JP3008785B2 - Rear wheel steering control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a 4WS vehicle,
The present invention relates to a rear-wheel steering control device that attempts to variably set steering characteristics according to changes in road traffic conditions and driving conditions.

[0002]

2. Description of the Related Art Conventionally, this type of apparatus estimates a characteristic index Z representing a driving characteristic of a driver by a neural network operation as disclosed in, for example, Japanese Patent Laid-Open No. 5-185947, and estimates the characteristic index. According to the characteristic index Z,
The gain of the rear wheel steering amount in proportion to the yaw rate γ as the driving state amount of the vehicle is automatically changed.

[0003]

However, when the driving characteristics of the driver suddenly change, the characteristic index Z also changes rapidly. Therefore, the gain of the rear wheel steering amount is changed based on the suddenly changing characteristic index Z. The disadvantage is that the behavior of the vehicle is disturbed. The present invention has been made in consideration of the above-described circumstances, and an object of the present invention is to prevent the behavior of the vehicle from being disturbed when the driving characteristics of the driver suddenly change, and to improve the driving feeling. To provide a rear wheel steering control device capable of

[0004]

SUMMARY OF THE INVENTION In order to achieve the above object, a rear wheel steering control device according to the present invention is provided with a vehicle running state parameter.
Based on the meter, as the road traffic situation at the time of vehicle travel
Road traffic condition recognition means for estimating any of the city area degree, congested road degree and mountainous road degree, and the vehicle traveling state parameter
Based on the driver's driving state , the operational crispness
And operating state recognition means for estimating the case, as the front wheel steering state
Steering state detecting means for detecting a steering angle and a steering angular velocity of the front wheels by multiplying the steering angle by a first control coefficient.
Multiplying a steering angle proportional component and the steering angular velocity by a second control coefficient
And the steering angular velocity proportional component obtained by
Rear-wheel steering drive means for driving the rear wheels according to the result ;
According to the estimated road traffic condition and the driving condition
Control means for variably setting the first and second control coefficients , wherein the control means controls the road traffic condition and the driving
The first and second controls for the rate of change of the state over time;
Should be used to change the coefficients or the first and second control coefficients
It is characterized in that the rate of change over time of the increase / decrease correction amount is limited.

[0005] In a preferred embodiment according to claim 1, the vehicle speed can further comprise Rukoto vehicle speed detecting means for detecting a said steering state detecting means detects a steering angle of the front wheels as a parameter of the steering state, the control means, The multiplication is performed based on a first control coefficient of an in-phase component that causes the rear wheel to be steered in the same phase direction as the front wheel with a characteristic that increases in accordance with an increase in the vehicle speed in a region equal to or higher than a predetermined vehicle speed, and Change of state
Rutotomoni moving the characteristics with respect to the vehicle speed of the first control coefficient to the change direction of the vehicle in accordance with the reduction, the road traffic-like況及
Characterized in that to limit the amount of movement in the direction of change in the vehicle speed with respect to changes in the fine operating conditions.

[0006] In a preferable embodiment of claim 1, wherein the control means, the steering angle of the front wheels only when zero, be made to change the said first and second control coefficient
I can .

[0007]

According to the rear wheel steering control device of the present invention, the first and second control coefficients are variably set by the control means in accordance with road traffic conditions and changes in the driving state of the driver. wheel steering drive means, respectively the first and second control coefficients these variably set, multiplied by the steering angle and the steering angular velocity, these <br/> multiplication results to further drive the steering of the rear wheels based on a result of addition I do. Further, the control means includes a road traffic condition and a driving condition.
By limiting the change in the first and second control coefficients or the amount of increase / decrease correction used for changing these control coefficients, the rapid change in the first and second control coefficients is suppressed.

[0008] Preferably , the control means multiplies the steering angle of the front wheel by a first control coefficient of an in-phase component, and drives the rear wheel based on the multiplication result. Then, the characteristic of the control coefficient with respect to the vehicle speed is moved in the vehicle speed changing direction according to the road traffic condition and the driving condition. The amount of movement in the direction in which the vehicle speed changes with this characteristic is used for the change in road traffic conditions and driving conditions.
As a result, the rapid fluctuation of the first control coefficient is suppressed.

More preferably , the control means changes each control coefficient only when the detected steering angle of the front wheel is zero. Therefore, that looks like a control coefficient is not changed while the rear wheel in accordance with the front wheel steering angle is steered driven.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the accompanying drawings. The rear wheel steering control device of the present invention includes:
Estimation means for estimating the road traffic condition and the vehicle driving operation state of the driver is provided. First, the estimation means (estimation method) will be described.

In the estimation method, a vehicle driving operation state by a driver is estimated based on a road traffic condition obtained based on a vehicle driving state parameter and a physical quantity representing the vehicle driving state. More specifically, as shown in FIG. 1, from the vehicle speed and the steering wheel angle, an average speed, a travel time ratio (a ratio of the travel time to the entire time including the vehicle travel time and the travel stop time) as vehicle travel state parameters, and An average lateral acceleration is determined. Then, by means of fuzzy inference based on these vehicle running state parameters, the degree of urban area, the degree of congested road, and the degree of mountain road are detected as parameters representing road traffic conditions.

On the other hand, as shown in FIG. 2, a physical quantity representing a vehicle driving state, such as an accelerator opening, a vehicle speed and a steering wheel angle, is detected, and a longitudinal acceleration is calculated from the vehicle speed, and a lateral acceleration is calculated from the vehicle speed and the steering wheel angle. Required by Then, the frequency distribution of each of the vehicle speed, the accelerator opening, the longitudinal acceleration, and the lateral acceleration as the vehicle driving parameters is obtained by the frequency analysis. Next, an average value and a variance of each frequency distribution are obtained as parameters characterizing the frequency distribution.

Further, parameters representing road traffic conditions (city level, congested road level and mountain road level) and parameters (average value and variance) characterizing the frequency distribution of the respective vehicle driving parameters are input to the neural network. Is done. In the neural network, a weighted sum of these parameters is obtained, and thereby an output parameter representing a vehicle driving operation state of the driver, for example, a degree of crispness in the vehicle driving operation of the driver is obtained.

As shown in FIG. 3, a controller 15 is mounted on a vehicle to which the estimation method according to the present embodiment is applied. Although illustration is omitted, this controller 15
Including a processor having a fuzzy inference function and a neural network function, a memory, and an input / output circuit,
Various control programs and various data are stored in the memory. The controller 15 is connected to a vehicle speed sensor 22, a steering wheel angle sensor 104, and a throttle opening sensor 23.

The processor of the controller 15 receives the vehicle speed signal, the steering wheel angle signal, and the throttle opening signal from the sensors 26, 16 and 104, and sequentially executes various routines to be described later to estimate the driver's crispness. It has become. "Running time ratio calculation routine" For example, after the engine is started, while the vehicle is in a driving state (including a running state and a running stopped state), the processor of the controller 15 executes the running time ratio calculation routine shown in FIG. It is repeatedly executed in the cycle of.

In each calculation routine execution cycle, the processor inputs the vehicle speed signal vel from the vehicle speed sensor 22 representing the actual vehicle speed, and determines whether or not the vehicle speed vel exceeds a predetermined vehicle speed (for example, 10 km / h). (Step S1). If the result of this determination is affirmative, the count value of a running time counter (not shown) built in the controller 15
"1" is added to rtime (step S2). on the other hand,
If the result of the determination in step S1 is negative, "1" is added to the count value stime of the traveling stop time counter (not shown) (step S3).

Step S following step S2 or S3
At 4, it is determined whether or not the sum of the running time counter value rtime and the running stop time counter value stime is equal to the value “200”. If the result of this determination is negative, the value obtained by dividing the running time counter value rtime by the sum of this value and the running stop time counter value stime is multiplied by the value “100”, and the running time ratio ratio (%) Is calculated (step S5).

On the other hand, if the determination in step S4 is affirmative, a value equal to the product of the running time counter value rtime and the value "0.95" is reset in the running time counter, and
A value equal to the product of the travel stop time counter value "stime" and the value "0.95" is reset to the travel stop time counter (step S6), and then the travel time ratio is calculated in step S5.

That is, when the vehicle is driven for 400 seconds corresponding to the value "200" from the start of the engine, the two counter values are reset, and thereafter, 1
Each time 0 seconds elapse, the counter value is reset. As a result, it is possible to calculate the running time ratio reflecting the vehicle driving situation before the current time even if a counter having a relatively small capacity is used. "Average Speed Calculation Routine" The processor of the controller 15 repeatedly executes the average speed calculation routine shown in FIG.

In each routine execution cycle, the processor reads the vehicle speed data vx from the vehicle speed sensor 22 and stores the stored values vxsum [i] (i = 1 to 5) of five accumulative speed registers built in the controller 10. Vehicle speed
vx is added (step S11). Next, the processor determines whether or not the value of the flag f_1m is “1” indicating the average speed calculation timing (Step S12). The flag f_1m takes a value “1” in a one-minute cycle. If the result of the determination in step S12 is negative, the processing in the current cycle ends.

If one minute elapses from the start of this routine and the determination result in step S12 becomes positive, the index jj
Is added to the value "1" to update the index jj, and the updated index
The accumulated speed register value vxsum [jj] corresponding to jj is changed to “15
Divide by 0 to calculate the average speed vxave,
xsum [jj] is reset to “0” (step S13).
Then, it is determined whether or not the updated index jj is "5" (step S14). If the determination result is negative, the process in the current cycle is ended.

Thereafter, the index jj is updated every one minute, and the accumulated speed register value corresponding to the updated index jj is updated.
An average speed vxave is obtained from vxsum [jj]. And 5
The index jj is reset to "0" every time the minutes elapse (step S15). As described above, the actual vehicle speed vx is added to each of the five cumulative speed register values vxsum [i] every two seconds, and the corresponding one of the five cumulative speed registers is output 150 times (for five minutes). Based on the stored value vxsum [jj] representing the detected total vehicle speed, the average speed vxave is 1
Calculated every minute. "Average lateral acceleration calculation routine" The processor of the controller 15 repeatedly executes the average lateral acceleration calculation routine shown in FIG. 6 at a cycle of, for example, 2 seconds.

In each routine execution cycle, the processor reads an output signal from the vehicle speed sensor 22 representing the vehicle speed vx and an output signal from the steering wheel angle sensor 104 representing the steering wheel angle steera, and refers to a map (not shown). A predetermined steering wheel angle gygain which is expressed as a function of the vehicle speed vx and gives a lateral acceleration of 1 (G) is obtained based on the vehicle speed vx. Next, the processor calculates the lateral acceleration gy by dividing the steering wheel angle steera by the predetermined steering wheel angle gygain, and stores the stored values gysum [i] (i = 1 to 5) of the five cumulative lateral acceleration registers built in the controller 10. The lateral acceleration gy is added to each of 5) (step S21). Next, the processor determines whether or not the value of the flag f_8s is “1” indicating the average lateral acceleration calculation timing (Step S22).
The flag f_8g takes a value “1” in an 8-second cycle. If the result of the determination in step S22 is negative, the processing in the current cycle ends.

If eight seconds have elapsed since the start of this routine and the result of the determination in step S22 becomes affirmative, the index jj
Is added to the value "1" to update the index jj, and the updated index
The cumulative lateral acceleration register value gysum [jj] corresponding to jj is set to "2
By dividing by 0, the average lateral acceleration gyave is calculated, and this register value gysum [jj] is reset to “0” (step S2).
3). Then, it is determined whether or not the updated index jj is "5" (step S24), and if the result of this determination is negative, the processing in this cycle is ended.

Thereafter, the index jj is updated every eight seconds, and the average lateral acceleration gyave is obtained from the accumulated lateral acceleration register value gysum [jj] corresponding to the updated index jj. Then, the index jj is reset to “0” every time 40 seconds elapse (step S25). As described above, the calculated lateral acceleration gy is added to each of the five cumulative lateral acceleration register values gysum [i] every two seconds, and the corresponding one of the five cumulative lateral acceleration registers, 20 times (40 The average lateral acceleration gyave is calculated every 8 seconds, based on the stored value gysum [jj] representing the total lateral acceleration calculated over a period of 2 seconds. In the present embodiment, the urban driving mode, the congested road driving mode, and the mountain driving mode are determined as the vehicle driving modes related to the estimation of the vehicle driving operation state by the driver. As the target, the degree of urban area, the degree of congested road, and the degree of mountainous road are determined.

The urban area degree and the congestion road degree are determined by fuzzy inference. In connection with this fuzzy inference, a membership function (FIGS. 7 and 8) representing a fuzzy subset in the entire space (set of cars) for the travel time ratio and the average speed, and nine fuzzy rules shown in the following table It is set in advance and stored in the memory of the controller 15.

[0027]

[Table 1]

The fuzzy rule setting shown in the above table is based on the fact that the average speed is low and the running time ratio is medium when traveling in an urban area, and the average speed is low and the running time ratio is low when traveling on a congested road. It is a thing. FIG.
Each of the symbols S, M, and B is a label that indicates a fuzzy set in a set of cars with respect to the travel time ratio. The membership function that defines the fuzzy set S has a fitness of “1” when the running time ratio is between 0% and 20%, and the fitness is “1” as the running time ratio increases from 20% to 40%. It is determined to decrease from “1” to “0”. Also, the membership function that defines the fuzzy set M increases the fitness from “0” to “1” as the running time ratio increases from 20% to 40%,
The fitness is “1” when the travel time ratio is between 40% and 65%, and the fitness is reduced from “1” to “0” as the travel time ratio increases from 65% to 85%. Stipulated. Then, the membership function defining the fuzzy set B changes the degree of conformity from “0” to “1” as the running time ratio increases from 65% to 85%.
The fitness is determined to be “1” when the running time ratio is 85% or more.

Referring to FIG. 8, the membership function defining the fuzzy set S in the platform set with respect to the average speed has a fitness of “1” when the average speed is between 0 km / h and 10 km / h, Is determined so that the fitness decreases from “1” to “0” as the speed increases from 10 km / h to 20 km / h. The membership function that defines the fuzzy set M has an average speed of 10 km / h to 20 km / h.
The fitness level increases from "0" to "1" as the speed increases to km / h, and the fitness level is "1" between 20 km / h and 40 km / h, and the average speed is 40 km / h. From 60
The fitness is determined to decrease from “1” to “0” as the speed increases to km / h. The membership function defining the fuzzy set B has an average speed of 40
The degree of conformity is “0” as it increases from km / h to 60 km / h
From "1" to "1", and the fitness is determined to be "1" when the average speed is 60 km / h or more.

The processor of the controller 10 determines the first to third combinations of the combination of the running time ratio (%) and the average speed (km / h) obtained by the calculation routines shown in FIGS.
The degree of adaptation adapt [i] to each of the ninth rules is obtained, and then the city area degree and the congestion road degree are calculated according to the following formulas. City area degree [city] = {(adap [i] x r_city [i]) ÷ adapt [i] (i = 1 to 9) Traffic jam degree [jam] = {(adap [i] x r_jam [i])} adapt [i] (i = 1 to 9) In detail, the processor obtains the fitness of the actual travel time ratio with respect to one of the fuzzy sets S, M, and B corresponding to the i-th rule among the travel time ratios, , I-th of the fuzzy sets S, M and B for the average speed
The fitness of the actual average speed for one corresponding to the rule is determined. Then, the smaller one of the two degrees of fitness is referred to as the i-th
The adaptation degree adapt [i] of the combination of the actual running time ratio and the actual average speed with respect to the rule is set.

Referring to the first rule, as shown in FIGS. 9 and 10, when the actual traveling time ratio is 30% and the actual average speed is 10 km / h, the driving time ratio fuzzy set S "0.5" is obtained as the fitness of the actual running time ratio 30%, and "1" is the fitness of the actual average speed 10 km / h for the average speed fuzzy set S.
Is found. Therefore, the degree of conformity ad to the first rule of the combination of the actual traveling time ratio 30% and the actual average speed 10 km / h.
apt [1] becomes “0.5”.

Next, the processor of the controller 15
The average lateral acceleration stored in the memory of the controller 15
With reference to the mountain road map, the mountain road is calculated based on the average lateral acceleration obtained in the routine of FIG. As illustrated in FIG. 11, in this map, the mountainous road degree becomes “0” when the average lateral acceleration is from 0 G to about 0.1 G, and the average lateral acceleration increases from about 0.1 G to 0.4 G. The mountain road degree increases from “0” to “100” as the distance from the vehicle increases, and the mountain road degree is set to “100” when the average lateral acceleration is 0.4 G or more. This map setting is based on the fact that the lateral acceleration integral value increases when traveling on a mountain road. "Frequency Analysis Routine" The processor of the controller 15 executes frequency analysis for each of the vehicle speed, the longitudinal acceleration, the lateral acceleration, and the accelerator opening at a cycle of, for example, 200 milliseconds, and obtains an average value and a variance of each physical quantity. FIG. 12 shows a frequency analysis routine for the vehicle speed. A frequency analysis routine (not shown) other than the vehicle speed has the same configuration as this routine.

The vehicle speed as a frequency analysis target parameter is
It is represented by an output signal from the vehicle speed sensor 22, and its input range is set to, for example, 0 to 100 km / h.
The accelerator opening tps (%) is calculated by the throttle opening sensor 23.
Is calculated according to the following equation based on the output signal of
Its input range is 0-100%. tps = (tdata-tpsoff) ÷ (tpson-tpsoff) × 100 Here, the symbol tdata represents the current throttle opening sensor output, and the symbols tpsoff and tpson are the throttle opening sensor in the accelerator off state and the accelerator fully open state. Represents the output respectively.

The processor samples the output of the vehicle speed sensor 22 at a cycle of, for example, 100 milliseconds and calculates the longitudinal acceleration gx (unit: G) according to the following equation. The input range of the longitudinal acceleration is, for example, 0 to 0.3G. gx = (vx−vx0) × 10 ÷ (3.6 × 9.8) Here, the symbol vx represents the current vehicle speed (km / h), and vx0 is 1
Indicates the vehicle speed (km / h) 00 milliseconds ago.

Further, an output signal from the vehicle speed sensor 22 indicating the vehicle speed vx and an output signal from the steering wheel angle sensor 104 indicating the steering wheel angle steera are read, and expressed as a function of the vehicle speed vx by referring to a map (not shown). A predetermined steering wheel angle gygain that gives the lateral acceleration of (G) is obtained based on the vehicle speed vx. Next, as shown in the following equation, the processor calculates the lateral acceleration gy (G) by dividing the steering wheel angle steera by the predetermined steering wheel angle gygain. The input range of the lateral acceleration is, for example, 0 to 0.5G.

Gy = steera ÷ gygain Referring to FIG. 12, the processor divides a vehicle speed signal vel as a frequency analysis target parameter (input data) into 10 equal parts within an input range of 0 to 100 km / h (INT (vel /
10)) is added with "1" to obtain a value dat (step S3).
1) It is determined whether or not this value dat is greater than "10" (step S32). If the determination result is affirmative, the value dat is reset to “10” in step S33, and the process proceeds to step S34. On the other hand, step S3
If the result of the determination in step 2 is negative, the process immediately proceeds from step S32 to step S34. In step S34,
As shown in FIG. 13, one of the populations of the input data
“1” is added to the corresponding one element number hist [dat] of the 0 arrays (the number of elements in the maximum value side array is 0 in FIG. 13).

In the next step S35, first to first
The sum num of the number of elements in the 0 array is obtained, and the sum sum of the product of the number of elements obtained for each array (the i-th array) and the value “i−1” is obtained. The processor sums the product s
The value obtained by dividing um by the sum num of the number of elements is further divided by the value “10” to obtain an average value ave of the input data (here, the vehicle speed) (step S36).

Next, the processor determines that the average value ave is "10
"0" (step S37).
If the determination result is affirmative, the average value is determined in step S38.
ave is reset to “100”, and the routine goes to Step S39. On the other hand, if the decision result in the step S37 is negative, the process immediately shifts from the step S37 to the step S39.
That is, the average value ave of the input data is limited to a value up to “100”.

In step S39, the average value ave is set to "1".
0 is subtracted from the value “i−1” ((i−
The product of the square of (1)-(ave / 10)) and the number of array elements hist [i] is obtained for each array, and then the sum sum2 of the products is calculated. Next, the processor calculates the variance var of the input data by further dividing the value obtained by dividing the sum sum2 by the sum num of the number of elements by the value “5” (step S40). Then, it is determined whether or not the variance var of the input data is greater than "100" (step S41). If the determination result is affirmative, the variance var of "10" is determined in step S42.
After resetting to “0”, the process proceeds to step S43. If the determination result in step S41 is negative, step S41 is performed.
The process immediately shifts from step S43 to step S43. That is, the variance var of the input data is limited to the value “100”.

In step S43, it is determined whether or not the sum num of the number of elements is larger than "256". If the result of this determination is negative, the processing in this cycle is terminated, while if the result of determination is positive. For example, the number of elements hist [i] of each of the first to tenth arrays is reset to a value obtained by multiplying the value by "15/16" (step S44), and the processing in the current cycle ends. That is, the number of elements num of the population is "256"
Is exceeded, the number of elements in each array is reduced by a factor of "15/16". Thereafter, the processing of FIG. 12 is repeated, and the average value and the variance of the vehicle speed vel as input data are periodically obtained.

Accelerator opening, which is other input data,
The average value and variance of each of the longitudinal acceleration and the lateral acceleration are obtained in the same manner. The average value and the variance of each input data increase as the driving method of the driver becomes crisp. However, road traffic conditions have a significant effect on the average vehicle speed. The processor of the “driving operation state calculation routine” controller 15 uses the neural network function to
Obtain the driving operation state by the driver. In this embodiment,
In addition to the average and variance of each of the vehicle speed, accelerator opening, longitudinal acceleration and lateral acceleration obtained by the above-described frequency analysis, the urban area degree, congestion road degree and mountain road degree obtained by the above fuzzy inference are stored in the neural network. Then, the degree of crispness as a driving operation state by the driver is obtained.

Conceptually, a neural network is
The processing elements (PE) shown in FIG. 14 are intricately entangled with each other as shown in FIG. 15, and each PE has a number of inputs x [i] and each weight w [j].
The sum of the products multiplied by [i] is input. And
In each PE, this sum is converted by a certain transfer function f, and an output y [i] is sent from each PE.

Referring to FIGS. 14 and 15,
The neural network used in this embodiment is an input layer 2
01 and an output layer 203 with one hidden layer 202 interposed, the input layer 201 is composed of 11 PEs,
The hidden layer 202 is composed of six PEs, and the output layer 203 is 1
Consists of two PEs. And the transfer function f of PE is f
(X) = x. Also, the weight w [j] [i] at the connection between the PEs is determined through a learning process. Note that an input 204 called a bias is additionally set in the neural network of this embodiment.

In this embodiment, the function of the above-described neural network is achieved by the controller 15. In order to achieve this neural network function, the processor of the controller 15 determines the average value and variance of the vehicle speed, the accelerator opening, the longitudinal acceleration and the lateral acceleration, the urban area degree, the congestion road degree and the mountain road degree (all output ranges thereof). With the input data of “0” to “100”), the crispness calculation routine shown in FIG. 16 is periodically executed.

In the routine shown in FIG. 16, the processor subtracts "100" from the product of the input data dd [i] and "2" to obtain 11 input data dd [i] (i = 1 to 11). The range is converted from “0 to 100” to “−100 to 100”, respectively.
n [i] is obtained (step S51). Next, the processor:
Sum dri of the product of input data din [i] obtained for each range-converted input data din [i] and weight coefficient nmap [i + 1]
ve and the same product for bias (nmap
[1] * 100) is obtained, and the product (nmap [1] * 100) related to the bias is added to the total drive related to the input data to obtain the output drive representing the degree of crispness (step S52).

Then, the crispness output drive is set to "100".
“100” is added to the value obtained by dividing by “00”, and the addition result is divided by “2” to change the range of the crispness output to “−1”.
“000000 to 100000” is converted to “0 to 100” (step S53), and the calculation of the crispness in one calculation cycle is thereby completed. As described above,
An output drive indicating the degree of crispness as a vehicle driving operation state by the driver is obtained. According to the test driving results, the estimated value of the driver's crispness represented by the output drive was in good agreement with the crispness evaluated and reported by the driver himself. This is based on the evaluation of the driving operation state of the driver, which is difficult to evaluate with physical quantities such as vehicle speed, based on the average and variance of physical quantities that characterize the frequency distribution of various physical quantities, and taking into account road traffic conditions. It is understood that it depends.

An embodiment of the rear wheel steering control device according to the present invention will be described below. The present embodiment intends to control the vehicle driving characteristics to those adapted to the road traffic condition and the vehicle driving operation state (crispness) estimated by the above estimation method, and a procedure for estimating the crispness. Is the same as that of the above-mentioned estimation method, and a description of the device configuration and the like for this is omitted.

Referring to FIG. 17, the left and right front wheels 1L, 1R of the vehicle are connected to a front wheel power steering device 2 via tie rods 3. The device 2 forms a four-wheel steering device (4WS) to which an embodiment of the present invention is applied in cooperation with various elements described later, and includes a rack and pinion mechanism (illustrated in FIG. ), And a front wheel steering actuator (not shown), which is connected to this and comprises a hydraulic cylinder.

The front wheel steering actuator is connected to one hydraulic pump 7 of the pump unit 6 via a front wheel steering valve 5 operated by a steering handle 4. The pump unit 6 is composed of a double pump driven by an engine 8, and the other hydraulic pump 9 is connected to a rear wheel steering actuator 11 via a rear wheel steering valve 10.
It is connected to the.

The rear wheel steering actuator 11 is also composed of a hydraulic cylinder, and the piston rod of the actuator 11 is connected to the left and right rear wheels 13L, 13R via tie rods 12. In FIG. 17, reference numeral 14 indicates a reservoir tank. The front wheel steering actuator operates according to the steering direction by supplying hydraulic oil from a hydraulic pump 7 via a front wheel steering valve 5 when the steering handle 4 is steered. On the other hand, the operation of the rear wheel steering actuator 11 is controlled by the controller 15. That is, when the steering handle 4 is steered, the controller 15 supplies an operation control signal suitable for the vehicle running state to the rear wheel steering valve 10, whereby the rear wheel steering actuator is transmitted from the hydraulic pump 9 via the valve 10. The operating hydraulic pressure supplied to the motor 11 is controlled.

The controller 15 is electrically connected to various sensors and meters 61 in connection with the above-described front and rear wheel steering actuator operation control. That is,
The controller 15 includes a vehicle speed V (corresponding to the above-described vx vehicle speed signal) from the vehicle speed sensor 26 and the meters 61 and sensor signals indicating the operating states of various devices, and the steering wheel angle sensor 1.
6, a sensor signal indicating the steering wheel angle θH (corresponding to the steering wheel angle steera described above) and a sensor signal indicating the actual rear wheel steering angle δRa from the rear wheel steering angle sensor 17 are supplied.

As shown in FIG.
Are functionally similar to the steering wheel angle sensor 16 and the vehicle speed sensor 2
6 and an input unit 30 for receiving data from the meter;
A mode determining unit 32 for determining a running mode of the vehicle based on data from the input unit 30; Further, the controller 15 includes an input unit 30 and a mode determination unit 32
A steering valve operation control unit 34 for calculating an operation control signal SR of the rear wheel steering valve 10 based on the data from the ECU 1 and an output for outputting the operation control signal SR calculated by the control unit 34 to the rear wheel steering valve 10 And a unit 35.

The mode determining unit 32 determines the steering mode of the rear wheels (for example, suspension of the control, large steering angle control of the rear wheels) based on the steering wheel angle θH, the vehicle speed V, and the data supplied from the meter to the input unit 30. , Rear wheel phase control). The steering valve operation control unit 34 calculates a rear wheel steering valve operation control signal SR based on the output data from the mode determination unit 32.
When the rear wheel phase control is selected, the rear wheel steering angle δR is calculated based on the steering wheel angle θH, the vehicle speed V, and the actual rear wheel steering angle δRa. The control unit 34 calculates the rear wheel steering angle δR
Is calculated based on, for example, the following known calculation formula.

ΔR = K 1 δF−K 2 (dδF / dt) Note that the in-phase coefficient K
1, negative phase coefficient K2, front wheel steering angle δF, front wheel steering angular velocity d
δF / dt. By changing the in-phase coefficient K1 and the anti-phase coefficient K2 according to the control rules shown in Table 2 in the above-described arithmetic expression, the vehicle characteristics can be changed to a luxury vehicle or a sporty vehicle. That is, if the in-phase coefficient K1 is increased and the anti-phase coefficient K2 is decreased, luxury vehicle characteristics are obtained. On the other hand, if the in-phase coefficient K1 is decreased and the anti-phase coefficient K2 is increased, sporty vehicle characteristics are obtained. .

[0055]

[Table 2]

The controller 15 calculates an in-phase coefficient K1 corresponding to the vehicle speed V in accordance with a map corresponding to the vehicle speed-in-phase coefficient characteristic shown by a solid line in FIG. The in-phase coefficient K1 represents a ratio between the front wheel steering angle and the rear wheel steering angle, and is a predetermined vehicle speed (for example, 60 km / h).
In the above-described vehicle speed range, the value increases as the vehicle speed increases.

Further, the controller 15 calculates a negative phase coefficient K2 corresponding to the vehicle speed V according to a map corresponding to the vehicle speed / negative phase coefficient characteristic indicated by a solid line in FIG. The negative phase coefficient K2 takes a value that increases and decreases as the vehicle speed increases in a predetermined vehicle speed region (for example, a vehicle speed region between 30 km / h and 125 km / h).

The increase / decrease speed V1 and the increase / decrease coefficient α are calculated according to the road traffic conditions (for example, the degree of urban road) and the driving state of the driver (the crispness) estimated by the above estimation method. First, the calculation of the increase / decrease speed V1 will be described. For example, as shown in FIG. 21, the road traffic condition includes a city level and a mountain level. Speed of increase / decrease Va from city level and crispness to city level
Speed V corresponding to the city level based on the characteristic map
a is required. Note that the characteristic line of the characteristic map is shifted according to the crispness, and the increase / decrease speed Va takes a smaller value as the value of the crispness increases.

Further, an increase / decrease speed Vb corresponding to the mountain road degree is obtained from the mountain road degree and the crispness based on a characteristic map of the increase / decrease speed Vb for the mountain road degree.
The characteristic line of the characteristic map is shifted according to the degree of crispness, and the increase / decrease speed Vb takes a larger value as the value of the crispness increases.

The increase / decrease speed V1 is calculated from the sum of the obtained increase / decrease speed Va and increase / decrease speed Vb, and the calculated increase / decrease speed V1 is calculated in the limit routine and calculated as the increase / decrease speed V1 '. . The controller processor repeatedly executes the increase / decrease speed limiting routine shown in FIG. 23 at a cycle of, for example, 10 ms.

In each calculation routine execution cycle, the processor inputs a signal from a steering wheel angle sensor representing the rotation angle of the steering wheel, and determines whether the steering wheel angle is in the range of the neutral position to determine the same phase. It is determined whether the steering amount is zero (step S5).
0). If this determination result is affirmative, the next step S51
Then, if the determination result is negative, the process in the current cycle is ended.

Therefore, when the result of the determination in step 50 is negative, that is, when the in-phase steering amount is not zero, the change in the increase / decrease speed is prohibited. In step S51,
It is determined whether or not a value obtained by subtracting the previous output value of the increase / decrease speed V1 'from the calculated increase / decrease speed V1 is larger than a predetermined value, for example, "0.01". When this determination result is affirmative, step S52 is executed, and when the determination result is negative, step S53 is executed.

In step S52, if the change amount of the increase / decrease speed per unit time exceeds the predetermined value "0.01" as a result of the determination in step S51, the increase amount of the increase / decrease speed is kept constant. That is, the previous increase / decrease speed V
A value obtained by adding a predetermined value, for example, "0.01" to 1 'is calculated as the current increase / decrease speed V1'. When the result of the determination in step S51 is negative, in step S53, the change speed V1 'which is the previous output value is calculated from the calculated change speed V1.
It is determined whether or not the value obtained by subtracting is smaller than a predetermined value, for example, “0.01”. When this determination result is affirmative, step S54 is executed, and when the determination result is negative, step S55 is executed.

In step 54, if the amount of change of the increase / decrease speed per unit time is smaller than the predetermined value "0.01" as a result of the determination in step S53, the decrease amount of the increase / decrease speed is kept constant. That is, a value obtained by subtracting a predetermined value, for example, "0.01" from the previous increase / decrease speed V1 'is used as the current increase / decrease speed V1'.
Calculate as 1 '. When the result of the determination in step S53 is negative, that is, when the amount of change in the increase / decrease speed per unit time is within a predetermined range, "-0.01 <V1-V1 '<0.
01 ", the calculated increase / decrease speed V1 is adopted as the current increase / decrease speed V1 'in step S55.

In this restriction routine, when the in-phase steering amount is not zero, the change in the increase / decrease speed is prohibited because when the increase / decrease speed changes during the in-phase steering, even if the change is very small, the driver cannot change the speed. This is because it becomes uncomfortable. Next, the calculation of the increase / decrease coefficient α will be described. As shown in FIG. 22, as in the case of the increasing / decreasing speed V, the city traffic level and the mountain road level are used as the road traffic conditions. An increase / decrease coefficient αa corresponding to the city level is calculated based on the characteristic map of the increase / decrease coefficient αa for the city level from the city level and the crispness. Note that the characteristic line of the characteristic map is shifted according to the crispness, and the increase / decrease coefficient αa takes a larger value as the value of the crispness increases.

Further, an increase / decrease coefficient αb with respect to a mountain road degree from the mountain road degree and the crispness as the road traffic condition.
The increase / decrease coefficient αb corresponding to the degree of the mountainous road is obtained based on the characteristic map shown in FIG. The characteristic line of the characteristic map is shifted according to the degree of crispness, and as the crispness increases, the increase / decrease coefficient α
b has a large value.

Then, an increase / decrease coefficient α1 is calculated from the sum of the obtained increase / decrease coefficient αa and increase / decrease coefficient αb, and the calculated increase / decrease coefficient α1 is processed in a restriction routine to be calculated as an increase / decrease coefficient α1 ′. . The processor of the controller repeatedly executes the increase / decrease coefficient limiting routine shown in FIG. 24 at a cycle of, for example, 10 ms.

First, the processor determines whether or not a value obtained by subtracting the previous output value of the increase / decrease coefficient α1 ′ from the calculated increase / decrease coefficient α1 is larger than a predetermined value, for example, “0.002” (step S56). ). When this determination result is affirmative, step S57 is executed, and when the determination result is negative, step S58 is executed. In step S57, based on the determination result in step S56, the amount of change in the increase / decrease coefficient per unit time is a predetermined value, for example, “0.002”.
In the case where it exceeds, the increase amount of the increase / decrease coefficient is kept constant. That is, a value obtained by adding a predetermined value, for example, “0.002” to the previous increase / decrease coefficient α1 ′ is calculated as the current increase / decrease coefficient α1 ′.

When the result of the determination in step S56 is negative, in step S58, the value obtained by subtracting the previous output value of the increase / decrease coefficient α1 ′ from the calculated increase / decrease coefficient α1 is smaller than a predetermined value, for example, “0.002”. It is determined whether or not it is smaller. If this determination result is affirmative, step S59 is performed.
Is executed, and if the determination result is negative, step S60 is executed.
Is executed.

In step S59, if the amount of change in the increase / decrease coefficient per unit time is smaller than a predetermined value, for example, "0.002", as a result of the determination in step S58, the amount of decrease in the increase / decrease coefficient is kept constant. That is, a value obtained by subtracting a predetermined value, for example, “0.002” from the previous increase / decrease coefficient α1 ′ is calculated as the current increase / decrease coefficient α1 ′. Step S5
8 is negative, that is, the change amount of the increase / decrease coefficient is within the predetermined range “−0.002 <α1−α1 ′ <
If it is "0.002", in step S60, the calculated increase / decrease coefficient α1 is adopted as the current increase / decrease coefficient α1 ′.

Accordingly, the in-phase coefficient K1 is corrected as shown by the dotted line in FIG. 19 according to the increase / decrease speed V1 calculated based on the road traffic condition and the driving state of the driver.
That is, as the increase / decrease speed V1 takes a positive value,
The common mode coefficient K1 is corrected so that the common mode coefficient K1 takes a small value. In other words, the characteristic line of the map is moved such that the rising start speed “60 + V1” of the in-phase coefficient K1 increases as the increase / decrease coefficient takes a positive value. As a result, an in-phase coefficient K1 suitable for road traffic conditions, driving conditions, and vehicle speed is obtained.

The negative phase coefficient K2 is an increase / decrease coefficient α1 calculated based on the road traffic condition and the driving state of the driver.
Is corrected as shown by the dotted line in FIG. That is, as the increase / decrease coefficient α takes a value larger than “1”, the inverse phase coefficient K2 becomes larger so that the inverse phase coefficient K2 takes a larger value.
Is corrected. In other words, the characteristic line is moved by α times based on the obtained increase / decrease coefficient α. As a result, the inverse phase coefficient K adapted to road traffic conditions, driving conditions and vehicle speed
2 is found.

Thereafter, the steering valve operation control unit 34 performs an operation control according to a deviation between the target rear wheel steering angle δR calculated as described above and the actual rear wheel steering angle δRa from the rear wheel steering angle sensor 17. The signal SR is output to the rear wheel steering valve 10 via the output unit 35, whereby the rear wheel steering actuator 11 matches the actual steering angles of the rear wheels 13L and 13R with the rear wheel steering angle δR. It works to.

As described above, in the rear wheel steering control apparatus according to the embodiment, as shown in Table 3, the steering characteristics according to the road traffic condition and the driving state can be realized, and the driving feeling at the time of rear wheel steering can be realized. Ring improves.

[0075]

[Table 3]

The increase / decrease speed V for correcting the in-phase coefficient K1
Since the amount of change per unit time of the increase / decrease coefficient α for correcting the negative phase coefficient K2 is limited to a predetermined amount, even if the road traffic condition or the driving state of the driver changes suddenly, the in-phase coefficient K1 and the negative phase coefficient K2 do not change suddenly, and the driver does not feel uncomfortable at the time of steering.

[0077]

According to the rear wheel steering control device of the present invention, the control coefficient for rear wheel steering does not change abruptly even when the estimated road traffic condition and the driving condition of the driver suddenly change. Therefore, the disturbance of the behavior of the vehicle due to the rear wheel steering can be effectively prevented, and the rear wheel steering control according to the road traffic condition and the driving state can be realized. Road traffic conditions and
And fine-grained control according to the driving conditions of the driver.
Excellent, such as being able to effectively prevent disturbances in vehicle behavior
It has the effect.

In the present invention, in the case where control for in-phase steering of the rear wheels is performed according to the steering angle and the vehicle speed of the front wheels, the vehicle speed at which the rear wheel steering is started can be variably set, and the road traffic condition and the driver's When moving the characteristic of the control coefficient with respect to the vehicle speed in the vehicle speed direction according to the change in the driving state, the amount of movement in the vehicle speed direction is limited to a predetermined value or less. The effect of the variable control can be sufficiently obtained, and at the same time, the disturbance of the behavior of the vehicle can be effectively prevented.

According to another preferred embodiment , the control coefficient is not changed while the rear wheel is being steered according to the steering angle of the front wheel, and the occupant does not feel uncomfortable.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram showing a procedure for grasping a road traffic situation in an embodiment of a method for estimating a vehicle driving operation state according to the present invention.

FIG. 2 is a conceptual diagram showing a driving operation state grasping procedure in the embodiment.

FIG. 3 is a schematic block diagram showing a controller and a sensor for performing the estimation method according to the embodiment.

FIG. 4 is a flowchart of a running time ratio calculation routine executed by the controller shown in FIG. 3;

FIG. 5 is a flowchart illustrating an average speed calculation routine executed by a controller.

FIG. 6 is a flowchart of an average lateral acceleration calculation routine executed by a controller.

FIG. 7 is a graph showing a membership function that defines a fuzzy set for a travel time ratio.

FIG. 8 is a graph showing a membership function defining a fuzzy set for average speed.

FIG. 9 is a graph showing a calculation example of a degree of adaptation of an actual traveling time ratio to a traveling time ratio fuzzy set.

FIG. 10 is a graph showing a calculation example of a degree of adaptation of an actual average speed to an average speed fuzzy set.

FIG. 11 is a graph illustrating an average lateral acceleration / mountain road degree map;

FIG. 12 is a flowchart of a frequency analysis routine executed by the controller of FIG. 3;

FIG. 13 is a graph showing an array constituting a population of input data as a frequency analysis target.

FIG. 14 is a conceptual diagram showing processing elements constituting a neural network.

FIG. 15 is a conceptual diagram of a neural network constituted by the processing elements shown in FIG.

FIG. 16 is a flowchart illustrating a crispness calculation routine executed by the controller of FIG. 3;

FIG. 17 is a schematic diagram showing an automobile equipped with a four-wheel steering device for implementing the rear-wheel steering control device of the present invention.

FIG. 18 is a functional block diagram showing a controller in detail.

FIG. 19 is a graph showing a characteristic graph of an in-phase coefficient K1 with respect to a vehicle speed.

FIG. 20 is a graph showing a characteristic graph of the inverse phase coefficient K2 with respect to the vehicle speed.

FIG. 21 is a schematic block diagram until an increase / decrease speed is calculated.

FIG. 22 is a schematic block diagram until an increase / decrease coefficient is calculated.

FIG. 23 is a flowchart showing a routine for limiting the increase / decrease speed.

FIG. 24 is a flow chart showing a routine for limiting an increase / decrease coefficient.

[Explanation of symbols]

 Reference Signs List 4 steering handle 15 controller 16 handle angle sensor 34 steering valve operation control unit 60 yaw rate sensor 201 input layer of neural network 202 output layer of neural network

────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁷ Identification code FI B62D 137: 00 (56) References JP-A-5-185947 (JP, A) JP-A-62-181964 (JP, A) JP-A-2-227381 (JP, A) JP-A-1-195184 (JP, A) JP-A-63-162372 (JP, A) JP-A-2-175471 (JP, A) JP-A-5-213221 ( JP, A) JP-A-7-105474 (JP, A) JP-A-7-101272 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B62D 6/00 B62D 7/14

Claims (1)

(57) [Claims] 1. A based on the vehicle running condition parameter, urban degree as road traffic conditions while the vehicle is running, a congested road degree及
Road traffic condition recognizing means for estimating any one of the road conditions , a driving state recognizing means for estimating the degree of operational crispness as the driving state of the driver based on the vehicle driving state parameter, and a steering state of the front wheels. A steering state detecting means for detecting a steering angle and a steering angular velocity of a front wheel, and a steering angle ratio obtained by multiplying the steering angle by a first control coefficient.
An example component is obtained by multiplying the steering angular velocity by a second control coefficient.
Rear wheel steering drive means for driving the rear wheels according to the addition result, and according to the estimated road traffic condition and the driving condition.
Control means for variably setting the first and second control coefficients , wherein the control means controls the road traffic condition and the driving condition.
With respect to the rate of change over time, the first and second control coefficients
Or the increase or decrease to be used for changing the first and second control coefficients.
A rear wheel steering control device for limiting a positive temporal change rate .