JP2021060103A - Lock-up control device for automatic transmission - Google Patents

Abstract

To suppress a feeling of incompatibility that a driver has in drive lock-up fastening by correcting a drive learning value on which clutch deterioration is reflected in the drive lock-up fastening.SOLUTION: A lock-up controller 100 comprises a drive learning control unit 102 and a coast learning control unit 103. The drive learning control unit 102 acquires, in a drive state, a drive learning value based upon a meet point where clutch capacity begins to be generated in the middle of a rise in hydraulic pressure to a lock-up clutch 2a. The coast learning control unit 103 acquires, in a coast state, a coast learning value based upon a slip point at which a clutch slip begins to be generated in the middle of a decrease in hydraulic pressure to the lock-up clutch 2a. The drive learning control unit 102 acquires a drive learning corrected value by correcting the drive learning value based upon deterioration of the lock-up clutch 2a in the coast learning calculated with the coast learning value from the coast learning control unit 103.SELECTED DRAWING: Figure 2

Description

The present invention relates to a lockup control device for an automatic transmission mounted on a vehicle.

Conventionally, hydraulic learning is executed during lockup control when the speed ratio (ratio of turbine rotation speed to engine rotation speed) after the vehicle starts becomes a predetermined value or more. Execution of slip start control is prohibited until this hydraulic learning is executed a predetermined number of times. A vehicle control device that executes slip start control at the next vehicle start when the hydraulic learning is executed a predetermined number of times and the slip start control execution permission condition is satisfied is disclosed (see Patent Document 1).

Japanese Unexamined Patent Publication No. 2012-047254

In the device disclosed in Patent Document 1, in the piston stroke learning of the lockup clutch, in the transient state where the torque, rotation, and torque capacitance coefficient τ change, learning cannot be performed in the lockup scene immediately after starting. In some cases. If the piston stroke learning cannot be performed, there is a problem that the behavior changes when the lockup clutch is engaged, the engagement takes a long time, and the driver may feel uncomfortable.

The present invention has been made by paying attention to the above problems, and an object of the present invention is to suppress a sense of discomfort given to the driver when the lockup is engaged by correcting a drive learning value that reflects clutch deterioration when the drive lockup is engaged. To do.

In order to achieve the above object, the present invention includes a lockup clutch included in a torque converter arranged between a traveling drive source and a transmission mechanism, and a lockup controller for controlling engagement / release of the lockup clutch. ..
In the lockup control device of this automatic transmission, the lockup controller includes a drive learning control unit and a coast learning control unit.
The drive learning control unit is based on the meet point at which the clutch capacity starts to be generated while the oil pressure to the lockup clutch is rising when the driving force of the driving drive source is transmitted to the drive wheels. Get the drive learning value.
The coast learning control unit is at the slip point where clutch slip starts while the oil pressure to the lockup clutch is decreasing when the driving drive source is driven around by the driving force from the drive wheels. Obtain the coast learning value based on.
The drive learning control unit acquires the drive learning correction value corrected by the drive learning value based on the deterioration allowance of the lockup clutch calculated from the coast learning value from the coast learning control unit.

Since the above-mentioned solution is adopted, it is possible to suppress the discomfort given to the driver when the lockup is engaged by correcting the drive learning value which reflects the deterioration of the clutch when the drive lockup is engaged.

FIG. 5 is an overall system diagram showing an engine vehicle equipped with an automatic transmission to which the lockup control device of the first embodiment is applied. It is a detailed block diagram which shows the lock-up control system of an automatic transmission. It is explanatory drawing which shows the shelf pressure instruction value in the lock-up piston stroke which is the object which uses the detailed structure of the lock-up clutch and the learning result. It is a flowchart which shows the flow of the lockup control process executed by a lockup controller. It is a comparison characteristic diagram which shows the lockup torque estimated value characteristic and the actual lockup torque characteristic used in the lockup learning control process which obtains the initial pressure learning value in the background technique. It is a problem explanatory diagram which shows the initial error and the error after deterioration of the clutch torque capacity of the lockup clutch in the background technique. It is a time chart which shows the drive learning correction value by the deterioration correction using the deterioration allowance of the coast learning value at the time of drive learning convergence. It is a time chart which shows the drive learning correction value by the deterioration correction using the deterioration allowance of the coast learning value at the time of drive learning unconvergence. It is a shelf pressure instruction width characteristic figure which shows the OK width of the shelf pressure instruction value during a lockup piston stroke with respect to the change of the accelerator opening degree. It is an effect explanatory view which shows the initial error of the clutch torque capacity of the lockup clutch in Example 1, the error after F / T, and the error after deterioration.

Hereinafter, a mode for implementing the lockup control device for the automatic transmission of the present invention will be described with reference to the first embodiment shown in the drawings.

The lock-up control device of the first embodiment is applied to an engine vehicle (an example of a vehicle) equipped with an automatic transmission by shift-by-wire and park-by-wire having gear stages of 9 forward gears and 1 reverse gear. It is a thing. Hereinafter, the configuration of the first embodiment will be described separately as “overall system configuration”, “detailed configuration of lockup control system”, and “lockup engagement control processing configuration”.

[Overall system configuration (Fig. 1)]
As shown in FIG. 1, the drive system of the engine vehicle includes an engine 1, a torque converter 2, an automatic transmission 3, a propeller shaft 4, and drive wheels 5. A control valve unit 6 including a spool valve for shifting, a hydraulic control circuit, a solenoid valve, and the like is attached to the transmission case of the automatic transmission 3.

The torque converter 2 is a fluid coupling having a function of amplifying an input torque from the engine 1 and a function of absorbing fluctuations in transmitted torque by sliding. The torque converter 2 has a built-in lockup clutch 2a that directly connects the crankshaft of the engine 1 and the input shaft of the gear train 3a by engaging the clutch.

The automatic transmission 3 incorporates a gear train 3a and a park gear 3b. The gear train 3a has a plurality of planetary gear trains and friction elements, and achieves 9th forward gear and 1st reverse gear by combining fastening elements among the plurality of friction elements. The park gear 3b fixes the output shaft (= transmission output shaft) of the gear train 3a by gear meshing.

The control valve unit 6 has, as a solenoid valve, a clutch solenoid 20 provided for each friction element, a line pressure solenoid 21, a lubrication solenoid 22, and a lockup solenoid 23 provided in an oil passage from a hydraulic source. Each of these solenoid valves has a three-way linear solenoid structure, and operates in pressure adjustment in response to a control command from the transmission control unit 10.

As shown in FIG. 1, the electronic control system of the engine vehicle includes a transmission control unit 10 (abbreviation: "ATCU"), an engine control module 11 (abbreviation: "ECM"), and a CAN communication line. 70 and. Here, the transmission control unit 10 is started / stopped by an ignition signal (IGN signal) from the sensor module unit 71 (abbreviation: “USM”).

The transmission control unit 10 is provided integrally with mechatronics on the upper surface position of the control valve unit 6, and the main board temperature sensor 31 and the sub board temperature sensor 32 are provided on the unit board by a redundant system while ensuring independence from each other. .. That is, the main board temperature sensor 31 and the sub board temperature sensor 32 transmit the sensor value information to the transmission control unit 10, but unlike the well-known automatic transmission unit, the transmission hydraulic oil (ATF) is installed in the oil pan. Sends temperature information that is not in direct contact with. The transmission control unit 10 also inputs signals from the lockup oil pressure sensor 12, the turbine rotation sensor 13, the output shaft rotation sensor 14, and the third clutch oil pressure sensor 15. Further, signals from the shifter control unit 18, the intermediate shaft rotation sensor 19, and the like are input.

The lockup oil pressure sensor 12 detects the lockup oil pressure of the lockup clutch 2a and transmits a signal indicating the lockup oil pressure PLU to the transmission control unit 10. The turbine rotation sensor 13 detects the turbine rotation speed (= transmission input shaft rotation speed) of the torque converter 2 and transmits a signal indicating the turbine rotation speed Nt to the transmission control unit 10. The output shaft rotation sensor 14 detects the output shaft rotation speed of the automatic transmission 3 and transmits a signal indicating the output shaft rotation speed No. (= vehicle speed VSP) to the transmission control unit 10. The third clutch oil pressure sensor 15 detects the clutch oil pressure of the third clutch K3 and transmits a signal indicating the third clutch oil pressure PK3 to the transmission control unit 10.

The shifter control unit 18 determines the range position selected by the driver's select operation on the shifter 181 and transmits the range position signal to the transmission control unit 10. The shifter 181 has a momentary structure, has a P range button 181b on the upper portion of the operation unit 181a, and has an unlock button 181c (only when N → R) on the side portion of the operation unit 181a. The range positions include an H range (home range), an R range (reverse range), a D range (drive range), and N (d) and N (r) (neutral range). The intermediate shaft rotation sensor 19 detects the rotation speed of the intermediate shaft (intermediate shaft = rotating member connected to the first carrier C1), and transmits a signal indicating the intermediate shaft rotation speed Nint to the transmission control unit 10.

The transmission control unit 10 performs lockup control by releasing / engaging / engaging the lockup clutch 2a. Furthermore, by monitoring changes in the driving point (VSP, APO) due to the vehicle speed VSP and the accelerator opening APO on the shift map (not shown),
1. Auto upshift (due to vehicle speed increase while maintaining accelerator opening)
2. Foot release upshift (by accelerator foot release operation)
3. Foot return upshift (by accelerator return operation)
4. Power on / downshift (due to a decrease in vehicle speed while maintaining the accelerator opening)
5. Small opening sudden downshift (depending on the small amount of accelerator operation)
6. Large opening sudden downshift (depending on the amount of accelerator operation: "kickdown")
7. Slow downshift (due to slow accelerator operation and increased vehicle speed)
8. Coast downshift (due to a decrease in vehicle speed when the accelerator is released)
Shift control is performed according to a basic shift pattern called.

The engine control module 11 inputs signals from the accelerator opening sensor 16, the engine rotation sensor 17, and the like.

The accelerator opening sensor 16 detects the accelerator opening caused by the driver's accelerator operation, and transmits a signal indicating the accelerator opening APO to the engine control module 11. The engine rotation sensor 17 detects the rotation speed of the engine 1 and transmits a signal indicating the engine rotation speed Ne to the engine control module 11.

In the engine control module 11, in addition to various controls of the engine itself, engine torque limit control and the like are performed by cooperative control with the transmission control unit 10. Since the transmission control unit 10 is connected to the transmission control unit 10 via a CAN communication line 70 capable of exchanging information in both directions, when an information request is input from the transmission control unit 10, the accelerator opening APO and the engine speed are reached. Ne information is output to the transmission control unit 10. Further, the information of the engine torque Te and the turbine torque Tt calculated by estimation is output to the transmission control unit 10. Further, when an engine torque limit request based on the upper limit torque is input from the transmission control unit 10, engine torque limit control is executed in which the engine torque is limited by a predetermined upper limit torque.

[Detailed configuration of lockup control system (Figs. 2 and 3)]
As shown in FIG. 2, the lockup control system of the automatic transmission 3 includes a lockup clutch 2a, a lockup controller 100, and a lockup solenoid 23. In the following description, "LU" is an abbreviation for "lockup".

The lockup clutch 2a is a multi-plate friction clutch provided in the torque converter 2 arranged between the engine 1 (driving drive source) and the gear train 3a (transmission mechanism). As shown in FIG. 3, the lockup clutch 2a has a lockup piston 2b, a lockup oil chamber 2c, and a return spring 2d, and strokes by supplying lockup hydraulic pressure to the lockup oil chamber 2c. It is fastened by the lockup piston 2b. The lockup oil pressure is adjusted by the lockup solenoid 23 according to a lockup control instruction (fastening instruction / release instruction) from the lockup controller 100.

The lockup controller 100 includes a turbine rotation speed Nt from the turbine rotation sensor 13, a vehicle speed VSP from the output shaft rotation sensor 14, an accelerator opening APO, an engine torque Te, an engine rotation speed Ne, and a lockup hydraulic pressure from the engine control module 11. Information such as the lockup hydraulic PLU from the sensor 12 is input.

As shown in FIG. 2, the lockup controller 100 includes an initial drive learning execution unit 101, a drive learning control unit 102, a coast learning control unit 103, a learning value storage unit 104, and a lockup control unit 105. Have.

Here, "drive learning" means that when the driving force of the engine 1 is transmitted to the drive wheels 5, the clutch capacity is started to be generated while the oil pressure to the lockup clutch 2a is rising. Control to acquire drive learning value based on points. Drive learning implementation conditions include clutch release condition, engine torque stabilization condition, throttle opening condition, throttle opening stabilization condition, oil temperature condition, oil amount balance judgment condition, etc., and when all of these conditions are satisfied Drive learning is carried out. The "drive learning value" refers to the LU indicated value when the meet point is detected or the LU piston stroke value. The "meet point" is defined as the point at which the lockup hydraulic PLU starts to rise sharply when the change characteristic of the lockup hydraulic PLU from the lockup hydraulic sensor 12 is monitored.

"Coast learning" is a slip point where clutch slip starts to occur while the oil pressure to the lockup clutch 2a is decreasing when the engine 1 is driven by the driving force from the drive wheels 5. It refers to the control to acquire the coast learning value based on. The coast learning implementation conditions include clutch engagement conditions, accelerator opening zero conditions, fuel cut conditions, oil temperature conditions, oil amount balance judgment conditions, etc., and coast learning is performed when all of these conditions are satisfied. .. The "coast learning value" refers to the LU indicated value when the slip point is detected or the LU piston stroke value. The "slip point" monitors the engine speed Ne from the engine speed sensor 17 and the turbine speed Nt from the turbine speed sensor 13, and shifts from Ne = Nt (lockup state) to Ne <Nt (slip state). Let it be a point.

The initial drive learning implementation unit 101 depends on the number of learning experiences that can be set as the drive learning implementation condition and acquire the drive learning convergence value at the time of the final test (F / T) conducted before shipping to the market. Carry out drive learning. The initial drive learning value acquired by the initial drive learning execution unit 101 is output to the drive learning control unit 102 and the coast learning control unit 103.

The drive learning control unit 102 acquires the drive learning correction value corrected by the drive learning value based on the deterioration allowance of the lockup clutch 2a in the coast learning calculated from the coast learning value from the coast learning control unit 103. Here, the deterioration allowance of the lockup clutch 2a in drive learning is acquired by inputting a deterioration correction value from the coast learning control unit 103 upon request and converting the input deterioration correction value into learning value for drive learning. Deterioration correction conversion value.

When the drive learning control unit 102 determines that the drive learning value has converged, the drive learning correction value is acquired by correcting the drive learning convergence value after the convergence determination with the deterioration correction conversion value. On the other hand, when it is determined that the drive learning value has not converged, the drive learning correction value is acquired by converting the coast learning convergent value after the convergence test from the coast learning control unit 103 into the drive learning. ..

The coast learning control unit 103 sets the deterioration allowance of the lockup clutch 2a in coast learning as a deterioration correction value obtained by subtracting the coast learning convergence value before deterioration from the coast learning average value after deterioration. Then, the information of the deterioration correction value is output to the drive learning control unit 102. The "coast learning average value after deterioration" is calculated by adding the value obtained by multiplying the difference of the latest learning values by the weight (gain) to the previous coast learning average value, starting from the value at the time of convergence. To do.

When the initial drive learning value is input from the initial drive learning execution unit 101, the coast learning control unit 103 acquires the initial coast learning value by converting the initial drive learning value into the learning value for coast learning. Then, during the implementation of the initial drive learning, the coast learning convergence value before deterioration is acquired based on the initial coast learning value.

The learning value storage unit 104 stores the drive learning correction value calculated by the drive learning control unit 102 and the coast learning value calculated by the coast learning control unit 103 while updating the latest learning value information. The drive learning value and the coast learning value stored in the learning value storage unit 104 are read out in response to a request from the lockup control unit 105.

The lockup control unit 105 starts outputting a lockup engagement instruction to the lockup solenoid 23 when the vehicle speed VSP reaches the lockup start vehicle speed set in the low vehicle speed range in the lockup released state when the vehicle starts in the first speed in the D range. To do. Then, lock-up convergence control is performed to converge the actual slip amount toward zero by the target slip rotation change rate, in which the lowering gradient becomes larger as the clutch difference rotation (the difference between the engine speed Ne and the turbine speed Nt) becomes larger. In this lockup convergence control, as shown in FIG. 3, the drive learning value is used as information for determining the magnitude of the shelf pressure indicated value (for example, drive learning value-offset value) during the lockup piston stroke following the initial pressure indicated value. Is used.

After the actual slip amount converges to zero by the lockup convergence control, the slip amount of the lockup clutch 2a is set to zero (fastened state) instead of the completely engaged state in which slip is not allowed against the transmission torque fluctuation. Continue to keep zero slip control. Here, "zero slip control" is a control for transmitting torque up to the engine torque Te in the clutch engaged state by matching the engagement torque capacity of the lockup clutch 2a with the engine torque Te which is the input torque. Is.

That is, when the vehicle speed VSP reaches the lockup start vehicle speed in the low vehicle speed range when starting from the 1st speed in the D range, the clutch engagement is started, and when the clutch is engaged, zero slip control is performed regardless of the upshift or downshift in the gear train 3a. continue. As described above, the basic control for engaging the lockup is a control that prioritizes the fuel efficiency performance of the engine 1. Then, for shifting in the gear train 3a where the input torque to the lockup clutch 2a fluctuates, the shifting behavior is changed not by releasing the lockup but by slipping (clutch slip) in the zero slip control. I try to suppress it.

The lockup control unit 105 instructs the lockup solenoid 23 to release the lockup when the vehicle speed VSP reaches the lockup release vehicle speed set in the low vehicle speed range while the lockup is engaged during coast deceleration by releasing the accelerator foot in the D range. Starts outputting. In this lockup release control, the coast learning value is used as information for determining the magnitude of the lockup release pressure instruction value (for example, coast learning value + offset value).

The purpose of learning and controlling the coast capacity is to reduce the lockup oil pressure as much as possible during coast deceleration (fuel cut state) in order to improve the lockup release responsiveness. That is, the lock-up oil pressure at the time of coast has a high initial learning value (unlearned) at the time of shipment from the factory, but the learning is performed in the normal riding by the user, and the coast has a capacity that is balanced with the engine torque (negative torque) at the time of coast. When the lockup is lowered to hydraulic pressure.

In this way, the performance effect obtained by experiencing the update of the learning value from the initial learning value to the learning value by the coast learning control and lowering the lock-up oil pressure during coast to a low lock-up oil pressure is the prevention of engine stall (during sudden deceleration). ), Improvement of LU release behavior change (during slow deceleration), improvement of chip-in behavior change (during re-acceleration from the coast), etc.

[Lockup fastening control processing configuration (Fig. 4)]
The lockup engagement control process shown in FIG. 4 has already been subjected to initial drive learning by the final test, and is started by turning on the ignition switch in the market. In addition, in order to compensate for variations and deterioration in the market, drive learning is carried out to acquire a new drive learning correction value every time the drive learning implementation condition is satisfied, and a new coast learning value is executed every time the coast learning implementation condition is satisfied. Coastal learning is carried out.

In step S1, following the start, it is determined whether or not the drive learning value has converged in the initial drive learning by the final test. If YES (the drive learning value has converged), the process proceeds to step S2, and if NO (the drive learning value has not converged), the process proceeds to step S6.

In step S2, following the determination that the drive learning value in S1 has converged, the coast learning control unit 103 inputs the deterioration correction value of the coast learning value, and the process proceeds to step S3. Here, the "deterioration correction value" is a value obtained by subtracting the coast learning convergence value before deterioration from the coast learning average value after deterioration in the coast learning control unit 103.

In step S3, following the input of the deterioration correction value in S2, the deterioration correction conversion value is calculated by the learning value conversion process of converting the deterioration correction value by coast learning into the deterioration correction value by drive learning, and the process proceeds to step S4. Here, in the learning value conversion process, for example, the conversion characteristics of the coast learning value and the drive learning value determined in advance by an experiment or the like are used.

In step S4, following the calculation of the deterioration correction conversion value in S3, the drive learning correction value at the time of determining the convergence of the drive learning value is calculated, and the process proceeds to step S5. Here, as the drive learning correction value at the time of determining the drive learning value convergence, the formula of drive learning value = (drive learning convergence value-deterioration correction conversion value) is used.

In step S5, following the calculation of the drive learning correction value in S4, the calculated drive learning correction value is stored and updated as the drive learning value, and the process proceeds to step S8.

In step S6, following the determination that the drive learning value in S1 has not converged, the coast learning control unit 103 inputs the coast learning convergent value, and the process proceeds to step S7. Here, the "coast learning convergence value" is a value calculated by the initial drive learning by the final test in the coast learning control unit 103, and is set as the coast learning convergence value before deterioration.

In step S7, following the input of the coast learning convergence value in S6, the drive learning correction value at the time of determining that the drive learning value has not converged is calculated, and the process proceeds to step S8. Here, the drive learning correction value at the time of determining that the drive learning value has not converged is calculated by a learning value conversion process that converts the coast learning convergent value into a drive learning convergent value, and in the learning value conversion process, for example, it is determined in advance by an experiment or the like. The conversion characteristics of the obtained coast learning value and drive learning value are used.

In step S8, following the calculation of the drive learning correction value in S7, the calculated drive learning correction value is stored and updated as the drive learning value, and the process proceeds to step S8.

In step S9, following the storage of the drive learning value in S5 or S8, it is determined whether or not the vehicle speed VSP is equal to or higher than the lockup start vehicle speed in the drive state in which the lockup clutch 2a is released. If YES (vehicle speed VSP ≥ lockup start vehicle speed), the process proceeds to step S10, and if NO (vehicle speed VSP <lockup start vehicle speed), the process proceeds to step S11.

In step S10, following the determination that VSP ≥ the lockup start vehicle speed in S9, the learning value storage unit 104 is stored in the learning value storage unit 104 for determining the lockup fastening instruction value (shelf pressure instruction value during the LU piston stroke). The drive learning value is used, and the process proceeds to step S13.

In step S11, following the determination that VSP <lockup start vehicle speed in S9, it is determined whether or not the vehicle speed VSP is equal to or less than the lockup release vehicle speed in the coast state in which the lockup clutch 2a is engaged. .. If YES (vehicle speed VSP ≤ lockup release vehicle speed), the process proceeds to step S12, and if NO (vehicle speed VSP> lockup release vehicle speed), the process proceeds to step S13.

In step S12, following the determination that the vehicle speed VSP ≤ the lockup release vehicle speed in S11, the coast learning value stored in the learning value storage unit 104 is used to determine the lockup release instruction value, and the process proceeds to step S13. move on.

In step S13, following the output of the LU conclusion instruction value in S10, the determination of NO in S11, or the output of the LU release instruction value in S12, it is determined whether or not the ignition is off. If YES (ignition off), the process proceeds to the end, and if NO (ignition on), the process returns to step S1. Even if the ignition is turned off, the latest updated values of the drive learning value and the coast learning value stored and saved in the learning value storage unit 104 are saved as they are.

Next, "problems of background technology and problem-solving measures" will be described. Then, the operation of the first embodiment will be described separately by dividing it into an "initial drive learning execution action" and a "lock-up learning control action".

[Problems of background technology and problem-solving measures (Figs. 5 and 6)]
The lockup learning control in the background technology is an estimated lockup transmission torque based on the difference between the ^{engine torque Te and the torque converter transmission torque τNe 2} when the lockup clutch shifts from the non-engaged state to the engaged state during driving. Calculate Tlu ^# . Then, it is known that the initial pressure learning value is the tightening start oil pressure when it is determined that the lockup transmission torque estimated value Tlu ^{# has entered an upward trend.}

However, the actual lockup transmission torque Tlu is
Tlu ＝ Te−τNe ^ 2-iω'
Therefore, in the original physical model, the rotational inertia component iω'is also taken into consideration, but the lockup transmission torque estimate Tlu ^# is not sufficient for consideration. Therefore, as shown in FIG. 5, the lockup transmission torque estimated value Tlu ^# (dotted line) becomes larger than the actual lockup transmission torque Tlu (solid line), and an error occurs. The cause of the error between the actual lockup transmission torque Tlu and the estimated lockup transmission torque Tlu ^# is that the response delay of τ (variation in the time direction) and the response delay of Ne (variation in the time direction) contribute greatly. It depends.

Therefore, in the case of an automatic transmission that starts lockup engagement in the 1st speed of the D range, the rotation change and the accelerator opening degree change at the time of lockup engagement are large. Therefore, the initial pressure when the lockup learning control of the background art to obtain the learned value, the actual lock-up transmission torque TLU lockup transmission torque estimate TLU ^# error of directly using the lock-up transmission torque estimate TLU ^# It will be reflected.

Next, in performing the learning correction, the initial variation of the lockup clutch, the variation after the final test, and the variation after deterioration were examined. As a result of the variation examination, it was found that the variation width of the initial clutch torque capacity is directly reflected in the variation width after deterioration as shown in FIG. 6 unless the initial variation is corrected in the final test. In other words, it was found that the cause of the clutch torque capacity variation exceeding the performance establishment range of the LU piston stroke is the deterioration of the lockup clutch.

If the clutch torque capacity varies to a side higher than the performance establishment width of the LU piston stroke, the behavior changes in the vehicle front-rear direction due to the lock-up behavior change when the lock-up is fastened. On the other hand, if the clutch torque capacity varies to the side lower than the performance establishment width of the LU piston stroke, it takes time to complete the engagement when the lockup is engaged, giving the driver a feeling of lag. Therefore, since drive learning is an important performance related to the change in behavior when the lockup is engaged, it is necessary to correct the clutch torque capacity (= lockup differential pressure) according to the deterioration of the lockup clutch.

On the other hand, the present inventors carry out the initial drive learning in the final test, but after that, in order to guarantee the variation in the market, it is necessary to take a measure capable of correcting the deterioration of the lockup clutch. Considered a specific measure. And by examining the measures
(A) The deterioration allowance of the lockup clutch is included in both the drive learning value and the coast learning value.
(B) The dependence of μ deterioration differs between the drive scene and the coast scene, but it can be adjusted by grasping the sensitivity to each deterioration.
(C) Coast learning has a higher learning frequency than drive learning, and coast learning values tend to converge compared to drive learning values.
I focused on that point.

Based on the above points of interest, the present disclosure includes a lockup clutch 2a included in a torque converter 2 arranged between an engine 1 and a gear train 3a, a lockup controller 100 that controls engagement / release of the lockup clutch 2a, and a lockup controller 100. To be equipped. In the lockup control device of the automatic transmission 3, the lockup controller 100 includes a drive learning control unit 102 and a coast learning control unit 103. The drive learning control unit 102 is based on a meet point that starts generating clutch capacity while the oil pressure to the lockup clutch 2a is rising when the driving force of the engine 1 is transmitted to the drive wheels 5. To get the drive learning value. The coast learning control unit 103 starts generating clutch slip while the oil pressure to the lockup clutch 2a is decreasing when the engine 1 is driven by the driving force from the drive wheels 5 in the coast state. Obtain the coast learning value based on. The drive learning control unit 102 adopts a means of acquiring a drive learning correction value corrected by the drive learning value based on the deterioration allowance of the lockup clutch 2a calculated from the coast learning value from the coast learning control unit 103.

That is, only the deterioration allowance of the lockup clutch 2a is separated from the coast learning values, and the drive learning correction value is acquired by reflecting the deterioration allowance obtained by the coast learning in the drive learning value. At this time, since the coast learning for acquiring the deterioration allowance reflected in the drive learning has a higher learning frequency than the drive learning frequency in which the learning implementation conditions are strict, it is possible to deal with the deterioration correction in the market without any problem.

Therefore, the variation in the clutch torque capacity due to the deterioration allowance of the lockup clutch 2a can be suppressed, and the clutch torque capacity can be kept within the range of the LU piston stroke performance establishment range when the lockup is engaged. As a result, by correcting the drive learning value that reflects the deterioration of the clutch when the drive lockup is engaged, it is possible to suppress a change in behavior and a sense of discomfort due to lag when the lockup is engaged.

[Initial drive learning implementation action (Fig. 2)]
The initial drive learning implementation unit 101 depends on the number of learning experiences that can acquire the drive learning convergence value by setting the test condition to the drive learning implementation condition at the time of the final test conducted before shipping to the market. Drive learning is carried out. Then, the initial drive learning value acquired by the initial drive learning execution unit 101 is output to the drive learning control unit 102 and the coast learning control unit 103.

The drive learning control unit 102 for inputting the initial drive learning value determines whether or not the drive learning value has converged, and when the drive learning value has been determined to converge, the drive learning convergence value is being executed during the initial drive learning. Is obtained. As a general rule, the drive learning convergence value is acquired during the initial drive learning, but if the drive learning value convergence test is not determined due to the variation in the learning value, after adding the experience of drive learning in the market. , The drive learning convergence value will be acquired.

The coast learning control unit 103 for inputting the initial drive learning value acquires the initial coast learning value by converting the input initial drive learning value into the learning value for coast learning. Then, during the implementation of the initial drive learning, the coast learning convergence value before deterioration is acquired based on the initial coast learning value.

As described above, the lockup controller 100 has an initial drive learning execution unit 101 that performs initial drive learning based on the number of learning experiences that can acquire a drive learning convergence value before shipping to the market.

For example, if the initial drive learning is not performed, the drive learning convergence value cannot be obtained unless the drive learning in the market is experienced many times, and it takes time to obtain the drive learning convergence value. The learning value cannot be corrected. On the other hand, by performing the initial drive learning based on the number of learning experiences that can acquire the drive learning convergence value, the drive learning value using the drive learning convergence value can be obtained at an early timing immediately after shipping to the market. The correction can be carried out.

When the initial drive learning value is input from the initial drive learning execution unit 101, the coast learning control unit 103 acquires the initial coast learning value by converting the initial drive learning value into the learning value for coast learning. Then, during the implementation of the initial drive learning, the coast learning convergence value before deterioration is acquired based on the initial coast learning value.

For example, in the final test conducted before shipping to the market, the initial drive learning based on the number of learning experiences that can acquire the drive learning convergence value and the number of learning experiences that can acquire the coast learning convergence value. It is conceivable to carry out initial coastal learning by. However, in this case, the learning load in the final test is increased. On the other hand, drive learning and coast learning utilize the fact that they have an adjustable correlation with each other, and by acquiring the initial coast learning value by the learning value conversion process, the coast learning converges without performing the initial coast learning. You will be able to get the value.

[Lock-up learning control action (FIGS. 4, 7 to 10)]
When the drive learning value has converged, the process proceeds in the order of S1 → S2 → S3 → S4 → S5 in the flowchart of FIG. In S2, the deterioration correction value of the coast learning value is input from the coast learning control unit 103. In S3, the deterioration correction conversion value is calculated by the learning value conversion process that converts the deterioration correction value by coast learning into the deterioration correction value by drive learning. In S4, the drive learning correction value at the time of determining the drive learning value convergence is calculated using the formula of drive learning value = (drive learning convergence value-deterioration correction conversion value). In S5, the calculated drive learning correction value is stored and updated as the drive learning value.

When the drive learning value has not converged, the process proceeds from S1 → S6 → S7 → S8 in the flowchart of FIG. In S6, the coast learning convergence value is input from the coast learning control unit 103. In S7, the drive learning correction value at the time of determining that the drive learning value has not converged is calculated by the learning value conversion process that converts the coast learning convergent value into the drive learning convergent value. In S8, the calculated drive learning correction value is stored and updated as the drive learning value.

Then, in S9, when it is determined that the vehicle speed VSP is equal to or higher than the lockup start vehicle speed in the driving state in which the lockup clutch 2a is released, the process proceeds to S10. In S10, the drive learning value stored in the learning value storage unit 104 is used when determining the lockup fastening instruction value (shelf pressure instruction value during the LU piston stroke).

Although the lockup engagement condition is not satisfied in S9, if it is determined in S11 that the vehicle speed VSP is equal to or less than the lockup release vehicle speed in the coastal state in which the lockup clutch 2a is engaged, the process proceeds to step S12. In S12, the coast learning value stored in the learning value storage unit 104 is used to determine the lockup release instruction value. While it is determined in S13 that the ignition is on, the processes S1 to S12 are repeated.

As described above, the coast learning control unit 103 sets the deterioration allowance of the lockup clutch 2a in coast learning as a deterioration correction value obtained by subtracting the coast learning convergence value before deterioration from the coast learning average value after deterioration.

That is, the deterioration allowance (CST) of the lockup clutch 2a in coast learning is a deterioration correction value obtained by subtracting the convergence value from the average value, as shown in FIG. That is, the coast learning convergence value before deterioration is used as the reference value, and the deviation width due to the learning value gradually deviating from the reference value as the deterioration progresses is used as the deterioration correction value. Therefore, the deterioration allowance of the lockup clutch 2a in coast learning can be accurately acquired with reference to the coast learning convergence value before deterioration.

The drive learning control unit 102 uses the deterioration allowance of the lockup clutch 2a in drive learning as the deterioration correction conversion value acquired by converting the deterioration correction value from the coast learning control unit 103 into the learning value for drive learning. There is.

That is, the drive learning value is used at the time of LU stroke and the coast learning value is used at the time of coast fuel cut, and since the scenes are different, the dependence on the friction coefficient μ, torque, etc. is different, and the correction value after deterioration in each scene is different. The absolute amount is also different. That is, as shown in FIG. 7, the deterioration allowance (Drive) of the lockup clutch 2a in the drive learning is smaller than the deterioration allowance (CST) of the lockup clutch 2a in the coast learning. Therefore, it is considered that the deterioration correction allowance has a certain degree of sensitivity, and it is defined as a conversion function or a correction coefficient so that it can be corrected. Therefore, the deterioration allowance of the lockup clutch 2a in the drive learning can be accurately acquired based on the deterioration correction value in the coast learning.

When the drive learning control unit 102 determines that the drive learning value has converged, the drive learning control unit 102 acquires the drive learning correction value by correcting the drive learning convergence value after the convergence test with the deterioration correction conversion value (FIG. 4). S1 → S2 → S3 → S4).

That is, when the drive learning has converged, as shown in FIG. 7, the drive learning correction value can be obtained by correcting the drive learning convergence value after the convergence determination with the deterioration correction conversion value. However, when the deterioration correction value in coast learning is reflected in the drive learning in the state where the drive learning is unconverged, as shown in FIG. 8, the correction is such that the unconverged state is maintained, and the correction to the convergent side is performed. Not done. Therefore, the deterioration correction value in the coast learning is reflected in the drive learning on the premise that both the drive learning value and the coast learning value are converged.

Therefore, the variation in the clutch torque capacity due to the deterioration allowance of the lockup clutch 2a is suppressed, and as shown in FIG. 9, the clutch torque capacity should be within the range of the LU piston stroke performance establishment width (OK width) when the lockup is engaged. Can be done. The LU piston stroke performance establishment width is when the shelf pressure indicated value during the stroke of the lockup piston in FIG. 9 is below the upper limit (dotted line) and above the lower limit (solid line), and when it is higher than the upper limit (dotted line). If the change in vehicle behavior is in the NG region and is lower than the lower limit (solid line), the time until the start of the inertia phase is long (lag) is in the NG region. Therefore, assuming that both the drive learning value and the coast learning value have converged, the deterioration correction value in the coast learning is reflected in the drive learning, and the correction is given to the driver when the lockup is concluded. It is possible to suppress discomfort due to behavior changes and lag.

For example, in the first embodiment, the drive learning is performed in the final test for the initial variation, and after the final test, the deterioration correction value in the coast learning is reflected in the drive learning. Therefore, as shown in FIG. 10, it was found that the variation width of the initial clutch torque capacity is within the OK width range both after the final test and after deterioration.

When the drive learning control unit 102 determines that the drive learning value has not converged, the drive learning control unit 102 processes the coast learning convergence value after the convergence determination from the coast learning control unit 103 into a learning value conversion process to drive learning. The learning correction value is acquired (S1 → S6 → S7 in FIG. 4).

For example, after the final test is carried out, the number of experience of drive learning may be insufficient depending on the product variation, etc., and the drive learning value may not reach convergence. At this time, if the premise operating condition is not satisfied and the drive learning value is not corrected at all, the performance of the LU piston stroke is only deteriorated from the initial state when the lockup is fastened. On the other hand, as a countermeasure when the drive learning value does not converge, the value obtained by correcting the coast learning convergence value is reflected in the drive learning value from the relationship between the coast learning value and the drive learning value. That is, it is not the deterioration correction but the reflection of the coast learning convergence value on the drive learning side. Therefore, when the drive learning value has not converged, the coast learning convergent value is reflected on the drive learning side to prevent the LU piston stroke performance from deteriorating due to deterioration from the initial state when the lockup is concluded. Will be possible.

As described above, in the lockup control device of the automatic transmission 3 of the first embodiment, the effects listed below can be obtained.

(1) A torque converter 2 arranged between a driving drive source (engine 1) and a transmission mechanism (gear train 3a), a lockup clutch 2a, and a lockup controller 100 that controls engagement / release of the lockup clutch 2a. In the lockup control device of the automatic transmission 3 provided with,
The lockup controller 100
Based on the meet point where the clutch capacity starts to be generated while the oil pressure to the lockup clutch 2a is rising when the driving force of the driving drive source (engine 1) is transmitted to the drive wheels 5. The drive learning control unit 102 that acquires the drive learning value,
At the slip point where clutch slip starts while the oil pressure to the lockup clutch 2a is decreasing when the driving drive source (engine 1) is driven around by the driving force from the drive wheels 5. The coast learning control unit 103, which acquires the coast learning value based on the coast learning control unit 103, is provided.
The drive learning control unit 102 acquires the drive learning correction value corrected by the drive learning value based on the deterioration allowance of the lockup clutch 2a in the coast learning calculated from the coast learning value from the coast learning control unit 103.
Therefore, by correcting the drive learning value that reflects the deterioration of the clutch when the drive lockup is engaged, it is possible to suppress the discomfort given to the driver when the lockup is engaged.

(2) The coast learning control unit 103 sets the deterioration allowance of the lockup clutch 2a in coast learning as a deterioration correction value obtained by subtracting the coast learning convergence value before deterioration from the coast learning average value after deterioration.
Therefore, the deterioration allowance of the lockup clutch 2a in coast learning can be accurately obtained with reference to the coast learning convergence value before deterioration.

(3) The drive learning control unit 102 performs deterioration correction obtained by converting the deterioration allowance of the lockup clutch 2a in drive learning and the deterioration correction value from the coast learning control unit 103 into the learning value for drive learning. Use as a conversion value.
Therefore, the deterioration allowance of the lockup clutch 2a in the drive learning can be accurately acquired based on the deterioration correction value in the coast learning.

(4) When the drive learning control unit 102 determines that the drive learning value has converged, the drive learning control unit 102 acquires the drive learning correction value by correcting the drive learning convergence value after the convergence determination with the deterioration correction conversion value.
Therefore, assuming that both the drive learning value and the coast learning value have converged, the deterioration correction value in the coast learning is reflected in the drive learning, and the correction is given to the driver when the lockup is concluded. It is possible to suppress discomfort due to behavior changes and lag.

(5) When the drive learning control unit 102 determines that the drive learning value has not converged, the drive learning control unit 102 processes the coast learning convergence value after the convergence test from the coast learning control unit 103 to the learning value conversion process to drive learning. To get the drive learning correction value.
Therefore, when the drive learning value has not converged, the coast learning convergent value is reflected on the drive learning side to prevent the LU piston stroke performance from deteriorating due to deterioration from the initial state when the lockup is concluded. Can be done.

(6) The lockup controller 100 has an initial drive learning execution unit 101 that performs initial drive learning based on the number of learning experiences that can acquire a drive learning convergence value before shipping to the market.
Therefore, it is possible to correct the drive learning value using the drive learning convergence value at an early timing immediately after the shipment to the market.

(7) When the coast learning control unit 103 inputs the initial drive learning value from the initial drive learning execution unit 101, the coast learning control unit 103 acquires the initial coast learning value by converting the initial drive learning value into the learning value for coast learning.
During the initial drive learning, the coast learning convergence value before deterioration is acquired based on the initial coast learning value.
Therefore, by utilizing the fact that drive learning and coast learning have an adjustable correlation with each other and acquiring the initial coast learning value by the learning value conversion process, the coast learning convergence value without performing the initial coast learning. Can be obtained.

The lockup control device for the automatic transmission of the present invention has been described above based on the first embodiment. However, the specific configuration is not limited to the first embodiment, and design changes and additions are permitted as long as the gist of the invention according to each claim is not deviated from the claims.

In the first embodiment, an example is shown in which the drive learning control unit 102 acquires the drive learning correction value by correcting the drive learning convergence value after the convergence determination with the deterioration correction conversion value. However, the drive learning control unit does not use, for example, the drive learning convergence value or the deterioration correction conversion value, but grasps the deterioration amount of the lockup clutch in the coast learning by the ratio to the coast learning value, and obtains this as the drive learning value. It may be an example to be reflected in. In short, it suffices to acquire the drive learning correction value obtained by correcting the drive learning value based on the deterioration allowance of the lockup clutch in the coast learning calculated from the coast learning value.

In the first embodiment, as an automatic transmission, an example of an automatic transmission 3 having 9 forward speeds and 1 reverse speed is shown. However, the automatic transmission may be an example of an automatic transmission having a stepped speed change other than the forward 9-speed reverse 1-speed, or has an auxiliary transmission that combines a belt-type continuously variable transmission and a multi-speed transmission. It may be a continuously variable transmission or a belt type continuously variable transmission.

In the first embodiment, an example of the lockup control device of the automatic transmission 3 mounted on the engine vehicle is shown. However, it can be applied not only to an engine vehicle but also as a lockup control device for an automatic transmission of a hybrid vehicle, an electric vehicle, or the like.

1 Engine (driving drive source)
2 Torque converter 2a Lock-up clutch 2b Lock-up piston 2c Lock-up oil chamber 2d Return spring 3 Automatic transmission 3a Gear train 4 Propeller shaft 5 Drive wheel 10 Transmission control unit 100 Lock-up controller 101 Initial drive learning execution unit 102 Drive learning control Unit 103 Coast learning control unit 104 Learning value storage unit 105 Lockup control unit 11 Engine control unit 12 Lockup oil pressure sensor 13 Turbine rotation sensor 14 Output shaft rotation sensor 16 Accelerator opening sensor 17 Engine rotation sensor 23 Lockup solenoid

Claims (7)

In a lockup control device for an automatic transmission, which comprises a lockup clutch in a torque converter arranged between a driving drive source for traveling and a transmission mechanism, and a lockup controller for controlling engagement / release of the lockup clutch.
The lockup controller is
When the driving force of the traveling drive source is transmitted to the drive wheels, the drive learning value is calculated based on the meet point at which the clutch capacity starts to be generated while the oil pressure to the lockup clutch is rising. Drive learning control unit to acquire and
When the driving drive source is driven around by the driving force from the driving wheels, the coast is based on the slip point at which the clutch slip starts to occur while the oil pressure to the lockup clutch is decreasing. Equipped with a coast learning control unit that acquires learning values,
The drive learning control unit acquires a drive learning correction value obtained by correcting the drive learning value based on the deterioration allowance of the lockup clutch in coast learning calculated from the coast learning value from the coast learning control unit. Lock-up control device for automatic transmissions. In the lockup control device for the automatic transmission according to claim 1,
The coast learning control unit automatically sets the deterioration allowance of the lockup clutch in the coast learning as a deterioration correction value obtained by subtracting the coast learning convergence value before deterioration from the coast learning average value after deterioration. Transmission lockup control device. In the lockup control device for the automatic transmission according to claim 2.
The drive learning control unit obtains a deterioration correction conversion value obtained by converting the deterioration allowance of the lockup clutch in drive learning and the deterioration correction value from the coast learning control unit into a learning value for drive learning. A lockup control device for an automatic transmission. In the lockup control device for the automatic transmission according to claim 3.
When the drive learning control unit determines that the drive learning value has converged, the drive learning control unit acquires the drive learning correction value by correcting the drive learning convergence value after the convergence determination with the deterioration correction conversion value. Lock-up control device for automatic transmissions. In the lockup control device for the automatic transmission according to claim 3.
When the drive learning control unit determines that the drive learning value has not converged, the drive learning control unit converts the coast learning convergent value after the convergence test from the coast learning control unit into a learning value conversion process to drive learning. An automatic transmission lockup control device characterized by acquiring the drive learning correction value. In the lockup control device for the automatic transmission according to any one of claims 1 to 5.
The lockup controller is characterized by having an initial drive learning execution unit that performs initial drive learning based on the number of learning experiences that can acquire the drive learning convergence value before shipping to the market. Lockup controller. In the lockup control device for the automatic transmission according to claim 6.
When the coast learning control unit inputs the initial drive learning value from the initial drive learning execution unit, the coast learning control unit acquires the initial coast learning value by converting the initial drive learning value into the learning value for coast learning.
A lock-up control device for an automatic transmission, characterized in that a coast learning convergence value before deterioration is acquired based on the initial coast learning value during initial drive learning.

Images