JPS6223805A - Adaptive control device for drive comfortability of automobile - Google Patents

Abstract

PURPOSE:To enhance the drive comfortability of an automobile, by automatically finding out a corresponding control rule in accordance with a correcting instruction from a terminal device when the result of control does not satisfy requirements relating to the comfortability of the passengers, so that the drive comfortability may be corrected. CONSTITUTION:A detector 40 detects environment 10 such as weather, road surface condition, etc. and the situation of a system 20 such as, the inclination of a vehicle, the speed thereof, etc. and delivers its output to a control condition comparing device 60 which compares the present detected conditions and conditions stored in a known condition memory device 80. If both kinds of conditions are coincident with each other, a fuzzy control device 150 reads out a corresponding one of rules stored in a control rule memory device 140 to calculate a control value. However, if both kinds of conditions are not coincident with each other, an evaluation device 10 evaluates the resemblance between both kinds of conditions, and a resemblance control device 130 estimates a control value in accordance with the thus evaluated resemblance and the corresponding rule read out from the memory device 140. The control rule may be optionally corrected in accordance with the operation of a correction and demand input device 210.

Description

[Detailed description of the invention]

[Field of Application of the Invention] The present invention relates to the control of the riding comfort of an automobile, and in particular, when the control result is not in accordance with the riding comfort characteristics of the driver or passengers (hereinafter referred to as occupants), the present invention is directed to controlling the riding comfort of an automobile. The present invention relates to a ride comfort adaptive control device that can automatically find and modify applicable control rules based on the basis.

[Background of the invention]

Conventionally, the method of modifying the control law was described in 1 by Kanno.
``Application of Fuzzy Sets and Logic to Control'' Measurement and Control, V
Ol, 18, Nα・2. PP, 150-160 (February 1978). Here, methods for adding, deleting, and replacing control laws are described, but it is assumed that the control laws to be modified are clear and play a large role in determining control results (control values). No consideration is given to cases in which control rules must be searched. [Object of the Invention] The object of the present invention is to solve the problem when the control result does not meet the purpose of the passenger, that is, when it does not satisfy the requirements regarding riding comfort.
The purpose of the present invention is to provide a device that controls the characteristics of an automobile in accordance with requests from passengers, such as making the suspension a little softer or a little more stiff. [Summary of the Invention] In order to achieve such an object, the present invention automatically determines a control rule that plays a large role in determining a control result (control value) from a combination result of a plurality of control rules. /(), and automatically adjusts the above control rules based on the passenger's adjustment from the terminal device + F1 line command, and controls the suspension. An embodiment will be explained in detail with reference to Figs. 1 to 3. Fig. 1 shows the configuration of an embodiment of a system for realizing a ride comfort adaptive control device for NoJ city according to the present invention. 1, a detector or a human 40 detects a signal 30 from an environment 10 and a system 1520, and sends the signal 50 to a control condition comparison device 60. In the control condition comparison device 60, the current detection condition is stored in a known condition storage device. It is compared with the signal 70 whether the condition is included in the condition stored in 80. As a result, if the condition is included, the signal 50 is sent to the fuzzy control device 150 by the signal 90. If the condition is not included, the signal 50 is sent to the fuzzy control device 150. The signal 1-00 is sent to the similarity evaluation device 1]0.The similarity evaluation device 1]0 evaluates the similarity kf?f between the detected control condition and the stored known condition, and calculates the similarity. is sent to the analogy control device 1:30 by the signal 120.The analogy control device 1;30 obtains the rule from among the plurality of rules stored in the similarity and control rule storage device 140 by the signal 170. , the control value is estimated by analogy, and the controlled device 200 is operated according to the borrowed value of 1.80. 11 of the rules are obtained by the signal 160, and the control value is calculated by 7tl.The controlled device 200 is operated on that value by the corresponding signal 190. That is, the fuzzy
The control device Ft150 is a control device when the detected control condition is known, and the analog control device 1:30 is a control device when the detected control condition is known.
This can be said to be a control device for when the control conditions are unknown. Shu 11:
The request input device 21-0 indicates that the occupant is fully satisfied with the control results! ,
This is used when L/ is not available, and allows input such as making the suspension or shock absorber a little softer or a little harder. When the signals 220 corresponding to these inputs are input to the control rule search device 230, the control rule search device 230 detects an unsatisfactory control result (control value).
Search for control rules that have a large weight in the decision. When the searched control rule is sent to the control rule modification device 250 by signal 24.0, the control rule is modified in accordance with a modification command from the passenger. The modified control rules are sent by signal 2.60 to the control device 140 and stored there. Hereinafter, the operation of each block will be explained in detail when the system 20 is a car. Et 1141. o is the weather, road condition, etc. These situations are detected by a detector 40 through a signal 30. Detector 40 may be a human. The detected weather conditions include heavy rain, heavy snowfall, and weak crosswinds, and the road surface is likely to be uneven, such as gravel or asphalt. For system 20 (car), tilt, speed. These include acceleration vibration, steering wheel angle, and vehicle height. These are sent by signal 50 to the control condition comparison device 6o. In the control condition comparison device 60, if the condition is stored in the known condition storage device 80 and sent to the device 60 by the signal 70, for example, ``if the unevenness is large, all the springs are made stiffer and the damping force of the absorber is increased. ′, ``if it is a sharp curve to the left, the spring on the right side will be stiffened'' (generally written as lfx→hatheny-)B), i.e., the control condition of X→Δ Compare the currently detected control condition Z → (If x = z and A = C, send z − + C to the fuzzy control device] 50 by the signal 90. If not, send the signal 90 to the fuzzy control device). 100 similarity evaluation device 2j
110, and the similarity is evaluated. The degrees of similarity between X and z, and between A and C, may be given by humans, but one method of calculating them by comparing detection conditions and known conditions will be shown below. Define similarity ff between
It can be thought of as an attribute of a concept (the concept of speed is composed of the dimensions of length and time), and g is defined by the overlap of the dimensions. Here, g(x+z) is the degree of similarity between X and Z, f (x) is the number of dimensions that constitute concept X, and f (x, z) is the degree of similarity between X and Z.
It is the number of dimensions common to the dimension of concept 2 and the dimension of concept 2. Regarding h, we have already proposed by R.R. Yager (2).
The following extensions can be used for variables. Here h At. is the similarity between A and C, and A(u) is f
The membership function for the ugly set, C(%l
) is a membership function when C is a fuzzy set. Further, it is assumed that A(u) and C(w) are standardized. The values of g(x+Z) and hAtQ are sent to analogy controller 130 by signal 1.20. The analogy control devices 1 and 30 infer a control value by analogy using the degree of similarity and a control rule such as If x-+A then y→B obtained from the signal 170 from the control rule storage device 140. In other words, the membership function D (y) of the fuzZy set for determining the control value is D (y) = B (y) g′””°hAc
(3) becomes. Control value y. can be obtained by the following equation. The above is a case where only one control rule is given, but
Generally there are several. In this case, each control rule R0,i
=1 t 2 F ”'I From equation (1) to equation (
3) to find D□(y), D(y) = uiDt
(yL i=1, 2. . . . , n, and calculate y in the same way as Equation (4). As a simple example, the case where there are two control rules, R1 and R2, will be shown below. These control rules is given in the form of a membership function as shown in Figure 2. The observable quantities are shown on the left side of the K column in Figure 2. Then, equations (1) to ( 3) from R1 and K using D (y)
02(y) can be obtained from R2 and K as shown in FIG. Furthermore, D (y)=01 (y)UO
D(y) is found from 3(y). Therefore, yD is expressed as (
4). y obtained in this way. The controlled device 200 is then controlled by the signal 180. On the other hand, if the observed quantity K matches the condition stored in the known condition storage device 80, the signal 90 and the control rule storage device 14
y by the fuzzy control device 150 using a signal 160 signifying the control rule Rj in question for K stored in 0. can be determined. That is, it is sufficient if a control result that satisfies the occupants is obtained using this yIl, but if this is not the case, it becomes necessary to modify the given control rule. The method for modifying the control rules is shown below. A control rule modification request is input from the modification request input device [2]0. For input, use a computer terminal device so that you can input, for example, a little softer or a little harder. When the signal 220 corresponding to the modification request is input to the control rule search device 230, the control rule search device 230 executes '
1. When determining io. /"D(y)occupies the maximum area in dy, and finds (y). In the example of Fig. 2, it becomes (y). Therefore, the control rule to be modified is R1
It can be seen that it is. R1 has the relevant signal and is "a little softer". 1' When the corresponding signal 240 is input to the control rule modification device 250, such as "make it a little harder," the control rule modification device 25
0, the control rule is modified by moving -Bt(y) to the left or right by a certain width. This result is sent by signal 260 to control rule storage 140 and stored therein. FIG. 3 is a flowchart showing an example of the processing procedure of the present invention. Block 300 is an input representing the environment and system status, such as the weather, road surface condition, vibration and tilt of the vehicle, as described above. Block 310 is the stored control condition Ifx-)Atheny-+B and the known control condition Ifx-)A which is the first half thereof. It is determined in condition determination block 320 whether these conditions are the same. If the detected control conditions are included in the known control conditions, block 3
40 operations are performed, and if it is not included, block 33
Performs 0 operation. At block 330, equation (1) and equation (
2) to find the degree of similarity. In block 340, the formula (
3) and the control value y by equation (4). seek. Next, block 350 determines whether the control results are satisfactory. This is done by the crew. If the result is satisfactory, the control rule is left unchanged; if the result is not satisfactory, the control rule is modified. The modification method is as described above, in which a control rule to be modified is first searched for in block 360 and modified in block 370. The modified results are stored in block 310. (]1) In addition, in section 1-, the control rule to be modified is f r) (y)
Dt(y) which occupies the maximum area in dy), ? ,
: Although it was set as Y, I occupy the second and third areas.
It is also possible to sequentially obtain ``, 1, + (y) and Dh(y), add way 1- between them and correct R1, and R. According to this embodiment, ( 1) When controlling a car, even if unstored control conditions are input, if it has similarity to the stored control conditions, the control is possible. (2) Since control is performed by analogy, conventional control is possible. Requires fewer control rules to be memorized than methods, (3) Control is possible even when control conditions are ambiguous, (4) Control is possible even when vague evaluation factors Y, such as ride comfort, etc. are included. (5) If the control result does not meet the passenger's wishes, the control rule that has a large weight in determining the control result can be automatically searched and corrected based on the control rule correction command from the occupant. The present invention applies to public systems such as water supply, sewage, transportation, etc. Also in the control of electric power systems, nuclear couplants, etc.
By creating the aforementioned control rules according to each system and plan 1, control can be performed using the same procedure. [Effects of the Invention] According to the present invention, it is possible to provide a device that automatically searches and improves the control characteristics of a vehicle, particularly the control rules, according to the occupant's requests, and it is possible to realize a vehicle having characteristics closer to the occupant's preferences. There is a result.

[Brief explanation of drawings]

Fig. 1 is a block diagram of an embodiment of an adaptive control device according to the present invention, Fig. 2 is an explanatory diagram of an example of fuzzy control rules and the case where control values are determined using these control rules, and Fig. 3 is a processing diagram. Flowchart 67 Figure 1 showing an example of the procedure

Claims (1)

[Claims] In a control device comprising a device that inputs a parameter in response to a change in the parameter, and a device that generates an output in response to the parameter input, the step of calculating the similarity of the parameters and outputting different outputs depending on the similarity are provided. What is claimed is: 1. A driving comfort adaptive control device for an automobile, comprising a step of generating a desired output, and a step of modifying a control rule so that the output is in a desired state.