JPH10274577A - Contact sensation evaluation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To quantify a sitting sense of a person at the time of coming in contact with an object, with high accuracy. SOLUTION: Contact pressure distribution at the time of a testee being seated on a vehicular seat is measured, and the measured sitting pressure distribution is converted into sensitive pressure distribution of the testee by a conversion expression taking account of a spatial masking characteristic of skin sensation of a human body. A plural kinds of feature quantities are computed from the sensitive pressure distribution, and with the functional evaluation value of sitting sensation of the testee as an object parameter, multiple regression analysis is repeated while changing the feature quantity used as an explanation parameter out of the computed feature quantities, thus obtaining a multiple regression expression (evaluation expression) maximum in the contribution, rate. The sitting sensation of the vehicular seat is quantified for evaluation using the evaluation expression.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a touch sensation evaluation method, and more particularly to a touch sensation evaluation method for evaluating a touch sensation when a human body surface comes into contact with a target object such as a vehicle seat.

[0002]

2. Description of the Related Art The performance of a vehicle seat greatly affects comfort when the vehicle is in a vehicle. Because comfort is a human feeling,
In determining the specifications of the vehicle seat, in addition to performing physical measurements such as the amount of deflection and vibration characteristics by trial manufacture of the seat, the subject is seated on the prototype seat, and the contact formed by the subject by sitting down By performing a so-called sensory evaluation, the sensation (more specifically, the sitting sensation) is evaluated for each evaluation item related to the sitting sensation such as, for example, a “seat sensation”, a “soft sensation”, a “tension”, and a “fit”. It is common to evaluate the seating sensation of a vehicle seat. However, the result of the sensory evaluation depends on the subject's subjectiveness, and thus there is a problem that the reliability of the evaluation result of the sitting sensation based on the sensory evaluation is low.

In connection with the above, "Study of Sitting Comf
ort of Automotive Seats ", FISITA'94 Technical Paper
s 945243, Engineering for the Costmers, p90 (1994)
(Presenter: Tsinghua Univ. (China)) measured the contact pressure distribution while the subject was sitting on the vehicle seat, and measured the features obtained from the contact pressure distribution and the seating of the subject on the vehicle seat. A technology that directly associates sensory evaluation values representing sensation with statistical methods to obtain an evaluation formula for quantifying the sitting sensation of a occupant sitting on a vehicle seat from the contact pressure distribution is described. Have been.

As described above, the reliability of the evaluation result in the evaluation of the sitting sensation based on the sensory evaluation is obtained by quantitatively obtaining the sitting sensation of the occupant of the vehicle seat from the contact pressure distribution using the evaluation formula. In order to improve the performance, it is not necessary to perform a complicated operation such as performing a sensory evaluation of the vehicle seat by a large number of subjects.

[0005]

However, the contact pressure distribution when the occupant is seated on the seat is considerably different from the pressure distribution actually felt by the occupant. Since the seating sensation of the subject is formed based on the sensation of the pressure distribution and the like actually felt by the subject, the feature amount obtained from the contact pressure distribution in a state where the subject is seated on the vehicle seat, The correlation with the sensory evaluation value indicating the feeling of sitting on the vehicle seat by the subject is low. For this reason, even if an evaluation formula is obtained by directly associating the characteristic amount obtained from the contact pressure distribution with the sensory evaluation value representing the subject's sitting sensation with respect to the vehicle seat, the accuracy of quantifying the sitting sensation with respect to the vehicle seat is high. Was inadequate.

[0006] In addition, the seating sensation of the subject is described in more detail below.
While the sensation such as the distribution of pressure actually felt by the subject is formed by comparing it with a psychological scale such as a preference for seating sensation and evaluation criteria, the above-described technique quantifies the sensation of sitting on the vehicle seat. In determining the evaluation formula for the conversion, the variance of the psychological scale such as individual occupant's preference and evaluation criteria for the sitting sensation is not taken into account, so the evaluation formula is calculated using the sensory evaluation value by a specific subject. , And even if the seating sensation for the vehicle seat is quantified using the evaluation formula, the quantified sitting sensation is different from the specific subject in the psychological scale of the sitting sensation. And there was a big difference.

Further, in the above technique, the contact pressure distribution is measured after a seated person sits on the vehicle seat and becomes stationary (steady state). Since the change in the contact pressure distribution during the period up to (transient state) is not considered at all, the seating sensation for the vehicle seat based on the pressure change in the transient state felt by the subject is not considered. There was also a problem that it was not reflected in the quantification of

[0008] The problem described above is not limited to quantification of a seating sensation with respect to a vehicle seat or other seats. For various objects that are expected to come into contact with the human body, the human body contacts the object. This is a problem that similarly arises when quantifying the contact sensation when doing.

The present invention has been made in view of the above facts, and has as its object to provide a contact sensation evaluation method capable of quantifying the contact sensation of a human when touching a target object with high accuracy. is there.

[0010]

In order to achieve the above object, a contact sensation evaluation method according to the present invention measures a contact pressure distribution when a human body comes into contact with a target object, and calculates the contact pressure distribution. It converts into the sensed pressure distribution of the human body in consideration of the spatial masking characteristics of the skin sensation of the human body, and from the sensed pressure distribution obtained by the conversion, the sensory evaluation and correspondence of the contact sensation when the human body comes into contact with the target object Calculate the feature quantity that can be attached,
The calculated feature amount is associated with a sensory evaluation of a contact sensation when the human body contacts the target object, and an evaluation formula for quantifying a contact sensation when the human body contacts the target object is determined. The contact sensation is quantified and evaluated using the evaluation formula.

When the stimulus for pressing the skin is applied to two points on the skin at the same time (simultaneous two-point pressure stimulation), the human skin senses the localization of each of the two points ambiguously depending on the distance between the two points. Characteristics, so-called spatial masking characteristics. For this reason, it is considered that the spatial masking characteristic also acts when the skin of the human body senses the contact pressure distribution on the contact surface when the human body comes into contact with the target object as pressure sense (sensed pressure). Then, the distribution of the sensed pressure received is collated with psychological measures such as preferences and evaluation criteria for the contact sensation,
A contact sensation is formed for a person in contact with the target object.

For this reason, according to the present invention, the contact pressure distribution when the human body comes into contact with the target object is measured, and the measured contact pressure distribution is taken into consideration in consideration of the spatial masking characteristic of the human body's skin sensation. It converts the distribution into a distribution, and calculates a feature amount that can be associated with the sensory evaluation of the contact sensation when the human body comes into contact with the target object from the distribution of the sensed pressure obtained by the conversion. The sensed pressure distribution is a pressure distribution that the human body actually feels when the human body surface comes into contact with the target object. Therefore, the correlation with the contact sensation (and the sensory evaluation of the contact sensation) when the human body comes into contact with the target object is extremely high. The characteristics of the sensitive pressure distribution calculated from the sensitive pressure distribution are also very high.
The correlation with is very high.

According to the present invention, the above-mentioned feature amount, which has a very high correlation with the contact sensation when the human body comes into contact with the target object, is associated with the sensory evaluation of the contact sensation when the human body comes into contact with the target object. Then, an evaluation formula for quantifying a contact sensation when the human body comes into contact with the target object is obtained, and the contact sensation is quantified and evaluated using the evaluation formula. Therefore, it is possible to quantify the human sense of contact when touching the target object with high accuracy.

The conversion from the contact pressure distribution to the sensitive pressure distribution may be performed, for example, on the entire contact surface between the human body and the target object. If the seat is the target object, a line along the line extending from the position below the tail bone of the human body, from the position below the ischial tuberosity to the position below the knee, or the line passing through the position below the left and right ischial tuberosities of the human body , Etc.), the contact pressure distribution may be converted to obtain the sensitive pressure distribution. Since the sensitive pressure increases in places where the contact pressure is high,
Even if a sensitive pressure distribution is obtained only for a portion where the contact pressure is high, and the evaluation formula for quantifying the contact sensation by calculating a feature amount from the sensitive pressure distribution, the quantification accuracy of the contact sensation decreases. There is almost no case, and the calculation can be simplified without lowering the accuracy of quantifying the contact sensation.

When converting a contact pressure distribution along a predetermined line having a high contact pressure on a contact surface between a human body and a target object into a sensitive pressure distribution, for example, a plurality of measurement points i (i) arranged along the predetermined line = 1, 2, ... N, where the N is a representative of the contact pressure distribution along said predetermined line by the contact pressure P _i in each of the several) of measuring points of the contact pressure, contact pressure P _i at each measurement point Based on the following equation (1), each of the sensing pressures SP _j at a number of virtual points j (j = 1, 2,... M, where M is the number of virtual points) is calculated on the predetermined line. By doing so, it is possible to convert the contact pressure distribution along the predetermined line into the sensitive pressure distribution along the predetermined line.

[0016]

(Equation 1) The above equation (1) is determined based on the spatial masking characteristic of the skin sensation of the human body, where x _j is the position of a virtual point on the line with respect to a predetermined position, and x _i is the predetermined position. Indicates the position of the reference measurement point. Σ is a constant (standard deviation) representing the spread of the pressure sensitivity distribution, and is a two-point discrimination threshold of skin sensation when two points of simultaneous pressure stimulation are applied to the skin of the human body (two points when stimulation is given. The distance is determined based on the shortest distance that is felt separately. According to the above equation (1), the measured contact pressure distribution can be accurately converted into the sensed pressure distribution of the human body. Further, by transforming the expression (1), it is possible to convert the contact pressure distribution within a predetermined range having a two-dimensional spread into a sensitive pressure distribution within the predetermined range.

In consideration of the accuracy of quantification of the contact sensation by the evaluation formula, it is preferable to use a plurality of types of feature amounts, for example, a feature amount representing the maximum value of the sensed pressure in the sensed pressure distribution, A feature amount (for example, a half-value width or the like) representing the spread of a portion where the pressure distribution is equal to or larger than a predetermined value, a feature amount representing the gradient of the received pressure distribution, or the like can be used.

When a plurality of types of feature amounts are used as described above, the degree of correlation with the contact sensation (and the sensory evaluation of the contact sensation) differs for each feature amount, and the correlation with the sensory evaluation of the contact sensation is different. It is conceivable that there is no feature amount or an extremely low feature amount. Therefore, when calculating the evaluation formula by associating the feature amount with the sensory evaluation of the contact sensation, the feature amount (contact The evaluation formula may be obtained by extracting only the feature amount having a correlation with the sensory evaluation of the sensation, or without extracting the feature amount that can be associated with the sensory evaluation of the touch sensation. The evaluation formula may be determined so that the effect on the value obtained by quantifying the contact sensation is zero or extremely small for a feature amount having no or extremely low correlation between the two.

Further, in the present invention, in measuring the contact pressure distribution, the contact pressure is measured for at least a period from when the contact between the human body and the target object is started to when the contact between the human body and the target object is in a steady state. The pressure distribution is measured (it may be measured intermittently a plurality of times or may be measured continuously), and the measured contact pressure distribution is converted into the pressure distribution of the human body. It is preferable to calculate a characteristic amount related to a temporal change of the distribution, and to obtain an evaluation expression for quantifying the contact sensation using the characteristic amount.

The sensory evaluation of the contact sensation is generally performed for a plurality of evaluation items (for example, "seat feeling", "soft feeling", "tension feeling", "fit feeling", etc. for a seat). Among the plurality of evaluation items, there is an evaluation item for which a sensory evaluation is performed in consideration of a change in the pressure sensitivity during a period (transient state) until the contact between the human body and the target object becomes a steady state. There are many cases. On the other hand, as described above, the characteristic amount related to the temporal change of the received pressure distribution is also calculated, and the evaluation formula is calculated using the characteristic amount. Also for the evaluation item to be evaluated, it is possible to accurately quantify the contact sensation corresponding to the evaluation item.

Further, the measurement of the contact pressure distribution and the acquisition of the sensory evaluation of the contact sensation are performed for a plurality of subjects or a plurality of target objects, and a large number of data of the sensory evaluation of the contact pressure distribution and the contact sensation are collected. After that, when an evaluation formula for quantifying the contact sensation is obtained from this data group, it is necessary to determine whether the psychological scale of the contact sensation is similar for each of the data of the subjects or the distribution of the sensed pressure (or the distribution of the contact pressure). It is preferable that the data group is grouped for each data having a similar pattern, and an evaluation formula for quantifying a contact sensation is obtained for each group.

Psychological scales such as preferences and evaluation criteria for the contact sensation are different from subject to subject, but subjects with similar psychological scales for the contact sensation have similar touch sensations themselves. Presumed. In addition, if the pattern of the contact pressure distribution is significantly different due to factors such as the physique of the subject or the size of the target object, the pattern of the sensed pressure distribution changes accordingly, and the tendency of sensory evaluation of the contact sensation tends to increase. May significantly vary depending on the contact pressure distribution pattern.

On the other hand, a data group is grouped for each data having a similar psychological measure of the contact sensation, or for each data having a similar pattern of the sensed pressure distribution (or the contact pressure distribution). If an evaluation formula for quantifying sensation is obtained in group units, the degree to which the tendency of the contact sensation based on patterns such as the subject's psychological scale and the distribution of the sensed pressure on touch sensation is reflected in the evaluation formula for each group Will be higher. Therefore, by quantifying the contact sensation using the evaluation formula obtained for each group, the contact sensation can be quantified with higher accuracy for each group.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings. In the following, the seating sensation is obtained from the contact pressure distribution in the seat cushion when the subject is seated on the vehicle seat as the target object and the result of the sensory evaluation of the subject's sitting sensation, which is one of the contact sensations of the present invention. An example will be described in which an evaluation formula for quantifying is determined, and the seating sensation is quantified and evaluated using the evaluation formula. In the following, description will be made using numerical values that do not hinder the present invention, but the present invention is not limited to the numerical values described below.

FIG. 1 shows a seat sensation sensation evaluation apparatus 10 to which a touch sensation evaluation method according to the present invention can be applied. The seat occupant sensation evaluation device 10 includes a pressure sensor array 12 in which a large number of pressure sensors are arranged in a matrix on a sheet-shaped base material. The pressure sensor array 12 is provided on the seat cushion portion 14 of the vehicle seat 14.
A is placed on the bearing surface of A. The pressure sensor array 12 is connected to an input / output port 1 of a computer 18 via a driver 16.
8D, a signal representing the contact pressure output from each pressure sensor of the pressure sensor array 12 is transmitted to the driver 1
6 and converted into digital data by the computer 1
8 is output.

The computer 18 has a CPU 18A, an RO
M18B, RAM18C, input / output port 18D,
These are connected to each other via a bus including a data bus and a control bus. I / O port 18D
Has a display 20 for displaying various information;
A keyboard 22 for an operator to input data and various instructions, a storage device 24 having a hard disk and other non-volatile storage media, and a printer 26 for printing processing results and the like are connected.

The ROM 18B contains the computer 1
8 is stored as a program for causing the device 8 to function as a seating sensation evaluation device. This program is read from the ROM 18B and executed by the CPU 18A when the power of the computer 18 is turned on and an instruction to execute a feature amount calculation / sensory evaluation value storage process or an evaluation formula creation process described later is given.

Next, as an operation of the present embodiment, referring to the flowchart of FIG. 2, each time the subject is seated on the vehicle seat 14 and the sensory evaluation of the seating sensation is performed, the characteristic amount calculation The sensory evaluation value storage processing will be described. In step 100, the vehicle seat 14
It is determined whether or not the subject is seated at. This determination may be made based on, for example, whether or not the detected value of the contact pressure by any one of the pressure sensors of the pressure sensor array 12 has become greater than 0, or the subject starts sitting on the vehicle seat 14. At the same time, the determination can be made based on whether or not a predetermined key of the keyboard 22 pressed by the operator is pressed.

If the determination in step 100 is affirmative, the process proceeds to step 102. Step 102 to step 11
In 2, data representing the distribution of the contact pressure when the subject is seated on the vehicle seat 14 is taken in. As is clear from FIG. 4, the contact pressure when the subject is seated on the vehicle seat 14 is: The value fluctuates greatly until a predetermined time elapses from the start of sitting (transient state), and changes to a substantially constant value after a predetermined time elapses from the start of sitting (steady state).
The change in the contact pressure in the transient state (more specifically, the change in the perceived pressure due to the change in the contact pressure) is expressed as a “soft feeling” or a “tension” among the evaluation items of the sensory evaluation of the sitting sensation. Relatedly, the contact pressure in a steady state (more specifically, the sensed pressure) is related to “fitness” among the evaluation items of the sensory evaluation of the sitting sensation.

Therefore, in the next step 102 and subsequent steps,
As shown in FIG. 4, a state in which a predetermined time t _{1 has} elapsed from the start of sitting (hereinafter, referred to as state 1), a state in which a predetermined time t _{2 has} elapsed (hereinafter, referred to as state 2), and a state in which a predetermined time t _{3 has} elapsed (hereinafter, state) , State 3) is taken in (each state 1 and state 2 correspond to a transient state, and state 3 corresponds to a steady state).

That is, in step 102, the test subject has been seated on the vehicle seat 14 for a predetermined time t _1.
Is determined. If the determination is affirmative, the process proceeds to step 104, in which data representing the contact pressure detected by each pressure sensor of the pressure sensor array 12 is captured via the driver 16, and the captured data is captured by the seat cushion portion 14A in state 1. (An example of the contact pressure distribution on the seat surface in the state 1 is shown in FIG. 5A) in the storage device 24. In FIG. 5,
The distribution of the contact pressure is represented by an image in which the density is reduced as the contact pressure increases.

In the next step 106, it is determined whether or not a predetermined time t ₂ has elapsed since the subject started sitting on the vehicle seat 14. If the determination is affirmative, step 10
8, the data representing the contact pressure detected by each pressure sensor of the pressure sensor array 12 is fetched, and the fetched data is stored in the contact pressure distribution on the seat in state 2 (on the seat in state 2). An example of the contact pressure distribution is shown in FIG.

In the next step 110, it is determined whether or not a predetermined time t ₃ has elapsed since the test subject started sitting on the vehicle seat 14. If the determination is affirmative, step 11
2, the data representing the contact pressure detected by each pressure sensor of the pressure sensor array 12 is fetched, and the fetched data is used as the contact pressure distribution on the seat in state 3 (on the seat in state 3). An example of the contact pressure distribution is shown in FIG. As described above, the data representing the contact pressure distribution in the state 1, the state 2, and the state 3 are stored in the storage device 24, respectively.

Note that steps 100 to 112 described above are performed.
Corresponds to “measurement of contact pressure distribution when a human body comes into contact with a target object” in the present invention.

By the way, the contact pressure distribution in a state where a human is seated on a seat (steady state) is a portion of the contact portion between the human body and the seat on the seat surface corresponding to the skeleton of the human body (particularly, Since the contact pressure increases at the ischial tuberosity (the portion corresponding to FIGS. 6A and 7), the portion corresponding to the skeleton of the human body is considered to be a main part of the contact pressure distribution.

Therefore, in step 114, data representing the contact pressure distribution in state 3 is fetched from the storage device 24, and the portions corresponding to the left and right ischial tuberosities (lower part of the ischial tuberosities) of the subject from the contact pressure distribution in state 3 Position on the seat, corresponding to the left and right knees of the subject (lower knee)
Are determined on the bearing surface. Specifically, it is determined that the portion where the contact pressure is maximized within a predetermined range on the bearing surface where the left or right ischial tuberosity is estimated to be present is determined to be the ischial tuberosity, and the left and right ischial tuberosities are determined. It is performed for each lower part of the nodule, and it is determined that the part where the contact pressure is maximum within a predetermined range on the seat surface where the left or right lower part of the knee is estimated to be present is determined to be the lower part of the left and right knees. For each.

In the next step 116, as the main part (line) of the contact pressure distribution on the seat surface of the seat cushion portion 14A, the lower part of the left ischial tuberosity from the part corresponding to the coccyx (lower coccyx) of the subject. And the line extending from the lower part of the coccyx to the lower part of the right knee through the lower part of the right ischial tuberosity (the front-rear line on the right side of the seat) Then, a line (left-right direction line) extending through the lower part of the left and right ischial tuberosities is set (see FIG. 6B).

After the lower part of the ischial tubercle is determined in steps 114 and 116, the opening angle of the foot is constant within a predetermined range on the seat surface where it is estimated that the left or right lower part of the knee is present. The lower part of the knee may be determined by setting the left and right front-rear direction lines so as to pass through the lower part of the ischial tuberosity according to the angle.

In the next step 118, the value of the counter n is set to 1, and in the next step 120, data representing the contact pressure distribution in the state n (initial state 1) is fetched from the storage device 24 and set in the previous step 116. The contact pressure distribution along the front-rear direction line and the left-right direction line on the left and right sides of the seat is extracted. Thereby, as an example, a contact pressure distribution as shown in FIG. 7A is extracted as a contact pressure distribution along the front-back direction line, and as a contact pressure distribution along the left-right direction, as shown in FIG. Data representing the contact pressure distribution as shown is extracted.

The data representing the contact pressure distribution extracted above includes a large number of measuring points i (i = 1, 2,... N, where N is a measuring point of the contact pressure) arranged along the front-back direction line and the left-right direction line. Is the data representing the contact pressure distribution along the line by the contact pressure P _{i in} each of
In 20, based on the contact pressure P _i at each of a number of measurement points i along the longitudinal direction line and the left-right direction line, in accordance with the following equation (2), a number of virtual points arrayed in the front-rear direction line and the horizontal direction line on j (j = 1,2, ... M , where M is the number of virtual points, which may be a same number as N) by each calculating sensitive pressure SP _j in the contact pressure distribution along the longitudinal direction line and the horizontal direction line Are respectively converted into a pressure-sensitive distribution.

[0041]

(Equation 2) The above equation (2) is determined based on the spatial masking characteristic of the skin sensation of the human body, where x _j is the position of a virtual point with reference to a predetermined position on the front-back direction line or the left-right direction line, x _i Represents the position of the measurement point with reference to the predetermined position. Further, σ is a constant (standard deviation) representing the spread of the received pressure distribution, and its value is determined based on the two-point discrimination threshold of the skin sensation when two-point simultaneous pressure stimulation is applied to the skin of the human body. In the present embodiment, the two-point discrimination threshold at the buttocks of the human body is assumed to be 6 cm, and σ = √10 based on this value.

As described above, the contact pressure distribution along the front-back direction line and the left-right direction line can be accurately converted into the distribution of the subject's sensed pressure along the front-rear direction line and the left-right direction line. FIG. 8 shows, as an example, a contact pressure distribution represented by a contact pressure P _i at each of a large number of measurement points i along the longitudinal line, and a large number of along the longitudinal line calculated from the equation (2). Sensing pressure S at each of the virtual points
An example of the distribution of the received pressure represented by P _j is shown. It can be understood that the distribution of the received pressure is a smooth change compared to the distribution of the contact pressure due to the spatial masking characteristic of the skin sensation. The vertical axis in FIG. 8 indicates the contact pressure P _i and the sensed pressure SP _j
And "seat pressure".

In the next step 124, state 1 to state 3
It is determined whether or not the sensed pressure distribution along the front-rear direction line and the left-right direction line has been determined for each state. If the determination is negative, the routine proceeds to step 126, where "1" is added to the counter n, and the routine returns to step 120. Therefore, the processing of steps 120 and 122 is executed for all of the states 1 to 3, and the pressure-sensitive pressure distribution along the front-back direction line and the left-right direction line in each state is calculated.

Note that steps 114 to 126 described above are performed.
Corresponds to "converting the measured contact pressure distribution to the human body's perceived pressure distribution in consideration of the spatial masking characteristic of the skin sensation of the human body" in the present invention.

If the determination in step 124 is affirmative, the process proceeds to step 128, and based on the detected pressure distribution along the front-rear direction line and the left-right direction line in each of the states 1 to 3 obtained above, According to the equations (3) to (10),
Eight types of feature values PP, PS, PV,
TV, PSD, PVT, TVT, and PSDT are calculated.

PP = f ₁ (Pmax) = (Pmax _Left + Pmax _Right ) / 2 (3) PS = f ₂ (Ps) = (Ps _Left1 + Ps _Left2 + Ps _Left3 + Ps _Left4 + Ps _Right1 + Ps _Right2 + Ps _Right3 + Ps _Right4 ) / 8 (4) PV = f ₃ (Pmax _, Pv) = (πPv _Left1 · Pv _Left2 / Pmax _Left + πPv _Right1 · Pv _Right2 / Pmax _Right ) / 8 (5) TV = f ₄ (Pmax, Ppk, Lpk) = Ppk _Left / (Lpk _Left · Pmax _Left ) + Ppk _Right / (Lpk _Right · Pmax _Right ) (6) PSD = f ₅ (Pmax, Plr, Llr) = Plr / Llr / max (Pmax _Left , Pmax) _{Right) ... (7) PVT =} f 6 (PV, t) = PV (t 2) -PV (t 1) ... (8) TVT = f 7 (TV, t) = TV (t 2) -TV (t ₁ )... (9) PSDT = f ₈ (PSD, t) = PSD (t ₃ ) −PSD ( t ₁ ) (10) The suffixes “ _Left ” and “ _Right ” in each of the above equations
Represents the left and right sides of the seating surface, respectively.
The characteristic amount PP obtained by the equation (3) represents the pressure felt by the occupant near the ischial tuberosity, and Pmax in the equation (3)
_Left is the maximum pressure sensitivity at the left ischial tuberosity, Pma
x _Right is the maximum value of the pressure sensitivity at the right ischial tuberosity (see FIG. 9B).

Further, the feature quantity PS obtained by the equation (4) is
The pressure to support the pelvis of the elderly, expressed in equation (4)
Ps_Left1, Ps_Left2Is based on the left ischial tuberosity
10cm in the front-back direction (direction along the front-back direction line)
Sensing pressure value at position, Ps _Right1, Ps_Right2Is on the right
At a distance of 10 cm in the anterior-posterior direction with respect to the ischial tuberosity
Pressure value, Ps_Left3, Ps_Left4Is the left ischial tuberosity
Left and right direction (based on the horizontal line
Pressure) at 5cm position, Ps_Right3, Ps
_Right4Is 5cm left and right with reference to the right ischial tuberosity
(A) and (B) in FIG.
reference).

The feature value PV obtained by the equation (5) represents the spread of the pressure felt by the _{occupant near the ischial tuberosity. In} equation (5), Pv _Left1 is the front-to-back direction from the left _ischial tuberosity. Pv _Right1 is the width of the two points at which the pressure sensitivity is half the maximum value along the anteroposterior direction from the right _ischial tuberosity from the right _ischium tubercle.
v _Left2 is the width of two points where the sensed pressure is half of the maximum from the left _ischial tuberosity along the left and right direction, and Pv _Right2 is the half of the maximum pressure is along the left and right directions from the right _ischial tuberosity. (See FIGS. 9A and 9B).

The feature value TV obtained by the equation (6) is a pressure gradient between the ischial tuberosity and the lower part of the knee (when the value of the feature value TV is large, the pressure is concentrated on the ischial tuberosity, and when the value is small, the pressure is applied to the entire seat surface. Ppk is the difference between the maximum value of the pressure sensitivity at the ischial tuberosity and the maximum value of the pressure sensitivity at the thigh near the lower part of the knee, and Lpk is the distance therebetween (FIG. 9).
(A)).

The feature PSD obtained by the equation (7) represents the balance between left and right, Plr is the difference between the maximum values of the sensed pressures at the left and right ischial nodes, and Llr is the distance between the right and left ischial nodes. (See FIG. 9B). The feature quantity PVT obtained by the equation (8), the feature quantity TVT obtained by the equation (9), and the feature quantity PSDT obtained by the equation (10) are feature quantities related to the temporal change of the distribution of the pressure sensitivity. The amount of change in the characteristic amount PV, the characteristic amount TV, and the characteristic amount PSD at the moment of the transition (in the transient state of sitting). t represents the elapsed time from the start of sitting, and t ₁ , t ₂ , and t ₃ represent state 1, state 2,
This corresponds to state 3 (see FIG. 4).

The above-described step 128 calculates “a feature quantity that can be associated with a sensory evaluation of a contact sensation when the human body comes into contact with the target object from the sensed pressure distribution obtained by the conversion” in the present invention. It corresponds to that.

After calculating the characteristic amount relating to the distribution of the sensed pressure as described above, the process proceeds to the next step 130, where a message requesting the input of the sensory evaluation value of the seating sensation is displayed on the display 20, so that the operator is notified. The user is prompted to input a sensory evaluation value, and in the next step 132, it is determined whether a sensory evaluation value has been input. If the determination is negative, the process returns to step 130, and the message is continuously displayed until a sensory evaluation value is input.

In the present embodiment, the SD method is used for the sensory evaluation of the seating sensation of the vehicle seat, and as shown in the following Table 1,
With respect to a plurality of evaluation items relating to the seating sensation (in Table 1, for example, “seat sensation”, “soft sensation”, “tension sensation”, and “fit sensation”), the subject was asked to evaluate which of the seven levels was used. And the sensory evaluation value is determined.

[0054]

[Table 1] The sensory evaluation value determined by the subject
2 and the like, the determination in step 132 is affirmed, and the process proceeds to step 134, where the input sensory evaluation value is calculated using the feature amounts PP and P calculated in step 128.
S, PV, TV, PSD, PVT, TVT, PSDT, and a single data file corresponding to the storage device 24.
And the process ends.

Each time the subject is seated and the sensory evaluation of the seating sensation is performed, the above-described feature amount calculation and sensory evaluation value storage processing are executed to obtain the sensory evaluation results (sensory evaluation values by subject and sensory pressure distribution Is stored in the storage device 24 as a data file.

Next, when an operator gives an instruction to create an evaluation formula for quantifying the feeling of sitting, the computer 18
Will be described with reference to the flowchart of FIG. In step 150, a message requesting the designation of the data file used for creating the evaluation formula is displayed on the display 20, thereby urging the operator to designate the data file used for creating the evaluation formula. In the next step 152, it is determined whether or not a data file to be used for creating an evaluation formula is specified, and the process waits until the determination is affirmed.

If the determination in step 152 is affirmative, the process proceeds to step 154, in which the data of the data file specified by the operator (the sensory evaluation value by the subject and the characteristic amount related to the pressure distribution) are fetched from the storage device 24.
In the next step 156, a specific evaluation item among the evaluation items related to the sitting sensation is set as an objective variable of the multiple regression analysis.

In the next step 158 and subsequent steps, step 1
Based on the data in the data file fetched in 54, a feature amount having a high correlation with a specific evaluation item set as an objective variable is selected, and an evaluation formula for quantifying the specific evaluation item is created. That is, in step 158, a feature value to be used as an explanatory variable of the multiple regression analysis is selected from the eight types of feature values.
The regression coefficient of the multiple regression equation is calculated by multiple regression analysis using the sensory evaluation value of the specific evaluation item included in the data file captured in 54 and the value of the feature amount selected in step 158. In step 162, the multiple regression equation obtained by the above calculation is tested, and in step 164, it is determined whether or not the multiple regression equation with the maximum contribution rate (minimum error) has been obtained. If the determination in step 164 is negative, the feature amount used as the explanatory variable is changed in step 166, and the process returns to step 160 to repeat the calculation of the regression coefficient of the multiple regression equation.

As described above, various methods are known as a method of selecting an optimum explanatory variable from a plurality of explanatory variable candidates. For example, 2 ^p − A method of examining one regression model (brute force method), a method of starting from a case where no explanatory variables are included and increasing the explanatory variables one by one (forward selection method),
Any method such as a method of starting from a state in which all of the explanatory variables are included and decreasing the explanatory variables one by one (backward elimination method), a method of increasing or decreasing the explanatory variables (sequential method), and the like may be adopted. .

By repeating the above steps 160 to 166, the optimal feature value having the maximum contribution rate is selected and extracted, and the optimal multiple regression equation is obtained.

As an example, the following Table 2 shows the results of the sensory evaluation (sensory evaluation value for each evaluation item) of the sitting sensation when a certain subject sits on 10 types of vehicle seats.
And calculated in each case from the sensed pressure distribution.
Table 3 shows the values of the types of feature amounts.

[0062]

[Table 2]

[0063]

[Table 3] For example, as data used to create an evaluation formula,
Assuming that the data of the case is designated, when "seating feeling" is set as the objective variable among the four types of evaluation items, the results shown in Table 4 below are finally obtained.

[0064]

[Table 4] As is evident from Table 4 above, as a result of performing a multiple regression analysis on the evaluation item “seating sensation” and obtaining a multiple regression equation with the largest contribution rate (minimum error), a feature highly correlated with the “seating sensation” As quantities, feature quantities PS, PV, and TV were extracted. It was confirmed that the prediction accuracy was very high when the contribution ratio R ² = 90% due to these features.

When the above result is obtained, step 1
The multiple regression equation in which the determination at step 64 is affirmative and the contribution rate is maximized in step 168 is stored as an evaluation equation for a specific evaluation item. For example, when the results as shown in Table 4 are obtained, the evaluation formula for the feeling of sitting is: (Singing feeling) = 0.2609 × PS−2.7207 × PV−57.3148 ×
TV-2.4931. By using this evaluation formula, the evaluation item “seating feeling” of the various evaluation items related to the sitting sensation can be quantified by using three types of characteristic amounts PS, PV, and TV.

This step 168, together with steps 158 to 166 described above, is performed by associating the calculated feature value with the sensory evaluation of the contact sensation when the human body comes into contact with the target object. To obtain an evaluation formula for quantifying the contact sensation when the object comes into contact with the target object. "

In the next step 174, it is determined whether or not all the evaluation items relating to the sitting sensation have been processed. If the determination is negative, the process proceeds to step 172, where an unprocessed evaluation item (for example, “fitness”) is set as a target variable, and the process returns to step 158.
The subsequent processing is repeated. As a result, for all the evaluation items related to the sitting sensation, the selection of the feature amount and the creation of the evaluation formula are performed. If the determination in step 170 is affirmative, the result of the processing is displayed on the display 20, and the evaluation formula creation processing ends.

In the above description, the contact pressure distribution detected by the pressure sensor array 12 is converted into a sensitive pressure distribution in consideration of the spatial masking characteristic of the human skin sensation, and the characteristic amount (human being) calculated from the sensitive pressure distribution is calculated. The evaluation formula for quantifying the sensation of each evaluation item is calculated using the characteristic amount that has a high correlation with the sitting sensation. An evaluation expression that can be obtained can be obtained.

Among the evaluation items related to the sitting sensation, evaluations such as “soft feeling” and “tension” are based on how the sensed pressure changes in a transient state from the start of sitting until a predetermined time elapses. Depends greatly on On the other hand, in the present embodiment, among the eight types of feature amounts, the feature amounts PVT, TVT, P related to the temporal change of the distribution of the sensed pressure.
Since SDT is included, the above-mentioned feature amounts PVT, TVT,
By extracting at least one of the PSDTs,
With respect to evaluation items such as "soft feeling" and "tension", which are greatly affected by the manner in which the sensing pressure changes in the transient state, an evaluation formula that can quantify the feeling of this evaluation item with high accuracy can be obtained. .

For quantifying each evaluation item relating to the seating sensation of, for example, a newly-produced vehicle seat or the like, for example, steps 100 to 126 of the feature value calculation / sensory evaluation value storage processing shown in FIG. In the same manner as in the above, the pressure distributions in the state 1 to the state 3 are obtained, and based on the pressure distribution in each state, the characteristic amount corresponding to the explanatory variable in the evaluation formula of each evaluation item (sensory evaluation of the evaluation item) By calculating only the characteristic amounts that can be associated with the values and substituting the calculated characteristic amounts into the evaluation formulas of the respective evaluation items, the seating sensation when the user sits on the newly prototyped vehicle seat or the like is evaluated. It can be quantified for each item, and based on the evaluation value obtained by the quantification, the sitting sensation can be evaluated with high accuracy for each evaluation item.

The quantification and evaluation of the sitting sensation described above corresponds to “quantifying and evaluating the contact sensation using an evaluation formula” in the present invention.

Table 4 shows, as an example, the results of performing multiple regression analysis using data of one subject.
In general, in order to generalize the evaluation formula, it is necessary to create the evaluation formula using data of a plurality of subjects and data of various types of vehicle seats. However, there are various types of vehicle seats, and psychological scales such as preferences for seating sensation and evaluation criteria are different for each subject. Therefore, data of a plurality of subjects and various types of vehicle seats are used. If a simple multiple regression analysis is performed to obtain an evaluation formula using all of the above data, there is a problem that a highly accurate evaluation formula for quantifying the sitting sensation cannot be obtained.

Therefore, when an evaluation formula is obtained using data of a plurality of subjects or data of various types of vehicle seats, the direction of the psychological scale of the subject regarding the pattern of the sensed pressure distribution and the seating sensation is determined. It is preferable to create an evaluation formula by dividing data into groups.

For example, the pattern of the received pressure distribution changes depending on the manner of contact between the human body and the target object. In the pressure-sensitive distribution along the front-rear direction line when seated on the vehicle seat, there are two cases (see the solid line in FIG. 10) and one case (see the broken line in FIG. 10). In some cases, humans feel that the number of contact points is two when the number of protrusions is two, and that the number of contact points is one when the number of protrusions is one. The degree of contact strength is expressed by the maximum pressure-sensitive pressure of the convex portion at the lower part of the knee when there are two convex portions, and the magnitude of the pressure-sensitive pressure at the lower part of the knee when there is one convex portion. It is represented by

The manner of contact between the lower part of the knee and the vehicle seat and the degree of the strength greatly affect the feeling of sitting on the vehicle seat. If they are mixed, these data are divided into different groups and multiple regression analysis is performed separately to create an evaluation formula. In this case, first, classification is performed according to whether there are two or one projections, and if further grouping is required, classification may be performed stepwise according to the degree of contact strength.

For grouping of seating sensation according to the direction of the psychological scale, if an evaluation formula is created for each individual subject using only the data of each individual subject, the sign of the feature amount expressed in the evaluation formula can be obtained. (More specifically, the sign of the regression coefficient), it is possible to determine the direction of the psychological scale of the sitting sensation for each subject, and based on the judgment result, the direction of the psychological scale of the sitting sensation. Can be grouped for each subject having similar.

Therefore, the prediction accuracy is high (for example, a contribution rate of 95
%), Data is roughly grouped according to the pattern of the pressure-sensitive distribution (the number of projections), and the evaluation formula is determined for each individual subject. Judgment of subjects having similar orientations of the psychological scale, and dividing the data of the group into a plurality of small groups consisting only of data of subjects having similar orientations of the psychological scale for sitting sensation. An evaluation formula may be created for each individual small group.

As an example, the data of the ten cases shown in Tables 2 and 3 are grouped according to the pattern of the pressure-sensitive distribution along the front-back direction line, and the data of the group having two convex portions (7 in this example). Assuming that the evaluation item “seated feeling” is the target variable using only the data for the cases), the results shown in Table 5 below are finally obtained by the multiple regression analysis.

[0079]

[Table 5] As is apparent from Table 5 above, when multiple regression analysis is performed after data grouping according to the pattern of the pressure-sensitive pressure distribution, the contribution ratio R ² = 99%, and multiple regression analysis is performed without performing data grouping. Is performed (in the case of Table 4), the quantification accuracy of the “seated feeling” is improved. The evaluation formula for the feeling of sitting when the results in Table 5 were obtained is: (Sit of feeling) = 0.21313 × PS−2.73639 × PV + 1.684
77 × PVT + 32.21183 × PSD−7.42862 Therefore, by using the above-mentioned evaluation formula, it is possible to quantify the “seating feeling” in the case where the pattern of the received pressure distribution is a pattern having two convex portions with higher accuracy.

As is clear from the above evaluation formula, there are four types of feature amounts extracted as feature amounts having a high correlation with the “feeling of sitting”, namely, feature amounts PS, PV, PVT, and PSD. Although the type of the feature amount has changed compared to the case where no grouping was performed, the evaluation item "seated feeling"
The directionality (sign of the regression coefficient) of the feature quantities PS and PV, which are the feature quantities having a particularly high correlation with (i.e., the absolute value of the standard regression coefficient is large) has not changed. Among the evaluation items related to sensation, PS or P that has a high correlation with this “seating
It is believed that the V orientation provides guidance in the design of the vehicle seat.

In the above description, a case has been described in which the feature quantity is automatically selected as an explanatory variable in the evaluation formula creation processing (FIG. 3). However, the present invention is not limited to this. Each time a feature value is selected and multiple regression analysis is performed, the analysis result may be displayed on the display 20 or the like, and the operator may select a feature value to be used as an explanatory variable based on the displayed analysis result.

In the above, in the evaluation formula creation processing (FIG. 3), only the feature quantity having a high correlation with the evaluation item as the objective variable is extracted as the explanatory variable so that the multiple regression equation having the largest contribution rate is obtained. However, the present invention is not limited to this. For example, assuming a multiple regression model (see the following equation (3)) using all feature amounts as explanatory variables, (evaluation item x) = a ₁ × PP + a _{2 × PS + a 3 × PV} + a 4 × TV + a 5 × PSD + a 6 × PVT + a 7 × TVT + a 8 × PSDT + a 9 ... (3) without performing the extraction of the evaluation items with a high feature quantity correlated as objective variables, the regression coefficients a ₁ ~ it may be obtained a _9. In this case, for the feature quantity having a low correlation with the evaluation item as the objective variable, 0 or a value close to 0 is set as the regression coefficient a. However, the true regression coefficient is 0
If the feature quantity is included as an explanatory variable in the multiple regression equation, the estimation accuracy as the regression equation decreases, so only feature quantities that are highly correlated with the evaluation item as the objective variable are extracted and extracted. It is preferable to obtain a multiple regression equation (evaluation equation) using only the above.

In the above description, the distribution of the received pressure along the front-rear direction line and the left-right direction line is obtained. However, the present invention is not limited to this, and the distribution of the received pressure is within a predetermined range having a two-dimensional spread. May be calculated, a characteristic amount may be calculated from the distribution of the received pressure, and an evaluation formula may be created for each evaluation item.

Further, in the above description, the sensation of sitting when sitting on the automobile seat is quantified as the contact sensation when contacting the target object. However, the present invention is not limited to this. It is needless to say that the present invention can be applied to quantification of a seating sensation when sitting on a seat for another transportation such as an aircraft or a contact sensation when contacting with any object that comes into contact with a human body.

[0085]

As described above, the present invention converts the contact pressure distribution when the human body comes into contact with the target object into the human body's perceived pressure distribution in consideration of the spatial masking characteristic of the human skin sensation, Calculate the feature amount that can be associated with the sensory evaluation of the contact sensation when the human body comes into contact with the target object, and associate the calculated feature amount with the sensory evaluation of the contact sensation when the human body comes into contact with the target object, Since the evaluation formula for quantifying the contact sensation when the human body comes into contact with the target object is obtained, it is possible to quantify the human touch sensation when contacting the target object with high accuracy. It has an excellent effect.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of a seat sitting sensation evaluation device according to an embodiment.

FIG. 2 is a flowchart illustrating a feature value calculation / sensory evaluation value storage process.

FIG. 3 is a flowchart illustrating an evaluation formula creation process.

FIG. 4 is a diagram showing a transition of a contact pressure on a seat surface after starting to sit on a vehicle seat.

5A is an image diagram showing an example of a contact pressure distribution in a state 1, FIG. 5B is an image diagram showing an example of a contact pressure distribution in a state 2, and FIG.

6A is an image diagram showing a human skeleton related to a contact pressure distribution, and FIG. 6B is an image diagram showing right and left front-rear lines and left-right lines for extracting the contact pressure distribution, respectively.

7A is a diagram illustrating an example of a contact pressure distribution along left and right front-rear direction lines, and FIG. 7B is a diagram illustrating an example of a contact pressure distribution along a left and right direction line.

FIG. 8 is a diagram showing an example of a contact pressure distribution along a front-back direction line and an example of a sensitive pressure distribution calculated from the contact pressure distribution.

FIG. 9A is a view for explaining various characteristic amounts related to the received pressure distribution, FIG.
(B) is a diagram which shows the distribution of the received pressure along the left-right direction line.

FIG. 10 is a diagram showing two types of patterns of pressure-sensitive distribution along a front-back direction line.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Seat sitting feeling evaluation device 12 Pressure sensor array 14 Vehicle seat 18 Computer

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tetsuzo Inoue 41 Toyoda Central Research Institute, Inc. Inside Kanto Automobile Industry Co., Ltd. (72) Naoya Oda, Inventor Naoya, Taura Port Town, Kanagawa Prefecture Inside Kanto Automobile Industry Co., Ltd. Inside

Claims (1)

[Claims] A contact pressure distribution when a human body comes into contact with a target object is measured, and the measured contact pressure distribution is converted into a human body sensitive pressure distribution in consideration of a spatial masking characteristic of a human body's skin sensation. Calculating, from the sensed pressure distribution obtained by the conversion, a feature amount that can be associated with a sensory evaluation of a touch sensation when the human body comes into contact with the target object; The evaluation formula for quantifying the contact sensation when the human body comes into contact with the target object is found in association with the sensory evaluation of the contact sensation when contacting with Touch sensation evaluation method.