JPH07167750A - Learning method of driving of automatic operation device - Google Patents

Abstract

PURPOSE:To effect the operation of a pedal etc., surely by an actuator by compensating the amount of control based on the result which is confirmed before vehicle speed reaches a stable level after an automatic operation start is initiated. CONSTITUTION:A robot 1 is provided with an actuator 4 etc., for an accelerator pedal 8 at a body 3 which is mounted to a substrate 2. The actuator 4 is constituted by providing a slide arm 4a which is moved in reciprocating motion by a servo motor at a support member 4b and a stroke sensor and a contact sensor 4d are mounted. Then, by the time when vehicle speed reaches a stable level after automatic operation is initiated, the amount of control for pedal operation required for automatic operation is confirmed and the amount of control for pedal operation is compensated based on the confirmation result, thus detecting and correcting deviation and error in the position relationship between an automatic operation device and the pedal due to the positional deviation of a device which is operated automatically and the wear of the pedal every time when automatic operation is started. Therefore, a desired driving performance can be surely performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a driving simulation of an actual vehicle in which a driving wheel is mounted on a roller of a chassis dynamometer to drive the vehicle, and a dynamic driving performance test of the vehicle is performed indoors. The present invention relates to a running learning method for an automatic driving device that automatically drives a car or an engine.

[0002]

2. Description of the Related Art Conventionally, a dynamic driving test of an automobile has been performed by a chassis dynamometer to perform a simulated driving operation of an actual vehicle. Recently, the simulated driving operation of the actual vehicle includes hydraulic pressure, air pressure or DC. Multiple actuators are operated individually by a motor, etc., and this actuator operates the accelerator pedal, brake pedal,
BACKGROUND ART Automatic vehicle-driving robots capable of stepping on a clutch pedal or switching a shift lever have been used.

In order to carry out a program operation of an automobile to be tested based on a predetermined traveling pattern by such an automatic vehicle driving robot, actuators for operating the pedals and shift levers are firmly attached to the vehicle body. After that, the operating state of each pedal or shift lever is created manually, and the vehicle automatic driving robot is made to store the operation amount and coordinates for operating each pedal or shift lever. It was controlling the actuator.

[0004]

By the way, the above-mentioned automatic vehicle driving robot is provided on the driver's seat of the vehicle by an appropriate supporting member by facing each actuator so as to correspond to each pedal or shift lever of the vehicle. However, during repeated test runs, displacement of the actuator and disturbance of the posture occur due to vibration of the vehicle body, scratches when operating the shift lever, and excessive depression of the pedals. May deviate from its original position.

As described above, when the actuator is displaced, a difference occurs between the stored coordinates and the actual coordinates, the operation amount for operating the actuator becomes improper, and proper acceleration, deceleration, and clutch are performed. The control cannot be performed properly, such as the inability to turn on and off the power well, and as a result, the operation of the program will be seriously hindered.

On the other hand, in the past, the positional relationship between the automatic vehicle driving robot and the pedals and shift levers of the vehicle was determined separately from the program driving. However, due to the wear of the pedals, etc. It was difficult to completely correct the positional relationship.

The present invention has been made in consideration of the above-mentioned matters. Even if the pedals are worn, the positional relationship between the automatic driving device such as the automatic vehicle driving robot and each pedal of the vehicle is accurately grasped, and the position is determined. It is an object of the present invention to provide a running learning method for an automatic driving device, which can correct the deviation of the relationship to cause the actuator to operate the pedal or the like without fail.

[0008]

In order to achieve the above object, the present invention provides a method for automatically driving a vehicle by an automatic driving device equipped with actuators for operating pedals of an accelerator, a brake and a clutch of the vehicle, respectively. After starting automatic operation, before the vehicle speed rises,
Check the control amount for pedal operation required for automatic operation,
The control amount for pedal operation is corrected based on the confirmation result.

[0009]

[Operation] The engine of the car is started for program operation. Before the vehicle speed rises, that is, when the vehicle speed is zero, first, the brake pedal actuator is operated to obtain the full stroke of the brake, and this is compared with the reference full stroke set during learning. , When there is a difference between the two, depending on the brake pedal force,
Rewrite the brake characteristics. Then, the slip point of the clutch is obtained, and the value of the slip point is rewritten. Furthermore, the dominant starting point of the accelerator is obtained, and the value of the dominant starting point is rewritten.

In this way, every time the automatic operation is performed, the pedal operation amount is confirmed immediately after the start of the automatic operation, and the control amount for pedal operation is corrected based on the confirmation result. The car can be driven automatically.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, FIGS. 3 and 4 show an example of a vehicle automatic driving robot (hereinafter, simply referred to as a robot) as an automatic driving device used in the method of the present invention. In these figures, 1 is a robot, The main body 3 attached to the base 2 is provided with an accelerator pedal actuator 4, a brake pedal actuator 5, a clutch pedal actuator 6, and a shift lever actuator 7.

Each of the accelerator pedal actuator 4, the brake pedal actuator 5, and the clutch pedal actuator 6 is reciprocally moved by a servomotor (not shown) or the like, and slide arms 4a, 5 are provided.
The support members 4b, 5b and 6b are provided with a and 6a. A stroke sensor 4 such as an encoder for detecting the stroke of the slide arms 4a, 5a, 6a is provided on each of the pedal actuators 4, 5, 6.
c, 5c, 6c are attached.

Further, contact sensors 4d, 5 such as proximity switches for detecting that the slide arms 4a, 5a, 6a respectively come into contact with the accelerator pedal 8, the brake pedal 9, and the clutch pedal 10, respectively.
d and 6d are attached.

The shift lever actuator 7 has an upper end of the shift lever 11 at the tip of a support arm 12 configured to move the shift lever 11 in the X-axis direction, the Y-axis direction, and the Z-axis direction in order to move the shift lever 11. A gripping body 13 having an inverted concave cross-section for gripping the portion is attached and configured.

FIG. 5 schematically shows an example of the chassis dynamometer. In this figure, 14 is an automobile used for driving simulation driving, and its driving wheels 15 are placed on and abutted on the rollers 16. The robot 1 is mounted on the chassis dynamometer 17 in a closed state, and the robot 1 is mounted and fixed on the driver's seat 18. Reference numeral 19 is a detector for detecting the rotation speed of the roller 15. Reference numeral 20 is a control device to which the output signals of the stroke sensors 4c, 5c, 6c, the contact sensors 4d, 5d, 6d, the rotation detector 18 and other necessary data are input and stored. 21
Is a driver for driving the actuators 4 to 7,
It is configured to exchange signals with the control device 20.

Then, the robot 1 having the above structure is made to learn the necessary items as follows. That is, first, the upper body of the shift lever 11 is gripped by the gripping body 13 of the shift lever actuator 7, and the shift lever actuator 7 is operated by the manual remote controller (not shown) according to the teaching screen (not shown). , And learn each shift position such as neutral.

Next, as shown in FIG. 6A, when the accelerator pedal actuator 4 in which the slide arm 4a is housed in the support member 4b is actuated, the contact sensor 4d.
As shown in FIG. 7B, the accelerator pedal 8 is contacted and a signal is output.
It is stored in the reference data memory in the control device 20 as the depression start position of the accelerator pedal actuator 4 with respect to. Then, as shown in FIG. 3C, when the accelerator pedal actuator 4 fully depresses the accelerator pedal 8 and the slide arm 4a becomes unable to slide, the stroke sensor 4c uses the depression start position as a base point. The stroke amount of the accelerator pedal 8 by the accelerator pedal actuator 4 is detected and stored in the reference data memory.

As for the brake pedal actuator 5 and the clutch pedal actuator 6, their contact sensors 5d, 6d and stroke sensor 5c,
The stroke amounts of the actuators 5 and 6 are stored in the reference data memory according to the output signals of 6c. Further, the brake pedal actuator 5 gradually depresses the brake pedal 9 from the stepping start position to the maximum stepping position, and the brake signal is output by the stroke sensor 5c and the output signals of the operating force (operating current) of the brake pedal actuator 5. The relationship (brake characteristic) between the stepping stroke amount of the pedal actuator 5 and the operating force is stored in the memory of the control device 20.

Further, the engine is started, the clutch pedal 10 is depressed by the clutch pedal actuator 6 to the maximum depressed position, and the shift lever actuator 7 is depressed.
The shift lever 11 is moved from the neutral position to the 1st position with, and then the slide arm 7a of the shift lever actuator 7 is gradually retracted to move the clutch pedal 1
Return 0. Then, when the clutch is connected, the roller 15 of the chassis dynamometer 16 starts rotating,
The rotation detector 19 detects it and outputs a signal,
According to the signal, the clutch pedal actuator 6
The clutch connection start position (slip point) of is detected by the stroke sensor 6c and stored in the reference data memory.

The shift lever actuator 7
After making the shift lever 11 neutral with, gradually depress the accelerator pedal 8 with the accelerator pedal actuator 4 from the depression start position to the maximum depression position,
The relationship between the stroke amount of the accelerator pedal actuator 4 and the engine speed is stored in a memory (not shown) of the control device 20 by the output signals of the stroke sensor 4c and the engine speed detecting means (not shown) during that time. Remember.

As described above, the stroke sensor 4c, 5c, 6c and the contact sensor 4 are operated by sequentially operating the accelerator pedal actuator 4, the brake pedal actuator 5, and the clutch pedal actuator 6, respectively.
The learning required for operating the accelerator pedal 8, the brake pedal 9, and the clutch pedal 10 is performed by the output signals of the sensors such as d, 5d, 6d, the rotation detector 18, the engine speed detecting means, and the like.

As described above, after the robot 1 initially learns necessary items, the robot 1 drives the vehicle 14
According to the present invention, the pedal operation amount required for automatic operation is confirmed after the automatic driving is started and before the vehicle speed rises, and the control amount for pedal operation is based on the confirmation result. I am trying to correct. Therefore, the position learning that is performed during the execution of the program from immediately after the engine starts until the vehicle speed rises will be described with reference to FIGS. 1 and 2.

First, the brake pedal actuator 5
Is operated to depress the brake pedal 9 to obtain its full stroke (step S1).

When there is a difference between the full stroke and the full stroke in the above-mentioned initial learning,
In order to absorb the change, the brake characteristics are rewritten by the operating force of the brake or the full scale of the brake is rewritten (step S2).

Next, the clutch pedal actuator 6
Is operated to depress the clutch pedal 10 to obtain the slip point of the clutch (step S3).

When there is a difference between the slip point and the slip point in the initial learning, the value of the slip point is rewritten to absorb the difference (step S).
4).

Then, the accelerator pedal actuator 4 is operated to depress the accelerator pedal 4 to obtain the accelerator starting point (step S5).

When there is a difference between the dominant starting point and the dominant starting point in the initial learning, the value of the dominant starting point is rewritten in order to absorb the difference (step S).
6).

By the way, when the actual vehicle running simulation operation is performed, the time width from the start of the engine (time T ₁ in FIG. 2 to the start of running the vehicle 14 (time T ₃ in FIG. 2) is specified to be, for example, 20 seconds. Therefore, the position learning from step S1 to step S6 must be performed from the engine start time T ₁ to the vehicle speed rising time T ₃ , and in particular, the position learning is performed at the vehicle speed rising time T 1. _It has to be completed at time T ₂ which is 5 seconds before _3. That is, the position learning is 1 after engine start.
Programmed to occur in less than 5 seconds.

When the position learning is performed, there is no difference in the full stroke of the brake pedal, the slip point of the clutch, and the starting point of the accelerator hand from the set value (reference set value) set in the initial learning. In this case, it goes without saying that the operations of steps S2, S4 and S6 may be skipped.

Five seconds after the position learning is completed as described above, the vehicle speed rises and a predetermined program run is performed.

The present invention is not limited to the above-mentioned embodiments, but can be applied to the case where an engine performance test is carried out.

[0033]

As described above, according to the present invention, an automatic driving device is operated when a device for automatically driving an automobile or the like is misaligned or various pedals provided in the automobile or the like are worn. Even if there is a deviation or deviation in the positional relationship between the pedal and the pedal, it can be detected and corrected each time automatic driving starts, so that the test object such as an automobile can be controlled well and the desired running performance can be obtained. The test can be reliably performed.

[Brief description of drawings]

FIG. 1 is a flowchart showing an example of the method of the present invention.

FIG. 2 is a diagram showing an example of a traveling pattern in a traveling performance test.

FIG. 3 is a front view showing an example of a robot used in the method of the present invention.

FIG. 4 is an enlarged plan view of the robot.

FIG. 5 is a diagram schematically showing an example of a chassis dynamometer.

FIG. 6 is a diagram for explaining how the actuator depresses a pedal.

[Explanation of symbols]

1 ... Robot, 4, 5, 6 ... Pedal actuator,
8 ... accelerator pedal, 9 ... brake pedal, 10 ... clutch pedal, 14 ... automobile.

Claims (1)

[Claims] 1. A method of automatically driving a vehicle by an automatic driving device including actuators for operating pedals of an accelerator, a brake and a clutch of the vehicle, wherein the automatic driving is performed after the automatic driving is started and before the vehicle speed rises. A method for learning during running of an automatic driving device, characterized in that a control amount required for pedal operation is confirmed, and the control amount for pedal operation is corrected based on the confirmation result.