JP3199334B2 - Control gain changing device for automotive control unit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control gain changing device for changing a control gain of a control device for an automobile by learning control, and more particularly to a control gain changing device for improving a disadvantage of the learning control.

[0002]

2. Description of the Related Art In a conventional automobile, a control gain of a driving characteristic is set so that a large number of unspecified drivers can obtain a certain degree of satisfaction regardless of where and under what kind of use environment and condition. It is common that it is. However, depending on the preference of each driver, a power mode and a normal mode can be selected, a desired one of a control mode, a hard mode, and a soft mode in an active suspension device can be selected, and a four-wheel steering device can be selected. Only a specific small number of control gains can be selected, such as selecting between a sports mode and a normal mode.

However, in an automobile in which the control gain of the driving characteristic is set for an unspecified number of drivers, or in an automobile in which only a specific control gain can be set via a manual switch, it is difficult to satisfy all drivers. It is impossible at all. Therefore, a learning control vehicle has been proposed in which learning data is created by learning the driving characteristics of a driver, and the control gain of the running characteristics can be changed based on the learning data.

[0004] For example, Japanese Patent Publication No. 3-44029 discloses a steering angular velocity, a steering angle, a yaw rate, and the like during steering.
Lateral acceleration and the like are sampled, the characteristics of the driver's steering are extracted based on the average value within a predetermined time, and learning data is created. Based on the learning data, the front wheels and / or the rear wheels with respect to the steering angle of the steering wheel are generated. There has been proposed a learning control vehicle that performs learning control so as to change the steering angle ratio.

[0005]

[0005] Problems to be Solved by the Invention [0004] By the way, as a road on which a driver drives, there is a road on which ordinary asphalt is paved and the driving state is good, a rough road with severe unevenness and a mountain road with continuous sharp curves. There is a road with a bad driving condition. In the conventional learning control vehicle, when the traveling road travels on these rough roads or mountainous roads, when the friction state of the road is poor, when driving at night when lighting is required, or when driving in traffic congestion Even when the driving condition is poor, the learning data is updated and the learning control is performed each time the vehicle is driven, so that there is a problem that the reliability of the learning control is reduced.

SUMMARY OF THE INVENTION It is an object of the present invention to control a control apparatus for a vehicle which can perform highly reliable learning control by updating learning data only under a general driving condition excluding a special driving condition. It is to provide a gain changing device.

[0007]

According to a first aspect of the present invention, there is provided a control gain changing device for a vehicle control device, wherein the control gain of the vehicle control device is changed using learning data. Vehicle speed detection means, and learning data update means for receiving a vehicle speed signal from the vehicle speed detection means and updating learning data for learning the driver's steering characteristics when the vehicle speed is within a predetermined vehicle speed section, Control gain changing means for changing the control gain by using the learning data updated by the learning data updating means; and road surface μ detection for detecting a frictional state of a traveling road.
Means, receiving a detection signal from the road surface μ detecting means,
In the case of, the learning data updating means
And a first update canceling means for canceling the update of the
is there.

According to a second aspect of the present invention, there is provided a control gain changing device for a vehicle control device, wherein the control gain of the vehicle control device is set to a learning data.
In a control gain changing device that changes using a
Vehicle speed detecting means for detecting the vehicle speed of the vehicle, and vehicle speed detecting means
A predetermined vehicle speed section where the vehicle speed is received and the vehicle speed is set in advance
Learning to learn the driving characteristics of the driver when inside
Learning data updating means for updating learning data;
Control learning using the learning data updated by the data updating means.
Receiving a control gain changing means for changing the down, and night running detecting means for detecting that the time zone to run row is nighttime, the detection signal from the night running detecting means, the learning data updating when driving at night A second update canceling means for canceling the update of the learning data to the means.

[0009]

In the control gain changing device according to the first aspect, the learning data updating means receives a vehicle speed signal from the vehicle speed detecting means and, when the vehicle speed is within a predetermined vehicle speed section set in advance, the learning data updating means. Since the learning data for learning the steering characteristics is updated, the control gain changing means changes the control gain using the learning data updated by the learning data updating means. As described above, the learning data is updated only when the vehicle speed is within a predetermined vehicle speed section set in advance, and learning control is performed based on the learning data. Therefore, the reliability of learning control can be improved .

In addition, the first update canceling unit receives the detection signal from the road surface μ detecting unit and, when the road is low μ, causes the learning data updating unit to stop updating the learning data. The training data is not updated. Therefore, the update of the learning data is stopped on the low μ road, so that the reliability of the learning control can be further improved.

In the control gain changing device according to the second aspect , the learning data updating means includes a vehicle speed signal from the vehicle speed detecting means.
When the vehicle speed is within the predetermined vehicle speed section set in advance
To learn the driving characteristics of the driver
Therefore, the control gain changing means updates the learning data
Control gain using the learning data updated by the
To change. As described above, the vehicle speed is set to a predetermined value.
The learning data is updated only during the vehicle speed
Since learning control is performed based on learning data, the reliability of learning control
Performance can be improved.

In addition, the second update canceling means receives the detection signal from the night running detecting means and stops the learning data updating means from updating the learning data during night driving. Data is not updated. Therefore, the update of the learning data is stopped during the night driving, so that the reliability of the learning control can be further improved.

[0013]

Hereinafter, embodiments of the present invention will be described with reference to the drawings. First, an outline of the control gain changing device of the vehicle control device according to the present embodiment will be described.
When the driving speed (vehicle speed) when driving the car is within a general vehicle speed section of 40 to 80 Km / h, the control gain is changed by the learning control while updating the learning data. When the vehicle travels on a road, a low μ road, or a rough road with severe unevenness, the updating of the learning data is stopped to improve the reliability of the learning data, that is, the reliability of the learning control.

Next, details of the overall configuration of the vehicle, the control system, the control gain change control by learning control, and the learning data update control will be described in order. As shown in FIG. 1, an automobile A includes at least an engine 1, an automatic transmission 2, front and rear wheels 3, a steering wheel 5, an active suspension device 6, a rear wheel steering device 7, a power steering device 8, A wheel brake device 9 and the like are provided.

In the vehicle A, as each control unit included in the vehicle control unit, an engine control unit that controls at least the intake amount, ignition timing, fuel injection amount, and the like of the engine 1 with a set control gain. 10 (EGI)
And the active suspension device 6 for the front wheel 3 and the rear wheel 4
And an active suspension control device 11 (ACS) for controlling the rear wheel steering device 7 for steering the rear wheel 4 with the set control gain.
Wheel steering control device 12 (4WS) and power steering control device 13 for controlling power steering device 8 for assisting steering wheel 5 with a set control gain
(P / S). Note that the respective control target devices controlled by the respective unit control devices 10 to 13 and the control for them are general ones and will not be described. A base control gain is set in each of the control devices 10 to 13 in advance.
The base control gains are respectively changed by the control gain change signal from the CPU, and the control based on the changed control gains is executed.

Next, a control system of the control gain changing device will be described with reference to FIG. Control gain changing device 20
Performs control for setting the control gain of each of the control devices 10 to 13 and learning control. The control device 21 of the control gain changing device 20 includes an input / output interface including a conversion circuit for performing A / D conversion of signals from sensors and a shaping circuit for shaping a waveform, a microcomputer, and the like. The control device 21 is connected to various control devices 10 to 13, a ROM 22 and a RAM 23, sensors 24 to 29, and a headlight switch 30 as shown in the figure.

The sensors include a vehicle speed sensor 24 for detecting the vehicle speed V of the vehicle A, a steering angle sensor 25 for detecting the steering angle θh, a lateral acceleration sensor 26 for detecting the lateral acceleration Gh acting on the vehicle A, and a road surface sensor. Μ sensor 27 for detecting a friction state (low μ, high μ), yaw rate sensor 28 for detecting yaw rate ψv acting on car A, car A
Vertical G sensor 29 for detecting vertical acceleration Gv acting on
Etc. are provided.

The ROM 22 has various control units 10 to 10
A control program for calculating four correction coefficients K (Ke, Ka, Kw, and Kp) for correcting the 13 base control gains and base data V0 and Dθh0 to be described later are stored in advance. Learning data V
There are provided a learning data memory for storing b and Dθhb, a buffer for storing a predetermined number of (current) detection data Vi and θhi, and the like.

Each of the correction coefficients K is K = (1.0 + Δ)
When the control gain is not changed, the correction component Δ is set to Δ = 0, and when the control gain is changed based on the learning control, the correction component Δ is set to the control data described later. By using V, Dθh and base data V0, Dθh0, (V−V0) and (Dθh−Dθh0) are obtained.
The calculation is performed by applying the correction coefficient calculation control to a predetermined calculation expression or map of the control program. A control signal corresponding to the correction coefficient K is transmitted to each of the control devices 10 to 13.
, And the control signal changes the base control gain.

Basically, from the viewpoint of preventing a significant change in the characteristics of the vehicle, the control gain correction coefficient is set to 0.8 to 1.
2, but is not limited to this range. Here, regarding the control gain of the EGI 10, the control gain “small” is in the low fuel consumption direction, the control gain “large” is in the power increasing direction, and the control gain “small” is in the riding comfort increasing direction (soft direction). ), The control gain “large” is the steering stability increasing direction (hard direction), and the control gain “small” is the small turning up direction (negative phase gain increasing direction) with respect to the control gain of 4WS12.
The control gain “large” is a direction in which the steering stability is increased (in-phase gain is increased).
The control gain “small” is a direction in which the steering force is reduced, and the control gain “large” is a direction in which the steering force is increased.

Next, the control gain change control routine will be described with reference to the flowchart of FIG. In the drawing, Si (i = 10, 11,...) Indicates each step. When this control is started when the ignition switch of the vehicle A is turned ON, when the stop flag SF for stopping the learning control is reset, that is, when the learning control is permitted (S10: No), the RAM 2
3, the learning data Vb and Dθhb are read (S1
1) Then, the control data V and Dθh are V = Vb,
Dθh is set to Dθhb (S12), and then the process proceeds to S15.

However, when the stop flag SF is set, that is, when the learning control is prohibited (S
10: Yes), base data V0, D from ROM 22
θh0 is read (S13), and the control data V and Dθh are set to V = V0 and Dθh = Dθh0 (S14), and then the process proceeds to S15. The reason for setting V = V0 and Dθh = Dθh0 is to control each of the control devices 10 to 13 with the base control gain without changing the control gain based on the learning data.

Next, in S15, S12 or S14
As described above, based on the control data V and Dθh and the base data V0 and Dθh0, the correction coefficients K (K
e, Ka, Kw, and Kp) are calculated based on the control program of the correction coefficient calculation control in the ROM 22. Next, S1
In 6, the control signal corresponding to each correction coefficient K is
The data is output to each of the control devices 10 to 13 respectively, and thereafter, in S16
Then, the process proceeds from S10 to S10, and the process from S10 is repeated every minute time.

Next, the routine of the learning data update control will be described with reference to the flowchart of FIG. This learning data update control is started when the ignition switch is ON, and is executed in parallel with the control gain change control. In the drawing, Si (i = 30, 31,...)
Indicates each step. First, various signals (current vehicle speed Vi, current steering angle θhi, signals of other sensors) are read (S30), and then, learning data update determination processing (see FIG. 5) is executed (S31). ).

The learning data update determination processing will be described with reference to FIG. 5. If the current vehicle speed V is 40 to 80K
It is determined whether the vehicle is within a predetermined vehicle speed section of m / h (S4).
0), then, it is determined whether or not the headlight switch 30 is ON, that is, whether or not the vehicle is traveling at night (S41), and then, it is determined whether or not the road surface state is a low μ road (S42). It is determined whether or not the road is a mountain road (S44). S4, provided that the vehicle speed V is within the predetermined vehicle speed section, the headlight switch 30 is not ON, the vehicle is not on a low μ road, the vehicle is not in a bad road condition, and the vehicle is not on a mountain road.
In step 5, the learning control suspension flag SF is reset in order to permit the learning control, and the process proceeds to S32. On the other hand, when all of the above conditions are not satisfied, S4
The process proceeds to S6, where the learning control suspension flag SF is set in order to suspend the updating of the learning data and the learning control, and then proceed to S32.

Next, in step S32 of FIG. 4, the suspension flag SF is determined. If the suspension flag SF is set, the process returns as it is. If the suspension flag SF is reset, the current steering angle θhi and the previous steering angle are used. Then, the current steering angular velocity Dθhi is calculated (S33), and the current data Vi and Dθhi are stored in the buffer of the RAM 23.

Next, in S35, it is determined whether or not a predetermined number of data has been stored in the buffer of the RAM 23.
In the case of o, the process returns as it is, and in the case of Yes, the learning data Vb and Dθhb are read from the learning data memory of the RAM 23 and the average values Vi and Dθhi of a predetermined number of data Vi and Dθhi stored in the buffer of the RAM 23 are calculated. For example, the learning data Vb,
Dθhb is calculated and updated (S36), and the learning data Vb and Dθhb are stored in the learning data memory of the RAM 23 (S37), and thereafter, the process returns to S30 and S30.
Subsequent steps are repeatedly executed at minute intervals.

As described above, under the general driving condition where the vehicle speed V is in the predetermined vehicle speed section (40 to 80 km / h), the update of the learning data and the change of the control gain by the learning control are performed. When the vehicle speed V is out of the predetermined vehicle speed section, or even when the vehicle speed V is within the predetermined vehicle speed section, at the time of running at night, at the time of running on a low μ road, at the time of running on a bad road, or at the time of running on a mountain road Sometimes, because of a special driving situation, the stop flag SF is set and the current data Vi, D
Since θhi is not stored in the buffer of the RAM 23 and the learning data is not updated, the reliability of the learning data, that is, the reliability of the learning control can be further improved.

The above-mentioned predetermined vehicle speed section is merely an example,
The minimum speed and the maximum speed of the vehicle speed section may be slightly changed. In the embodiment, the learning data Vb, Dθ
The example in which the control gain of each of the control devices 10 to 13 is changed using hb has been described, but the learning data Vb,
Dθhb is merely an example, and it goes without saying that the present invention can be similarly applied to a control gain changing device that changes the control gain using learning data other than these, and in the above-described embodiment, EGI10, ACS11,
Although an example in which the control gains of the 4WS12 and the P / S13 are changed has been described, various control gains such as an antilock braking control device, a traction control device, and an air conditioner control device may be used instead of or in addition to these. It is needless to say that the present invention can be similarly applied to a control gain changing device that changes the control gain using learning data.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a drive system, a suspension system, and a steering system of an automobile according to an embodiment.

FIG. 2 is a configuration diagram of a control system of the control gain changing device.

FIG. 3 is a flowchart of a control gain change control routine.

FIG. 4 is a flowchart of a learning data update control routine;

FIG. 5 is a schematic flowchart of a subroutine of a learning data update determination process in FIG. 4;

[Explanation of symbols]

A Automobile 1 Engine 6 Active suspension device 7 Rear wheel steering device 8 Power steering device 9 Braking device 10 Engine control device (EGI) 11 Active suspension control device (AC
S) 12 Four-wheel steering control device (4WS) 13 Power steering control device (P / S) 20 Control gain changing device of control device of automobile 21 Control device 22 ROM 23 RAM

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Kiyoshi Sakamoto 3-1 Shinchi, Fuchu-cho, Aki-gun, Hiroshima Mazda Co., Ltd. (72) Inventor Tetsuya Tachiya 3-1 Shinchi, Fuchu-cho, Aki-gun, Hiroshima Mazda (56) References JP-A-2-3755 (JP, A) JP-A-4-22737 (JP, A) JP-A-4-244407 (JP, A) JP-A-63-149230 (JP, A A) (58) Field surveyed (Int. Cl. ⁷ , DB name) B60R 16/02 660 B60K 28/02 F02D 45/00 340 G05B 13/02

Claims (1)

(57) [Claims] 1. A control gain changing device for changing a control gain of a control device of an automobile using learning data, comprising: vehicle speed detecting means for detecting a vehicle speed of the automobile; A learning data updating unit that updates learning data for learning a driver's steering characteristics when the vehicle speed is within a set predetermined vehicle speed section; and the control gain is adjusted by using the learning data updated by the learning data updating unit. Control gain changing means for changing, and road surface μ detecting means for detecting a friction state of a traveling road
When receives the detection signal from the road surface μ detecting means, preparative low μ road
At this time, the learning data updating means updates the learning data.
A control gain changing device for a control device for a vehicle , comprising a first update canceling means for canceling a new one . 2. A learning de control gain of a motor vehicle control device
In the control gain changing device for changing using chromatography data, and vehicle speed detecting means for detecting a vehicle speed of the vehicle, receiving a vehicle speed signal from said vehicle speed detecting means, the set vehicle speed in advance
The driver's steering characteristics are within the specified speed range.
Learning data to update learning data for learning
Using the new data and the learning data updated by the learning data updating device.
A control gain changing means for changing the control gain Te, and night running detecting means for detecting that the time zone to run row is nighttime, receives a detection signal from the night driving detecting unit, when driving at night are the learning data and the second update cancellation means for stopping the updating of the learning data to update unit, a control gain changing device of automobile control system you characterized in that a.