JP2014176179A - Control unit of motor-driven vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain both assurance of a learning frequency and improvement of correction precision by implementing learning control, which uses learning correction quantity characteristics exhibiting different slopes for respective gear ratios with respect to transmission input torque, during implementation of μ slip control.SOLUTION: A control unit of an FR hybrid vehicle includes a motor generator, automatic transmission, second clutch, AT controller, and integrated controller. The second clutch is intervened in a power transmission path leading from the motor generator MG to right and left rear wheels, and is fully or slipperily engaged as an element other than gear change elements of the automatic transmission. The AT controller uses second clutch learning correction quantity characteristics, which exhibit different slopes for respective gear ratios with respect to transmission input torque, during implementation of μ slip control under an EV non-gear change state, implements learning control under which a misalignment occurring with the transmission input torque is computed, and thus acquires a learning correction quantity with which the second clutch learning correction quantity characteristics for the respective gear ratios are rewritten.

Description

  The present invention relates to a control device for an electric vehicle in which a motor and an automatic transmission are mounted in a drive system, and a frictional engagement element is provided in a power transmission path to a drive wheel.

  In an electric vehicle equipped with a motor and an automatic transmission in the drive system and provided with a second clutch in the power transmission path to the drive wheels, the target transmission torque capacity of the second clutch, the input torque and the input rotation speed during μ-slip control There is known a learning correction based on the size of (see Patent Document 1).

JP 2012-90491 A

  The learning correction during the conventional μ-slip control has no problem when the second clutch at the shift stage showing the correction amount characteristic having no gain sensitivity (inclination) with respect to the transmission input torque is targeted. However, if the learning correction is performed by applying the conventional learning correction as it is to the second clutch at the shift stage having the correction amount characteristic having the gain sensitivity with respect to the transmission input torque, the correction for the change in the transmission input torque is performed. The slope of the quantity characteristic is not considered. For this reason, the learning correction amount obtained by the learning control causes mislearning that deviates from the true deviation, so that high correction accuracy cannot be obtained, and accurate engagement capacity control of the second clutch cannot be ensured. There was a problem.

  The present invention has been made paying attention to the above problem, and during μ-slip control, by performing learning control using learning correction amount characteristics having different slopes for each gear position with respect to transmission input torque, An object of the present invention is to provide a control device for an electric vehicle that achieves both learning frequency and improvement of correction accuracy.

In order to achieve the above object, a control device for an electric vehicle according to the present invention includes a motor included in a drive source, an automatic transmission, a friction engagement element, a μ slip control means, and a correction amount learning control means.
The automatic transmission is interposed between the motor and the drive wheel, and switches a plurality of shift stages.
The friction engagement element is interposed in a power transmission path from the motor to the drive wheel, and is completely engaged or slip-engaged as an element other than the transmission element of the automatic transmission.
The μ slip control means implements μ slip control for maintaining minute slip rotation (μ slip rotation) of the frictional engagement element by motor rotation speed control while the automatic transmission is running in a non-shifting state of the motor.
The correction amount learning control means uses, as the correction amount characteristic of the element transmission torque of the friction engagement element during the μ slip control, a learning correction amount characteristic having a different slope for each shift stage with respect to the transmission input torque, A learning correction amount for rewriting the learning correction amount characteristic for each gear is acquired by performing learning control for calculating a deviation amount in the transmission input torque at that time.

Therefore, during μ-slip control performed by motor rotation speed control, the learning correction amount characteristic having a different slope for each shift stage with respect to the transmission input torque is used, and the deviation at the transmission input torque at that time is calculated. Learning control is performed. By performing this learning control, a learning correction amount for rewriting the learning correction amount characteristic for each gear position is acquired.
That is, during μ-slip control, learning control that calculates the deviation of the transmission input torque at that time is performed, so that learning control is performed without being subjected to learning control execution restrictions due to the magnitude of transmission input torque. The frequency of learning experienced frequently is secured.
In the learning control, since the actual torque of the friction engagement element can be estimated from the detected motor torque value during the motor speed control, the difference between the actual torque and the target torque is calculated from the difference between the target friction engagement element torque and the detected motor torque value. Can be estimated. By rewriting the deviation as the learning correction amount by offsetting the learning correction amount characteristic by the learning correction amount, the inclination of the learning correction amount characteristic is reflected as it is, and the correction accuracy is improved.
As described above, during μ-slip control, learning control using learning correction amount characteristics having different slopes for each gear position with respect to transmission input torque can be performed to ensure learning frequency and improve correction accuracy. Both can be achieved.

1 is an overall system diagram showing an FR hybrid vehicle (an example of an electric vehicle) by rear wheel drive to which a control device in Embodiment 1 is applied. It is a figure which shows an example of the EV-HEV selection map set to the mode selection part of the integrated controller of Example 1. It is a skeleton figure which shows an example of the automatic transmission which incorporated the 2nd clutch used as the object of learning control in the control apparatus in Example 1. FIG. 3 is an engagement operation table showing an engagement state of each friction engagement element for each shift stage in the automatic transmission according to the first embodiment. It is a figure which shows an example of the shift map of the automatic transmission set to the AT controller in Example 1. FIG. It is a control block diagram which shows the structure of the learning control required information calculating part which the integrated controller in Example 1 has. FIG. 3 is a control block diagram illustrating a configuration of a learning control unit for a CL2 learning correction amount characteristic included in the AT controller according to the first embodiment. FIG. 6 is a learning correction amount calculation explanatory diagram illustrating learning correction amount calculation contents in learning control using the CL2 learning correction amount characteristic of the first embodiment. 6 is a flowchart illustrating a flow of a CL2 learning correction amount characteristic learning control process executed by the integrated controller and the AT controller according to the first embodiment. It is a contribution characteristic figure to the output torque which shows the contribution of an input torque and a clutch torque with respect to an output torque when one fastening element is slip-engaged in each gear stage of the automatic transmission of Example 1. CL2 torque (TTCL2), target input torque (ETTMIN), CL2 differential rotation, mu slip control, final learning execution permission, mu slip control correction amount when learning control of CL2 learning correction amount characteristics of Example 1 is executed It is a time chart which shows each characteristic of (wvTRCL2ln), learning correction amount, intermediate calculation addition amount, temporary storage, and learning storage processing.

  Hereinafter, the best mode for realizing an electric vehicle control apparatus of the present invention will be described based on Example 1 shown in the drawings.

First, the configuration will be described.
The configuration of the control apparatus for the electric vehicle in the first embodiment includes “the overall system configuration”, “the schematic configuration of the automatic transmission”, “the learning control configuration of the CL2 learning correction amount characteristic”, “the learning control processing of the CL2 learning correction amount characteristic” The description will be divided into “configuration”.

[Overall system configuration]
FIG. 1 shows a rear-wheel drive FR hybrid vehicle (an example of an electric vehicle) to which the control device according to the first embodiment is applied, and FIG. 2 shows an EV-HEV selection set in a mode selection unit of the integrated controller 10. An example of a map is shown. The overall system configuration will be described below with reference to FIGS.

  As shown in FIG. 1, the drive system of the FR hybrid vehicle includes an engine Eng, a first clutch CL1, a motor / generator MG (motor), a second clutch CL2 (friction engagement element), and an automatic transmission AT. , Transmission input shaft IN, propeller shaft PS, differential DF, left drive shaft DSL, right drive shaft DSR, left rear wheel RL (drive wheel), right rear wheel RR (drive wheel), Have. M-O / P is a mechanical oil pump, S-O / P is an electric oil pump, FL is a left front wheel, FR is a right front wheel, and FW is a flywheel.

  The first clutch CL1 is a fastening element provided between the engine Eng and the generator MG, and is engaged by an urging force of a diaphragm spring or the like when the CL1 hydraulic pressure is not applied, and counteracts this urging force. This type is a so-called normally closed clutch that is released by applying CL1 hydraulic pressure.

  The automatic transmission AT is a stepped transmission that automatically switches between the seventh forward speed and the first reverse speed according to the vehicle speed, the accelerator opening, and the like. The second clutch CL2 interposed in the power transmission path from the motor / generator MG to the left and right rear wheels RL, RR is not a new dedicated clutch independent of the automatic transmission AT, but an automatic transmission. Friction engagement elements (clutch and brake) for shifting AT are used. That is, the second clutch CL2 is a frictional engagement element selected as an element suitable for the engagement condition among a plurality of frictional engagement elements that are engaged at each gear stage of the automatic transmission AT. The first clutch hydraulic unit 6 and the second clutch hydraulic unit 8 are built in an AT hydraulic control valve unit CVU attached to the automatic transmission AT.

  In this FR hybrid vehicle, there are electric vehicle mode (hereinafter referred to as “EV mode”), hybrid vehicle mode (hereinafter referred to as “HEV mode”), and drive torque control mode (hereinafter referred to as “EV mode”). And “WSC mode”).

  The “EV mode” is a mode in which the first clutch CL1 is disengaged and the drive source is only the motor / generator MG, and includes a motor drive mode (motor power running) and a generator power generation mode (generator regeneration). This “EV mode” is selected, for example, when the required driving force is low and the battery SOC is secured.

  The "HEV mode" is a mode in which the first clutch CL1 is engaged and the drive source is the engine Eng and the motor / generator MG. The motor assist mode (motor power running), engine power generation mode (generator regeneration), and deceleration regeneration It has a power generation mode (generator regeneration). This “HEV mode” is selected, for example, when the required driving force is high or when the battery SOC is insufficient.

  The "WSC mode" is driven in the "HEV mode", but the torque transmission of the second clutch CL2 is maintained while maintaining the second clutch CL2 in the slip engagement state by controlling the rotation speed of the motor / generator MG. This mode controls the capacity. The torque transmission capacity of the second clutch CL2 is controlled so that the driving force transmitted after passing through the second clutch CL2 becomes the required driving force that appears in the accelerator operation amount of the driver. The “WSC mode” is selected in a region where the engine speed is lower than the idle speed, such as when starting in the “HEV mode” selection state.

  As shown in FIG. 1, the control system of the FR hybrid vehicle includes an engine controller 1, a motor controller 2, an inverter 3, a battery 4, a first clutch controller 5, a first clutch hydraulic unit 6, and an AT controller. 7, a second clutch hydraulic unit 8, a brake controller 9, and an integrated controller 10.

  The controllers 1, 2, 5, 7, 9 and the integrated controller 10 are connected via a CAN communication line 11 that can exchange information with each other. In addition, 12 is an engine speed sensor, 13 is a resolver, 15 is a first clutch stroke sensor that detects the stroke position of the piston 14a of the hydraulic actuator 14, 19 is a wheel speed sensor, and 20 is a brake stroke sensor.

  The AT controller 7 inputs information from an accelerator opening sensor 16, a vehicle speed sensor 17, an inhibitor switch 18 for detecting a selected range position (N range, D range, R range, P range, etc.), and the like. . Then, when driving with the D range selected, the optimum shift stage is searched based on the position where the driving point determined by the accelerator opening APO and the vehicle speed VSP exists on the shift map (see FIG. 5), and the searched shift The control command to obtain the gear is output to the AT hydraulic control valve unit CVU. In addition to this shift control, based on a command from the integrated controller 10, control of complete engagement (HEV mode) / slip engagement (engine start) / release (EV mode) of the first clutch CL1 is performed. The second clutch CL2 is fully engaged (HEV mode) / μ slip engagement (EV mode) / rotational difference absorption slip engagement (WSC mode) / variable torque cutoff slip engagement (engine start / stop mode).

  Here, the control for maintaining the minute slip rotation (μ slip rotation) of the second clutch CL2 while the automatic transmission AT is traveling in the EV mode in the non-shift state is referred to as “μ slip control”. This “μ slip control” is performed by motor rotation speed control that controls the actual motor rotation speed of the motor / generator MG so as to match the target motor rotation speed at which the second clutch CL2 performs minute slip rotation. Since the motor torque during the motor rotation speed control depends on the load received by the motor / generator MG by the second clutch CL2, the CL2 actual torque can be estimated from the detected motor torque value during the motor rotation speed control. In addition, “μ slip control” is performed in the EV non-shifting state and the target drive torque is greater than the specified value (set in consideration of the region where slip cannot be performed due to friction or the region where accuracy cannot be secured due to low hydraulic pressure). . When the target drive torque is less than the specified value, a capacity safety factor is secured so that the second clutch CL2 does not slip. Therefore, immediately after the EV shift, immediately after the mode transition from the HEV mode to the EV mode, the second clutch CL2 is slipped in when the target drive torque is depressed from the low torque, and the μ slip control is activated (μ slip control means).

  The integrated controller 10 manages the energy consumption of the entire vehicle and has a function for running the vehicle with the highest efficiency. The motor rotation number sensor 21 for detecting the motor rotation number Nm and other sensors and switches 22 Necessary information and information are input via the CAN communication line 11. The integrated controller 10 includes a mode selection unit that selects a mode in which an operating point determined by the accelerator opening APO and the vehicle speed VSP is searched based on a position on the EV-HEV selection map shown in FIG. 2 as a target mode. Then, engine start control is performed when the mode is switched from the “EV mode” to the “HEV mode”. In addition, engine stop control is performed when the mode is switched from the “HEV mode” to the “EV mode”.

[Schematic configuration of automatic transmission]
FIG. 3 shows an example of the automatic transmission AT according to the first embodiment in a skeleton diagram, FIG. 4 shows the engagement state of each friction engagement element for each shift stage in the automatic transmission AT, and FIG. 5 shows the AT controller. 7 shows an example of a shift map of the automatic transmission AT set to 7. Hereinafter, a schematic configuration of the automatic transmission AT will be described with reference to FIGS.

  The automatic transmission AT is a stepped automatic transmission with 7 forward speeds and 1 reverse speed. As shown in FIG. 3, the driving force from at least one of the engine Eng and the motor / generator MG is input to the transmission. The rotational speed is changed by a transmission gear mechanism having four planetary gears and seven frictional engagement elements, which is input from the shaft Input, and is output from the transmission output shaft Output.

  As the transmission gear mechanism, the first planetary gear set GS1 by the first planetary gear G1 and the second planetary gear G2 and the second planetary gearset GS2 by the third planetary gear G3 and the fourth planetary gear G4 are coaxially arranged. , Are arranged in order. In addition, the first clutch C1 (I / C), the second clutch C2 (D / C), the third clutch C3 (H & LR / C), and the first brake B1 (Fr / B), the second brake B2 (Low / B), the third brake B3 (2346 / B), and the fourth brake B4 (R / B) are arranged. Further, a first one-way clutch F1 (1stOWC) and a second one-way clutch F2 (1 & 2OWC) are arranged as engagement elements for machine operation.

  The first planetary gear G1, the second planetary gear G2, the third planetary gear G3, and the fourth planetary gear G4 are a sun gear (S1 to S4), a ring gear (R1 to R4), and both gears (S1 to S4). , (R1 to R4) and a carrier (PC1 to PC4) for supporting pinions (P1 to P4) meshing with each other.

  The transmission input shaft Input is connected to the second ring gear R2 and inputs rotational driving force from at least one of the engine Eng and the motor generator MG. The transmission output shaft Output is connected to the third carrier PC3 and transmits the output rotational driving force to the driving wheels (left and right rear wheels RL, RR) via a final gear or the like.

  The first ring gear R1, the second carrier PC2, and the fourth ring gear R4 are integrally connected by the first connecting member M1. The third ring gear R3 and the fourth carrier PC4 are integrally connected by the second connecting member M2. The first sun gear S1 and the second sun gear S2 are integrally connected by a third connecting member M3.

  FIG. 4 is a fastening operation table. In FIG. 4, ◯ indicates that the friction engagement element is hydraulically engaged in the drive state, and (◯) indicates that the friction engagement element is hydraulically engaged (drive state) in the coast state. Indicates a one-way clutch operation), and no mark indicates that the frictional engagement element is in a released state. In addition, the frictional engagement element in the engaged state indicated by hatching indicates an element used as the second clutch CL2 at each shift stage.

  Regarding the shift to the adjacent shift stage, among the above-described friction engagement elements, one of the friction engagement elements that have been engaged is released, and one of the friction engagement elements that has been released is engaged. As shown in FIG. 4, it is possible to realize a shift speed of the first reverse speed with the seventh forward speed. Further, when the shift speed is the first speed and the second speed, the second brake B2 (Low / B) is set as the second clutch CL2. When the shift speed is the third speed, the second clutch C2 (D / C) is the second clutch CL2. When the shift speed is the fourth speed and the fifth speed, the third clutch C3 (H & LR / C) is the second clutch CL2. When the shift speed is 6th speed and 7th speed, the first clutch C1 (I / C) is the second clutch CL2. When the shift speed is the reverse speed, the fourth brake B4 (R / B) is the second clutch CL2.

  FIG. 5 is a shift map, and when the operating point on the map specified by the vehicle speed VSP and the accelerator opening APO crosses the upshift line, an upshift command is output. For example, when the shift speed is the first speed, if the driving point (VSP, APO) crosses the 1 → 2 up shift line due to the increase in the vehicle speed VSP, a 1 → 2 up shift command is output. FIG. 5 shows only the up shift line, but of course, the down shift line is also set with hysteresis for the up shift line.

[Learning control configuration of CL2 learning correction amount characteristics]
FIG. 6 shows the configuration of the learning control necessary information calculation unit included in the integrated controller 10, FIG. 7 shows the configuration of the learning control unit of the CL2 learning correction amount characteristic included in the AT controller 7, and FIG. 8 shows the CL2 learning correction. The learning correction amount calculation content in the quantity characteristic learning control processing is shown. Hereinafter, the learning control configuration of the CL2 learning correction amount characteristic will be described with reference to FIGS.

  As shown in FIG. 6, the integrated controller 10 includes a torque deviation calculation unit 10a, a torque F / B control unit 10b, a learning (correction) amount calculation unit 10c, a CL2 torque calculation unit 10d, and a μ slip learning permission. And a determination unit 10e.

  The torque deviation calculator 10a calculates the torque deviation by subtracting the estimated MG torque (torque after rotation speed F / B control) from the target MG torque (≈target drive torque during EV).

  The torque F / B control unit 10b receives the torque deviation from the torque deviation calculation unit 10a and calculates a torque F / B correction amount (F / B compensation value) so as to cancel the torque deviation.

  The learning (correction) amount calculation unit 10c inputs the target MG rotation speed, the torque F / B control deviation, and the torque F / B correction amount, calculates the correction amount wvTRCL2ln during μ slip control by inertia torque correction, and performs AT Output to the controller 7. This correction amount wvTRCL2ln during μ-slip control corresponds to the difference between the target CL2 torque (@transmission input end conversion value) and the motor torque detection value. In addition, the application of 3rd speed or higher is added as a condition.

  The CL2 torque calculator 10d calculates the CL2 torque TTCL2 by adding the target CL2 torque (≈target drive torque during EV) and the torque F / B correction amount, and outputs them to the AT controller 7. The target CL2 torque is output to the AT controller 7 as (transmission) input torque ETTMIN.

  The μ slip learning permission determination unit 10e receives a mode signal indicating the currently selected mode, determines the μ slip learning permission by inputting the EV mode, and outputs the learning permission flag fTTCL2ln to the AT controller 7. To do. In addition, the application of 3rd speed or higher is added as a condition.

  As shown in FIG. 7, the AT controller 7 includes a filter processing unit 71, a μ slip learning amount calculation unit 72, a μ slip learning amount switching unit 73, a learning value, as a learning control unit for the D1 speed / R range. A calculation unit 74 and an upper / lower limit processing unit 75 are included. Note that the second brake B2 (Low / B), which is the second clutch CL2 at the second speed, is less affected by the wraparound caused by the shift speed, so apply the learning value at the first speed to the second speed. Correspond with.

  The filter processing unit 71 filters the correction amount wvTRCL2ln during μ slip control input from the integrated controller 10, and outputs the correction amount wvTRCL2ln during μ slip control after the filter processing. The μ slip control correction amount wvTRCL2ln after the filtering process corresponds to the learning amount sampled by the learning amount average value calculation unit 78b.

  The μ slip learning amount calculation unit 72 calculates the μ slip learning amount for the D1 speed and the R range by learning correction during μ slip control described in Japanese Patent Application Laid-Open No. 2012-90491, which is a prior art. Specifically, it includes a torque correction amount calculation unit 72a, a rotation speed correction amount calculation unit 72b, a gain integration unit 72c, a gain integration unit 72d, and an addition unit 72e. The torque correction amount calculation unit 72a receives the input torque estimated value and the ATF temperature (transmission hydraulic fluid temperature), and calculates the torque correction amount using the friction torque correction map. The rotational speed correction amount calculation unit 72b inputs the AT input rotational speed (INPREV) and the ATF temperature, and calculates the rotational speed correction amount using the rotational speed correction map. Then, after gains Kf and Ki are accumulated in gain accumulating units 72c and 72d, an addition unit 72e calculates a μ slip learning amount by an equation of (Kf × torque correction amount + Ki × rotational speed correction amount).

  The μ slip learning amount switching unit 73 outputs the μ slip learning amount calculated by the μ slip learning amount calculating unit 72 to the learning value calculating unit 74 in the D1 speed and the R range.

  The learning value calculation unit 74 calculates the learning value by adding the correction amount wvTRCL2ln during μ slip control from the filter processing unit 71 and the μ slip learning amount from the μ slip learning amount switching unit 73.

  The upper and lower limit processing unit 75 performs a process of limiting the learning value output from the learning value calculation unit 74 by the upper limit value and the lower limit value, and determines this as the final learning value in the D1 speed / R range, This is reflected in the engagement capacity control of the two-clutch CL2.

  As shown in FIG. 7, the AT controller 7 includes an upper / lower limit processing unit 76, a learning permission / inhibition unit 77, a learning value calculation unit 78, and a μ as a learning control unit for D3,4,5,7th speed. A slip learning amount reflecting unit 79. In addition, since the 6th speed uses the same 1st clutch C1 (I / C) as the 7th speed as the 2nd clutch CL2, it corresponds by applying the learning value in the 7th speed to the 6th speed.

  The upper / lower limit processing unit 76 performs a process of limiting the correction amount wvTRCL2ln during μ slip control from the filter processing unit 71 by the upper limit value and the lower limit value (third speed or more), and outputs this to the learning value calculation unit 78. .

  The learning permission / inhibition unit 77 permits the learning amount calculation in the learning value calculation unit 78 when the learning control start condition is satisfied, and the learning amount in the learning value calculation unit 78 when the learning control prohibition condition is satisfied. The operation is prohibited.

The establishment of the learning control start condition is:
(1) In EV mode and automatic transmission AT is not geared
(2) Target drive torque exceeds specified value
(3) The second clutch CL2 is slipping
(4) Continuous judgment time is maintained when the difference between the target CL2 torque (transmission input end converted value) and motor torque detection value during μ-slip control is less than the specified value.
(5) Judgment is made by satisfying all the above conditions (1) to (5), ie, output of the learning permission flag.

The establishment of the learning control prohibition condition is determined by setting a prohibition flag. Note that if the permission flag changes to the prohibition flag during the learning amount calculation, the learning amount calculation is prohibited and the previous value is taken over. And even if it changes from a prohibition flag to a permission flag again, learning amount calculation remains prohibited. Here, when the prohibition flag is set as an example,
・ When a failure / diagnostic error is detected (when parameters used for learning cannot be used)
・ When the input torque is outside the low torque range (for example, outside the EV operating range)
-During EV shift (during the shift, the second clutch CL2 is fully engaged)
The time when the input torque fluctuates over the threshold during the learning amount calculation, the time when the input rotation fluctuates over the threshold during the learning amount calculation, and the like.

  The learning value calculation unit 78 calculates a learning value (μ slip learning amount + update addition amount) at the D3,4,5,7th speed. The learning value calculation unit 78 includes an input torque average value calculation unit 78a, a learning amount average value calculation unit 78b, an update ratio setting unit 78c, a μ slip learning amount calculation unit 78d, and a μ slip learning amount selection unit 78e. Have.

  The input torque average value calculation unit 78a calculates an input torque average value γ that is an average value of the input torque (ETTMIN) from the start of learning control to the sampling time.

  The learning amount average value calculation unit 78b calculates the difference between the target CL2 torque (@transmission input end converted value) from the learning control start to the sampling time and the detected motor torque (= μ slip control correction amount wvTRCL2ln after filtering). The learning amount average value, which is the average value of the learning amount by () is calculated.

  The update rate setting unit 78c multiplies the learning amount average value calculated by the learning amount average value calculation unit 78b and the gain Kf indicating the update rate to obtain a learning value that is a learning amount average value for calculation (for calculation). B is calculated.

  The μ slip learning amount calculation unit 78d calculates the μ slip learning amount by the following calculation processing method. Hereinafter, the μ slip learning amount calculation processing method will be described based on the CL2 learning correction amount characteristic shown in FIG.

In the μ slip learning amount calculation process, the μ slip learning amount has a gain sensitivity with respect to the input torque, so it does not mislearn in the following (1) and (2) scenes, and does not affect the performance when applied. Therefore, it has a function to consider the torque sensitivity by providing continuity.
(1) The average input torque γ at the time of learning (μ slip control) and the average learning value (for calculation) at that time are calculated. In the CL2 learning correction amount characteristic of FIG. 8, the input torque (ETTMIN) average value γ and the learning value (for calculation) B during learning correspond.
(2) Using (1), the learning addition amount Z when the TM input torque is zero is calculated, and the upper and lower limit processing is performed.
Learning addition amount Z = −max (min (learning value (for calculation) B− (learning value gain A × input torque during learning (ETTMIN) average value γ, learning update upper limit value), learning update lower limit value)
Here, as shown in the CL2 learning correction amount characteristic of FIG. 8, the learning value gain A is a constant that represents the inclination of the learning storage value with respect to the input torque, and this inclination increases as the (transmission) input torque increases. This indicates that the gripping amount of the second clutch CL2 is large. Further, as the learning value gain A, a value for each gear position determined in advance by an experiment or the like is used. Further, in the arithmetic expression of the learning addition amount Z, (learning value gain A × input torque during learning (ETTMIN) average value γ, learning update upper limit value), learning update lower limit value) is the CL2 learning correction amount characteristic of FIG. Corresponds to the previous learning amount C.
(3) The offset of (2) is added to the previous learned reflected value, the upper / lower limit process is performed, and the temporarily stored value of the learned stored value E is stored.
Learning storage value E = last learning storage value D + learning addition amount Z
It is. The upper and lower limit processing of the learning amount storage value E is processing that is limited by the CL2 learning storage upper limit value and the CL2 learning storage lower limit value as shown in the CL2 learning correction amount characteristic of FIG.
(4) The learning addition amount Z is temporarily stored, and the stored value is overwritten every time learning is performed.
(5) The learning addition amount Z is stored simultaneously with the reflection at the learning reflection timing.

  If the sampling time validity determination is not made, the μ slip learning amount selection unit 78e selects the previous value of the μ slip learning amount, and if the sampling time validity determination is made, the μ slip learning amount calculation unit 78d calculates the μ. Slip learning amount This value is selected.

  The μ slip learning amount reflecting unit 79 generates a final learning amount for CL2 control in order to reflect the μ slip learning amount calculated by the learning value calculating unit 78 in the engagement capacity control of the second clutch CL2. When the shift speed is 1st speed and 2nd speed, the second brake B2 (Low / B) is the second clutch CL2, and when the shift speed is 6th speed and 7th speed, the first clutch C1 (I / C) is the second clutch CL2. Therefore, when the second brake B2 (Low / B) becomes the second clutch CL2 and participates in the shift, the torque sharing ratio is calculated, and the torque sharing ratio is added to the μ slip learning amount calculated by the learning value calculation unit 78. Is the final learning amount for CL2 control. Similarly, when the first clutch C1 (I / C) becomes the second clutch CL2 and is involved in the shift, the torque sharing ratio is calculated, and the torque sharing is calculated to the μ slip learning amount calculated by the learning value calculation unit 78. The product of the ratio is the final learning amount for CL2 control. In other cases, the μ slip learning amount calculated by the learning value calculation unit 78 is used as the final CL2 control learning amount.

[Learning control processing configuration of CL2 learning correction amount characteristics]
FIG. 9 shows the flow of the learning control process of the CL2 learning correction amount characteristic executed by the integrated controller 10 and the AT controller 7 of the first embodiment. Hereinafter, each step of the flowchart of FIG. 9 representing the learning control processing configuration of the CL2 learning correction amount characteristic will be described.

In step S1, following the determination of NO in steps S1, S2, and S11, it is determined whether or not the μ slip-in start condition is satisfied. If YES (μ slip-in start condition is satisfied), the process proceeds to step S2. If NO (μ slip-in start condition is not satisfied), the determination in step S1 is repeated.
Here, the establishment of the μ slip-in start condition is
(1) In EV mode and automatic transmission AT is not geared
(2) Judgment is made by satisfying all the above conditions (1) and (2) that the target drive torque is not less than the specified value.

In step S2, following the determination that the μ slip-in start condition is satisfied in step S1, it is determined whether the μ slip control start condition is satisfied. If YES (μ slip control start condition is satisfied), the process proceeds to step S3. If NO (μ slip control start condition is not satisfied), the process returns to step S1.
Here, the establishment of the μ slip control start condition is
(1) In EV mode and automatic transmission AT is not geared
(2) Target drive torque exceeds specified value
(3) Determination is made by satisfying all the above conditions (1) to (3) that the second clutch CL2 is in the slip state.

In step S3, it is determined that the μ slip control start condition is satisfied in step S2, or that the learning control start condition is not satisfied in step S3, or that the CL2 slip state is in step S11. Following the determination, it is determined whether or not a learning control start condition is satisfied. If YES (learning control start condition is satisfied), the process proceeds to step S4. If NO (learning control start condition is not satisfied), the determination in step S3 is repeated.
Here, the establishment of the learning control start condition is
(4) Continuous judgment time is maintained when the difference between the target CL2 torque (transmission input end converted value) and motor torque detection value during μ-slip control is less than the specified value.
(5) Judgment is made by satisfying all the above conditions (4) and (5), ie, output of the learning permission flag.

  In step S4, learning control is started following the determination that the learning control start condition is satisfied in step S3, and the process proceeds to step S5.

  In step S5, following the learning control start in step S4 or the takeover of the temporarily stored value of the previous learning amount in step S6, the sampling number set from the start of learning control is obtained as the sampling number of the input torque and learning amount. It is determined whether or not the sampling time necessary for the measurement has elapsed. If YES (sampling time has elapsed), the process proceeds to step S7. If NO (before sampling time has elapsed), the process proceeds to step S6.

  In step S6, following the determination that the sampling time has not elapsed in step S5, the temporary storage value of the previous learning amount is inherited (in the case of the first time from the start of learning control, the previous learning value is inherited), and the process proceeds to step S5. Return.

In step S7, following the determination that the sampling time has elapsed in step S5 or the determination that the learning amount calculation permission condition is satisfied in step S8, calculation of the learning amount is started, and the process proceeds to step S8.
Here, after the learning amount calculation starts, the learning amount of the sampling time is always calculated until the learning amount calculation ends. The learning amount calculation is performed by the following “calculation input processing”, “learning amount calculation processing”, and “learning amount storage value upper and lower limit processing”.
<Calculation input processing>
(1-1) Integration of motor torque detection value during sampling time
(1-2) Accumulate CL2 torque deviation amount during sampling time (target CL2 torque @ difference between transmission input end converted value and motor torque detection value) <Learning amount calculation>
Calculate motor torque average value using (2-1) (1-1) (calculation formula: (1-1) / sampling time)
Using (2-2) and (1-2), calculate the CL2 torque deviation average value (calculation formula: (1-2) / sampling time)
(2-3) Calculate CL2 learning amount using (2-1) and (2-2) (calculation formula: (2-1)-(2-2) x gain K (fixed constant value))
<Upper and lower limit processing of learning amount storage value>
(2-3) should be between the upper and lower limits of the stored learning value
Min {Upper limit learning value, (2-3)}
Max {Lower learning value, (2-3)}
In step S8, following the learning amount calculation in step S7, it is determined whether or not a learning amount calculation permission condition is satisfied. If YES (learning amount calculation permission condition is satisfied), the process returns to step S7. If NO (learning amount calculation permission condition is not satisfied), the process proceeds to step S9.
Here, the satisfaction of the learning amount calculation permission condition is
(1) In EV mode and automatic transmission AT is not geared
(2) Target drive torque exceeds specified value
(3) The second clutch CL2 is slipping
(4) Continuous judgment time is maintained when the difference between the target CL2 torque (transmission input end converted value) and motor torque detection value during μ-slip control is less than the specified value.
(5) Judgment is made by satisfying all the above conditions (1) to (5), ie, output of the learning permission flag.

  In step S9, following the determination that the learning amount calculation permission condition is not satisfied in step S8, the learning amount calculation is ended, and the learning amount (final value) at the end of the learning amount calculation is temporarily stored (the temporarily stored value is The process proceeds to step S10.

  In step S10, following the end of the learning amount calculation and the temporary storage of the final value in step S9, it is determined whether or not it is a non-shift state in the EV mode. If YES (EV non-shift state), the process proceeds to step S11. If NO (other than EV non-shift state), the process proceeds to step S12.

  In step S11, it is determined whether or not the second clutch CL2 is in a slip state following the determination in step S10 that the vehicle is in the EV non-shift state. If YES (CL2 slip state), the process returns to step S3. If NO (CL2 non-slip state), the process returns to step S1.

In step S12, following the determination in step S10 that the vehicle is not in the non-EV shift state, the learning control is terminated, the temporarily stored learning amount (final value) is determined as the learning value, and the second clutch CL2 thereafter is determined. Reflect to the fastening capacity control and proceed to the end.
Here, when the second clutch CL2 is subjected to the engagement displacement control, a learning stored value (FIG. 8) corresponding to the magnitude of the TM input torque is used as a correction value for the CL2 torque command value obtained by calculation or the like. That is, in the CL2 learning correction amount characteristic shown in FIG. 8, the vertical axis represents a negative correction value (the upper side means a large negative correction value). Therefore, a value obtained by subtracting the correction value (a larger value as the TM input torque is higher) from the CL2 torque command value is used as the final CL2 torque command value and reflected in the engagement capacity control. As a result of reflecting in this engagement capacity control, it is possible to prevent the second clutch CL2 from being gripped too much.

Next, the operation will be described.
The effects of the control device for the FR hybrid vehicle according to the first embodiment are as follows: “reason that the CL2 learning correction amount characteristic has gain sensitivity (slope) with respect to the input torque”, “learning control processing operation of the CL2 learning correction amount characteristic”, The explanation will be divided into “Learning control action of CL2 learning correction amount characteristic”.

[Reason why CL2 learning correction amount characteristic has gain sensitivity (slope) to input torque]
The present invention uses the CL2 learning correction amount characteristic for the transmission input torque during the μ slip control performed by the motor rotation speed control, and performs the learning control to perform the CL2 learning correction amount for the transmission input torque for each gear position. A characteristic learning correction amount is acquired.

  That is, even if a clutch independent of the automatic transmission AT is set as the second clutch CL2, the inclination of the CL2 learning correction amount characteristic with respect to the transmission input torque differs for each shift stage of the automatic transmission AT. For this reason, learning control is performed using the CL2 learning correction amount characteristic with respect to the transmission input torque. In particular, in the case of the first embodiment, the CL2 element is selected from a plurality of frictional engagement elements built in the automatic transmission AT for each shift stage. In this case, the CL2 learning correction amount characteristic is provided for each shift stage. Not only is the slope different, but the slope of the CL2 learning correction amount characteristic is different for each selected CL2 element. Hereinafter, the reason why the CL2 learning correction amount characteristic in the automatic transmission AT according to the first embodiment has gain sensitivity (gradient) with respect to the input torque will be described with reference to FIG.

FIG. 10 shows the relationship that when one element slips at each shift speed, the contribution of the input torque to the output torque differs from the contribution of the clutch torque.
When the low brake Low / B (second clutch CL2) slips at the first speed (1st) and the second speed (2nd), the contribution to the output torque is mostly 90% or more.
When the direct clutch D / C (second clutch CL2) slips at the third speed (3rd), the contribution to the output torque is about 60% for the clutch torque and about 40% for the input torque.
When the high & low reverse clutch H & LR / C (second clutch CL2) slips at 4th gear (4th) and 5th gear (5th), the contribution to the output torque is around 50%. Input torque accounts for around 50%.
When the input clutch I / C (second clutch CL2) slips at 6th speed (6th) and 7th speed (7th), the contribution to the output torque is about 60% of the clutch torque and the input torque is It occupies around 40%.
And when the 2nd clutch CL2 is slipped, there exists a tendency for real CL2 torque to come out with respect to clutch capacity, so that the occupation ratio of input torque is high with respect to the contribution to output torque.

  Therefore, in the first gear (1st) and the second gear (2nd), when the low brake Low / B (second clutch CL2) is slipped, the occupation ratio of the input torque is very large with respect to the contribution to the output torque. Therefore, the actual CL2 torque against the clutch capacity can be suppressed. In other words, at the first speed (1st) and the second speed (2nd) where the low brake Low / B is the second clutch CL2, the CL2 learning correction amount is substantially constant with respect to the change of the input torque. Therefore, the learning correction accuracy during the μ slip control can be ensured without performing the learning control using the CL2 learning correction amount characteristic for the input torque.

  On the other hand, in the third speed (3rd) to the seventh speed (7th), when the second clutch CL2 is slipped, the occupation ratio of the input torque to the contribution to the output torque is the first speed (1st) and Since it is much higher than the 2nd speed (2nd), the actual CL2 torque tends to increase with respect to the clutch capacity, the CL2 learning correction amount characteristic with respect to the input torque has a slope, and the slope of the characteristic is Depends on gear and CL2 factor. In other words, at the third speed (3rd) to the seventh speed (7th), the CL2 learning correction amount characteristic having an inclination with respect to the input torque is used for each shift speed, and the learning control is performed. If the inclination (gain sensitivity) is not taken into consideration, the learning correction accuracy during μ slip control cannot be secured.

  For the above reasons, in the first embodiment, the learning correction during the μ-slip control of the prior art is adopted in the first gear, the second gear, and the R range, and the CL2 learning for the input torque is performed in the third gear to the seventh gear. The correction amount characteristic learning control is employed.

[Learning control processing effect of CL2 learning correction amount characteristics]
As described above, in the first embodiment, the learning control of the CL2 learning correction amount characteristic with respect to the input torque is performed at the third to seventh speeds among the respective gear speeds. Hereinafter, the learning control processing operation of the CL2 learning correction amount characteristic will be described based on the flowchart shown in FIG.

  First, when the μ slip-in start condition, the μ slip control start condition, and the learning control start condition are all established, the process proceeds from step S1 to step S2 to step S3 to step S4 in the flowchart of FIG. 9, and learning control is started. .

  Until the sampling time elapses after the learning control is started, the flow from step S5 to step S6 is repeated in the flowchart of FIG. In this step S6, in the case of the first time from the start of learning control, the previous learning value is taken over.

  Then, when the sampling time elapses, the process proceeds from step S5 to step S7 to step S8 in the flowchart of FIG. 9, and while the learning amount calculation permission condition is satisfied in step S8, the process proceeds from step S7 to step S8. The forward flow is repeated. In this step S7, calculation of the learning amount is started, and after the learning amount calculation is started, the learning amount of the sampling time is constantly calculated until the learning amount calculation ends. Here, that the learning amount of the sampling time is always calculated means that the calculation of the learning correction amount is continued until the calculation is completed by updating the sampling data in which the oldest data is deleted and the latest data is added.

  If the learning amount calculation permission condition is not satisfied in step S8, the process proceeds from step S8 to step S9 to step S10. In step S9, the learning amount calculation is terminated, and the learning amount (final value) at the end of the learning amount calculation is temporarily stored by overwriting.

  If it is determined in step S10 that the EV is not in a gear shift state and the CL2 slip state is maintained, the process proceeds from step S10 to step S11 to step S3, and whether or not the learning control start condition is satisfied again. To be judged. If it is determined in step S10 that the EV is not gear-shifted and the CL2 is not slipped, the process proceeds from step S10 to step S11 to step S1, and again whether the μ slip-in start condition is satisfied. It is determined whether or not.

  If it is determined in step S10 that the vehicle is in the EV shift state, the process proceeds from step S10 to step S12 to the end. In this step S12, the learning control is terminated, the temporarily stored learning amount (final value) is determined as the learning value, and is reflected in the subsequent engagement capacity control of the second clutch CL2.

[Learning control action of CL2 learning correction amount characteristics]
As described above, the learning control action of the CL2 learning correction amount characteristic will be described based on the time chart of FIG. 11 showing a specific example when the learning control processing of the CL2 learning correction amount characteristic is performed during the μ slip control. To do.

  When the CL2 torque is reduced at time t0 and the μ slip determination is made at time t1, CL2 differential rotation occurs and the flag during μ slip control is set. When the μ slip control is started at time t1, and the learning control start condition is satisfied at time t2, the learning calculation process is started from time t2, and the generation of the correction amount during μ slip control is started.

  When the learning calculation process is started at time t2, the learning correction amount calculation process starts awaiting the sampling time from time t2 until time t3 elapses. When the time t3 has elapsed, the learning correction amount calculation process is started, and the learning correction amount is calculated until time t4 when the learning calculation process ends. This calculation of the learning correction amount is not limited by the upper limit value and the lower limit value.

  When the learning calculation process ends at time t4, the learning correction amount at time t4 is moved as the intermediate calculation addition amount, and “previous learning stored value + intermediate calculation addition amount” is temporarily stored. This temporary storage value is overwritten every time learning is performed.

  When the learning calculation process is started again at a time t5 when a predetermined time has elapsed from the time t4, the learning correction amount calculation process starts awaiting the sampling time until the time t6 elapses from the time t5. When the time t6 has elapsed, the learning correction amount calculation process is started, and the learning correction amount is calculated until time t7 when the learning calculation process ends. The calculation of the learning correction amount is limited by the lower limit value.

  When the learning calculation processing is completed at time t7, the learning correction amount (= lower limit value) at time t7 is moved as an intermediate calculation addition amount, and “previous learning stored value + intermediate calculation addition amount” is temporarily Saved.

  The learning calculation process starts again at time t8 when a predetermined time has elapsed from time t7, but when the learning calculation process ends at time t9 before the sampling time elapses, the learning correction amount is calculated. Of course, neither movement nor temporary storage as a calculation addition amount during the process is performed.

  When the learning calculation process is started again at a time t10 when a predetermined time has elapsed from the time t9, the learning correction amount calculation process starts awaiting the sampling time until the time t11 elapses from the time t10. When the time t11 has elapsed, the learning correction amount calculation process is started, and the learning correction amount is calculated until time t12 when the learning control ends. The calculation of the learning correction amount is limited by the lower limit value.

  When learning control ends at time t12, the learning correction amount (= lower limit value) at time t12 is moved as an intermediate operation addition amount, and “previous learning stored value + intermediate operation addition amount” is an upper limit value and lower limit value. It is temporarily saved while being restricted by the value. Furthermore, the temporarily stored value (final value) is determined as a learning value to be reflected in the engagement capacity control of the second clutch CL2 by the learning storage process.

  As described above, in the first embodiment, during the μ slip control performed by the motor rotational speed control, the CL2 learning correction amount characteristic having a different slope for each shift stage with respect to the transmission input torque is used, and the transmission at that time Learning control for calculating a deviation in the input torque is performed. And the structure which acquires the learning correction amount which rewrites the CL2 learning correction amount characteristic for every gear stage (each 3rd gear-7th gear) was implemented by implementing this learning control.

  That is, during μ-slip control, learning control that calculates the deviation of the transmission input torque at that time is performed, so that learning control is performed without being subjected to learning control execution restrictions due to the magnitude of transmission input torque. The learning frequency that is frequently experienced is ensured (see FIG. 11).

  In the learning control, since the actual torque of the second clutch CL2 can be estimated based on the detected motor torque value during the motor speed control, the difference between the actual torque and the target torque is calculated from the difference between the target CL2 torque and the detected motor torque value. Can be estimated. By using this deviation as a learning correction amount and rewriting the CL2 learning correction amount characteristic by offsetting the learning correction amount, the inclination of the CL2 learning correction amount characteristic is reflected as it is, and the correction accuracy is improved (FIG. 8). reference).

Next, the effect will be described.
In the control device for the FR hybrid vehicle of the first embodiment, the effects listed below can be obtained.

(1) A motor (motor / generator MG) as a drive source;
An automatic transmission AT that is interposed between the motor (motor / generator MG) and driving wheels (left and right rear wheels RL, RR), and switches a plurality of shift stages;
Friction engagement that is interposed in the power transmission path from the motor (motor / generator MG) to the drive wheels (left and right rear wheels RL, RR) and is completely engaged or slip-engaged as an element other than the transmission element of the automatic transmission AT An element (second clutch CL2);
The μ slip control is performed by the motor rotation speed control to maintain the micro slip rotation (μ slip rotation) of the friction engagement element (second clutch CL2) while the automatic transmission AT is running in the non-shifting state of the motor. Control means (integrated controller 10);
As the correction amount characteristic of the element transmission torque of the friction engagement element (second clutch CL2) during the μ slip control, a learning correction amount characteristic (CL2 in FIG. 8) having a different slope with respect to the transmission input torque for each shift stage. Learning correction amount characteristics), and learning control for calculating the deviation of the transmission input torque at that time is performed, so that the learning correction amount characteristics (CL2 learning correction amount characteristics in FIG. 8) for each of the gears are obtained. Correction amount learning control means (FIG. 9) for acquiring the learning correction amount to be rewritten;
Is provided.
For this reason, during μ-slip control, learning control using learning correction amount characteristics (CL2 learning correction amount characteristics) with different slopes for each shift stage with respect to transmission input torque is performed to ensure learning frequency. Both improvement in correction accuracy can be achieved.

(2) When the learning control ends, the correction amount learning control means (FIG. 9) stores the learning correction amount temporarily stored at the end of the learning control as a final learning value when the transmission input torque is zero. .
For this reason, in addition to the effect of (1), the learning correction amount can be shared and the learning value storage memory can be reduced in a plurality of shift steps that do not affect the learning correction amount due to the different shift steps. it can. In the case of the first embodiment, the sixth-stage learning correction amount is shared with the seventh-speed learning correction amount using the same CL2 element.

(3) The correction amount learning control means (FIG. 9) continues the determination time that the difference between the target frictional engagement element torque (target CL2 torque) during the μ slip control and the detected motor torque is equal to or less than a specified value. This is the learning control start condition of the learning correction amount characteristic (CL2 learning correction amount characteristic in FIG. 8).
For this reason, in addition to the effects of (1) or (2), highly accurate learning control can be performed in a stable region in which fluctuations in torque deviation are suppressed and stable during μ slip control.

(4) When the sampling time elapses from the start of learning control, the correction amount learning control means (FIG. 9) starts learning correction amount calculation processing, and after the learning correction amount calculation processing starts, the learning correction amount calculation processing ends. While the sampling data is updated, the learning correction amount calculation for the sampling time is continued.
For this reason, in addition to the effect of (3), highly accurate learning control can be performed by continuing the learning correction amount calculation from the start to the end of the learning correction amount calculation processing.

(5) The μ slip control means (integrated controller 10) is configured to perform μ based on motor rotation speed control when the automatic transmission AT is running in a motor in a non-shifting state and the target drive torque is in a region greater than a specified value. Slip control is performed (FIG. 1).
For this reason, in addition to the effects of (1) to (4), learning is started in the learning control start condition that μ-slip control is being performed. A reduction in control accuracy can be prevented.

(6) The correction amount learning control means (FIG. 9) is a shift stage (third speed stage) having a learning correction amount characteristic having an inclination with respect to the transmission input torque among the plurality of shift stages of the automatic transmission AT. ˜7th speed) is selected, and the learning correction amount characteristic (CL2 learning correction amount characteristic in FIG. 8) for each selected gear stage (3rd speed to 7th speed) is the target of learning control.
For this reason, in addition to the effects (1) to (5), not all the gears of the automatic transmission AT but the selected gears (third gear to seventh gear) are subject to learning control. Thus, it is possible to carry out highly accurate learning control while achieving simplification of learning calculation processing.

  As mentioned above, although the control apparatus of the electric vehicle of this invention has been demonstrated based on Example 1, it is not restricted to this Example 1 about a concrete structure, The invention which concerns on each claim of a claim Design changes and additions are permitted without departing from the gist of the present invention.

  In the first embodiment, the shift element in the automatic transmission AT is used as the friction engagement element, and the element selected from the three engagement elements that are engaged at each shift speed is the second clutch CL2. However, as the frictional engagement element, an automatic transmission such as a second clutch interposed between the motor and the input shaft of the automatic transmission or a second clutch interposed between the output shaft of the automatic transmission and the drive wheel is used. It is good also as a 2nd clutch provided independently from the machine.

  In the first embodiment, an example in which the control device for an electric vehicle according to the present invention is applied to an FR hybrid vehicle having one motor and two clutches is shown. However, the control device of the present invention can be applied not only to a 1-motor 2-clutch FF hybrid vehicle, but also to a parallel-type hybrid vehicle having a power split mechanism other than a 1-motor 2-clutch. . Furthermore, the present invention can be applied to an electric vehicle or the like provided with a stepped automatic transmission in a drive system.

Eng engine
CL1 1st clutch
MG motor / generator (motor)
CL2 2nd clutch (friction engagement element)
AT automatic transmission
IN Transmission input shaft
C1 1st clutch
C2 2nd clutch
C3 3rd clutch
B1 First brake
B2 Second brake
B3 3rd brake
B4 Fourth brake 7 AT controller 76 Upper / lower limit processing unit 77 Learning permission / prohibition unit 78 Learning value calculation unit 79 μ slip learning amount reflecting unit 10 Integrated controller

Claims (6)

A motor as a drive source;
An automatic transmission that is interposed between the motor and the drive wheel and switches a plurality of shift stages;
A friction engagement element that is interposed in a power transmission path from the motor to the drive wheel, and is completely or slip-engaged as an element other than the transmission element of the automatic transmission;
Μ slip control means for performing μ slip control by motor rotation speed control for maintaining minute slip rotation (μ slip rotation) of the friction engagement element during motor running in the non-shift state of the automatic transmission;
During the μ-slip control, as a correction amount characteristic of the element transmission torque of the friction engagement element, a learning correction amount characteristic having a different slope for each shift stage with respect to the transmission input torque is used. Correction amount learning control means for acquiring a learning correction amount for rewriting the learning correction amount characteristic for each shift stage by performing learning control for calculating the deviation of
An electric vehicle control device comprising: In the control device of the electric vehicle according to claim 1,
When the learning control ends, the correction amount learning control means stores the learning correction amount temporarily stored at the end of the learning control as a final learning value when the transmission input torque is zero. Control device. In the control apparatus of the electric vehicle according to claim 1 or 2,
The correction amount learning control means starts learning control of the learning correction amount characteristic that the difference between the target frictional engagement element torque during μ-slip control and the detected motor torque value continues below a specified value for a continuous determination time. A control device for an electric vehicle, characterized in that: In the control apparatus of the electric vehicle according to claim 3,
The correction amount learning control means starts the learning correction amount calculation process when the sampling time elapses from the start of the learning control, and updates the sampling data until the learning correction amount calculation process ends after the learning correction amount calculation process starts. A control device for an electric vehicle, characterized in that the learning correction amount calculation of the sampling time is continued. In the control apparatus of the electric vehicle described in any one of Claim 1 to 4,
The μ slip control means performs μ slip control by motor rotation speed control when the automatic transmission is running in a motor in a non-shifting state and the target drive torque is in a region equal to or greater than a specified value. A control device for an electric vehicle. In the control apparatus of the electric vehicle described in any one of Claim 1 to 5,
The correction amount learning control means selects a gear stage having a learning correction amount characteristic having an inclination with respect to a transmission input torque from among a plurality of gear stages of the automatic transmission, and learn correction for each selected gear stage. A control device for an electric vehicle characterized in that a quantity characteristic is an object of learning control.