JPH06106971A - Catch preventing device for power window - Google Patents

Abstract

PURPOSE:To appropriately prevent catching of a power window, and to enhance the safety and the usability. CONSTITUTION:A catch preventing device for a power window comprises an MRE sensor 20 for detecting a travel position of a wind glass pane, a counter 31, a timer 32 for detecting a physical quantity relating to a travel load of the window glass pane, a latch circuit 35 and a timer 33, a fuzzy inference device 34 receiving a travel position and a travel load for fuzzy-inferring a moving travel distance of the window glass pane by use of a membership function and a fuzzy rule, a selector 60 for moving the window glass pane in its opening direction in accordance with a value computed by the fuzzy inference device 34 if a switch 70 is in such a condition that a closing movement is instructed, and a window glass pane opening and closing device 10 are provided. Further, A kind of an obstacle is appropriately determined by the fuzzy inference device 34, and thereby it is possible to instruct an appropriate movement.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pinch prevention device for preventing an obstacle from being pinched in a power window for electrically moving a window glass of a vehicle or the like.

[0002]

2. Description of the Related Art Conventionally, there is known a device for preventing an obstacle from being caught in a power window of a vehicle or the like (Japanese Patent Application Laid-Open No. 50-47318). According to this device, in the upper window frame, the foreign matter detection switch is provided at a position where the upper side of the window does not contact, and when the foreign matter detection switch is turned on, the electric motor is rotated in the reverse direction by a predetermined rotation amount and the window glass Prevents obstacles such as human hands from getting caught in the window frame.

[0003]

In the above device, since the window is lowered only after the foreign matter detection switch is turned on, it is necessary to apply sufficient pressure to the foreign matter detection switch to turn it on. However, since the foreign matter detection switch is turned on by receiving the rising pressure of the window through the obstacle, when the obstacle is elastic such as a human arm, neck, etc., the rising pressure of the window causes the obstacle. It is absorbed by the object, and the time when the foreign matter detection switch turns on is delayed. Therefore, by the time the power window drops, considerable pressure is applied to the soft obstacle,
The crew will feel pain.

On the other hand, for example, when the size of an obstacle such as a finger is small, it is possible to easily prevent the finger from being caught by slightly lowering the power window. It is more convenient to raise.
On the other hand, if it is an obstacle such as a neck, the power window cannot be prevented from being pinched unless the power window is largely lowered.

The present invention has been made to solve the above problems, and an object thereof is to discriminate the type of an obstacle to be caught in a power window and open the power window according to the type. It is to improve the safety and usability of the power window by determining the movement amount.

[0006]

The structure of the invention for solving the above problems relates to a position detecting means for detecting a moving position of a window glass and a moving load of the window glass in a device for preventing pinching of a power window. The physical quantity detected by the load-related physical quantity detection means for detecting the physical quantity, the movement position detected by the position detection means, and the physical quantity detected by the load related physical quantity detection means are input, and the preset membership function and fuzzy rule are set. Using the fuzzy inference means for fuzzy inference of the opening movement amount of the window glass and the opening movement amount calculated by the fuzzy inference means, when the switch means is in the closing movement command state, the window glass is opened. And an opening control means for moving in the direction.

[0007]

The position detecting means detects the position of the window glass, and the current distance between the upper side of the window glass and the upper window frame can be known. The physical quantity related to the moving load of the power window is detected by the load-related quantity detection means. In the closed moving period of the power window, when there is no obstacle, the moving load of the power window is smaller than the predetermined value. On the other hand, when the obstacle is sandwiched between the upper side of the window glass and the window frame, the moving load of the power window increases according to the reaction force received by the window glass from the obstacle. That is,
The moving load of the power window is small at the beginning,
It gradually increases greatly. The increase rate of the moving load with respect to the moving amount, the increasing amount of the moving load at a predetermined moving amount, and the moving amount to the lock change according to the hardness of the sandwiched obstacle. Therefore, the hardness of the obstacle can be determined from the characteristics of the physical quantity related to the detected moving load.

Next, the detected movement position of the window glass and the movement load-related physical quantity are input to the fuzzy inference means. That is, the presence / absence of an obstacle and the type of the obstacle are determined from the relationship between the moving position and the physical quantity associated with the moving load. Then, when the window glass is in the closed movement state, the window glass is moved in the opening direction by the optimum opening movement amount according to the type of the obstacle which is the discrimination information.

[0009]

As described above, the present invention inputs the detected moving position of the window glass and the physical quantity related to the moving load, and uses the preset membership function and fuzzy rule to detect the window glass. Fuzzy inference means for fuzzy inferring the opening movement amount and opening control means for moving the window glass in the opening direction based on the calculated opening movement amount are provided. Therefore, the type of the obstacle that is sandwiched is fuzzy inferred, and the window glass is moved in the opening direction by the optimum amount according to the type of the obstacle, thus improving safety and usability. It

[0010]

EXAMPLES The present invention will be described below based on specific examples. FIG. 1 is a block diagram showing the configuration of the pinch prevention device according to this embodiment. The window glass 12 is opened and closed, that is, driven downward and upward by a motor 11 which is rotationally driven by the window glass opening / closing device 10. M
The RE sensor 20 is a sensor that outputs a time-series pulse signal (hereinafter referred to as an MRE signal) shown in FIG. 7 according to the rotation of the motor 11. The counter 31 detects the motor back electromotive force detector 8 when the window is in the fully closed position or the fully open position.
It is a device that inputs a reset pulse output from 0, resets the value, and inputs an MRE signal to add and count the number of pulses. The value of the counter 31 measures the current position P of the window glass. .

The current position P of the window glass 12 described above.
Since the distance L between the upper side of the window glass 12 in the fully opened state and the window frame is known, LP represents the opening width of the window glass 12. Therefore, if an obstacle exists between the upper side of the window glass 12 and the window frame, the size of the obstacle can be determined from the current position P.

The timer 32 inputs a clock pulse having a sufficiently short period according to the period of the MRE signal and the MRE signal, resets the count value in synchronization with the rising t _{1 of the} MRE signal as shown in FIG. 8, and resets the clock pulse. Are added and counted, and the clock pulse is subtracted and counted in synchronization with the rising t ₂ of the MRE signal. Latch circuit 35 is MRE
It is a circuit that latches the value of the timer 32 in synchronization with the falling edge t _{1 of the} signal. Therefore, the value of the latch circuit 35 is MR
The time difference T _D between the L level period T _L and the H level period T _H of the E signal is shown.

The time difference T _D represents the amount of decrease in the rotation speed of the motor 11 per unit time, that is, the deceleration. In other words, it represents the increase amount of the mobile load per unit time at the present time. Therefore, since the increase speed of the moving load represents the hardness of the obstacle, the hardness of the obstacle can be determined from the time difference T _D.

The timer 33 is a circuit which inputs the MRE signal and the clock pulse, resets the count value in synchronization with the falling and rising of the MRE signal, and counts the clock pulse by addition. Therefore, the value of timer 33 is MR
The elapsed time T _C from the level transition timing of the E signal is shown.

When the obstacle is very hard, the window glass immediately stops due to the obstacle, so that the pulse width of the MRE signal does not gradually increase as shown in FIG. 8, but as shown in FIG. The level of the MRE signal is immediately fixed to one level. Therefore, when the elapsed time T _C exceeds the predetermined value, it can be determined that the obstacle is extremely hard.

Thus, the hardness of the obstacle can be determined by using the time difference T _D and the elapsed time T _C.

The fuzzy inference device 34 inputs the current position P output from the counter 31, the time difference T _D output from the latch circuit 35, and the elapsed time T _C output from the timer 33, and according to a predetermined membership function and fuzzy rule. Fuzzy inference is performed to open the window open command signal S1.
And an open movement amount signal S2.

The selection circuit 60 selects the manual command signal S3 from the manual opening / closing switch 70 or the open movement amount signal S2 from the fuzzy reasoning device 34 according to the level of the window opening movement command signal S1 from the fuzzy reasoning device 34. Is selected to output a command signal to the window glass opening / closing device 10. Then, the window glass opening / closing device 10 causes the motor 11 to normally rotate (open movement) and reversely rotate (close movement) according to the command signal. The motor back electromotive force detector 80 is a device that detects the back electromotive force generated in the motor 11 and outputs a reset signal to the counter 31 when the detected back electromotive force exceeds a threshold value.

Next, the fuzzy inference device 34 will be described. Fuzzy rules are if I ₁ = L ₁ (k ₁ ) and I ₂ = L
₂ (k ₂ ) and, ─, and I _n = L _n (k _n ) then O ₁ = OL ₁ (J ₁ ), and
It is given in the IF-THEN form of O ₂ = OL ₂ (J ₂ ), and, ─, and O _m = OL _m (j _m ). I ₁ , ─, I _n are input variables,
n is the number of input variables. In this example, the input variables are
The current position P, the time difference T _D , the elapsed time T _C , and n is 3
Is. Also, O ₁ , ─, O _m are output variables, and m is the number of output variables. In this embodiment, the output variables are the entrapment risk X and the opening movement amount Y, and m is 2. Also, L ₁ (k ₁ ) ~
L _n (k _n ), OL ₁ (J ₁ ) to OL _m (j _m ) are symbols representing each class, that is, a membership function. The membership function is a function that represents the relationship between the input variable I or the output variable O and the goodness-of-fit M with which each of these variables can be determined to belong to each class.

FIG. 2 is an input membership function that defines the relationship between the current position P of the window glass and the goodness of fit M _P. The input membership function is a function for inferring the fitness M _P of the size of the obstacle from the current position P, and six kinds of functions representing the class of the size of the next obstacle are provided. FG: Finger AM: Arm NK: Neck HD: Head BD: Body OT: Other That is, the order in which the current position P of the window glass becomes smaller,
That is, the position of the maximum value of the fitness that is determined to be a finger, an arm, a neck, a head, a torso, or another object whose size increases in order is set as the opening of the window glass increases.

FIG. 3 shows an input membership function that defines the relationship between the time difference T _D which is an input variable and the goodness of fit M _D. The input membership function is the fitness M of the hardness class of the obstacle from the increasing speed of the moving load of the window glass.
It is a function that infers _D , and there are five kinds of functions related to the class indicating the hardness of the next obstacle. NO: Nothing with no trapped obstacles (Nothing) SST: Class with slightly obstructed obstacles (Small Sof
t) ST: Class where obstacles are soft (Soft) SHD: Class where obstacles are a little hard (Small Hard) HD: Class where obstacles are hard (Hard) That is, time difference T _D (increase in moving load) In order of increasing speed, none, slightly softer, softer, slightly harder, harder, the position of the maximum value of the fitness of the class is set.

FIG. 4 shows an input membership function that defines the relationship between the elapsed time T _C and the goodness of fit M _C. This input membership function is a function for inferring the fitness M _C of the hardness class of the obstacle from the increasing speed of the moving load of the window glass, and the following two types of hardness-related functions are provided. VSH: A class where obstacles are very hard (Very Supper
Hard) NVS: A class where obstacles are not very hard (Not Very
Supper Hard)

FIG. 5 shows an entrapment danger level X which is an output variable,
It is an output membership function that defines the relationship with the goodness of fit M _X that represents the degree to which each variable value belongs to each class, and the following two types of functions are provided. The degree of pinching risk X is
It is a value determined from the fitness of each fuzzy rule that is determined according to the fitness of the obstacle size and hardness class.
It represents the degree of truth that indicates whether the obstacle is truly a finger, arm, neck, head, torso, etc. SL: Class that the entrapment risk is low (Small) BG: Class that the entrapment risk is high (Big) That is, as the entrapment risk X increases, the fitness M _X of the function SL decreases, As the entrapment risk X decreases, the fitness M _X of the function BG increases.

FIG. 6 is an output membership function which defines the relationship between the open movement amount Y of the window glass which is an output variable and the goodness of fit M _Y which represents the degree to which the value of the variable belongs to each class. 4 types of output functions are provided.
The open movement amount Y is a value determined from the conformity of each fuzzy rule determined according to the conformity of the size of the obstacle and the hardness class, and the obstacle is a finger, an arm, a neck, or a head. , The optimum amount of opening movement of the window glass, which is determined based on the degree of truth such as the body, that is, the optimum amount of return. ZR: A class for which the amount of opening movement is zero (Zero) SBK: A class for slightly returning the amount of opening movement (Small Back) BK: A class for returning the amount of opening movement (Back) BBK: When a large amount of opening movement is returned The position of the maximum value of the fitness of the class to be set to 0, a little back, a back, and a large back is moving in the direction in which the open movement amount Y increases.

Next, an example of fuzzy rules will be shown. Rule1: IF P = NK and T _D = ST THEN X = B
G and Y = BBK Rule2: IF P = FG and T _D = SHD THEN X =
BG and Y = SBK Rule3: IF T _C = VSH THEN X = BG and Y =
ZR Rule4: IF T _D = NO THEN X = SL

Rule1 is the current position P (size of obstacle)
Is in the neck class NK, and the time difference T _D (hardness of the obstacle) is in the soft class ST, the degree of truth that the obstacle is the neck is high. A class B in which the opening movement amount Y of the
It means to be BK. Rule2 is the current position P
(Size of obstacle) belongs to the finger class FG, and the time difference T _D
If (hardness of obstacle) belongs to class SHD that is a little hard,
Since the degree of truth that the obstacle is a finger is high, it means that the entrapment risk X is set to a large class BG, and the opening movement amount Y of the window is set to a class SBK that rotates slightly backward. Also, R
If the elapsed time T _C (hardness of the obstacle) belongs to the class VSH where the elapsed time T _C is very hard, ule3 means that the window is suddenly locked, so the pinching risk X is set to the large class BG,
This means setting the class ZR in which the opening movement amount Y of the window is stopped. In Rule 4, if the time difference T _D (hardness of obstacles) belongs to the class NO, the moving load of the window glass is a normal load and no obstacles exist, so the pinching risk X is small. This means SL.

The fuzzy inference operation is executed as follows. For one rule, the value (fitness M _P , M _D , M _C ) of the membership function defined by the rule is calculated for each input variable. Then, the value of the membership function is the smallest among all the input variables (min-operation)
Is the fitness M of the fuzzy rule. Such a calculation is executed for all fuzzy rules, and each goodness of fit M is calculated for each fuzzy rule. Next, for each fuzzy rule, in the graph showing the output membership function defined in the then part, the area of the output variable having a value equal to or less than the calculated fitness M is determined. Such an operation is executed for all rules, and the maximum area is determined by superimposing the areas obtained from each rule (MAX- operation). Then, the value of the output variable corresponding to the position of the center of gravity of the maximum area is output. Fuzzy inference is one example of performing such an arithmetic operation.

This fuzzy control will be described more specifically. In the following example, it is assumed that each variable has a maximum value of 1 and a minimum value of 0, and the fitness is normalized to a value of 1 or less. Regarding the input variables, the current position P is 0.84, the time difference T _D is 0.38,
The elapsed time T _C is 0.12. Fuzzy rules are applied in sequence. When the current position P is 0.84, the current position P is the class N
It belongs only to K or AM, not to any other class. Therefore, in the fuzzy rule that defines that the current position P belongs to a class other than the classes NK and AM, the value of the membership function of the current position P is 0. So in this case,
As a result of the min-operation, the goodness of fit M of the fuzzy rule becomes 0 regardless of the values of other input variables.

When the time difference T _D is 0.38, the time difference T
_D belongs to the classes ST and SHD and does not belong to any other class. Therefore, in the fuzzy rule that defines that the time difference T _D belongs to a class other than the classes ST and SHD, the time difference T _D
The value of the membership function of _D is 0. Therefore, in this case, the goodness of fit M of the fuzzy rule becomes 0 as a result of the min-calculation, regardless of the values of other input variables.

When the elapsed time T _C is 0.12, the elapsed time T
_C belongs to NVS, not VSH. Therefore, in the fuzzy rule that defines that the elapsed time T _C belongs to VSH, the value of the membership function of the elapsed time T _C is 0. Therefore, in this case, the goodness of fit M of the fuzzy rule becomes 0 as a result of the min-calculation, regardless of the values of other input variables.

Therefore, when each input variable has the above value, the current position P belongs to the class NK or AM and the time difference T _D
Belongs to the class ST or SHD, and only the fuzzy rule that defines that the elapsed time T _C belongs to NVS is the object of calculation of the goodness of fit.

For the current position P = 0.84, the value of the input membership function NK is 0.65, and the input membership function A is
The value of M is 0.35. Also, regarding the time difference T _D = 0.38,
The value of the input membership function ST is 0.5 and the value of the input membership function SHD is 0.5. Also, the elapsed time T _C
For = 0.12, the value of the input membership function NVS is 1. The suitability of each rule is calculated based on these values. The goodness of fit of each rule is determined by the minimum value of the input membership functions in the input variable (MIN- operation). For example, if the IF section is P = NK and T _D
The suitability of the rule defined as = ST is 0.5, which is the minimum value of 0.65 and 0.5. IF part is P = NKand T _D = SH
The goodness of fit of the rule defined as D is 0.5, which is the minimum of 0.65 and 0.5. The IF part has a minimum conformity of 0.35 and 0.5 for the rule that P = AM and T _D = ST.
It is 0.35. In the IF part, the conformity of the rule that P = AM and T _D = SHD is 0.35 and the minimum value of 0.5, which is 0.
35.

Next, in each output membership function defined in the THEN part of each rule, a region equal to or less than the goodness of fit of the rule is set. For example, IFP = NK
Regarding the rule that andT _D = ST THEN X = BG and Y = BBK, since the goodness of fit is 0.5, the function BG,
In BBK, regions S1 and A1 of 0.5 or less are set as shown in FIGS. Also, IFP = AM and
Regarding the rule of T _D = ST THEN X = BG and Y = BK, the goodness of fit is 0.35. Therefore, in the functions BG and BK, regions S2 and A2 of 0.35 or less are set as shown in FIGS. 5 and 6. It In this way, regions are set for the output membership function of each rule.

Next, the union of these areas (MAX-operation)
Is calculated, and the barycentric positions g1 and G2 of the union region are calculated as shown in FIGS. Center of gravity position g
From 1, G2, the value of pinching risk X and opening movement amount Y is 0.65, 0.
Determined as 74.

The pinching risk X thus determined is 0.5.
At the above time, the open movement command signal S1 is output to the switch 60. Further, an open movement amount signal S2 representing the open movement amount Y is output to the switch 60. The switch 60 is a manual opening / closing switch 70.
If the opening movement command signal S1 is input while the manual opening movement command signal S3 is input from the
Is output to the window glass opening / closing device 10. As a result, the window glass opening / closing device 10 rotates the motor 11 in the opening direction by the commanded value, and the window glass 12 is moved in the opening direction by the commanded value.

Next, the second embodiment will be described. First
In the embodiment, the output signals of the fuzzy inference device 34 are the open movement command signal S1 and the open movement amount signal S2.
As shown in 0, the speed command signal S4 may be output to the window glass opening / closing device 10 so as to increase the rotation speed of the motor 11 in proportion to the entrapment risk X.
In this case, the window glass 12 is moved in the opening direction at a higher speed as the risk of the obstacle getting caught is higher.

Next, a third embodiment will be described. First
In the embodiment, as a means for detecting the size of the obstacle, the motor 11 is used to obtain the position of the window glass 12.
Was indirectly determined from the number of pulses using the MRE sensor 20 from the rotation amount of the. However, in addition to the MRE sensor 20,
It may be obtained from another means for measuring the rotation amount. In addition, as a means for directly obtaining the position of the window glass 12, an optical sensor or an ultrasonic sensor is arranged at the upper end of the window frame to generate a sound wave,
Light is applied to the upper piece of the window glass 12, reflected sound waves,
You may make it detect the present position of the window glass 12 by reflected light.

Next, a fourth embodiment will be described. First
In the embodiment, in order to detect the hardness of the obstacle, the hardness is measured based on the increase amount of the pulse width of the MRE signal per unit movement amount from the MRE sensor 20, as shown in FIG. Alternatively, the load current flowing through the motor 11 may be detected from the motor back electromotive force detector 80, and the hardness of the obstacle may be detected from the magnitude of the load current.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a pinching prevention device according to a first specific example of the present invention.

FIG. 2 is a characteristic diagram showing an input membership function relating to the size of an obstacle used in the fuzzy inference device of the jamming prevention device according to the first example.

FIG. 3 is a characteristic diagram showing an input membership function relating to the hardness of an obstacle used in the fuzzy inference device of the jamming prevention device according to the first example.

FIG. 4 is a characteristic diagram showing an input membership function relating to the hardness of an obstacle used in the fuzzy inference device of the jamming prevention device according to the first example.

FIG. 5 is a characteristic diagram showing an output membership function relating to a risk of pinching used in the fuzzy inference device of the pinching prevention device according to the first example.

FIG. 6 is a characteristic diagram showing an output membership function relating to an opening movement amount used in the fuzzy inference device of the jamming prevention device according to the first example.

FIG. 7 is a characteristic diagram showing an MRE signal for obtaining the current position of the window glass in the pinching prevention device according to the first example.

FIG. 8 is a characteristic diagram showing a change characteristic of the MRE signal for obtaining the hardness of an obstacle in the jamming prevention device according to the first example.

FIG. 9 is a characteristic diagram showing change characteristics of an MRE signal for obtaining hardness of an obstacle in the pinch prevention device according to the first example.

FIG. 10 is a block diagram showing a configuration of a pinch prevention device according to a second specific example of the present invention.

FIG. 11 is a block diagram showing a configuration of a pinching prevention device according to a fourth specific example of the present invention.

[Explanation of symbols]

 20 ... MRE sensor (position detection means) 31 ... Counter (position detection means) 32 ... Timer (load-related material amount detection means) 33 ... Timer (load-related material amount detection means) 34 ... Fuzzy inference device (fuzzy inference means) 35 ... Latch Circuit (load-related material amount detecting means) 60 ... Selector (opening control means) 10 ... Window glass opening / closing device (opening control means) 11 ... Motor 70 ... Manual opening / closing switch (switch means) 36 ... Load current detector (load-related material amount) Detection means)

Claims (1)

[Claims] 1. A pinch prevention device for preventing an obstacle from being pinched between the window glass and the window frame in a power window that opens and closes a window glass of a vehicle by operating a switch means by rotating a motor. Position detecting means for detecting a moving position of the window glass, load-related material amount detecting means for detecting a physical quantity related to a moving load of the window glass, the moving position detected by the position detecting means, and the load Fuzzy inference means for inputting the physical quantity detected by the related matter quantity detecting means, and fuzzy inferring the opening movement amount of the window glass by using a preset membership function and fuzzy rule, and the fuzzy inference means. Based on the opening movement amount calculated by When in the state of the movement command, the safety device of the power window, characterized in that it comprises an open control means for moving the window glass in the opening direction.