JPH11108167A - Hydraulic control device for automatic transmission - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent vibration and reduction of a converging property in association with aging effect of a frictional engaging element, and enable correct feed back control for a long period by providing a learning control means for collecting a prescribed gain so as to approximate a value on the basis of a real change to a prescribed target value. SOLUTION: A signal of each sensor for engine rotating speed, a throttle opening, T/M input shaft rotating speed, car speed, and oil temperature, is inputted in a control unit 1, and is outputted to the linear solenoid valves SLS an SLU of a hydraulic circuit. The control unit 1 is provided with a feed back control means 1a and learning control means 1b, a hydraulic control signal is outputted by the feed back control means 1a while carrying out modification on the basis of a control deviation between real change and target change and prescribed gain so as to set a variation changed in association with advance of a shift, for example, an input rotating speed variation to a target prescribed change, and the prescribed gain is corrected by the learning control means 1b so as to approximate a value on the basis of real change to the prescribed target value. The correcting amount is set on the basis of at least one of an oil temperature, output rotating speed, and engine rotating speed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a hydraulic control device for an automatic transmission mounted on an automobile, and more particularly, to a hydraulic control device having feedback control.

[0002]

2. Description of the Related Art Conventionally, as disclosed in Japanese Utility Model Laid-Open Publication No. Sho 62-194960, hydraulic pressure control of an automatic transmission in which the hydraulic pressure of a starting clutch is feedback-controlled so that the rate of change in turbine (input) speed becomes a target value. A device has been proposed.
The hydraulic control device calculates a turbine rotational speed change rate from a difference between the actual turbine rotational speed and the previous turbine rotational speed,
Further, the turbine rotational speed change rate is compared with a target value of a predetermined characteristic to obtain a deviation, and learning control is performed by updating the deviation.

[0003]

In the feedback control, a proportional operation (P operation) correction value obtained from a predetermined proportional gain for compensating a control delay based on a deviation obtained by the learning control, The correction is made by an integral operation (I operation) correction value obtained from a predetermined integral gain for eliminating the deviation and a differential operation (D operation) correction value obtained from the differential gain for response time compensation. However, since each of the gains is a fixed value, a hydraulic pressure value corresponding to the deviation may not be obtained due to a change over time of the friction engagement element, and as a result, vibration and convergence may be reduced in feedback control. May be caused.

Therefore, the present invention prevents the vibration and the convergence due to the aging of the frictional engagement element from deteriorating by learning-controlling the gain at the time of the feedback control, thereby providing accurate feedback control over a long period of time. It is an object of the present invention to provide a hydraulic control device for an automatic transmission, which enables the following.

[0005]

According to a first aspect of the present invention, there is provided an input shaft to which power from an engine output shaft is input;
An output shaft connected to wheels, a plurality of friction engagement elements for changing a power transmission path between the input shaft and the output shaft, a hydraulic servo for disconnecting / engaging the friction engagement elements, Hydraulic control means (SLS, SL
U) and a control unit (1) that receives a signal from each sensor based on a vehicle running condition and outputs a hydraulic control signal (PA) to the hydraulic control unit. In the apparatus, the control unit may determine that an actual change (ΔNt) that changes with the progress of the shift is a target predetermined change (w_t).
control error (Error) and a predetermined gain (GP) between the actual change and the target change so that
Feedback control means (1a) for outputting the oil pressure control signal (PA) while correcting the predetermined gain (GP) based on the actual change (Error)
And a learning control means (1b) for correcting the value so as to approach a predetermined target value (GrdMax, GrdMin).

According to a second aspect of the present invention, the learning control means (1b) includes a value (Erro) based on the actual change.
r) maximum value (Emaxi) and minimum value (Eminj)
Is calculated, and if the difference between the maximum value and the minimum value is larger than a predetermined target range (GradMax), the difference (Emax) is calculated.
The hydraulic control device for an automatic transmission according to claim 1, wherein the predetermined gain (GP) is corrected so that (i-Eminj) falls within the target range.

According to a third aspect of the present invention, in the learning control means (1b), the value based on the actual change does not take the maximum value or the minimum value (i = 0, j = 0), and When the value (ErrorEnd) based on the change of the distance is more than a predetermined amount (GrdMin) from a predetermined reference value, the predetermined gain is corrected so that the value falls within a target range including the predetermined amount. Item 4. The hydraulic control device for an automatic transmission according to item 1 or 2.

According to a fourth aspect of the present invention, in the hydraulic control apparatus for an automatic transmission according to any one of the first to third aspects, the actual change is a change in the input rotation speed (ΔNt) with respect to the output rotation speed. is there.

According to a fifth aspect of the present invention, in the hydraulic control apparatus for an automatic transmission according to any one of the first to fourth aspects, the correction amount (ΔGP) of the predetermined gain is set based on at least an oil temperature. is there.

According to a sixth aspect of the present invention, in the hydraulic control of the automatic transmission according to the fifth aspect, the correction amount (ΔGP) of the predetermined gain is set based on at least one of an output speed and an engine speed. In the device.

[Operation] Based on the above structure, the hydraulic servo hydraulic pressure (w_target) is set so that a change accompanying the actual shift progression, such as a change in the input shaft speed (ΔNt) with respect to the output speed, becomes a target predetermined change (w_target). PA) is feedback-controlled, and the downshift or the upshift is advanced. At this time, the difference between the actual change and the target change (w_
target-ΔNt) control error (Error)
And a predetermined gain (GP) such as a proportional gain.

The learning control means (1b) provides a maximum value (E) of a value based on an actual change such as a control deviation (Error).
maxi) and a plurality of minimum values (Eminj) are calculated, and the maximum value (Emaxi−Emax) of the difference among them is calculated.
When j) is larger than the predetermined target value (GrdMax), it is determined that the hunting occurs because the gain is too large, and the predetermined gain (GP) is corrected to be small. The value does not take the maximum value or the minimum value, and is a predetermined amount (Grd) larger than the reference value.
If the distance is more than (Min), it is determined that the gain is too small and does not converge, and correction is made so that the predetermined gain becomes large.

The reference numerals in parentheses are for comparison with the drawings, but do not affect the configuration of the present invention at all.

[0014]

According to the present invention, the gain is learned and corrected so that the value based on the actual change in the feedback control approaches the predetermined target value. Variations in control can be prevented, and accurate feedback control can be maintained over a long period of time to improve shift feeling.

According to the present invention, when the difference between the maximum value and the minimum value of the value based on the actual change is larger than a predetermined target range, it is determined that hunting is occurring and the predetermined gain is corrected. Therefore, hunting in the feedback control can be reliably prevented.

According to the third aspect of the present invention, when the maximum value or the minimum value is not taken and the value based on the actual change is far from the target range, it is determined that the convergence has not occurred and the predetermined gain is increased. Since the correction is made, it is possible to prevent a decrease in convergence in the feedback control.

According to the present invention, the feedback control is performed based on the change in the input rotation speed (gear ratio) with respect to the output rotation speed, and the learning control is directly performed based on the change in the input rotation speed. Learning control can be performed.

According to the fifth aspect of the present invention, since the gain correction amount is set based on the oil temperature, it is possible to correct a difference in hydraulic responsiveness due to a change in oil temperature and perform appropriate learning control. .

According to the sixth aspect of the present invention, the amount of gain correction is set in accordance with the start-up characteristics of the engine and the transmission, so that accurate learning control can always be performed.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present automatic transmission has a number of frictional engagement elements such as clutches or brakes, and an automatic transmission in which the transmission path of a planetary gear is selected by appropriately connecting and disconnecting these frictional engagement elements. A mechanism (not shown) is provided, and the input shaft of the automatic transmission mechanism is connected to an engine output shaft via a torque converter, and the output shaft is connected to drive wheels.

FIG. 1 is a block diagram showing electric system control. Reference numeral 1 denotes a control unit (ECU) comprising a microcomputer (microcomputer), an engine rotation sensor 2 and a throttle opening for detecting a driver's accelerator pedal depression amount. Each signal from the degree sensor 3, the sensor 5 for detecting the input shaft rotation speed (= turbine rotation speed) of the transmission (automatic transmission mechanism), the vehicle speed (= automatic transmission output shaft rotation speed) sensor 6, and the oil temperature sensor 7 And is output to the linear solenoid valves SLS and SLU of the hydraulic circuit. The control unit 1 controls the change amount that changes as the shift progresses, for example, the input rotation speed change amount (ΔNt), to the target predetermined change (w_
control error (Error) between the actual change and the target change and a predetermined gain (G
P), the feedback control means 1a which outputs the hydraulic control signal, and corrects the predetermined gain so that a value (Error) based on the actual change approaches a predetermined target value (GrdMax, GrdMin). Learning control means 1b.

FIG. 2 is a diagram schematically showing a hydraulic circuit.
A plurality of friction engagement elements having the two linear solenoid valves SLS and SLU and switching a transmission path of a planetary gear unit of the automatic transmission mechanism to achieve, for example, a forward fourth speed or a fifth speed and a reverse first speed. A plurality of hydraulic servos 9 for connecting and disconnecting (clutches and brakes)
It has 0. Further, the linear solenoid valve S
Solenoid modulator pressure is supplied to the input ports a ₁ and a ₂ of the LS and SLU, and the control oil pressure from the output ports b ₁ and b _{2 of} these linear solenoid valves is applied to the control oil chambers of the pressure control valves 11 and 12, respectively. 11a and 12a. The pressure control valves 11 and 12 supply line pressures to the input ports 11b and 12b, respectively, and the pressure regulation from the output ports 11c and 12c regulated by the control hydraulic pressure is applied to the shift valves 13 and 15 respectively. It is supplied to the hydraulic servos 9 and 10 as needed.

This hydraulic circuit is for showing the basic concept. The hydraulic servos 9 and 10 and the shift valves 13 and 15 are symbolically shown.
A number of hydraulic servos are provided corresponding to the automatic transmission mechanism, and a number of shift valves for switching the hydraulic pressure to these hydraulic servos are also provided. Further, as shown in the hydraulic servo 10, the hydraulic servo has a piston 19 which is fitted to the cylinder 16 in an oil-tight manner by an oil seal 17, and the piston 19 is provided with a pressure control valve 12 acting on a hydraulic chamber 20. The outer friction plate 22 and the inner friction material 23 come into contact with each other based on the pressure-adjusted hydraulic pressure from the first and second springs, and move against the return spring 21. Although the friction plate and the friction material are shown by the clutch, it is needless to say that the friction plate and the friction material also correspond to the brake.

Next, power-on downshift will be described with reference to FIG. 3. First, control of the release hydraulic pressure PA will be described with reference to FIGS. Note that, specifically, a downshift (kickdown) in which the driver steps on the accelerator pedal to request a torque, and
2 shows a state in which two shifts are performed.
B4 brake, the hydraulic pressure PA of the hydraulic servo is
(For throttle pressure control) Linear solenoid valve SLT
Is regulated.

Throttle opening sensor 3 and vehicle speed sensor 6
When the control unit 1 determines a downshift based on the signal from the map, the timing is started and the shift control is started after a predetermined delay time from the shift determination (S1). At the start time point (t = 0), the release hydraulic pressure PA is at the engagement pressure, and the release friction engagement element is engaged. Then, the release torque T _A is calculated by a function of the input torque T _t (S2). The input torque _Tt is obtained by calculating an engine torque based on a throttle opening and an engine speed by a map, further calculating a speed ratio from an input / output speed of a torque converter, and obtaining a torque ratio by a map based on the speed ratio. It is determined by multiplying the engine torque by the above torque ratio. Further, the release-side torque T _A is obtained by involving the torque sharing ratio and the like in the input torque.

The waiting engagement pressure Pwt the release side from the release side torque T _A is calculated (S3), and outputs a control signal to the linear solenoid valve as disengagement side pressure PA is該待machine engagement pressure Pwt ( S4), the control of the release hydraulic pressure based on the input torque and the like is continued until a predetermined time tw has elapsed (S4).
5). The standby control is performed in steps S2 to S5, and the standby control time tw is changed by the input torque Tt.

Then, the predetermined release-side oil pressure P _AS and the release-side torque T _A are calculated in the same manner as described above (S7, S8), and the target oil pressure P _TA is calculated based on the release torque T _A (S9). Further, the margin ratio (degree of tie-up) S ₁₁ , S ₂₁
Thus, the release-side target oil pressure _PTA is calculated in consideration of the drive feeling (S10). Note that the above margin rate is
It is obtained from a large number of throttle opening / vehicle speed maps selected according to differences in oil temperature, and is generally S ₁₁ > 1.
0, S ₂₁ > 0.0.

Furthermore, the preset time t _TA, the gradient to the target hydraulic pressure _{P TA, [(P TA -P} AS) /
t _TA ], and the sweep down is performed by the gradient (S11). That is, in the power-on state, a sweep-down operation having a relatively steep gradient is performed, and the release-side hydraulic pressure PA continues until the release-side hydraulic pressure PA reaches the target hydraulic pressure _PTA immediately before the start of the inertia phase (S12). The oil pressure is set to the target oil pressure _PTA (S13). The target oil pressure P
_TA, the input shaft rotational speed change (gear ratio) to the output shaft rotational speed .DELTA.N (= [Delta] Nt) is continued until the shift start judgment rotation speed N _S (S14). The above-described steps S7 to S14 are the initial control.

Then, based on the detection of the input shaft speed (gear ratio) with respect to the output shaft speed, the downshift feedback control (S) is performed so that the input shaft speed becomes a predetermined change amount.
20) is performed, and the feedback control is performed by setting a gear ratio near the full rotation change speed of the gear ratio at which the downshift is completed.
2 [%], for example, 90 [%] (S2
1). It should be noted that until the servo activation control time t _SE ends (S23) and the engagement-side oil pressure PB becomes larger than the target oil pressure P _TB (S24) in relation to the control of the engagement-side oil pressure described below.
Then, the feedback control (S20) is continued.
Steps S20 to S24 correspond to feedback control.

When the shift up to a2 [%] is completed, a predetermined oil pressure change δP _FA having a relatively steep gradient is set, and the sweep down is performed at the gradient (S25).
When the release hydraulic pressure PA becomes 0, the release hydraulic control at the time of the downshift is completed (S26). Step S above
25 is the completion control.

Next, the control of the engagement side hydraulic pressure PB in the downshift will be described with reference to the flowcharts of FIGS. 6 and 7 and the time chart of FIG. Specifically, as described above, the shift is a 3-2 downshift, and therefore, the engagement-side friction engagement element is a B5 brake, and the hydraulic pressure PB of the hydraulic servo is linear (for lock-up control). The pressure is controlled by the solenoid valve SLU.

First, timing is started based on a downshift command from the control unit 1 (S30), and a predetermined signal is output to the linear solenoid valve SLU so that the engagement side hydraulic pressure PB becomes the predetermined pressure P _S1 (S31). . The predetermined pressure P _S1 is set to a hydraulic pressure required to fill the hydraulic chamber 20 of the hydraulic servo, and is maintained for a predetermined time t _SA . When the predetermined time t _SA has elapsed (S32), the engagement side hydraulic pressure PB sweeps down at a predetermined gradient [(P _S1 −P _S2 ) / t _SB ] (S3).
3) When the engagement side hydraulic pressure PB becomes the predetermined low pressure P _S2 (S3
4), the sweep-down is stopped and the predetermined low pressure _PS2 is maintained (S35). The predetermined low pressure P _S2 is set to a pressure that is equal to or higher than the piston stroke pressure and does not cause a change in the rotation of the input shaft, and the predetermined low pressure P _S2 is maintained until the time t reaches a predetermined time t _SE (S36). ). Steps S31 to S36 correspond to the servo start control.

[0033] Then, the engaging torque T _B is disengagement side pressure P
A and the function of the input torque Tt [T _B = f _TB (PA, T
is calculated by t)] (S37), and further consideration of the margin, the engaging torque T _{B is, [T B = S 1D ×} T B + S
_2D ] (S38). Then, the engagement side pressure PB is calculated from the engagement side torque T _B [PB = f
_PB (T _B )] (S39). Steps S37 to S39 above
Is the engagement control. Then, control of the engagement hydraulic pressure PB based on the engagement side input torque T _B (depending on the disengagement side pressure PA and the input torque Tt) by the step S39 is, a1 [%] of the total gear ratio downshift, e.g. 70
It continues until [%] (S40). That is, the input shaft rotational speed of the shift start time and N _TS, the rotational variation amount of .DELTA.N, pre-shift to g _i gear ratio, when the g i _{+ 1} and after shifting gear ratio, [(ΔN × 10
0) / N _TS (g _{i + 1} −g _i )] reaches a1 [%].

In step S40, the rotation change amount a
If it exceeds 1 [%], end control is started. First, the engagement-side target pressure P _TB is calculated from the engagement side input torque T _B (S4
1) Further, the engagement side hydraulic pressure PB at the time point of the rotation change amount a1 [%] is stored as _PLTB (S42). As a result, a predetermined gradient [(P _TB −P _LSB ) / t _LE ] is calculated based on a predetermined time t _LE that is set in advance, and the sweep-up is performed with the relatively gentle gradient (S43). This is continued until the engagement side oil pressure reaches the target oil pressure P _TB (S44). Further, a predetermined gradient δP _LB is set, and the sweep is performed at the gradient (S46). The sweep-up is continued until the input rotation speed change amount (ΔN) reaches a2 [%] of the entire gear ratio of the downshift, for example, 90 [%] (S47). The above-described steps S41 to S46 are the final control.

Further, an end time t _F of the end control is set (S48), a relatively steep gradient δP _FB is set, and the slope is swept up (S49).
The completion control time t _{FE is} continued (S50). The gradient δP _FB
When the power is on, the sweep-up of step S2
5 is set steeply in accordance with the release side hydraulic pressure δP _FA . Steps S48 and S49 are the completion control.

Next, the learning control of the gain in the feedback control according to the present invention will be described. The learning control is not limited to the feedback control (see S20) in the power-on / downshift, and can be applied to the feedback control in all the shifts, such as the feedback control in the upshift.

Referring to the feedback control flowchart of FIG. 8, first, a target turbine (input) rotational acceleration (rate of change) w_target and a target turbine (input) acceleration initial value ws are calculated (S51). The target turbine rotational acceleration w_target includes shiftR indicating the degree of progress of the shift and target rotational acceleration follow-up control (shift indicating the degree of the shift at the start of the feedback control.
Assuming that tRFBst,

[0038]

(Equation 1) Is required. Note that the degree of shift shiftR
Is Nt for turbine (input) speed, No for output (output) speed, and GEAR_BF (g
_{i + 1} ), and the gear ratio after shifting is GEAR_AF (g _i ),

[0039]

(Equation 2) Is required. In addition, the target turbine rotational acceleration initial value ws is obtained by assuming that the target shift time is Tfbend.

[0040]

(Equation 3) Required by

Next, a control error Error and a control cumulative error wError are calculated (S52). Control deviation E
rrr is a target turbine rotational acceleration (w_target)
t) and the current turbine rotational acceleration (rate of change) ΔNt (Error = w_target−ΔNt), and the control cumulative error wError is obtained by adding the current control error Error to the control cumulative error (wError) up to the previous time ( wError = wError + Error) is determined.

Further, the clutch oil pressure .DELTA.PTt for the amount of change in turbine torque and the clutch oil pressure .DELTA.P for the change in target rotational acceleration.
w is calculated (S53). The ΔPTt converts the difference (T−Tst) between the turbine (input) torque Tt and the turbine (input) torque Tst at the start of the feedback control into a change in the shared torque of the (disengaged) clutch, and converts this into the clutch hydraulic pressure. It is obtained by converting it into a change, and the above ΔPw
Is the difference (ws-ws) between the target turbine rotational acceleration initial value ws and the turbine rotational speed wst at the start of the feedback control.
t) is converted into a change in the shared torque of the disengagement side clutch, and is converted into a change in the clutch oil pressure to obtain the change. Also,
The control deviation clutch oil pressure ΔPwP and the control cumulative deviation clutch oil pressure ΔPwI are calculated (S54). The above ΔP
wP is obtained by converting the control error Error into an amount of change in the shared torque of the disengagement clutch, and converting it into a change in the clutch oil pressure. The ΔPwI is similarly obtained from the control cumulative error wError. The input torque is calculated based on the throttle opening and the engine speed by using a map, or by calculating with a computer, the engine torque is obtained. The clutch sharing torque is obtained from the torque sharing ratio and the allowance ratio based on the map.

Then, a clutch hydraulic pressure control value PAFB during the target rotational acceleration follow-up control is calculated (S55). The hydraulic pressure command value PAFB is calculated as follows: (the release-side hydraulic pressure command value PAst at the start of the feedback control, the turbine torque change amount clutch oil pressure ΔPTt, the target rotational acceleration change amount clutch oil pressure ΔPw, and the control deviation amount clutch oil pressure ΔPwP. Is multiplied by the proportional gain GP and the value obtained by multiplying the clutch hydraulic pressure ΔPwI by the integral gain GI by the control cumulative deviation, that is, PAFB = PAst + ΔPTt + ΔPw + GP × ΔPw
P + GI × ΔPwI.

Further, the local maximum value E of the control deviation (Error)
maxi (i = 1, 2, 3,...) are calculated (S56),
And the minimum value Eminj of the control deviation (j = 1, 2, 3,...)
Is calculated (S57). Then, the above step S51
When S57 to S57 are all calculated, the control end determination is established (S58), and the control error Error at that time is stored as the value ErrorEnd at the end of the target rotational acceleration follow-up (feedback) control of the control deviation (S59).

Then, as shown in FIG. 9 (a), the control deviation maximum value Emax and the minimum value Emin are calculated, the maximum value Emax is detected three times or more (i ≧ 3), and the minimum value Emin is calculated. Is detected three times (j = 3), and the maximum value of the difference between the local maximum value and the local minimum value exceeds a preset gain determination threshold (normal range) GradMax (Emaxi− Eminj ≧ GrdMa
x), it is determined that the hunting has occurred because the gain is too large (S60; YES), and learning correction is performed so that the proportional gain GP is reduced by dividing the proportional gain GP by the proportional gain correction amount ΔGP (S60; YES). GP = GP · ΔG
P) (S61). Note that the above three times of i ≧ 3 and j = 3 are one specific example, and a predetermined number of three or more or three or less can be arbitrarily selected.

On the other hand, as shown in FIG. 9 (b), the control error Error does not take the maximum value Emax or the minimum value Emin (i = 0 or j = 0) and the stored feedback control ends. Is smaller than the ideal curve (reference value) by the gain deviation determination threshold GrdMi.
n (Err)
orEnd ≧ GrdMin) It is determined that the gain is too small to converge (S62; YES), and learning correction is performed so that the proportional gain is increased by adding the proportional gain correction amount ΔGP to the proportional gain GP (GP = GP +
ΔGP) (S63).

FIG. 10 shows another embodiment.
Steps S56, S57 and S60 shown in FIG.
(A dash is used to distinguish them), and only the differences will be described. In the above-described embodiment, the minimum value and the maximum value are calculated by directly detecting the control deviation. However, in the present embodiment, the learning is indirectly performed using the hydraulic pressure.

That is, in steps S56 'and 57', the release hydraulic pressures Pmaxi and Pminj (i = 1, 2, 3) when the control error Error takes the maximum value and the minimum value.
.., J = 1, 2, 3,...). Then, in step S60 ', a predetermined number of local maximum values Pmaxi and local minimum values Pm
If the maximum value of the difference of inj is larger than a predetermined target range (large gain determination threshold) GrdMax ′, the proportional gain GP is corrected so that the difference (Pmaxi−Pminj) falls within the target range ( S61) Also, in step S62, the maximum value and the minimum value are not taken (i = 0,
j = 0) and the control deviation E at the end of the feedback control
When rrrEnd is separated from the reference value (ideal curve) by a predetermined amount (a small gain determination threshold) GrdMin,
The proportional gain GP is corrected so that the end-time control deviation ErrorEnd falls within the target range including the predetermined amount (S63).

As shown in FIG. 11, the proportional gain correction amount ΔGP in steps S61 and S63 is set by a map using the oil temperature region and the output rotation speed step as parameters. As a result, the difference in hydraulic response due to the oil temperature is appropriately corrected, and the difference in the change characteristic of the rotational speed on the clutch output side is corrected.

Further, as shown in FIG. 12, the proportional gain correction amount ΔGP is set by a map using the oil temperature region and the engine speed step as parameters. This corrects the difference in hydraulic response and the difference in engine startup characteristics due to the oil temperature.

Further, as shown in FIG. 13, the proportional gain ΔGP is obtained from a three-dimensional map of an oil temperature region, an output speed step, and an engine speed step.
Can also be set.

In the above embodiment, the index of the learning control of the gain in the feedback control is the control deviation due to the change rate (acceleration) of the input rotation speed (gear ratio) with respect to the output rotation speed, or the hydraulic pressure due to the control deviation. The change amount that changes with the progress of other shifts, such as the rate of change of the input shaft speed by the input shaft speed sensor, may be used.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an electronic control unit according to the present invention.

FIG. 2 is a diagram schematically showing a hydraulic circuit according to the present invention.

FIG. 3 shows a gear ratio indicating a power-on downshift.
4 is a time chart of a disengagement hydraulic pressure and an engagement hydraulic pressure.

FIG. 4 is a flowchart illustrating control of a release-side hydraulic pressure of a downshift.

FIG. 5 is a flowchart showing a continuation of FIG. 4;

FIG. 6 is a flowchart showing control of the engagement side hydraulic pressure of the downshift.

FIG. 7 is a flowchart showing a continuation of FIG. 6;

FIG. 8 is a flowchart showing gain learning control in feedback control according to the present invention.

FIG. 9A is a diagram showing a case where the gain is too large;
(b) is a diagram showing a case where the gain is too small.

FIG. 10 is a flowchart showing an embodiment of the gain learning control of the feedback control, which is partially changed.

FIG. 11 is a diagram showing an example of a map of a correction amount of gain learning control.

FIG. 12 is a diagram showing another example of a map of the correction amount of the gain learning control.

FIG. 13 is a diagram showing an example in which a map of a correction amount of gain learning control is further changed.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Control part 1a Feedback control means 1b Learning control means 2-7 Sensor 9,10 Hydraulic servo PA Release hydraulic pressure ΔN Actual change, rotation acceleration, rotation change rate SLS, SLU Hydraulic control means (phosphor solenoid valve) w_target Target Predetermined change (target turbine rotational acceleration) Error Value based on actual change (control deviation) Emaxi Local maximum value Eminj Local minimum value GP Predetermined (proportional) gain ΔGP Correction amount Grdmax, Grdmax ′ Predetermined target range (high gain determination threshold) Grdmin predetermined amount (small gain judgment threshold value)

 ──────────────────────────────────────────────────の Continued on the front page (72) Inventor Kenji Suzuki 10 Takane, Fujiimachi, Anjo, Aichi Prefecture Inside Aisin AW Co., Ltd. (72) Inventor Hiroshi Tsutsui 10 Takane, Fujiimachi, Anjo, Aichi Prefecture Aisin・ AW Co., Ltd.

Claims (6)

[Claims] An input shaft to which power from an engine output shaft is input, an output shaft connected to wheels, and a plurality of friction engagement elements for changing a power transmission path between the input shaft and the output shaft. A hydraulic servo for disconnecting and engaging these friction engagement elements, hydraulic control means for controlling the hydraulic pressure of the hydraulic servo,
Input the signal from each sensor based on the vehicle running situation,
A control unit that outputs a hydraulic control signal to the hydraulic control unit,
In the hydraulic control apparatus for an automatic transmission, the control unit includes a control deviation between the actual change and the target change such that the actual change that changes with the progress of the shift is a predetermined target change. And feedback control means for outputting the hydraulic control signal while correcting based on a predetermined gain; and learning control means for correcting the predetermined gain so that a value based on the actual change approaches a predetermined target value. A hydraulic control device for an automatic transmission, characterized in that: 2. The learning control means calculates a maximum value and a minimum value of the value based on the actual change, and when the difference between the maximum value and the minimum value is larger than a predetermined target range, the difference is calculated as the target value. The hydraulic control device for an automatic transmission according to claim 1, wherein the predetermined gain is corrected so as to fall within a range. 3. The learning control unit according to claim 1, wherein the value based on the actual change does not take a maximum value or a minimum value, and the value based on the actual change is separated by a predetermined amount or more from a predetermined reference value. The hydraulic control device for an automatic transmission according to claim 1, wherein the predetermined gain is corrected so that the value falls within a target range including the predetermined amount. 4. The hydraulic control apparatus for an automatic transmission according to claim 1, wherein the actual change is a change in an input rotation speed with respect to an output rotation speed. 5. The hydraulic control device for an automatic transmission according to claim 1, wherein the correction amount of the predetermined gain is set based on at least an oil temperature. 6. The hydraulic control device for an automatic transmission according to claim 5, wherein the correction amount of the predetermined gain is set based on at least one of an output speed and an engine speed.