JPH06321119A - Failure diagnostics and processing device of steering device for vehicle - Google Patents

Abstract

PURPOSE:To detect a failure of a steering device securely and take an appropriate measure for the steering device in accordance with level of failure securely. CONSTITUTION:It is a failure diagnostics and processing device FS of a steering device SS for vehicle which has a plural number of steering condition detectors S1 and a controller S3 which controls an actuator S2 in accordance with steering condition. It has a failure diagnostics device M2 in which connection coefficient between each neuron is studied and determined by the back propagation method by a teacher signal which is set in accordance with level of failure which corresponds to the case when abnormality occurs in each input value by specifying an output value of neural network which uses a value detected by a steering condition detector as an input value as a failure judgment level value and a failure countermeasure device M2 which selects and executes the predetermined failure countermeasure mode from among a plural number of failure countermeasure modes in accordance with failure judgment level value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a steering device such as an electric power steering device for a vehicle such as an automobile, and more particularly to a failure diagnosis processing device for the steering device.

[0002]

2. Description of the Related Art An electric power steering apparatus for a vehicle such as an automobile has heretofore generally been a motor for driving a steering linkage member such as a rack bar via a clutch and a steering gear box, and a controller for controlling the rotation of the motor. The control device is configured to generate a required steering assist force by controlling the motor based on the detection results of various sensors such as a torque sensor.

As one of such electric power steering devices, for example, as disclosed in Japanese Patent Application Laid-Open No. 63-175262, when the steering torque exceeds a predetermined value, it is determined that an abnormality has occurred in the power steering device. An electric power steering device configured to stop the supply of electric current to the power assist device to stop the power assist has been conventionally known. According to this type of power steering device, when an abnormality occurs in the torque sensor or the like, it is possible to reliably prevent improper control of the motor and improper power assist by stopping energization of the motor. It

[0004]

However, in the electric power steering apparatus, there are many places other than the torque sensor that may cause abnormalities. Therefore, all abnormalities are detected and judged only by the detected value of the torque sensor. You cannot do it. In addition, it is necessary to build an abnormality detection logic for all points where there is a risk of an abnormality occurring when trying to detect all abnormalities, and the degree of abnormality (fault level) at each point.
Since not only one but also the content of countermeasures depending on the degree of abnormality, the logic for dealing with abnormality detection becomes enormous and the power steering device becomes expensive. Furthermore, no matter how carefully the abnormality detection coping logic is created, if an unexpected abnormality occurs, the occurrence of the abnormality is not detected, and the power steering device continues to be unable to perform appropriate power assist. .

The above-mentioned various problems are not limited to the electric power steering device, but are also present in the four-wheel steering device for steering the rear wheels and controlling the steering angle of the rear wheels in accordance with the steering angle of the front wheels and the vehicle speed. Occurs similarly.

In view of the above-mentioned problems in the steering apparatus such as the conventional electric power steering apparatus, the present invention can reliably detect the occurrence of any failure in the steering apparatus. It is an object of the present invention to provide an inexpensive fault diagnosis processing device which is improved so that appropriate measures can be taken for a steering device according to the level of a fault.

[0007]

According to the present invention, a plurality of steering state detecting means S1 for detecting a steering state and a plurality of detected steering states are provided according to the present invention. The output value of the neural network having the detected value by the steering state detection means as the input value is used as the failure diagnosis processing device FS of the vehicle steering system SS having the control means S3 for controlling the actuator S2 for auxiliary steering according to Is a failure determination level value, and a failure diagnosis means M1 in which the coupling coefficient between neurons is learned and determined by the backpropagation method by a teacher signal set according to the failure level corresponding to the occurrence of an abnormality in each input value. And a failure handling means M2 for selecting and executing a predetermined failure handling mode from a plurality of failure handling modes according to the failure determination level value. It is achieved by the fault diagnosis process unit of the vehicle steering apparatus.

[0008]

According to the above configuration, the failure diagnosis processing device F
S has a failure diagnosing means M1 and a failure coping means M2, and the failure diagnosing means uses an output value of a neural network whose input value is a value detected by the steering state detecting means S1 as a failure determination level value to cause an abnormality in each input value. In this case, the coupling coefficient between the neurons is learned and determined by the back-propagation method by the teacher signal set according to the failure level corresponding to the case, and the failure coping means has a plurality of failure coping modes according to the failure judgment level value. It is configured to select and execute a more predetermined failure handling mode.

Therefore, when some failure occurs in the steering device and an abnormality occurs in the detected value of the related steering state detecting means, the output value of the neural network, that is, the failure determination level value of the failure diagnosing means changes according to the level of the failure. A predetermined failure-handling mode is selected and executed by the failure-handling means according to the failure determination level value, whereby a failure of the steering device is reliably detected, and a predetermined action is surely taken according to the failure level. To be

[0010]

Embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

FIG. 2 is a schematic block diagram showing a first embodiment of a failure diagnosis processing apparatus according to the present invention applied to an electric power steering apparatus, and FIG. 3 is an auxiliary steering electronic control apparatus shown in FIG. It is a block diagram shown.

In FIG. 2, reference numeral 10 denotes a steering wheel. The steering wheel 10 is a rack bar 1 as a steering linkage member via a steering shaft 12 and a steering gear box 14.
6 is driven. A power unit 20 is drivingly connected to the steering shaft 12 by a gear reduction mechanism 18. The power unit 20 is a motor 22
And an electromagnetic clutch 24 for selectively drivingly connecting the gear reduction mechanism 18 and the motor 22.

In the illustrated embodiment, the steering shaft 12 is provided with a steering angle sensor 26 for detecting the steering angle θ and a torque sensor 28 for detecting the steering torque T, and the outputs of these sensors are auxiliary. It is supplied to the steering electronic control unit 30 and the failure diagnosis processing electronic control unit 32. Further, the electronic control devices 30 and 32 also receive a signal indicating the rotation speeds Vfl and Vfr of the left and right front wheels detected by the wheel speed sensors 34 and 35 and a signal indicating the rotation angle φ of the motor 22 detected by the rotation angle sensor 36. It is supposed to be done.

As shown in detail in FIG. 3, the electronic control unit 30 includes a microcomputer 38, which is a central processing unit (CPU) 40.
, Read only memory (ROM) 42, random access memory (RAM) 44, and input port device 46
And an output port device 48, which are connected to each other by a bidirectional common bus 50.

The input port device 46 includes a steering angle sensor 26.
A signal indicating the detection value detected by the above is input. The input port device 46 appropriately processes the signal input thereto, and outputs the processed signal to the CPU and the RAM 44 according to the instruction of the CPU 40 based on the control program stored in the ROM 42. ROM 42 contains various maps shown in FIG.
The control program shown in FIG.

The CPU 40 is adapted to perform various operations and signal processing as will be described later based on the signal flow diagram shown in FIG. 4 and the control program shown in FIG. The output port device 48 outputs a control signal to the motor 22 via the drive circuit 52 according to an instruction from the CPU 40, and also outputs a control signal to the electromagnetic clutch 24 via the drive circuit 54, and when some failure occurs in the power steering device. The alarm lamp 56 is activated.

As shown in detail in FIG. 4, in the electronic control unit 30 of the illustrated embodiment, the basic assist amount Tab is calculated from the map 58 based on the steering torque T detected by the torque sensor 28. , The vehicle speed V is calculated by the vehicle speed calculator 60 from the rotational speeds Vfl and Vfr of the left and right front wheels detected by the wheel speed sensors 34 and 35, and the coefficient K1 based on the vehicle speed is calculated from the map 62 based on the vehicle speed. Then, the product Tab · K1 of the basic assist amount Tab and the coefficient K1 is calculated by the multiplier 64. Also, the differentiator 66 calculates the differential value Td of the steering torque T, and the coefficient multiplier 68
Derivative value Td of steering torque and coefficient K4 (positive constant)
And the product Td.K4 is calculated, and the product Tab
The sum of K1 and the product Td.T4 is calculated, and the signal indicating the calculation result is output to the plus terminal of the adder 72.

The steering angle velocity φd is calculated by differentiating the rotation angle φ of the motor detected by the rotation angle sensor 36 by the differentiator 74, and the damping amount Tm is calculated from the map 76 based on the steering angular velocity, and the damping amount is calculated. T
The signal indicating m is output to the multiplier 78. Also multiplier 7
A signal indicating the coefficient K2 calculated by the map 80 on the basis of the vehicle speed V calculated by the vehicle speed calculator 60 is input to 8, and the signal indicating the product φd · K2 calculated by the multiplier 78 is the minus of the adder 72. It is output to the terminal.

The steering angle θ detected by the steering angle sensor 26 is corrected by the neutral position detector 82 based on the vehicle speed V from the vehicle speed calculator 60 and the steering torque T from the torque sensor 28 so that the neutral position becomes zero. The steering wheel return amount Tr is calculated from the map 84 based on the corrected steering angle, and a signal indicating the return amount Tr is output to the multiplier 86. The multiplier 86 includes a vehicle speed calculator 60.
Coefficient K3 calculated from map 88 based on vehicle speed V
Is also input, whereby the multiplier 86 calculates the product Tr.multidot.K3 of the return amount Tr and the coefficient K3, and the signal indicating the calculation result is output to the minus terminal of the adder 72.

The output of the adder 72, that is, the signal indicating the assist amount Ta is output to the multiplier 90, and the coefficient K / A is input to the multiplier 90 from the coefficient unit 92. The numerator K of the coefficient K / A is increased / decreased by the coefficient unit 92 based on the signal from the failure diagnosis processing electronic control unit 32 as described later. Multiplier 90
, That is, the signal indicating the assist amount Taa corrected by the coefficient K / A is output to the plus terminal of the adder 94. A signal indicating the drive current supplied to the motor 22 from the drive circuit 52 and detected by the current sensor 37 is input to the minus terminal of the adder 94, so that the drive current supplied to the motor 22 becomes the assist amount Taa. Feedback controlled.

Next, the power assist control in the illustrated embodiment will be described with reference to the flow chart shown in FIG. The control by the electronic control unit 30 is started by closing an ignition switch not shown in FIG. Further, in the flowchart shown in FIG. 5, the flag Fs is a failure coping mode instruction flag set by the electronic control unit for failure diagnosis processing described later,
Reference numerals 1, 2, and 3 indicate that the power steering device has a failure with a small influence degree, a failure with a medium influence degree, and a failure with a high influence degree, respectively.

First, in step 10, the control signal is output to the electromagnetic clutch 24 via the drive circuit 54 to connect the clutch, and in step 20, the detected value detected by the steering angle sensor 26 or the like is detected. The signal indicating is read. In step 30, the steering assist amount Ta as described above according to the signal flow diagram shown in FIG.
Is calculated.

In step 40, it is determined whether the flag Fs is 0, that is, whether the power steering device is normal, and it is determined that Fs = 0 is not satisfied. Sometimes the routine proceeds to step 60, and when it is determined that Fs = 0, the numerator K of the coefficient K / A in the correction calculation of the assist amount Ta in step 180 is set to A in step 50. After step 180
Go to.

In step 60, it is judged whether the flag Fs is 1, that is, whether the power steering device has a malfunction of a small influence degree or not, and Fs = 1.
If it is determined that it is not, the process proceeds to step 110, and if it is determined that Fs = 1, the control signal is output to the alarm lamp 56 and the alarm lamp is activated in step 70. A warning is given to the driver of the vehicle that a minor impact has occurred, and step 8
At 0, it is determined whether or not the steering assist amount before correction Ta is positive. If it is determined that Ta> 0, the assist amount Ta is Tac (positive) at step 90. Constant) and the numerator K of the coefficient K / A is set to A, the routine proceeds to step 180, and when it is judged that Ta> 0 is not satisfied, the assist amount Ta is − in step 100. After Tac is set and the coefficient numerator K is set to A, the process proceeds to step 180.

In step 110, it is determined whether the flag Fs is 2, that is, whether the power steering device has a failure in the degree of influence, and Fs =
When it is determined that the value is 3, the process proceeds to step 160, and when it is determined that Fs = 2, the warning lamp 56 is activated in step 120 to affect the driver of the vehicle. An alarm is issued to the effect that an internal failure has occurred, and in step 130, the numerator K of the coefficient K / A is calculated.
Is decremented by 1.

In step 140, the coefficient numerator K is 0.
If it is determined that K = 0, the process proceeds to step 180. If it is determined that K = 0, the electromagnetic clutch 24 is determined in step 150. The clutch is released by stopping the output of the control signal to the power steering device is switched to the manual steering device,
Then, the control according to the flowchart shown in FIG. 5 is finished.

In step 160, the alarm lamp 56 is activated to give an alarm to the driver of the vehicle that a failure of great influence has occurred, and in step 170, the motor 22 is energized. Is stopped and the output of the control signal to the electromagnetic clutch 24 is stopped, the clutch is released, whereby the power steering device is switched to the manual steering device, and then the control according to the flowchart shown in FIG. 5 is performed. To finish.

In step 180, the corrected steering assist amount Taa is calculated by correcting the steering assist amount Ta according to the following equation 1, and in step 190, it corresponds to the corrected assist amount Taa. A voltage control signal is output to the motor 22 via the drive circuit 52, whereby the required power assist is performed, and then step 20 is performed.
Return to.

[Equation 1] Taa = (K / A) · Ta

As shown in detail in FIG. 6, the failure diagnosis processing electronic control unit 32 includes a microcomputer 98, and the microcomputer 98 has the same structure as the microcomputer 38 of the auxiliary steering electronic control unit 30. CPU 100, ROM 102, RAM 104
, An input port device 106, and an output port device 108, which are connected to each other by a bidirectional common bus 110.

The steering angle sensor 26 is provided in the input port device 96.
A signal indicating the detection value detected by the above is input. The input port device 96 appropriately processes the signal input thereto, and outputs the processed signal to the CPU and the RAM 94 according to the instruction of the CPU 100 based on the control program stored in the ROM 92. The ROM 92 stores the neural network 112 shown in FIG. 7 and the control programs shown in FIGS. 9 and 10.

The CPU 100 is a neural network 1
12 and various kinds of operations and signal processing as described later based on the control program shown in FIGS. 9 and 10, and thereby the failure judgment level value Ls of the neural network is calculated, and the failure is detected according to the level value Ls. The coping mode instruction flag Fs is set. The output port device 108 outputs a signal indicating the failure coping mode instruction flag Fs to the microcomputer 38 of the auxiliary steering electronic control unit 30 according to the instruction of the CPU 90.

As shown in FIG. 7, the neural network 112 has an input layer 114, an intermediate layer 116, and an output layer 118. The input layer 114 has a detection value detected by a sensor such as the steering angle sensor 26. And a command signal output from the microcomputer 38 of the electronic control unit 30 for auxiliary steering to the drive circuit 52 are input. The input layer 114 and the intermediate layer 116 have the same number of neurons as the number of signals input to the input layer, and the output layer 118 has one neuron.

Next, the failure level determination routine and the failure coping mode selection routine in the illustrated first embodiment will be described with reference to the flow charts shown in FIGS. 9 and 10, respectively.

In step 210 of the flow chart shown in FIG. 9, a signal indicating the steering angle θ detected by the steering angle sensor 26 is read, and in step 220 the steering angle sensor 26 detects it. The steering angle θ
s1, the steering torque T detected by the torque sensor 28 is Is2, the rotation speeds Vfl and Vfr detected by the wheel speed sensors 34 and 35 are Is3 and Is4, respectively, and the motor 22 of the motor 22 detected by the rotation angle sensor 36 is s1. The rotation angle φ is Is5, the drive current Im for the motor 22 detected by the current sensor 37 is Is6, the command current Isc output from the output port device 48 of the electronic control unit 30 to the drive circuit 52 is Is7, Msoi and Nsoi (i ＝
1, 2, ... 6, 7) respectively as weighting factors and threshold values for the input signal, and the output value Xsi (i =
1, 2 ... 6, 7) are calculated.

## EQU00002 ## Xsi = f (Msoi.Isi-Nsoi)

In step 230, the output value Ysj (j = 1, 2 ... 6, 7) of the intermediate layer 116 is calculated in accordance with the following expression 3, and in step 240, the failure judgment level is calculated according to the following expression 4. The value Ls is calculated.

## EQU00003 ## Ysj = f (.SIGMA.Mspij.Xsi-Nspj) (i = 1, 2 ... 6, 7)

## EQU00004 ## Ls = f (.SIGMA.Msqj.Ysj-Nsq) (j = 1, 2 ... 6, 7)

In the equations 2 to 4, the weighting factors Msoi and Msp
ij, Msqj, that is, the coupling coefficient between each neuron and the thresholds Nsoi, Nspj, Nsq are as shown in FIG.
Before the vehicle is shipped, the learning signal is given to the neurons of the output layer 118 offline before the vehicle is shipped, and the learning is determined by the back propagation method. The function f is a step function or a sigmoid function. Etc. functions.

[0037]

[Table 1] Device status Input signal status Teacher signal Normal status All input signals are normal 0 status Small impact failure Wheel wheel sensor abnormality, 0.3 Steering angle sensor abnormality etc. Medium impact failure Torque sensor offset, 0.6 Degradation of output of drive circuit, etc. Failure of large impact Torque sensor disconnection, 1.0 Auxiliary steering electronic control unit runaway, etc.

In step 310 of the flow chart shown in FIG. 10, it is judged whether or not the failure judgment level value Ls is less than 0.1, and it is judged that Ls <0.1. If it is found, in step 320 the flag Fs
Is set to 0, and when it is determined that Ls <0.1 is not satisfied, it is determined in step 330 whether Ls <0.4 is satisfied and Ls <0.4 is satisfied. When it is determined that the flag Fs is 1 in step 340.
Is set to.

At step 350, it is judged if the failure judgment level value Ls is less than 0.7, and Ls <
When it is determined that the value is 0.7, step 36
At 0, the flag Fs is set to 2, and when it is determined that Ls <0.7 is not established, the flag Fs is set to 3 at step 370. In step 380, the signal indicating the flag Fs set in step 320, 340, 360 or 370 is output to the microcomputer 38 of the auxiliary steering electronic control unit 30, and then the process returns to step 310.

Thus, according to the first embodiment shown in the figure, when each part of the power steering device is operating normally, the output of the neural network 112, that is, FIG.
The failure determination level value Ls calculated in step 240 of the flowchart shown in FIG.
Since the flag Fs is set to 0 in step 320 of the flow chart shown in FIG. 10, a yes determination is made in step 40 of the flow chart shown in FIG. 5, and the assist amount Taa is set to Ta. Then, the normal power assist is performed.

When a failure with a small influence on the power steering device occurs, such as an abnormality of the wheel speed sensor 34 or 35, the failure determination level value Ls becomes a value of 0.1 or more and less than 0.4, and the flag Fs is set. Since it is set to 1, a yes determination is made in step 60, an alarm is issued to the driver of the vehicle to indicate that a malfunction of a small degree of influence has occurred, and power assist is performed with a constant assist force. .

If a failure occurs during the influence of the power steering device such as the offset of the torque sensor 28,
Since the failure determination level value Ls becomes 0.4 or more and less than 0.7 and the flag Fs is set to 2, step 1
In step 10, a determination of yes is made, whereby a warning is issued to the driver of the vehicle that a failure in the influence degree has occurred, the assist force is gradually reduced, and finally the power steering device is manually operated. It is switched to the steering device.

Further, when a failure with a great influence occurs in the power steering device such as the disconnection of the torque sensor, the failure judgment level value Ls becomes a value of 0.7 or more and the flag Fs is set to 3, so that the step 110 is executed. At this time, a determination of "no" is made, whereby a warning is issued to the driver of the vehicle that a failure of great influence has occurred, and the power steering device is immediately switched to the manual steering device.

Therefore, according to the first embodiment, when a failure occurs in the power steering device, it is surely detected as a change in the failure judgment level value Ls,
Also, appropriate measures can be taken surely for the power steering device according to the value of the failure determination level value Ls, that is, according to the influence degree of the failure.

FIG. 11 is a schematic block diagram showing a second embodiment of the failure diagnosis processing apparatus according to the present invention applied to a four-wheel steering system, and FIG. 12 is a rear wheel steering electronic control system shown in FIG. It is a flow chart which shows a rear wheel steering control routine which is achieved.

As shown in FIG. 11, the left and right front wheels 1
30fl and 130fr are steered by a steering device 132 that responds to the rotation of the steering wheel 10. Left and right rear wheels 130rl and 130rr are steered by a rear wheel steering device 136 driven by a motor 134. It has become. The steering angle θ of the front wheels and the steering angle θr of the rear wheels are respectively determined by the steering angle sensors 26 and 138.
Is detected by the steering angle θ and θr
Is supplied to the rear wheel steering electronic control unit 140 and the failure diagnosis processing electronic control unit 142. Although not shown in FIG. 11, the electronic control unit 140
And 142 are constructed similarly to the electronic control units 30 and 32 in the first embodiment.

As shown in FIG. 11, the electronic control units 140 and 142 have a signal indicating the vehicle speed V detected by the vehicle speed sensor 144, a signal indicating the yaw rate δ of the vehicle detected by the yaw rate sensor 146, and a brake. A signal indicating whether or not the switch (BKSW) 148 is in the ON state is input. In particular, a signal indicating a failure coping mode instruction flag Fw, which will be described later, is input to the electronic control unit 140 from the electronic control unit 142, and the electronic control unit 142 receives the electronic control unit 140.
Further, a command signal output to a drive circuit for driving the motor 134 is input.

Although not shown in detail in the drawing, the ROM of the microcomputer of the electronic control unit 140 is shown in FIG.
The programs shown in FIG. 13 and the maps corresponding to the graphs shown in FIGS. 13 and 14 are stored. Further, the CPU of this microcomputer performs various calculations and processing of signals as described below based on the control program shown in FIG. 12, whereby the electronic control unit 140 causes the steering angle sensor 26 to detect the steering angle as described later. The motor 134 is controlled as necessary based on θ, etc., and the rear wheels 132fl and 132fl
When the fr is steered, and especially when some trouble occurs in the four-wheel steering system, the alarm lamp 150 is activated.

On the other hand, the ROM of the microcomputer of the electronic control unit 142 stores the neural network 152 shown in FIG. 15 and the control program shown in FIGS. In addition, C of this microcomputer
PU is the neural network 152 and FIG. 16 and FIG.
Based on the control program shown in FIG. 7, various calculations and signal processing are performed as will be described later to calculate the failure determination level value Lw of the neural network, and the failure coping mode instruction flag Fw according to the level value Lw. A signal indicating the failure coping mode instruction flag Fw is output to the microcomputer of the rear wheel steering electronic control unit 140.

As shown in FIG. 15, the neural network 152 is configured similarly to the neural network 112 in the first embodiment, has an input layer 154, an intermediate layer 156, and an output layer 158, and has an input layer 154. A signal indicating a detection value detected by a sensor such as the steering angle sensor 26 and a command signal output from a microcomputer of the rear wheel steering electronic control unit 140 to a drive circuit for driving the motor 134 are input to the layer 154. It is supposed to be done. The input layer 154 and the intermediate layer 156 have the same number of neurons as the number of signals input to the input layer, and the output layer 158 has one neuron.

Next, the rear wheel steering control in the illustrated embodiment will be described with reference to the flow chart shown in FIG. The control by the electronic control unit 140 is started by closing an ignition switch not shown in FIG. Further, the failure coping mode instruction flag Fw in the flowchart shown in FIG.
Indicates that the four-wheel steering system has a failure of small influence, a failure of medium influence, and a failure of high influence.

First, at step 410, it is judged if the flag Fw is 3, that is, if there is a failure of great influence on the four-wheel steering system, and Fw = 3 is not satisfied. If it is determined that the warning lamp is operated, the process proceeds to step 430. If it is determined that Fw = 3, the warning lamp is activated by outputting a control signal to the warning lamp 150 in step 420. As a result, a warning is given to the driver of the vehicle that a failure of great influence has occurred.

In step 430, it is determined whether the flag Fw is 2, that is, whether the four-wheel steering system has a failure in the influence degree, and it is determined that Fw = 2 is not satisfied. When the determination is made, the routine proceeds to step 450, where F
If it is determined that w = 2, step 44.
At 0, the target steering angle θrt of the rear wheels is set to 0, and then the routine proceeds to step 580.

In step 450, it is determined whether the flag Fw is 1, that is, whether or not the four-wheel steering system has a failure of a small influence degree, and Fw = 1. When the determination is made, the process proceeds to step 560 and Fw
If it is determined that = 1 is not established, step 46
At 0, a signal indicating the steering angle θ of the front wheels detected by the steering angle sensor 26 is read, and at step 470, the front and rear wheels are read from the map corresponding to the graph shown in FIG. 13 based on the vehicle speed V. The steering angle ratio R is calculated.

In step 480, the target steering angle θrt of the rear wheels is calculated according to the following equation 5, and in step 490, the deviation E of the steering angle of the rear wheels is calculated according to the following equation 6. The coefficient Kr in the equation 5 is a positive constant.

[Equation 5] θrt = R · θ + Kr · δ

[Equation 6] E = θrt−θr

In step 500, it is determined whether or not the brake switch 148 is in the on state. When it is determined that the brake switch is in the on state, the process proceeds to step 530 and the brake switch is turned off. When it is determined that the state is in the state, the flag F is reset to 0 in step 510, and in step 520, the control signal corresponding to the deviation E is sent to the motor through a drive circuit not shown in the figure. Output to 134, and then return to step 410.

In step 530, it is determined whether or not the flag F is 1, and when it is determined that F = 1, the process proceeds directly to step 550, and it is determined that F = 1 is not satisfied. When the determination is made, the flag F is set to 1 in step 540, and the deviation Ec of the steering angle of the rear wheels is set to the deviation E calculated in step 490. In step 550, the control signal corresponding to the deviation Ec is passed through a drive circuit (not shown) to the motor 1
It is output to 34, and then returns to step 410.

In step 560, the steering angle sensor 26
The steering angle θ of the front wheels detected by is read, and in step 570, the target steering angle θ of the rear wheels is calculated based on the steering angle θ of the front wheels from the map corresponding to the graph shown in FIG.
rt is calculated. In step 580, the steering angle θr of the rear wheels detected by the steering angle sensor 138 is read, and in step 590, the deviation E of the steering angle of the rear wheels is calculated according to the following equation 7, At 60, a control signal corresponding to the deviation E is output to the motor 134 via a drive circuit (not shown), and then the process returns to 410.

[Equation 7] E = θrt−θr

Next, the fault level determination routine and the fault coping mode selection routine in the second embodiment will be described with reference to the flow charts shown in FIGS. 16 and 17, respectively.

In step 610 of the flow chart shown in FIG. 16, a signal indicating the steering angle θ of the front wheels detected by the steering angle sensor 26 is read, and in step 620 the steering angle sensor 26 is read. The steering angle θ of the front wheels detected by the vehicle speed sensor 144 is set as Iw1, and the steering angle θr of the rear wheels detected by the steering angle sensor 138 is set as Iw2.
The vehicle speed V detected by the yaw rate sensor 1 is set as Iw3.
Let yaw rate δ of the vehicle detected by 46 be Iw4,
The signal from the brake switch 148 is Iw5, the command current Iwc output from the output port device of the electronic control unit 140 to the drive circuit for driving the motor is Iw7, and Mwoi and Nwoi (i = 1, 2, 5, 6) The output value Xwi (i = 1, 2 ... 5, 6) of the input layer 154 of the neural network 152 is calculated according to the following equation 8 with the respective weighting factors and threshold values for the respective input signals.

[Equation 8] Xwi = f (Mwoi · Iwi-Nwoi)

In step 630, the output value Ywj (j = 1, 2 ... 5, 6) of the intermediate layer 156 is calculated in accordance with the following equation 9, and in step 640, the failure judgment level is obtained according to the following equation 10. The value Lw is calculated.

[Formula 9] Ywj = f (ΣMwpij · Xwi-Nwpj) (i = 1, 2 ... 5, 6)

Lw = f (ΣMwqj · Ywj-Nwq) (j = 1, 2 ... 5, 6)

In the equations 8 to 10, the weighting factors Mwoi, M
The wpij, Mwqj, that is, the coupling coefficient between each neuron and the threshold values Nwoi, Nwpj, Nwq are the output layer 15 off-line before the shipment of the vehicle as in the case of the first embodiment.
It is determined by learning by the backpropagation method by giving the teacher signals shown in Table 2 below to the 8 neurons, and the function f is a function such as a step function or a sigmoid function.

[0063]

[Table 2] Device status Input signal status Teacher signal Normal status All input signals are normal 0 status Small impact failure Abnormal vehicle speed sensor, 0.3 Brake switch abnormality, yaw rate sensor abnormality, etc. Failure during operation Defects in front wheel steering angle sensor, 0.6 Deterioration of output of drive circuit, etc. Failures with high impact, Failure in rear wheel steering angle sensor, 1.0 Runaway of rear wheel steering electronic control unit, etc.

In step 710 of the flow chart shown in FIG. 17, it is judged whether or not the failure judgment level value Lw is less than 0.1, and it is judged that Lw <0.1. If it is found, the flag Fw is given in step 720.
Is set to 0, and when it is determined that Lw <0.1 is not satisfied, it is determined in step 730 whether Lw <0.4 is satisfied, and Lw <0.4 is determined. When it is determined that the flag Fw is 1 in step 740.
Is set to.

In step 750, it is judged whether the failure judgment level value Lw is less than 0.7, and Lw <
If it is determined that the value is 0.7, step 76.
The flag Fw is set to 2 at 0, and when it is determined that Lw <0.7 is not satisfied, the flag Fw is set to 3 at step 770. In step 780, the signal indicating the flag Fw set in steps 720, 740, 760 or 770 is the four-wheel steering system 14.
0 to the microcomputer, and then returns to step 710.

Thus, according to the illustrated second embodiment, when the respective parts of the four-wheel steering system are operating normally, the output of the neural network 152, that is, step 640 of the flow chart shown in FIG. The failure determination level value Lw calculated in this case becomes a value less than 0.1, and the flag Fw is set to 0 in step 720 of the flowchart shown in FIG. 17, so the flowchart shown in FIG. In steps 410 to 450 of No., a negative determination is made, whereby the steering angle θr of the rear wheels is controlled according to the steering angle θ of the front wheels, the vehicle speed V, and the yaw rate δ, so that the normal steering control of the rear wheels is performed. Done.

When a failure with a small influence occurs in the four-wheel steering system such as an abnormality of the vehicle speed sensor 144, the failure determination level value Lw becomes a value of 0.1 or more and less than 0.4, and the flag Fw.
Is set to 1, a yes determination is made in step 450, a warning is issued to the driver of the vehicle that a failure of a small degree of influence has occurred, and the steering angle θr of the rear wheels is set to that of the front wheels. It is controlled only according to the steering angle θ.

When a failure occurs in the degree of influence on the four-wheel steering system, such as an abnormality of the front wheel steering angle sensor 26, the failure determination level value Lw becomes a value of 0.4 or more and less than 0.7 and the flag Fw. Is set to 2, a yes determination is made in step 430, whereby a warning is issued to the driver of the vehicle that a failure in the degree of influence has occurred and the steering angle θr of the rear wheels is By controlling to 0, steering of the rear wheels is stopped.

Further, when a failure having a great influence on the four-wheel steering system occurs, such as an abnormality of the steering angle sensor 138 for the rear wheels, the failure judgment level value Lw becomes 0.7 or more and the flag Fw is set to 3.
Is set to No., a yes determination is made in step 410, which issues a warning to the driver of the vehicle that a failure of great influence has occurred, and the steering of the rear wheels by the four-wheel steering system. Is immediately discontinued.

Therefore, according to the second embodiment, if any failure occurs in the four-wheel steering system, it is surely detected as a change in the failure judgment level value Lw as in the case of the first embodiment. Further, appropriate measures are surely taken for the four-wheel steering system according to the value of the failure determination level value Lw, that is, according to the degree of influence of the failure.

Although the present invention has been described in detail with reference to specific embodiments, the present invention is not limited to these embodiments, and various other embodiments within the scope of the present invention. It will be apparent to those skilled in the art that

For example, in the above-described first and second embodiments, the failure diagnosis processing device for the vehicle steering system according to the present invention is applied to the electric power steering system and the four-wheel steering system, respectively. In the vehicle equipped with the electric power steering device and the four-wheel steering device, the failure diagnosis processing device of the present invention may be configured to comprehensively detect the failure of the two devices and deal with the failure.

[0073]

As is apparent from the above description, according to the present invention, the failure diagnosis processing device FS includes the failure diagnosis means M1.
And a failure coping means M2, and the failure diagnosing means uses the output value of the neural network whose input value is the value detected by the steering state detecting means S1 as a failure determination level value and corresponds to a failure corresponding to the occurrence of an abnormality in each input value. The coupling coefficient between each neuron is learned and determined by the backpropagation method by the teacher signal set according to the level, and the failure coping means has a predetermined failure coping mode from a plurality of failure coping modes according to the failure judgment level value. Is selected and executed, when some failure occurs in the steering device and an abnormality occurs in the detection value of the related steering state detecting means, the output value of the neural network, that is, the failure determination level value of the failure diagnosing means is changed. It changes according to the level of the failure, and the failure handling means selects a predetermined failure handling mode according to the failure determination level value. Is the row, thus it is possible to reliably detect a failure of the steering system can be devised reliably given treatment according to the level of failure.

Further, according to the present invention, it is not necessary to individually form an abnormality detection logic for all points where an abnormality may occur in the vehicle steering system, and it is not necessary to form an abnormality handling logic for each point. Therefore, the failure diagnosis processing device can be constructed at a lower cost than the conventional one.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing a configuration of a failure diagnosis processing device for a vehicle steering system according to the present invention, corresponding to the claims.

FIG. 2 is a schematic configuration diagram showing a first embodiment of a failure diagnosis processing device according to the present invention applied to an electric power steering device.

FIG. 3 is a block diagram showing the electronic control unit for auxiliary steering shown in FIG.

FIG. 4 is a signal flow diagram showing a signal flow in the electronic control device for auxiliary steering shown in FIGS. 2 and 3.

FIG. 5 is a flowchart showing a power assist control routine achieved by the electronic control device for auxiliary steering shown in FIGS. 2 and 3.

FIG. 6 is a block diagram showing an electronic control unit for failure diagnosis processing of the first embodiment shown in FIG.

7 is an explanatory diagram showing a neural network stored in the failure diagnosis processing electronic control device shown in FIGS. 2 and 6. FIG.

8 is an explanatory diagram showing a point of learning by a back propagation method for the neural network shown in FIG. 7. FIG.

9 is a flowchart showing a failure determination routine achieved by the electronic control unit for failure diagnosis processing shown in FIGS. 2 and 6. FIG.

10 is a flowchart showing a failure coping mode selection routine achieved by the failure diagnosis processing electronic control unit shown in FIGS. 2 and 6. FIG.

FIG. 11 is a schematic configuration diagram showing a second embodiment of a failure diagnosis processing device according to the present invention applied to a four-wheel steering device.

12 is a flowchart showing a rear wheel steering control routine that is achieved by the rear wheel steering electronic control device shown in FIG.

FIG. 13 is a graph showing a relationship between a vehicle speed V and a steering angle ratio R.

FIG. 14 is a graph showing a relationship between a front wheel steering angle θ and a rear wheel steering angle θr.

15 is an explanatory diagram showing a neural network stored in the failure diagnosis processing electronic control device shown in FIG. 11. FIG.

16 is a flowchart showing a failure determination routine achieved by the failure diagnosis processing electronic control unit shown in FIG. 11. FIG.

FIG. 17 is a flowchart showing a failure coping mode selection routine achieved by the failure diagnosis processing electronic control unit shown in FIG. 11.

[Explanation of symbols]

 10 ... Steering wheel 20 ... Power unit 26 ... Steering angle sensor 28 ... Torque sensor 30 ... Auxiliary steering electronic control device 32 ... Failure diagnosis processing electronic control device 38 ... Microcomputer 112 ... Neural network 132 ... Steering device 134 ... Motor 136 ... Rear wheel steering device 140 ... Rear wheel steering electronic control device 142 ... Fault diagnosis processing electronic control device 144 ... Vehicle speed sensor 146 ... Yaw rate sensor 148 ... Brake switch 152 ... Neural network

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masaaki Tsuboi, 1 Toyota Town, Toyota City, Aichi Prefecture, Toyota Motor Corporation (72) Inventor, Tomoyuki Watanabe 1 Toyota Town, Toyota City, Aichi Prefecture Toyota Motor Corporation

Claims (1)

[Claims] 1. A failure diagnosis processing device for a vehicle steering system, comprising: a plurality of steering state detecting means for detecting a steering state; and a control means for controlling an actuator for auxiliary steering according to the detected steering state. The output value of the neural network whose input value is the value detected by the steering state detecting means is used as the failure determination level value, and when the input value is abnormal, each teacher signal is set in accordance with the corresponding failure level. It has a failure diagnosis means in which the coupling coefficient between neurons is learned by the back propagation method, and a failure coping means for selecting and executing a predetermined failure coping mode from a plurality of failure coping modes according to the failure judgment level value. A failure diagnosis processing device for a vehicle steering system, comprising: