JP2794176B2 - Shift control device for automatic transmission for vehicles - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a shift control device for an automatic transmission for a vehicle, which ensures a shift during a gear shift. (Prior Art) In recent years, an automatic transmission for a vehicle has been developed, and a gear shift operation of the transmission is not manually performed by a driver, but an actuator operated by, for example, hydraulic pressure is operated by a command of an electronic control device mounted on the vehicle. Thus, the shift operation of the transmission is performed. In this kind of shift operation of the automatic transmission for a vehicle, for example, the completion of the operation is detected by a position sensor such as a detection switch to prevent an abnormal operation of the shift. There is a problem that adjustment is required in response to a difference in a shift operation position or the like. As a countermeasure against the above problem, a shift sensor for detecting the operating position of the shift actuator is provided, and a shift control device for an automatic transmission for a vehicle for controlling a shift operation in anticipation of an overshoot amount of the shift actuator obtained by a learning value A proposal is shown in Japanese Patent Application No. 60-199004. (Problems to be Solved by the Invention) In the above proposal, a new shift target value is set from the learning value based on the shift position, but the state of the learning value is not checked and learning is not performed. There is a problem in that a large amount of time is required for a learning operation, such as when the learning value is destroyed, and the shift operation is not performed smoothly. It is another object of the present invention to provide a device that does not cause any inconvenience even if an overshoot occurs in the operation of the shift actuator. The present invention has been made in view of the above problems,
The purpose is to reduce the shift operation time by checking the state of the learning value of the shift position (unlearned, initially learned, and normally learned) and performing a learning operation that matches the progress of each learning. It is in. (Means for Solving the Problems) The present invention provides a gear shift gear shift actuator,
A shift stroke sensor for detecting an operation amount of the gear shift actuator, and an electronic control for storing the shift amount as a learning value from a detection value of the shift stroke sensor and driving the gear shift actuator at the next shift using the learning value as a shift target value. Storage means for storing a fixed initial value corresponding to each gear position and rewriting and storing the previous shift amount of each gear position as a normal learning value, When the previous shift position learning state is unlearned, the fixed initial value is selected, and the previous shift position learning state is set to the initial shift target value S0 at the deepest position given for each gear cut. In the learning state, the normal learning value is selected, and when the destruction of the normal learning value is detected, the fixed initial value is set. Selecting means for selecting, and when updating the learning value of each gear position, comparing the previous learning value with the current learning value, and adopting only the learning value when the current learning value is on the shift completion side. Updating means for updating the learning value, and a predetermined value which is smaller than the target value by a stroke due to an overshoot from the time when the gear shift actuator is turned off when the gear shift actuator is driven with the learning value of each gear position as a target value. Shift control for an automatic transmission for a vehicle, comprising: setting means for setting a value smaller by a value α as a correction target value; and driving means for driving each of the gear actuators using the corrected correction target value. apparatus. (Operation) In the present invention, the state of the shift position learning value of the actuator for each gear position detected by the stroke sensor is stored by the storage means, and when the previous shift position learning state is not learned, the initial state is set at the time of shift-in. Judge that learning should be performed, judge that regular learning should be performed when the learning state is the initial learning state, and check the data if the shift position learning value data is destroyed by noise etc. When the learning value destruction is detected, the stored content is set to the unlearned state so that it can be returned to the original control normally, and even if the stored content is destroyed,
Similarly, the stored contents are set in an unlearned state so that normal control can be returned to normal. Further, the drive amount of the shift actuator based on the learning value is set to (learning value-α) in consideration of the overshoot amount, and the learning value is updated to be shifted from the previous learning value. (Embodiment) FIG. 1 shows a gear shift actuator 1 used in the present invention, which can be stopped at three positions.
It comprises a stepped cylinder 1a, a first piston 1b, and a cylindrical second piston 1c fitted to the first piston 1b, and a speed change is provided at an end of a piston rod 1d of the first piston 1b. The gear shift lever 1e of the machine is mounted. The piston rod 1d is in a neutral state where oil pressure acts on both oil chambers 1f and 1g of the stepped cylinder 1a. When oil pressure acts on the oil chamber 1g, the first piston 1b moves downward with the second piston 1c. Then, the gear is shifted to the gears R, 2nd and 4th gears shown in the schematic diagram. When hydraulic pressure acts on the oil chamber 1f, the first piston 1b
Only the gear moves upward, and the gears are shifted to the gears L, 3 and 5 speeds. Numerals 2 and 3 denote magnetic valves that connect the oil chambers 1g and 1f of the actuator to the hydraulic pressure source 4 or the tank 5. Reference numeral 6 denotes a gear shift fork shaft having a gear shift block 7 engaged with the end of the gear shift lever 1e, to which a gear shift fork 8 is attached. 9 is a sleeve that engages the gear shift fork end, 10 is an insert,
11 is a block ring for synchro moved by the insert 10, and 12 is a gear. Reference numeral 13 denotes a detent mechanism having a spring 13a and a steel ball 13b. The gear shift fork shaft 6 is provided with three positioning grooves 13c corresponding to, for example, L, neutral, and R gear shift positions, and one of these grooves 13c is provided with a steel spring held by a spring 13a. The ball 13b is pressed against and holds the gear shift stroke position of the gear shift fork shaft 6 at a constant pressure. 14 is a stroke sensor for detecting the stroke position of the piston rod 1d, 15 is an electronic control unit, which has a computer configuration including a central processing unit, an input / output device, various memories, etc. Based on various signals such as engine speed, vehicle speed, select lever, etc., the gear shift magnetic valves 2, 3 are selectively operated to shift the gear shift actuator to any of the three positions. is there. According to the present invention, a parameter G: LRN representing a shift position learning state, that is, three states of unlearned, initially learned, and normally learned is provided in a memory (not shown) of the electronic control device 15, and the parameter G: LRN is set to take a different value in each state. For example, a hexadecimal code "00" in an unlearned state, a hexadecimal code "F0" in an initial learned state, and a hexadecimal code "F" in a normally learned state.
F "is stored. Therefore, the parameter G: LRN is set to the code" 0 ".
When it is "0", it can be determined that the initial learning should be performed at the time of shift-in by referring to this parameter G: LRN. Similarly, when the parameter G: LRN is a code "F0" If there is, it is possible to determine that normal learning should be performed next, and if the data of the shift position learning value is destroyed due to noise or the like, the learning value destruction is detected by the data destruction detection means. At this point, that is, when the parameter G: LRN is detected using the check data, the parameter G: LRN is set to an unlearned state, that is, by resetting the code to “00”, so that the original control can be normally performed. When the parameter G: LRN itself is destroyed, it can be restored to normal by resetting the parameter G: LRN. The operation of the shift control device for an automatic transmission for a vehicle having the above configuration will be described with reference to the flowchart of Fig. 2. In the process shown in this flowchart, a shift command is issued. When a shift command is issued, first, the learning state G: LRN up to the previous time is checked (step ST1) .This check uses predetermined check data, and compares this check data with the parameter G: LRN. As a result of the check, it is determined whether or not the learning state G: LRN value is destroyed and becomes an abnormal value (step ST2) .If the value is abnormal, the shift learning value is invalid. Judge and reset the parameter G: LRN to an unlearned state (step ST3) If the parameter G: LRN is normal in step ST2 or if the parameter G: LRN is unlearned in step ST3 In step proceeds to ST4. Step ST4 to check the shift learning value when already made the shift learning, in determining whether the shift learning is already made. Step ST4, the parameters
When G: LRN is the code “00”, the shift learning value is not checked since the shift learning value is not checked, and the process proceeds to step ST8 to perform the initial learning. If the parameter G: LRN is not the code "00", it is the code "F0" or the code "FF", the process proceeds to Step ST5, and the shift learning value is checked (Step ST5). This check is made by comparing the check data with the shift learning value. If the learning value is abnormal, the process proceeds to step ST7, in which the parameter G: LRN is reset to an unlearned state (step ST7). If the learning value is normal or if the parameter G: LRN is reset in step ST7, the process proceeds to step ST8. In step ST8, the learning state of the gear is determined by the parameter G: LRN. When the parameter G: LRN is the code “00”, the process proceeds to step ST9 to shift to the initial learning.
In the initial shift in step ST9, the shift is performed toward the deepest shift position in design and manufacturing provided for each gear position, that is, the initial shift target value S0. When the shift instruction is carried out, always Nde reading the shift stroke, even if the elapsed time T ₀ sec determines whether reaches the target value S0 (step ST10). If the target value S0 has not been reached, the process proceeds to step ST11 to determine the number of shifts. If the number of shifts is n or more, the process proceeds to step ST12 where the shift position is S _L
To determine whether the range of to S _H. When the target value S0 has been reached in step ST10 or the shift position is in the range of S _{L to} S _H in step ST12, the operation proceeds to step ST15 in order to release the actuator and allow the detent mechanism to operate freely due to gear vibration and the like. move on. If the number of shifts is less than n times in step ST11 or if the shift position is not in the range of S _{L to} S _H in step ST12, the process proceeds to step ST13 to shift out and shift again to shift again.
Return to ST9. In step ST8, the parameter G: LRN is set to the code “0”.
If it is not "0", since the learning has been completed, the shift target value S is shifted by α (S-α) (step ST14).
This shift of α is performed in order to prevent the gear from being damaged due to the overstroke in the operation in anticipation of the overshoot from the time when the shift M / V is turned off. In the next step ST15, the shift target value S Shift M / V in front α
Turn off. As described above, the shift M / V is turned off immediately after passing the target value S-α, and the detent mechanism can be freely moved by gear vibration or the like as the actuator is free. Thereafter, the process proceeds to step ST16 in order to read the shift stroke each time the learning timing arrives under the shift learning condition (a condition in which the shift position is considered to be stable such as when the clutch is engaged). In step ST16, the shift position learning condition is checked. The learning condition is checked in step ST17, and if the learning condition is not satisfied, the process proceeds to step ST31 without performing any learning. When the learning condition is satisfied, the learning timing is checked in the next step ST18. When the learning timing is not satisfied, the process proceeds to step ST31 without performing any learning, and when the learning timing is satisfied, the shift stroke is read in step ST19. In step ST20, whether or not the shift stroke read in step ST19 is within a normal stroke range is compared with a preset range. If the comparison indicates an abnormal value, the process proceeds to step ST31 without learning as a reading error. When the value is a normal value, the process proceeds to step ST21 for learning. In step ST21, it is determined whether or not the parameter G: LRN is a code “00” or a code “00”. As a result of the judgment,
When the code is “00”, the state is in the unlearned state, and the process proceeds to step ST22 to perform the initial learning. If the code is not "00", it has already been learned and the codes are "F0" and "FF", so that the process proceeds to step ST26 to perform normal shift learning. In the initial learning, first, the shift stroke is substituted into the learning value S as an initial learning value (step ST22). Next, the learning value destruction detection check data is set according to the learning value to which the data is substituted (step ST23). Further, in step ST24, the parameter G: LRN is set to a code "F0" to indicate that the initial learning has been completed. In step ST25, the parameter G: LR corresponding to the code “F0” of the parameter G: LRN for detecting content destruction.
Create N check data and finish initial learning Step S
Proceed to T31. In step ST21, when the learning has already been performed and the parameter G: LRN is the code “F0” or “FF”, in step ST26 the past learning value is compared with the current shift stroke (the previous learning value and When the current learning value is compared with the current learning value, the time when the current learning value is on the shift completion side is defined as shift shift.) It is compared whether or not the current shift stroke is closer to the shift than the past learning value. As a result of the comparison, if the vehicle is not close to the shift, the process proceeds to step ST31 without learning the shift position. If the vehicle is nearing a shift, the process proceeds to step ST27 to perform shift position learning. In step ST27, a shift learning value update process is performed, and in step ST28, check data is created for detecting a learning value destruction according to the updated learning value. Next, in step ST29, the parameter G: LRN is set to a normally learned state, and the parameter G: L
Set the code “FF” to RN. Further, in step ST30,
Parameter G: LRN check data is created for content destruction detection according to the value of parameter G: LRN, and the process proceeds to step ST31. In step ST31, it is determined whether or not to continue the shift position. That is, when the shift-in state is maintained and the shift position is continued, the process returns to step ST16, and repeats the processing of steps ST16 to ST30. To prevent. If it is not the shift-in state in step ST31, the shift position is not continued, and the process ends. (Effects of the Invention) As described in detail above, the present invention adjusts the shift amount by learning without adjusting the shift amount for each individual vehicle in advance, and adjusts the shift amount due to aging. By always updating the normal learning value of the shift amount, the optimum shift amount can always be obtained, and when not learning, the shift actuator is driven by the initial value of the shift amount which is held in advance, the initial learning is performed, and the next After that, the shift actuator is driven based on the initial learning value, regular learning is performed, and the regular learning is repeatedly updated at predetermined intervals, and the driving amount of the shift actuator based on the learning value is an overshoot amount. (Learning value-α) in anticipation and the learning value is updated by adopting a shift shift from the previous learning value so that the optimal gear shift is achieved. The gears by further overshoot can row, without burdening the shift actuator, it is possible to always control the shift actuator to the optimal shift amount.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a system configuration diagram of a gear shift mechanism for explaining a gear shift control device of a vehicle automatic transmission according to the present invention;
FIG. 2 is a flowchart showing a flow of processing of the shift control device of the present invention. DESCRIPTION OF SYMBOLS 1 ... Gear shift actuator, 2,3 ... Magnetic valve, 4 ... Hydraulic source, 5 ... Tank, 6 ... Gear shift fork shaft, 7 ... Gear shift block, 8 ... Gear shift fork, 9 ... Sleeve, 10 ... Insert, 11 ... Block ring , 12 ... gear, 13 ... detention mechanism, 14 ... stroke sensor, 15 ... electronic control unit.

────────────────────────────────────────────────── (5) References JP-A-58-180327 (JP, A) JP-A-60-196452 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) F16H 61/00-61 / 14,61 / 28 F16H 59/00-59/10 B60K 41/04

Claims (1)

(57) [Claims] A gear shift actuator, a shift stroke sensor for detecting an operation amount of the gear shift actuator, and a shift amount stored as a learning value from a detection value of the shift stroke sensor, and the learned value is used as a shift target value at the next shift to drive the gear shift actuator. A shift control device for a vehicle automatic transmission having an electronic control device for performing the following operations: storing a fixed initial value corresponding to each gear position, and rewriting and storing a previous shift amount of each gear position as a normal learning value. When the previous shift position learning state is not yet learned, the fixed initial value is selected, and the previous shift position learning state is set to the initial shift target value S0 at the deepest position given for each gear cut.
Selecting means for selecting the regular learning value when the initial learning state is set to, and selecting the fixed initial value when destruction of the regular learning value is detected; and updating the learning value for each gear position. Updating means for comparing the previous learning value with the current learning value and updating the learning value by adopting only the learning value when the current learning value is on the shift completion side; and learning of each gear position. Setting means for setting a value smaller than a target value by a predetermined value α, which is a value smaller than a target value by a stroke due to overshoot from a point in time when the gear shift actuator is turned off, as a correction target value, when driving the gear shift actuator; And a driving means for driving each of the gear actuators using the corrected target value. A shift control device for an automatic transmission for a vehicle, comprising: