the oil temperature L which is the temperature of the working oil flowing in the hydraulic circuit 40.

[0038] Next, an overview of the control structure of the vehicle will be described. The vehicle 500 has a control device 90. The control device 90 can be configured as one or more processors that execute various processes according to a computer program (software). In addition, the control device 90 can also be configured as one or more dedicated hardware circuits such as an application-specific integrated circuit (ASIC) that executes at least a part of the various processes, or a circuitry including a combination thereof. The processor includes a CPU 91 and memories such as a RAM and a ROM 93. The memory stores program codes or instructions configured to cause the CPU 91 to execute processes. The memory, that is, the computer-readable medium includes all available media that can be accessed by a general-purpose or dedicated computer. In addition, the control device 90 has a storage device 95, which is a non-volatile memory that can be electrically rewritten. The CPU 91, the ROM 93, and the storage device 95 can communicate with each other via an internal bus 98.

[0039] Detection signals from various sensors installed in the vehicle 500 are input to the control device 90. Specifically, detection signals regarding the following respective parameters are input to the control device 90.

[0040] Vehicle speed SP detected by a vehicle speed sensor 58

[0041] Accelerator operation amount ACCP detected by an accelerator sensor 59

[0042] Longitudinal and lateral acceleration W detected by an acceleration sensor 64

[0043] Oil temperature L detected by an oil temperature sensor 56

[0044] The CPU 91 controls the internal combustion engine 10 and the automatic transmission 80, etc. by executing various programs stored in the ROM 93. Specifically, the CPU 91 calculates the target torque of the internal combustion engine 10 based on the accelerator operation amount ACCP and the vehicle speed SP. And, the CPU 91 controls the internal combustion engine 10 so that the output torque of the internal combustion engine 10 becomes the target torque. In addition, the CPU 91 calculates the target gear stage SFT of the automatic transmission 80 based on the accelerator operation amount ACCP and the vehicle speed SP. And, the CPU 91 controls the automatic transmission 80 so that the gear stage of the automatic transmission 80 becomes the target gear stage SFT.

[0045] The CPU 91 controls the lock-up clutch 30 by executing the program stored in the ROM 93. In controlling the lock-up clutch 30, the CPU 91 executes target determination processing, engagement processing, and stage determination processing. Hereinafter, these processes will be described in sequence.

[0046] Next, the target determination process will be described. The CPU 91 repeatedly performs the target determination process during the running of the vehicle 500. The target determination process is a process for determining the target operating state of the lock-up clutch 30. In the target determination process, the CPU 91 determines the target operating state of the lock-up clutch based on the accelerator operation amount ACCP, the vehicle speed SP, the target gear stage SFT, and the acceleration state of the vehicle 500. At that time, the CPU 91 refers to the lock-up map stored in the ROM 93.

[0047] As Figure 3 shown, the lock-up map (Japanese: ロックアップマップ) is a map (Japanese: マップ) that determines the target operating state for each operating region of the vehicle 500 specified by the vehicle speed SP and the accelerator operation amount ACCP. The lock-up map is prepared for each target gear stage SFT. In each lock-up map, generally speaking, the release state is set in the operating region where the accelerator operation amount ACCP is large. In the operating region where the accelerator operation amount ACCP is small, the release state is set on the low vehicle speed side, and the fully engaged state is set on the high vehicle speed side. Also, the semi-engaged state is set in the operating region between the release state and the fully engaged state and where the accelerator operation amount ACCP is greater than zero. In addition, as shown by the thick solid line of Figure 3 , the semi-engaged state is also set in the operating region where the accelerator operation amount ACCP is zero. That is, there is an operating region where the operating state of the lock-up clutch is set to the semi-engaged state during the deceleration of the vehicle 500.

[0048] When determining the target operating state of the lock-up clutch 30, the CPU 91 selects the lock-up map corresponding to the latest target gear stage SFT. And, in the selected lock-up map, the CPU 91 determines the target operating state corresponding to the latest accelerator operation amount ACCP and vehicle speed SP.

[0049] Next, the engagement process will be described. When the target operating state of the lock-up clutch 30 is switched from the release state to the engaged state in the above-described target determination process, the CPU 91 executes the engagement process. The engagement process is a process for switching the operating state of the lock-up clutch 30 to the engaged state and maintaining the operating state of the lock-up clutch 30 as the engaged state. In the engagement process, the CPU 91 repeatedly calculates the indicated pressure difference ΔPG, which is an indicated value of the lock-up pressure difference ΔP of the torque converter 20. When calculating the indicated pressure difference ΔPG, the CPU 91 outputs a control signal based on the indicated pressure difference ΔPG to the hydraulic circuit 40. According to this control signal, each solenoid valve 42 of the hydraulic circuit 40 operates, and the working oil is supplied and discharged to and from the control oil chamber 25, the front oil chamber 26, and the rear oil chamber 27. That is, the CPU 91 outputs the indicated pressure difference ΔPG of the torque converter 20 through the control signal, thereby supplying and discharging the working oil to and from each oil chamber of the torque converter 20.

[0050] The CPU 91 changes the indicated pressure difference ΔPG in the following manner during the engagement process. Hereinafter, the operation of the lock-up clutch 30 accompanying the engagement process will be described in accordance with the manner in which the indicated pressure difference ΔPG is changed.

[0051] like Figure 4 As shown, when the engagement process starts at time T1, the CPU 91 rapidly increases the indicated pressure difference ΔPG for the torque converter from the value P0 set in the released state to the application differential pressure (application differential pressure) P1. And, the CPU 91 maintains the indicated pressure difference ΔPG as the application differential pressure P1 until a predetermined first period has passed since the start of the engagement process. During this period, the locking clutch 30 is started to be pressed (Japanese: パック詰め). That is, the interval between the adjacent first friction plate 32 and the second friction plate 34 begins to be narrowed. The application differential pressure P1 is predetermined as the locking pressure difference ΔP required to start the pressing. In addition, the timing when the CPU 91 starts the engagement process is the timing when the lockup clutch 30 starts to switch from the released state to the engaged state.

[0052] The CPU 91 reduces the indicated pressure difference ΔPG to the standby pressure difference P2 at time T2 after the first period has passed since time T1. The CPU 91 maintains the indicated pressure difference ΔPG at the standby pressure difference P2 until a predetermined second period has passed since the indicated pressure difference ΔPG was reduced to the standby pressure difference P2. During this period, the locking clutch 30 is tightened. The standby pressure difference P2 is predetermined as the lockup pressure difference ΔP immediately before the adjacent first friction plate 32 and second friction plate 34 come into contact.

[0053] When it becomes time T3 after the second period has elapsed from time T2, CPU 91 ends maintaining the indicated pressure difference ΔPG at the standby pressure difference P2. Before the third period, which is predetermined, has elapsed since time T3, CPU 91 causes the indicated pressure difference ΔPG to continuously increase at a pressure difference change rate α. During this period, the slip amount of the lock-up clutch 30 gradually decreases. In addition, at the moment when the above-described second period has elapsed, CPU 91 calculates a final target pressure difference P3 based on the accelerator operation amount ACCP, vehicle speed SP, target shift stage SFT, oil temperature L, and longitudinal acceleration W of the vehicle 500. The final target pressure difference P3 is the lock-up pressure difference ΔP required to achieve the engaged state defined as the target operating state, that is, the fully engaged state or the semi-engaged state at a certain slip amount. When calculating the final target pressure difference P3, CPU 91 divides the value obtained by subtracting the standby pressure difference P2 from the final target pressure difference P3 by the third period, and calculates this value as the pressure difference change rate α. That is, the pressure difference change rate α is the change rate of the indicated pressure difference ΔPG per unit time for causing the indicated pressure difference ΔPG to reach the final target pressure difference P3 during the third period. CPU 91 causes the indicated pressure difference ΔPG to continuously increase based on this pressure difference change rate α.

[0054] When it becomes time T4 after the third period has elapsed from time T3, CPU 91 ends increasing the indicated pressure difference ΔPG. Then, until the target operating state of the lock-up clutch 30 switches from the engaged state to the released state after time T4, CPU 91 maintains the indicated pressure difference ΔPG at the final target pressure difference P3. Accordingly, the lock-up clutch 30 is maintained in the fully engaged state or the semi-engaged state at a certain slip amount.

[0055] In addition, when the target operating state of the lock-up clutch 30 switches from the engaged state to the released state, CPU 91 ends the engagement process. After that, as a process for switching the operating state of the lock-up clutch 30 to the released state, CPU 91 starts a process of decreasing the indicated pressure difference ΔPG. That is, the timing at which CPU 91 ends the engagement process is the timing at which the lock-up clutch 30 starts to switch from the engaged state to the released state.

[0056] Next, the stage determination process will be described. During the execution of the engagement process, CPU 91 repeatedly performs the stage determination process. The stage determination process is a process of determining stages obtained by dividing the execution process of the engagement process into a plurality of stages, and calculating a stage variable PHS corresponding to the determined stage. The stage variable PHS is a variable for determining each stage of the engagement process.

[0057] In the present embodiment, CPU 91 determines four stages. As Figure 4As shown, the first stage Q1 is the stage from time T1 to time T2 during a series of processes of card engagement processing in which the indicated pressure difference ΔPG is maintained as the applied pressure difference P1. The second stage Q2 is the stage from time T2 to time T3 in which the indicated pressure difference ΔPG is maintained as the standby pressure difference P2. The third stage Q3 is the stage from time T3 to time T4 in which the indicated pressure difference ΔPG rises at a pressure difference change rate α. The fourth stage Q4 is the stage after time T4 in which the indicated pressure difference ΔPG is maintained as the final target pressure difference P3. The CPU determines the stage in the card engagement processing with reference to the processing content executed in the card engagement processing.

[0058] When determining the stage, the CPU 91 calculates a stage variable PHS. In the present embodiment, the stage variable PHS is determined as a positive integer for identifying each stage. For example, it is "1" for the first stage Q1 and "2" for the second stage Q2. The CPU 91 calculates a value corresponding to the determined stage as the stage variable PHS.

[0059] Next, the control structure related to the calculation of the estimated pressure difference will be described. The control device 90 functions as a hydraulic estimation device that calculates an estimated pressure difference ΔPS, which is an estimated value of the lock-up pressure difference ΔP actually generated in the torque converter 20. That is, the control device 90 takes the torque converter 20 as a simulation object and simulates the actual lock-up pressure difference ΔP of the torque converter 20.

[0060] As Figure 1 shown, the storage device 95 stores mapping data D. The mapping data D is data that defines a mapping Dx that outputs an output variable by being input with an input variable. In the present embodiment, the input variables are an indicated pressure difference variable that is a variable representing the indicated pressure difference ΔPG, an acceleration variable that is a variable representing the longitudinal and lateral acceleration W, an accelerator variable that is a variable representing the accelerator operation amount ACCP, a vehicle speed variable that is a variable representing the vehicle speed SP, a shift variable that is a variable representing the gear ratio of the automatic transmission 80, and an oil temperature variable that is a variable representing the oil temperature L. The output variable is an estimated pressure difference variable that is a variable representing the estimated pressure difference ΔPS.

[0061] The mapping data D stored in the storage device 95 includes the mappings Dx for each stage of the card engagement processing. That is, the storage device 95 stores four mappings Dx corresponding to the first stage Q1 to the fourth stage Q4 obtained by dividing the progress of the card engagement processing into four.

[0062] The CPU 91 is capable of performing a pressure difference estimation process, which is a process of calculating an estimated pressure difference ΔPS during the execution of the card fitting process. The CPU 91 implements each process of the pressure difference estimation process by executing a program stored in the ROM 93. In the present embodiment, the CPU 91 and the ROM 93 constitute an execution device.

[0063] In the pressure difference estimation process, the CPU 91 performs an acquisition process, a selection process, and a calculation process. In the acquisition process, the CPU 91 acquires various input variables required for calculating the estimated pressure difference ΔPS such as the indicated pressure difference ΔPG. In addition, when the CPU 91 outputs the indicated pressure difference ΔPG to the torque converter 20 during the card fitting process being executed in the background, that is, when the CPU 91 outputs a control signal to the hydraulic circuit 40, the CPU 91 acquires various input variables. In the selection process, the CPU 91 selects a map Dx corresponding to the stage at which the indicated pressure difference ΔPG is output from the maps Dx at each stage stored in the storage device 95. In the calculation process, the CPU 91 calculates the value of the output variable by inputting the values of the input variables acquired in the acquisition process into the map Dx selected in the selection process.

[0064] Next, the pressure difference estimation process will be described. During the execution of the card fitting process, the CPU 91 repeatedly performs the pressure difference estimation process. As Figure 5 shown, when starting the pressure difference estimation process, the CPU 91 executes the process of step S10. In step S10, the CPU 91 acquires various variables required for the processes after step S20. Specifically, the CPU 91 acquires the stage variable PHS, the indicated pressure difference ΔPG, the acceleration identification value WD, the accelerator operation amount ACCP, the vehicle speed SP, the target gear stage SFT, and the oil temperature L.

[0065] Regarding the stage variable PHS, the CPU 91 acquires the latest value calculated in the stage determination process. Regarding the indicated pressure difference ΔPG, the CPU 91 acquires the latest value calculated in the card fitting process. In addition, the indicated pressure difference ΔPG is the above-mentioned indicated pressure difference variable. Regarding the accelerator operation amount ACCP, the CPU 91 acquires the latest value input from the accelerator sensor 59 to the control device 90. In addition, the accelerator operation amount ACCP is the above-mentioned accelerator operation amount variable. Regarding the vehicle speed SP, the CPU 91 acquires the latest value input from the vehicle speed sensor 58 to the control device 90. Regarding the target gear stage SFT, the CPU 91 acquires the latest value calculated for controlling the automatic transmission 80. In addition, the target gear stage SFT is the above-mentioned gear shift variable. Regarding the oil temperature L, the CPU 91 acquires the latest value input from the oil temperature sensor 56 to the control device 90. In addition, the oil temperature L is the above-mentioned oil temperature variable.

[0066] The acceleration identification value WD is an identification value used to identify whether the vehicle 500 is accelerating or decelerating. In the present embodiment, the acceleration identification value WD is determined to be "1" when the vehicle 500 is accelerating, "2" when the vehicle 500 is decelerating, and "3" when the vehicle 500 is traveling at a constant speed. The CPU 91 obtains the latest value input from the acceleration sensor 64 to the control device 90 in step S10. When the longitudinal acceleration W is positive, that is, when the vehicle 500 is accelerating, the CPU 91 calculates the acceleration identification value WD as "1". On the other hand, when the longitudinal acceleration W is negative, that is, when the vehicle 500 is decelerating, the CPU 91 calculates the acceleration identification value WD as "2". When the longitudinal acceleration W is zero, that is, when the vehicle 500 is traveling at a constant speed, the CPU 91 calculates the acceleration identification value WD as "3". The CPU 91 calculating the acceleration identification value WD is equivalent to the CPU 91 obtaining the acceleration identification value WD. In addition, the acceleration identification value WD is the above acceleration variable.

[0067] After obtaining various variables, the CPU 91 advances the process to step S20. In addition, the process of step S10 is an acquisition process. In step S20, the CPU 91 selects the map Dx used for calculating the estimated pressure difference ΔPS based on the phase variable PHS. The storage device 95 stores a selection map associating the phase variable PHS with the map Dx of each phase. The CPU 91 refers to the selection map and selects the map Dx of the phase corresponding to the phase variable PHS obtained in step S10 from the maps Dx of each phase stored in the storage device 95. When the map Dx is selected, the CPU 91 advances the process to step S30. In addition, the process of step S20 is a selection process.

[0068] In step S30, as a pre-process for calculating the estimated pressure difference ΔPS, the CPU 91 substitutes the values of various variables obtained in the process of step S10 into the input variables x(1) to x(6) for input to the map. Specifically, the CPU 91 substitutes the indicated pressure difference ΔPG into the input variable x(1). The CPU 91 substitutes the acceleration identification value WD into the input variable x(2). The CPU 91 substitutes the accelerator operation amount ACCP into the input variable x(3). The CPU 91 substitutes the vehicle speed SP into the input variable x(4). The CPU 91 substitutes the target gear stage SFT into the input variable (5). The CPU 91 substitutes the oil temperature L into the input variable x(6). After that, the CPU 91 advances the process to step S40.

[0069] In step S40 , the CPU 91 inputs the input variables x( 1 ) to x( 6 ) to the map Dx selected in step S20 , thereby calculating the output variable y. The output variable y is the estimated pressure difference ΔPS.

[0070] The mapping Dx is constructed as a fully connected feedforward neural network (Japanese: fully connected forward neural network) with one intermediate layer. The above-mentioned neural network includes an activation function h(x) as an input-side nonlinear mapping, which performs nonlinear transformation on the input-side coefficients wFjk (j=0~n, k=0~6) and the output of the input-side linear mapping as a linear mapping specified by the input-side coefficients wFjk. In this embodiment, the hyperbolic tangent "tanh(x)" is exemplified as the activation function h(x). In addition, the above-mentioned neural network includes an activation function f(x) as an output-side nonlinear mapping, which performs nonlinear transformation on the output-side coefficients wSj (j=0~n) and the output of the output-side linear mapping as a linear mapping specified by the output-side coefficients wSj. In this embodiment, the hyperbolic tangent "tanh(x)" is exemplified as the activation function f(x). In addition, the value n represents the dimension of the intermediate layer. The input side coefficient wFj0 ​​is a bias parameter and is a coefficient of the input variable x(0). The input variable x(0) is defined as "1". In addition, the output side coefficient wS0 is a bias parameter.

[0071] The mapping Dx is a learned model that has undergone machine learning using a transmission device with an internal combustion engine 10, a torque converter 20, an automatic transmission 80, a hydraulic circuit 40, etc. mounted on the vehicle 500 before being installed on the vehicle 500. During the learning of the mapping Dx, training data and teacher data are obtained in advance. That is, by installing the transmission device on a chassis dynamometer and simulating the vehicle's driving, training data and teacher data are produced. In producing the training data and teacher data, the engagement process is performed by setting the state of the transmission device to various states. The state of the transmission device is defined by a combination of the vehicle's acceleration state, accelerator operation amount ACCP, vehicle speed SP, gear stage, and oil temperature L. The vehicle's acceleration state can be determined as one of the states where the vehicle is accelerating, decelerating, or traveling at a constant speed. That is, for the three modes of the vehicle's acceleration state, the engagement process is performed by simulating the conditions where the values of the other above-mentioned parameters are combined in various ways, and the actual lock-up pressure difference (hereinafter referred to as the actual pressure difference) ΔPR at each condition is obtained as the teacher data. In addition, at the same timing as obtaining the actual pressure difference ΔPR, the values of various variables that are input variables for the mapping Dx are obtained as the training data. At this time, the values of the various variables are obtained in the same manner as the processing in step S10. In addition, a hydraulic sensor can be installed in the hydraulic circuit 40, and the actual pressure difference ΔPR can be calculated based on the detection value of this hydraulic sensor. Specifically, the hydraulic pressure of the oil passage connected to the control port 25a in the hydraulic circuit 40 is detected as the hydraulic pressure PC of the control oil chamber 25. In addition, the hydraulic pressure of the oil passage connected to the supply port 26a in the hydraulic circuit 40 is detected as the hydraulic pressure PF of the front oil chamber 26. And the actual pressure difference ΔPR can be calculated based on these detection values. In this way, after obtaining the training data and teacher data for each state of the transmission device, the mapping Dx is learned using the set of the training data and teacher data for each state of the transmission device. That is, for various states of the transmission device, the input-side variables and output-side variables are adjusted so that the difference between the value output by the mapping Dx with the training data as the input and the teacher data that is the actual pressure difference ΔPR becomes below a predetermined value. And it is set that: the learning is completed according to the above difference becoming below the predetermined value. In addition, the learning of the mapping Dx is performed with the stage corresponding to this mapping Dx as the object.

[0072] When calculating the estimated pressure difference ΔPS as the output variable y in step S40, the CPU 91 temporarily ends the series of processes of the pressure difference estimation process. And the CPU 91 re-executes the process of step S10 with the condition that the engagement process is being executed. In addition, the process of step S40 is a calculation process.

[0073] Next, the effects of the embodiment will be described. During the execution of the engagement process, the CPU 91 calculates the estimated pressure difference ΔPS of the torque converter 20 using the mapping data D. At that time, the CPU 91 selects the mapping Dx corresponding to each stage of the engagement process, and inputs various input variables to the selected mapping Dx to calculate the estimated pressure difference ΔPS.

[0074] Next, the effects of the embodiment will be described.

[0075] (1) As Figure 4 indicated by the double-dot chain line in, the actual pressure difference ΔPR in the torque converter 20 may deviate from the indicated pressure difference ΔPG. For example, the actual pressure difference ΔPR starts to change later than the indicated pressure difference ΔPG, or changes slowly with respect to the rapid change of the indicated pressure difference ΔPG. For this reason, in order to accurately calculate the estimated pressure difference ΔPS, it is necessary to appropriately reflect the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR in the calculation of the estimated pressure difference ΔPS.

[0076] In order to reflect the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR in the calculation of the estimated pressure difference ΔPS, it can be considered to derive a relational expression representing the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR and use it in the calculation of the estimated pressure difference ΔPS. However, when trying to derive a single relational expression suitable for various driving states of the vehicle 500, the content of the relational expression becomes very complex or lacks accuracy, so it is difficult to derive such a relational expression. On the other hand, even if the above relational expression is to be derived according to the driving state of the vehicle 500, it is necessary to analyze the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR one by one according to the driving state of the vehicle 500 and derive a relational expression suitable for each, which will take time.

[0077] Regarding this point, in the case of calculating the estimated pressure difference ΔPS using the mapping Dx as in the present embodiment, as long as appropriate training data and teacher data can be prepared, there is no need for the effort of deriving a complex relational expression, and the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR can be reflected in the calculation of the estimated pressure difference ΔPS. Further, when calculating the estimated pressure difference ΔPS using the mapping Dx, as long as a certain amount of data can be prepared as training data and teacher data, the accuracy of the estimated pressure difference ΔPS can be ensured. Furthermore, in the present embodiment, not only the indicated pressure difference ΔPG but also a plurality of variables are used as input variables. Therefore, in calculating the estimated pressure difference ΔPS, the correlation between the plurality of variables and the actual pressure difference ΔPR can be considered, and the estimated pressure difference ΔPS can be calculated. Therefore, the estimated pressure difference ΔPS can be calculated with high accuracy. In addition, the calculated estimated pressure difference ΔPS can be used for learning processing of a control program related to hydraulic supply to the lock-up clutch 30, modification of the control program, design of a lock-up clutch of other vehicles, and the like.

[0078] (2) In the present embodiment, according to each stage of the engagement process, the manner of providing the indicated pressure difference ΔPG is different. For example, in the first stage Q1, the indicated pressure difference ΔPG is rapidly increased, while in the second stage Q2, the indicated pressure difference ΔPG is rapidly decreased. Further, in the third stage Q3, the indicated pressure difference ΔPG is gradually increased. Corresponding to the differences in each stage of the manner of providing the indicated pressure difference ΔPG, the response manner of the actual pressure difference ΔPR also differs for each stage. Therefore, in accurately calculating the estimated pressure difference ΔPS, it is necessary to reflect the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR in each stage in the calculation of the estimated pressure difference ΔPS in each stage.

[0079] In the present embodiment, dedicated mappings Dx are used for each stage, and therefore, the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR in each stage can be reflected in the calculation of the estimated pressure difference ΔPS. Therefore, the calculation accuracy of the estimated pressure difference ΔPS in each stage becomes high.

[0080] (3) During acceleration and deceleration of the vehicle 500, the magnitudes of the torques acting on the torque converter 20 are different. Here, it is assumed that when the working state of the lock-up clutch 30 is the engaged state, torque transmission is performed with the same slip amount during acceleration and deceleration of the vehicle 500. During acceleration of the vehicle 500, the output of the internal combustion engine 10 is input to the input shaft 21 of the torque converter 20. And, the output shaft 22 rotates following the input shaft 21 of the torque converter 20. On the other hand, during deceleration of the vehicle 500, the output from the internal combustion engine 10 becomes smaller. And, in the torque converter 20, the input shaft 21 rotates following the output shaft 22. The magnitude of the torque transmitted between the input shaft 21 and the output shaft 22 during deceleration of the vehicle 500 becomes smaller in order unit than the magnitude of the torque transmitted during acceleration of the vehicle 500, that is, the magnitude of the torque corresponding to the output of the internal combustion engine 10. Due to such a difference, even when the same slip amount is achieved during acceleration and deceleration of the vehicle 500, the magnitude of the required lock-up pressure difference ΔP is different.

[0081] Therefore, during acceleration and deceleration of the vehicle 500, even if the slip amount of the target working state of the lock-up clutch is the same, the final target pressure difference P3 will be different. And, as the final target pressure difference P3 is different, the change manner of the indicated pressure difference ΔPG such as the pressure difference change rate α for making the indicated pressure difference ΔPG reach the final target pressure difference P3 is different. Due to the different change manners of the indicated pressure difference ΔPG, the correlation between the indicated pressure difference ΔPG and the actual pressure difference ΔPR may also be different during acceleration and deceleration of the vehicle 500.

[0082] In the present embodiment, one of the input variables of the map Dx includes the acceleration identification value WD. Accordingly, it is possible to calculate the estimated pressure difference ΔPS in consideration of the response characteristics of the actual pressure difference ΔPR corresponding to the respective change manners of the indicated pressure difference ΔPG during acceleration and deceleration of the vehicle 500.

[0083] (4) The magnitudes of the torques input to the torque converter 20 according to the accelerator operation amount ACCP are different. Therefore, similarly to the above (3), the change manner of the indicated pressure difference ΔPG is different according to the accelerator operation amount ACCP. As in the present embodiment, by making one of the input variables of the map Dx include the accelerator operation amount ACCP, it is possible to calculate the estimated pressure difference ΔPS in consideration of the response characteristics of the actual pressure difference ΔPR corresponding to the change manner of the indicated pressure difference ΔPG for each accelerator operation amount ACCP.

[0084] (5) In the lock-up map, the operating range where the target operating state of the lock-up clutch 30 becomes the fully engaged state is set on the high vehicle speed side compared to the operating range where the target operating state becomes the semi-engaged state. That is, the slip amount defined as the target operating state of the lock-up clutch 30 varies according to the vehicle speed SP. Therefore, the final target pressure difference P3 and the change pattern of the indicated pressure difference ΔPG until reaching the final target pressure difference P3 vary according to the vehicle speed SP. In the present embodiment, by including the vehicle speed SP as one of the input variables of the map Dx, it is possible to calculate the estimated pressure difference ΔPS in consideration of the response characteristics of the actual pressure difference ΔPR corresponding to the change pattern of the indicated pressure difference ΔPG at each vehicle speed SP.

[0085] (6) Similar to the vehicle speed SP, the slip amount defined as the target operating state of the lock-up clutch 30 varies according to the target gear stage SFT. Therefore, the change pattern of the indicated pressure difference ΔPG varies according to the target gear stage SFT. In the present embodiment, since one of the input variables of the map Dx includes the target gear stage SFT, it is possible to calculate the estimated pressure difference ΔPS in consideration of the response characteristics of the actual pressure difference ΔPR corresponding to the change pattern of the indicated pressure difference ΔPG at each target gear stage SFT.

[0086] (7) The lower the oil temperature L, the higher the viscosity of the working oil. Due to the correlation between the oil temperature L and the viscosity of the working oil, the oil temperature L may affect the responsiveness of the actual pressure difference ΔPR to the indicated pressure difference ΔPG. As in the present embodiment, by including the oil temperature L as one of the input variables, it is possible to calculate the estimated pressure difference ΔPS in consideration of the responsiveness of the actual pressure difference ΔPR to the indicated pressure difference ΔPG corresponding to the oil temperature L.

[0087] Next, a modified example of the present embodiment will be described. The modified example of the present embodiment is implemented by changing the above-described embodiment as follows. The present embodiment and the following modified examples can be implemented in combination with each other within a technically non-contradictory range.

[0088] Part of the pressure difference estimation process may also be performed by a computer outside the vehicle 500. For example, as Figure 6As shown, a server 600 may also be provided outside the vehicle 500. Also, the structure may be such that the selection process and the calculation process in the pressure difference estimation process are performed in the server 600. In this case, the server 600 may be configured as one or more processors that execute various processes according to a computer program (software). In addition, the server 600 may be configured as one or more dedicated hardware circuits such as an application specific integrated circuit (ASIC) that executes at least a part of the various processes, or a circuitry including a combination thereof. The processor includes a CPU 602 and a memory such as a RAM and a ROM 604. The memory stores program codes or instructions configured to cause the CPU 602 to execute processes. The memory, i.e., the computer-readable medium, includes all available media that can be accessed by a general or a dedicated computer. In addition, the server 600 has a storage device 606, which is a non-volatile memory that can be electrically rewritten. The storage device 606 stores the mapping data D described in the above embodiment, i.e., the mapping Dx at each stage. In addition, the server 600 has a communicator 610 for connecting to the outside of the server 600 through an external communication line network 700. The CPU 602, the ROM 604, the storage device 606, and the communicator 610 can communicate with each other through an internal bus 608.

[0089] In the case where the selection process and the calculation process in the pressure difference estimation process are performed by the server 600, the control device 90 of the vehicle 500 has a communicator 99 for communicating with the outside of the control device 90 through the external communication line network 700. In addition, aside from having the communicator 99, the structure of the control device 90 is the same as that of the above embodiment. Therefore, a detailed description of the control device 90 is omitted. In addition, in Figure 6 , parts that function in the same way as Figure 1 are labeled with the same reference numerals as Figure 1 The control device 90 and the server 600 together form a hydraulic pressure estimation device Z.

[0090] In the case where the selection process and the calculation process in the pressure difference estimation process are performed by the server 600, first, the control device 90 of the vehicle 500 performs an acquisition process, which is the process of step S10 of the above embodiment. When various variables are acquired through the process of step S10, the control device 90 transmits the values of the acquired various variables to the server 600. When receiving the values of the various variables, the CPU 602 of the server 600 calculates the estimated pressure difference ΔPS by performing the processes of step S20, step S30, and step S40 of the above embodiment. The CPU 602 of the server 600 performs the processes of step S20, step S30, and step S40 by executing the program stored in the ROM 604.

[0091] In the case where the differential pressure estimation process is performed by the control device 90 of the vehicle 500 and the server 600 as in this modification example, the CPU 91, ROM 93 of the control device 90 of the vehicle 500, and the CPU 602, ROM 604 of the server 600 constitute an execution device.

[0092] All processes of the differential pressure estimation process may also be performed outside the vehicle 500. For example, as in the above modification example, when the server 600 is provided outside the vehicle 500, the control device 90 of the vehicle 500 transmits the detection signals of various sensors installed in the vehicle 500 to the server 600. In addition, the control device 90 of the vehicle 500 also transmits other variables used in the differential pressure estimation process, such as the phase variable PHS and the target shift stage SFT, to the server 600. Then, the CPU 602 of the server 600 obtains the values of various variables by performing a process corresponding to step S10 of the above-described embodiment. After that, in the same manner as in the above modification example, the CPU 602 of the server 600 performs processes corresponding to step S20, step S30, and step S40. In such a configuration, the acquisition process, selection process, and calculation process are performed in the server 600.

[0093] When the differential pressure estimation process is performed outside the vehicle 500, the differential pressure estimation process may also be performed regardless of the timing at which the indicated differential pressure ΔPG is output in the card combination process. That is, the differential pressure estimation process may be performed independently instead of being performed in correspondence with the execution of the card combination process. For example, if time series data of various variables required for calculating the estimated differential pressure ΔPS is prepared in advance, the estimated differential pressure ΔPS can be calculated by performing only the differential pressure estimation process regardless of the card combination process.

[0094] In the above-described embodiment, the same map Dx may also be used in each stage of the card combination process to calculate the estimated differential pressure ΔPS. In each stage of the card combination process, the behavior such as the delay of the actual differential pressure ΔPR with respect to the indicated differential pressure ΔPG is presumed to be substantially similar. Therefore, even if the same map Dx is used in each stage, a certain degree of accuracy can be expected for the estimated differential pressure ΔPS.

[0095] The estimated pressure difference ΔPS can also be calculated when switching the operating state of the lock-up clutch 30 from the engaged state to the released state. In this case, either the map Dx dedicated to the case of switching the operating state of the lock-up clutch 30 to the released state or the same map Dx as that for the case of switching the operating state of the lock-up clutch 30 to the engaged state can be used. Additionally, the map Dx for each stage obtained by dividing the following period into multiple stages can also be used, where the period is the period of switching the operating state of the lock-up clutch 30 to the released state.

[0096] The determination method of the stage in the engagement process is not limited to the examples of the above-described embodiments. For example, when determining the end timing of the third stage Q3, the rotational speed difference obtained by subtracting the rotational speed of the output shaft 22 from the rotational speed of the input shaft 21 of the torque converter 20 can be used. Specifically, the timing when the rotational speed difference reaches a value specified by the target operating state can be determined as the end timing of the third stage Q3. In this case, a crank angle sensor for detecting the rotational angle of the crankshaft of the internal combustion engine 10 can be provided in the vehicle 500, and the rotational speed of the crankshaft based on the detection value of this sensor can be used as the rotational speed of the input shaft 21. Additionally, a rotational angle sensor for detecting the rotational angle of the input shaft of the automatic transmission 80 can be provided in the vehicle 500, and the rotational speed of the input shaft of the automatic transmission 80 based on the detection value of this sensor can be used as the rotational speed of the output shaft 22 of the torque converter 20.

[0097] The division method of the stage in the engagement process is not limited to the examples of the above-described embodiments. For example, the stage of maintaining the indicated pressure difference ΔPG as the applied pressure difference P1 and the stage of maintaining the indicated pressure difference ΔPG as the standby pressure difference P2 can be combined into one stage. When changing the division method of the stage, the map Dx for each changed stage can be prepared.

[0098] The temporal change of the indicated pressure difference ΔPG in the engagement process is not limited to the examples of the above-described embodiments. For example, the first period, the second period, and the third period can be set variably according to the driving state of the vehicle 500 or the like. Additionally, the applied pressure difference P1 and the standby pressure difference P2 can be set variably according to the driving state of the vehicle 500 or the like. Additionally, instead of making the pressure difference change rate α constant throughout the third period, the pressure difference change rate α can be changed in the middle of the third period. The temporal change of the indicated pressure difference ΔPG in the engagement process only needs to be an appropriate temporal change for switching the lock-up clutch 30 to the engaged state. When changing the temporal change of the indicated pressure difference ΔPG according to the content of the above-described embodiments, the division method of the stage and the determination method of the stage can be appropriately determined corresponding to this change.

[0099] As described in the effects of the embodiment, the magnitude of the torque acting on the torque converter 20 varies according to the acceleration state of the vehicle 500. Thus, the map Dx can also be prepared for each acceleration state of the vehicle 500. Specifically, dedicated maps Dx can also be prepared for the case where the vehicle 500 is accelerating, the case where the vehicle 500 is decelerating, and the case where the vehicle 500 is traveling at a constant speed. Also, the estimated pressure difference ΔPS can be calculated by the dedicated map Dx for each acceleration state of the vehicle 500.

[0100] The accelerator operation amount ACCP obtained in step S10 is not limited to the latest value at the time of executing the process of step S10. For example, the maximum value of the accelerator operation amount ACCP during the period from the previous timing of executing step S10 to the next execution of step S10 can also be obtained. In addition, instead of obtaining an instantaneous value, the average value of the accelerator operation amount ACCP for a certain period can be obtained. The same applies to the vehicle speed SP and the oil temperature L. Also, from the same viewpoint, when calculating the acceleration identification value WD, instead of calculating the acceleration identification value WD based on the instantaneous value of the longitudinal acceleration W, the acceleration identification value WD can be calculated based on the average value of the longitudinal acceleration W for a certain period or the like.

[0101] The variable used as the indicated pressure difference variable is not limited to the example of the above embodiment. For example, a value obtained by multiplying the indicated pressure difference ΔPG by a correction coefficient or the like suitable for accurately calculating the estimated pressure difference ΔPS can also be used as the indicated pressure difference variable. The indicated pressure difference variable only needs to be a variable representing the indicated pressure difference ΔPG.

[0102] The variable used as the acceleration variable is not limited to the example of the above embodiment. For example, the rotational speed difference obtained by subtracting the rotational speed of the output shaft 22 from the rotational speed of the input shaft 21 of the torque converter 20 can also be used as the acceleration variable. When the lock-up clutch 30 is in a semi-engaged state, during the acceleration of the vehicle 500, the rotational speed of the input shaft 21 is higher than the rotational speed of the output shaft 22. On the other hand, during the deceleration of the vehicle 500, the rotational speed of the output shaft 22 is higher than the rotational speed of the input shaft 21. Therefore, during the acceleration and deceleration of the vehicle 500, the sign of the rotational speed difference is reversed. Thus, the rotational speed value can be an index indicating the acceleration state of the vehicle 500. As described in the above modification example, regarding the rotational speed of the input shaft 21, the rotational speed of the crankshaft can be set. Regarding the rotational speed of the output shaft 22, the rotational speed of the input shaft of the automatic transmission 80 can be set. The acceleration variable only needs to be a variable representing the longitudinal acceleration W of the vehicle 500.

[0103] The variables used as the accelerator operation amount variables are not limited to the examples of the above-described embodiments. The throttle opening degree of the internal combustion engine 10 has a positive correlation with the accelerator operation amount ACCP. Thus, for example, the throttle opening degree of the internal combustion engine 10 may also be used as the accelerator operation amount variable. In this case, an opening degree sensor for detecting the throttle opening degree may be provided in the internal combustion engine 10. The accelerator operation amount variable may be a variable representing the accelerator operation amount ACCP.

[0104] The variables used as the vehicle speed variables are not limited to the examples of the above-described embodiments. For example, the rotational speed of the output shaft of the automatic transmission 80 may also be used as the vehicle speed variable. In this case, a rotational angle sensor for detecting the rotational angle of the output shaft of the automatic transmission 80 may be provided in the vehicle 500, and the rotational speed of the output shaft 22 of the automatic transmission 80 may be calculated based on the detection value of the sensor. The vehicle speed variable may be a variable representing the vehicle speed SP.

[0105] The variables used as the shift variables are not limited to the examples of the above-described embodiments. For example, the ratio of the rotational speed of the input shaft to the rotational speed of the output shaft of the automatic transmission 80, that is, the actual gear ratio, may also be used as the shift variable. The shift variable may be a variable representing the gear ratio of the transmission.

[0106] The variables used as the oil temperature variables are not limited to the examples of the above-described embodiments. For example, the oil temperature L may be divided into multiple stages of grades, and the value representing such grades may be used as the oil temperature variable. The oil temperature variable may be a variable representing the oil temperature L.

[0107] Similar to the oil temperature variable related to the modification example, multiple grades may also be set for other input variables according to their respective degrees, and the values representing such grades may be used. The types of input variables are not limited to the examples of the above-described embodiments. The input variables may also use other variables instead of the variables shown in the above-described embodiments, or use other variables on the basis thereof. In addition, the number of input variables may be reduced from that of the above-described embodiments. The number of input variables may be two or more. One of the two or more input variables may include an indication pressure difference variable.

[0108] The acceleration variable, accelerator variable, vehicle speed variable, shift variable, and oil temperature variable are not essential as input variables. Even if these variables are not included, as long as two or more variables including the indication pressure difference variable are used as the input variables, the estimated pressure difference ΔPS can be calculated with high accuracy accordingly.

[0109] As the input variable, variables other than those shown in the above-described embodiments may also be used. As the input variable, for example, a variable indicating the degree of deterioration of the hydraulic circuit 40 over time may also be used. Specifically, as the variable indicating the degree of deterioration of the hydraulic circuit 40 over time, the total driving distance of the vehicle 500 or the like may be used. Corresponding to the degree of deterioration of the hydraulic circuit 40 over time, the responsiveness of the actual pressure difference ΔPR to the indicated pressure difference ΔPG may vary. Therefore, if a variable indicating the degree of deterioration of the hydraulic circuit 40 over time is used as one of the input variables, the estimated pressure difference ΔPS can be calculated taking into account the degree of deterioration of the hydraulic circuit 40 over time.

[0110] The variable used for estimating the hydraulic variable is not limited to the examples in the above-described embodiments. For example, the estimated value of the flow rate difference when the lock-up pressure difference ΔP is converted into the flow rate difference of the working oil may also be used as the variable for estimating the hydraulic variable. The variable for estimating the hydraulic variable only needs to be a variable indicating the estimated pressure difference ΔPS.

[0111] The structure of the map Dx is not limited to the examples in the above-described embodiments. For example, the number of intermediate layers in the neural network may be two or more.

[0112] As the neural network, for example, a recurrent (Japanese: regression connection type) neural network may also be used. In this case, since the values of the past input variables are reflected when calculating the value of the output variable newly this time, it is suitable for calculating the estimated pressure difference ΔPS while reflecting the past history.

[0113] The method for obtaining the training data and the teacher data used in the learning of the map Dx is not limited to the examples in the above-described embodiments. For example, vehicles having the same specifications as the vehicle 500 may actually be driven to obtain the training data and the teacher data.

[0114] As shown in the following formula (1), the lock-up pressure difference ΔP may also be a value obtained by subtracting the average value of the hydraulic pressure PF in the front chamber 26 and the hydraulic pressure PB in the rear chamber from the hydraulic pressure PC in the control chamber 25.

[0115] ΔP = PC - ((PF + PB) / 2) · · · (1)

[0116] The lock-up pressure difference ΔP only needs to represent the pressure difference between the control chamber 25 and the front chamber 26.

[0117] The overall structure of the vehicle 500 is not limited to the examples in the above-described embodiments. For example, as the drive source of the vehicle 500, an electric generator may be provided instead of the internal combustion engine 10, or an electric generator may be provided in addition to the internal combustion engine 10. As the automatic transmission, a continuously variable transmission may also be used.

[0118] The connection method of the torque converter 20 to the internal combustion engine 10 and the automatic transmission 80 is not limited to the examples of the above embodiments. For example, the input shaft of the automatic transmission 80 may also be the output shaft 22 of the torque converter 20 itself.

[0119] The number of the first friction plate 32 and the second friction plate 34 is not limited to the examples of the above embodiments. The first friction plate 32 and the second friction plate 34 may each be one or more.

[0120] The content of the lock-up map is not limited to the examples of the above embodiments. The lock-up map only needs to be the content that can appropriately switch the working state of the lock-up clutch 30.

Claims

1. A hydraulic pressure estimation device that uses a torque converter as an analog object. The torque converter has two oil chambers and a lock-up clutch that switches between an engaged state and a released state according to the pressure difference between the two oil chambers. The hydraulic pressure estimation device calculates an estimated pressure difference based on an indicated pressure difference, which is an indicated value of the pressure difference of the torque converter, and the estimated pressure difference is an estimated value of the pressure difference generated in the torque converter. It is characterized in that, include: a storage device configured to store mapping data, the mapping data being data defining a mapping learned by machine learning that outputs an estimated pressure difference variable as an output variable by inputting an input variable, the estimated pressure difference variable being a variable indicating the estimated pressure difference, wherein the mapping includes an indicated pressure difference variable as one of the plurality of input variables, the indicated pressure difference variable being a variable indicating the indicated pressure difference; and an execution device configured to execute an acquisition process, wherein the acquisition process acquires the value of the input variable, and a calculation process, wherein the value of the input variable acquired by the acquisition process is input into the map to calculate the value of the output variable. The mapping is obtained by performing machine learning using previously acquired training data related to the input variables and teacher data related to the actual pressure difference, so that the difference between the value of the mapping output when the training data is used as input, that is, the value of the output variable and the teacher data, becomes less than a predetermined value.

2. The hydraulic pressure estimation device according to claim 1, characterized in that: The storage device and the execution device are provided in a vehicle equipped with the torque converter. The storage device is configured to store a plurality of maps for each stage obtained by dividing a period from when the lockup clutch starts switching to the engaged state to when the lockup clutch starts switching to the released state into a plurality of stages. The execution device is configured to execute the acquisition process, the selection process, and the calculation process, The acquisition process acquires the value of the input variable when the indicated pressure difference is output to the torque converter. The selection process selects the map corresponding to the stage when the indicated pressure difference is output from the plurality of maps in each stage, Furthermore, the calculation process calculates the value of the output variable by inputting the value of the input variable acquired in the acquisition process to the map selected in the selection process.

3. The hydraulic pressure estimation device according to claim 2, characterized in that: As one of the plurality of input variables, an acceleration variable is included. The acceleration variable is a variable indicating the acceleration of the vehicle.

4. The hydraulic pressure estimation device according to claim 2 or 3, characterized in that: As one of the plurality of input variables, an accelerator variable is included. The accelerator variable is a variable indicating an operation amount of an accelerator pedal of the vehicle.

5. The hydraulic pressure estimation device according to claim 2 or 3, characterized in that: As one of the plurality of input variables, a vehicle speed variable is included. The vehicle speed variable is a variable indicating the running speed of the vehicle.

6. The hydraulic pressure estimation device according to claim 2 or 3, characterized in that: As one of the plurality of input variables, a speed change variable is included. The speed change variable is a variable indicating a speed ratio of a transmission of the vehicle.

7. The hydraulic pressure estimation device according to claim 2 or 3, characterized in that: As one of the plurality of input variables, an oil temperature variable is included. The oil temperature variable is a variable indicating the temperature of the hydraulic oil supplied to the torque converter.

8. A non-transitory storage medium stores commands that can be executed by one or more processors included in an execution device of a hydraulic estimation device, and cause the one or more processors to execute the following functions, the functions including: An acquisition process that acquires a value of an input variable; and A calculation process that inputs the value of the input variable acquired by the acquisition process into a map, thereby calculating a value of an output variable, wherein the hydraulic estimation device uses a torque converter as an analog object, and has a storage device and the execution device, the torque converter has two oil chambers and a lock-up clutch that switches between an engaged state and a released state according to a pressure difference between the two oil chambers, wherein the storage device stores map data, the map data is data of a map that defines outputting a estimated pressure difference variable as the output variable by inputting the input variable, and has been learned by machine learning, the estimated pressure difference variable is a variable representing an estimated pressure difference, the estimated pressure difference is an estimated value of the pressure difference generated in the torque converter, herein, the map includes indicating a pressure difference variable as one of the plurality of input variables, the indicating pressure difference variable is a variable representing an indicated pressure difference, the indicated pressure difference is an indicated value of the pressure difference of the torque converter, the map is obtained by performing the machine learning using training data related to the input variable acquired in advance and teacher data related to the actual pressure difference, such that a difference between the value of the output variable, which is the value output by the map with the training data as the input, and the teacher data becomes equal to or less than a predetermined value.

Images