JP5308052B2 - Steering stability evaluation method - Google Patents

Description

  The present invention evaluates steering stability during steering using the lateral jerk gradient generated when the vehicle is driven according to the steering input pattern for steering stability evaluation and / or the amount of yaw rate overshoot. It relates to the evaluation method.

  When discussing the improvement of the vehicle's transient response performance to the driver's operation, it is natural to focus on the body movement itself such as yaw movement, lateral movement, roll movement, etc., and how the driver recognizes these movements. It is necessary to consider whether it is being evaluated or so-called motor sensation characteristics. Many studies have been made on the maneuvering stability index from the viewpoint of human characteristics, such as gain of yaw angular velocity, response time and ease of vehicle control, and slip angle and yaw response time.

  Then, for example, in the following Non-Patent Document 1, the physical property amounts of two vehicles, the peak value of the lateral jerk (lateral jerk) of the vehicle and the time required until the lateral jerk reaches the peak value, are used as indices. Describes that the feeling of steering stability can be evaluated.

Japan Society of Automotive Engineers Academic Lecture Preprints No. 148-07 326-20075730 "Vehicle transient response based on human sensitivity"

  However, in Non-Patent Document 1, an experiment is performed on various different vehicles, and the sensitivity evaluation of the steering stability of the vehicle and the physical properties of the vehicle (the peak value of the lateral jerk and the time until the peak value is reached). It does not include experiments on vehicles with different tires with tires with different characteristics.

  Therefore, the present inventor sequentially replaces and mounts various tires having different characteristics on one vehicle body, conducts a sensitivity evaluation experiment on steering stability for a vehicle having different tires, and shows an example of the experimental results. 11 shows. In the figure, eight types of tires A1 to A8 of size 225 / 55R17 with different characteristics are prepared, and each tire A1 to A8 is sequentially mounted on a sports passenger car (domestic FR car, 3.5L) to change the lane. The relationship between the sensitivity evaluation of steering stability when driving with the steering input pattern (shown in FIG. 2) and the amount of physical properties of the vehicle (the peak value of lateral jerk and the time required to reach the peak value) It is shown. The numbers in parentheses are the sensitivity evaluation values for steering stability. From the experimental results shown in the figure, the tendency as discussed in Non-Patent Document 1, that is, the peak value is large, and the shorter the time required to reach the peak value is, the better the sensory evaluation of steering stability is. For vehicles with different tires, the steering stability cannot be sensitively evaluated from the physical properties of the vehicle (the peak value of lateral jerk and the time required to reach the peak value). It has been found. This is presumably because in the case of a vehicle with different tires, the peak value of the lateral jerk greatly changes due to the tire tread rubber or the like.

  Therefore, the present invention is based on the use of the lateral jerk gradient Y ″ and / or the yaw rate overshoot amount instead of the peak value of the lateral jerk and the time required to reach the peak value. On the other hand, it aims at providing the steering stability evaluation method which can evaluate the steering stability at the time of steering more accurately.

In order to achieve the object, the invention of claim 1 of the present application includes a steering input setting step for setting a steering input pattern for steering stability evaluation;
A lateral acceleration measuring step of measuring lateral acceleration Y of the vehicle generated when the vehicle is driven according to the steering input pattern and obtaining lateral acceleration data Dy (t, Y) of the lateral acceleration Y over time;
The lateral acceleration data Dy (t, Y) is subjected to regression analysis to obtain a quadratic regression equation of the lateral acceleration data Dy (t, Y), and the lateral jerk obtained by differentiating the quadratic regression equation with time t. A step of calculating a lateral jerk gradient Y ″ that is a second derivative of the second regression equation by differentiating Y ′ by time t;
And an evaluation step for evaluating steering stability during steering based on the lateral jerk gradient Y ″.

In the invention of claim 2, the steering input pattern includes a change curve portion in which the steering angle θ extends in a substantially S shape from 0 ° to the maximum angle θA, and a straight line portion where the steering angle θ is a maximum angle θA. Including
The lateral acceleration data Dy (t, Y) is subjected to regression analysis in a region range including the maximum gradient position where the gradient of the tY curve composed of the lateral acceleration data Dy (t, Y) is maximum, and the quadratic regression equation. Is required.

  According to a third aspect of the present invention, in the evaluation step, the steering stability is evaluated based on the magnitude of the lateral jerk gradient Y ″.

According to the invention of claim 4, the evaluation step includes a preliminary steering stability evaluation value preliminary data Dk and a preliminary advance of the lateral jerk gradient Y ″ at that time by a preliminary steering stability evaluation test using the steering input pattern. And obtaining an estimation formula K = f (Y ″) of the steering stability evaluation value K based on the lateral jerk gradient Y ″ from the prior data Dk and Dy ″.
Using the estimation formula K = f (Y ″), a steering stability evaluation value is estimated and evaluated from the lateral jerk gradient Y ″ obtained by the calculation step.

In the invention according to claim 5, in the lateral acceleration measuring step, the yaw rate Z of the vehicle generated when the vehicle is driven according to the steering input pattern is measured, and the yaw rate Z when the steering angle θ reaches the maximum angle θA is measured. Including a yaw rate measuring step for obtaining an overshoot amount ΔZ of
The evaluation step is characterized in that steering stability during steering is evaluated based on the lateral jerk gradient Y ″ and the overshoot amount ΔZ of the yaw rate Z.

According to the invention of claim 6, the evaluation step includes a preliminary steering stability evaluation value preliminary data Dk and a preliminary advance of the lateral jerk gradient Y ″ at that time by a preliminary steering stability evaluation test using the steering input pattern. The data Dy ″ and the prior data Dz of the overshoot amount ΔZ of the yaw rate Z are obtained, and the steering stability evaluation value K based on the lateral jerk gradient Y ″ and the overshoot amount ΔZ is determined from the prior data Dk, Dy ″ and Dz. Deriving the estimation formula K = g (Y ″, Z),
Using the estimation formula K = g (Y ″, Z), the steering stability evaluation value K is estimated and evaluated from the lateral jerk gradient Y ″ obtained by the calculation step and the overshoot amount ΔZ. It is said.

The invention according to claim 7 is characterized in that the estimation equation K = g (Y ″, Z) is expressed by the following equation: a1, a2, and a3 are coefficients.
K = a1 · Y ″ + a2 · ΔZ + a3

  The invention of claim 8 is characterized in that the estimation equation K = g (Y ″, Z) is a multiple regression equation in which the coefficients a1, a2, and a3 are obtained by multiple regression analysis.

  According to the steering stability evaluation method of the present invention, the lateral jerk gradient Y ″ and / or the overshoot amount of the yaw rate are correlated with the steering stability during steering, and the vehicle travels according to a predetermined steering input pattern. By measuring the lateral jerk gradient Y ″ and / or the yaw rate overshoot amount, the lateral jerk gradient Y ″ and / or the yaw rate overshoot amount can be used for vehicles with different tires. The steering stability during steering can be evaluated.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a flowchart showing a steering stability evaluation method according to the first aspect of the invention, and includes a steering input setting step, a lateral acceleration measurement step, a calculation step, and an evaluation step.

  The steering input setting step is a step of setting a steering input pattern 1 for steering stability evaluation. As the steering input pattern 1, a pattern close to steering when performing a lane change is preferably employed. Specifically, as shown in FIG. 2, the steering input pattern 1 includes a change curve portion j1 in which the steering angle θ extends in a substantially S shape from 0 ° to the maximum angle θA, and the steering angle θ is the maximum angle θA. Those including a certain straight line portion j2 are preferable. In the figure, the straight portion j3 that travels with the steering angle θ of 0 ° and the steering in one steering direction (for example, right turn) until the steering angle θ reaches the maximum angle θA that is connected to the straight portion j3. The traveling curve portion j1 that travels while traveling, the straight portion j2 that travels while maintaining the steering angle θ of the maximum angle θA connected to the varying curve portion j1, and the steering angle θ that continues to the straight portion j2 The change curve portion j4 that travels while steering in the other steering direction (for example, the left side) until the maximum angle θB on the steering side is reached, and the steering angle θ is connected to the change curve portion j4 from the maximum angle θB to 0 °. What includes the change curve portion j5 to be returned is shown.

  Next, in the lateral acceleration measurement step, the vehicle is driven in accordance with the steering input pattern 1 and the lateral acceleration Y of the vehicle generated at that time is measured, whereby the lateral acceleration data Dy that is the time-dependent data of the lateral acceleration Y is measured. (T, Y) is obtained. Specifically, the lateral acceleration data Dy is obtained by automatically steering the vehicle along the steering input pattern 1 using an automatic steering device and measuring the lateral acceleration Y using a lateral acceleration sensor attached to the vehicle. (T, Y) can be obtained. The traveling speed at the time of measurement is not particularly limited as long as it is constant, but it is preferable to set the traveling speed in the range of 40 to 120 km / h, for example, in order to obtain a change in the lateral acceleration Y more clearly. FIG. 3 shows an example of a scatter diagram in which a part of the lateral acceleration data Dy (t, Y) measured when traveling at a speed of 80 km / h according to the steering input pattern 1 of FIG. .

Next, in the calculation step, first, by performing regression analysis on the lateral acceleration data Dy (t, Y) obtained in the lateral acceleration measuring step, a quadratic regression equation (1) of the lateral acceleration data Dy (t, Y) is obtained. ) That is, the regression coefficients c0, c1, and c2 in the quadratic regression equation (1) are obtained using a known regression analysis method (for example, the least square method).
Y = c0 · t ² + c1 · t + c2 --- (1)

  At this time, the lateral acceleration data Dy (t, Y) is subjected to regression analysis in the region range Q including the maximum gradient position Q0 where the gradient of the tY curve composed of the lateral acceleration data Dy (t, Y) is maximum. It is important to obtain a quadratic regression equation.

  For example, as shown in FIG. 4 exaggeratingly an example of a tY curve composed of the lateral acceleration data Dy (t, Y), when traveling along the steering input pattern 1, the steering angle θ is from 0 ° to the maximum angle. In the change curve portion j1 leading to θA, the gradient of the tY curve (the rate of change ΔY / Δt of the lateral acceleration Y with respect to time t) gradually increases with time, and the maximum gradient position Q0 at which the gradient becomes maximum. After that, it gradually decreased with the passage of time. The region range Q including the maximum gradient position Q0 is a stable region having the largest gradient and the smallest gradient change. Obtaining the quadratic regression equation (1) in this region range Q is the steering stability This is important for clearly and accurately expressing the difference in sensory evaluation of sex. The region range Q is preferably a range of ± 5 ° around the maximum gradient position Q0 and the steering angle θ.

In the calculating step, the lateral jerk Y ′ obtained by differentiating the quadratic regression equation (1) with time t is further differentiated with time t to obtain the second derivative of the quadratic regression equation (1). A certain lateral jerk gradient Y ″ is obtained.
Y ′ = c0 · 2 · t + c1 (2)
Y ″ = c0 · 2 −−− (3)

FIG. 3 shows a quadratic regression equation (1) (Y = 1) obtained from the lateral acceleration data Dy (t, Y) of the region range Q when the region range Q is 4.61 to 5.07 sec. .6831 · t2-14.393t + 30.767), and the contribution rate R ² of the quadratic regression equation (1) to the lateral acceleration data Dy (t, Y) was 0.9978. Further, from this quadratic regression equation (1), it is possible to obtain 3.36 as the lateral jerk gradient Y ″ that is the second derivative thereof. If the region range Q is too wide, the quadratic regression equation (1 ) contribution R ² is correlated and downward lowering of such a reliability Conversely area range Q is too small decreases, both the lateral jerk gradient Y "be obtained accurately becomes difficult. Thus the As contribution R ² is 0.9 or more, and more preferable to set the area range Q.

  Next, in the evaluation step, the steering stability during steering is evaluated based on the lateral jerk gradient Y ″ obtained in the calculation step.

  Here, FIG. 5 shows a scatter diagram of the lateral jerk gradient Y ″ based on the actual vehicle running experiment conducted by the present inventor and the sensitivity evaluation of steering stability. This actual vehicle running experiment is shown in FIG. 11 described above. Prepare the 8 types of tires A1 to A8 of size 225 / 55R17 with the same characteristics as the actual vehicle running experiment, and install each tire A1 to A8 on a sports passenger car (domestic FR car, 3.5L) in sequence. 2, the vehicle is driven according to the steering input pattern 1. At this time, the steering stability of each vehicle with different tires is evaluated based on a 10-point evaluation value K based on the sensory evaluation of the driver. The internal pressure is 230 kPa, the rim is 17 × 7.5 J, and the running speed is 80 km / h.

  Further, the lateral acceleration Y is measured over time during the travel, and the lateral jerk gradient Y "in each vehicle with different tires is obtained using the lateral acceleration measuring step and the calculating step, and the lateral jerk gradient Y" 5 and the data (Y ″, K) of the steering stability evaluation value K are shown in FIG. 5. Using the data (Y ″, K) of FIG. When the next regression equation was obtained, the correlation coefficient r with the data (Y ″, K) was 0.7787, and the significance probability p was 0.0228. That is, the lateral jerk gradient Y ″ and the steering stability There was a correlation with the evaluation value K, and it was confirmed that the steering stability could be evaluated using the lateral jerk gradient Y ″ at the level of p = 0.0228.

  As one method of the evaluation step, the magnitude of the lateral jerk gradient Y ″ obtained from the calculation step is compared. Thereby, the steering stability can be ranked.

  As another method of the evaluation step, a preliminary steering stability evaluation test using the steering input pattern 1 is performed, and the preliminary data Dk of the steering stability evaluation value K and the lateral jerk gradient Y ″ at that time are obtained. Prior data Dy ″ is obtained, and an estimation formula K = f (Y ″) of the steering stability evaluation value K based on the lateral jerk gradient Y ″ is derived in advance from the prior data Dk and Dy ″. By using this estimation formula K = f (Y ″), the steering stability evaluation value K is estimated from the lateral jerk gradient Y ″ obtained by the calculation step, and the steering stability is determined using this estimated value. The estimation equation K = f (Y ″) can be obtained by performing regression analysis on the prior data Dk and Dy ″ as in the linear regression equation of FIG.

  Next, in the steering stability evaluation method of the present invention, the steering stability can be evaluated with higher accuracy by using the lateral jerk gradient Y ″ and the overshoot amount ΔZ of the yaw rate Z in combination. First, as the lateral acceleration measuring step, the yaw rate Z of the vehicle generated when the vehicle is driven according to the steering input pattern 1 is measured, and the overshoot of the yaw rate Z when the steering angle θ reaches the maximum angle θmax is measured. The yaw rate measurement step for obtaining the amount ΔZ is included.As the evaluation step, the steering stability during steering is evaluated based on the lateral jerk gradient Y ″ and the overshoot amount ΔZ of the yaw rate Z.

  In the yaw rate measurement step, when measuring the lateral acceleration Y, the yaw rate Z is measured over time simultaneously with the lateral acceleration Y using a yaw rate sensor attached to the vehicle. FIG. 6 shows an example of measurement data of the yaw rate Z in comparison with the steering input pattern 1. As shown in the figure, immediately after the steering angle θ reaches the maximum angle θA, the yaw rate Z reaches the maximum value Zmax. However, in the steering input pattern 1, the yaw rate Z is once decreased after reaching the maximum angle θA, and the steering angle θ does not change, and then increases to a stable value. ZA will be reached. Therefore, in this specification, the amount of yaw rate Z overshooting and exceeding the stable value ZA, that is, the amount indicated by the difference (Zmax−ZA) between the stable value ZA and the maximum value Zmax is defined as the overshoot amount ΔZ. Define.

Here, FIG. 7 shows a dispersion diagram of the overshoot amount ΔZ and the lateral jerk gradient Y ″ of each vehicle with different tires measured based on the inventor's actual vehicle running experiment in FIGS. The numbers in parentheses indicate the steering stability evaluation value K. From this scatter diagram, the primary regression equation is obtained with the overshoot amount ΔZ as the dependent variable and the lateral jerk gradient Y ″ as the independent variable. As a result, the correlation coefficient r was −0.3340, and the significance probability p was 0.4188. That is, the overshoot amount ΔZ and the lateral jerk gradient Y ″ of the yaw rate Z have a very low correlation coefficient r and can be regarded as being independent of each other. Can be obtained as an independent variable, and a multiple regression equation (4) can be obtained with a steering stability evaluation value K as a dependent variable. The coefficients a1, a2, and a3 can be calculated and calculated by multiple regression analysis.
K = a1 · Y ″ + a2 · ΔZ + a3 (4)

When the data of FIG. 7 was subjected to multiple regression analysis, the following multiple regression equation (4a) with a1 = 0.50, a2 = -3.2, and a3 = 6.3 could be obtained.
K = 0.50 · Y ″ −3.2 · ΔZ + 6.3 −−− (4a)

Using this multiple regression equation (4a), the evaluation value K (estimated value) calculated from the overshoot amount ΔZ and the lateral jerk gradient Y ″ is compared with the actually measured evaluation value K (actual value). the results are shown in Figure 8. as the figure, the contribution rate ^{R 2} of the multiple regression equation (4a) is 0.9242, significant probability p is 0.0001, correlation between the actual measurement value and the estimated value From the lateral jerk gradient Y ″ and the overshoot amount ΔZ of the yaw rate Z, it can be confirmed that the steering stability can be evaluated with higher accuracy.

  As one method of the evaluation step, a preliminary steering stability evaluation test using the steering input pattern 1 is performed, and the preliminary data Dk of the steering stability evaluation value K and the lateral jerk gradient Y ″ at that time are obtained. The preliminary data Dy "and the preliminary data Dz of the overshoot amount ΔZ of the yaw rate Z are obtained, and the steering stability evaluation value based on the lateral jerk gradient Y" and the overshoot amount ΔZ is obtained from the preliminary data Dk, Dy "and Dz. An estimation equation K = g (Y ″, Z) of K is derived in advance. By using this estimation equation K = g (Y ″, Z), the lateral jerk gradient Y obtained by the calculation step is obtained. ”And the overshoot amount ΔZ, an evaluation value K of the steering stability is estimated, and the steering stability is evaluated using this estimated value.

  At this time, the following equation can be used as the estimation equation K = f (Y ″), and the coefficients a1, a2, and a3 are obtained, for example, by performing multiple regression analysis on the prior data Dk, Dy ″, and Dz. Can do. The coefficients a1, a2, and a3 can be derived empirically from, for example, sequentially accumulated data, in addition to the multiple regression analysis.

  As another method of the evaluation step, as shown in FIG. 7, data is mapped with the horizontal jerk gradient Y ″ as the vertical axis and the overshoot amount ΔZ as the horizontal axis, for example. As shown in FIG. As Y ″ increases and the overshoot amount ΔZ decreases, the steering stability tends to become better. Therefore, the steering stability can be ranked from the map diagram as shown in FIG. In such a case, there is an advantage that the preliminary steering stability evaluation test can be omitted.

  Next, the steering stability evaluation method of the second invention will be described. The steering stability evaluation method according to the second aspect of the invention evaluates the steering stability using only the overshoot amount ΔZ of the yaw rate Z instead of the lateral jerk gradient Y ″. As shown, the steering input setting step, the yaw rate measurement step, and the evaluation step are configured.

  The steering input setting step is substantially the same as the steering input setting step of the first invention, and sets a steering input pattern 1 as shown in FIG. The yaw rate measuring step is substantially the same as the yaw rate measuring step of the first invention. The yaw rate Z of the vehicle traveling according to the steering input pattern 1 is measured over time, and the yaw rate as shown in FIG. The overshoot amount ΔZ of Z is obtained.

  In the evaluation step, the steering stability during steering is evaluated based on the overshoot amount ΔZ obtained in the yaw rate measurement step.

  Here, the dispersion of the overshoot amount ΔZ of the yaw rate Z of each vehicle with different tires measured based on the same actual vehicle traveling experiment as the actual vehicle traveling experiment shown in FIGS. The figure is shown in FIG. From this scatter diagram, a linear regression equation was obtained with the sensitivity evaluation of steering stability as a dependent variable and the overshoot amount ΔZ as an independent variable. As a result, the correlation coefficient r was 0.7914, and the significance probability p was 0.0193. Met. That is, there is a correlation between the overshoot amount ΔZ of the yaw rate Z and the steering stability evaluation value K, and at the level of p = 0.0193, the steering stability can be evaluated using the overshoot amount ΔZ. Was confirmed.

  As one method of the evaluation step using the overshoot amount ΔZ, the magnitude of the overshoot amount ΔZ obtained from the yaw rate measurement step is compared. In this way, the steering stability is ranked.

  As another method of the evaluation step, a preliminary steering stability evaluation test using the steering input pattern 1 is performed, and preliminary data Dk of the steering stability evaluation value K and the overshoot amount ΔZ at that time In addition to obtaining the data Dz, an estimation formula Z = h (ΔZ) for the steering stability evaluation value K based on the overshoot amount ΔZ is derived in advance from the preliminary data Dk and Dz. Then, by using this estimation formula Z = h (ΔZ), an evaluation value K of steering stability is estimated from the overshoot amount ΔZ obtained by the yaw rate measurement step, and the steering stability is estimated using this estimated value. Evaluate. The estimation formula Z = h (ΔZ) can be obtained by performing regression analysis on the prior data Dk and Dz as in the linear regression formula of FIG.

  As described above, by using the lateral jerk gradient and / or the yaw rate overshoot amount as the physical property amount of the vehicle, the sensory evaluation of the steering stability can be quantified, and the good or bad can be clearly determined by the numerical value. It becomes possible. The lateral jerk gradient and / or the yaw rate overshoot amount may be obtained by simulation running on the tire of the design model using a computer in addition to the actual vehicle running. In such a case, the design model is controlled within the simulation. Sensory evaluation of stability can be performed.

  As mentioned above, although especially preferable embodiment of this invention was explained in full detail, this invention is not limited to embodiment of illustration, It can deform | transform and implement in a various aspect.

It is a flowchart which shows the steering stability evaluation method of 1st invention. It is a diagram which shows an example of a steering input pattern. It is the graph which plotted a part of lateral acceleration data. It is a diagram which exaggerates and shows an example of the tY curve which consists of lateral acceleration data. It is a scatter diagram of lateral jerk gradient obtained based on an actual vehicle running experiment and a sensitivity evaluation of steering stability. It is a diagram explaining the amount of overshoot of the yaw rate. It is a dispersion | distribution figure of the amount of overshoots obtained based on the actual vehicle running experiment, and lateral jerk gradient. It is a diagram which shows the relationship between the evaluation value K estimated from the overshoot amount and the lateral jerk gradient, and the actually measured evaluation value K. It is a flowchart which shows the steering stability evaluation method of 2nd invention. It is a scatter diagram of the amount of overshoot of the yaw rate obtained based on the actual vehicle running experiment and the sensitivity evaluation of steering stability. It is a scatter diagram of the peak value of lateral jerk and the time required to reach the peak value.

Explanation of symbols

1 Steering input pattern j1 Change curve part j2 Straight line part Q0 Maximum gradient position Q Area range

Claims (8)

A steering input setting step for setting a steering input pattern for steering stability evaluation;
A lateral acceleration measuring step of measuring lateral acceleration Y of the vehicle generated when the vehicle is driven according to the steering input pattern and obtaining lateral acceleration data Dy (t, Y) of the lateral acceleration Y over time;
The lateral acceleration data Dy (t, Y) is subjected to regression analysis to obtain a quadratic regression equation of the lateral acceleration data Dy (t, Y), and the lateral jerk obtained by differentiating the quadratic regression equation with time t. A step of calculating a lateral jerk gradient Y ″ that is a second derivative of the second regression equation by differentiating Y ′ by time t;
A steering stability evaluation method during steering, comprising: an evaluation step for evaluating steering stability during steering based on the lateral jerk gradient Y ″. The steering input pattern includes a change curve portion in which the steering angle θ extends in a substantially S shape from 0 ° to the maximum angle θA, and a linear portion in which the steering angle θ is constant at the maximum angle θA.
The lateral acceleration data Dy (t, Y) is subjected to regression analysis in a region range including the maximum gradient position where the gradient of the tY curve composed of the lateral acceleration data Dy (t, Y) is maximum, and the quadratic regression equation. The steering stability evaluation method according to claim 1, wherein:   3. The steering stability evaluation method according to claim 2, wherein the evaluation step evaluates steering stability based on the magnitude of the lateral jerk gradient Y ″. The evaluation step obtains the preliminary data Dk of the steering stability evaluation value and the preliminary data Dy ″ of the lateral jerk gradient Y ″ at that time by a preliminary steering stability evaluation test using the steering input pattern, and From this preliminary data Dk, Dy ″, an estimation formula K = f (Y ″) of the steering stability evaluation value K based on the lateral jerk gradient Y ″ is derived,
3. The steering stability evaluation according to claim 2, wherein the steering stability evaluation value is estimated and evaluated from the lateral jerk gradient Y ″ obtained by the calculation step using the estimation formula K = f (Y ″). Sex assessment method. The lateral acceleration measuring step measures a yaw rate Z of the vehicle that is generated when the vehicle is driven according to the steering input pattern, and obtains an overshoot amount ΔZ of the yaw rate Z when the steering angle θ reaches the maximum angle θA. Including measurement steps,
The steering stability evaluation method according to claim 2, wherein the evaluation step evaluates steering stability during steering based on the lateral jerk gradient Y ″ and an overshoot amount ΔZ of the yaw rate Z. In the evaluation step, the preliminary data Dk of the steering stability evaluation value, the preliminary data Dy ″ of the lateral jerk gradient Y ″ at that time, and the yaw rate Z are obtained by a preliminary steering stability evaluation test using the steering input pattern. Preliminary data Dz of the overshoot amount ΔZ is obtained, and an estimation formula K = g (Y ″) of the steering stability evaluation value K based on the lateral jerk gradient Y ″ and the overshoot amount ΔZ from the prior data Dk, Dy ″, Dz. , Z)
Using the estimation formula K = g (Y ″, Z), the steering stability evaluation value K is estimated and evaluated from the lateral jerk gradient Y ″ obtained by the calculation step and the overshoot amount ΔZ. The steering stability evaluation method according to claim 5. The steering stability evaluation method according to claim 6, wherein the estimation formula K = g (Y ″, Z) is expressed by the following formula.
K = a1 · Y ″ + a2 · ΔZ + a3
(A1, a2, and a3 are coefficients)   8. The steering stability evaluation method according to claim 7, wherein the estimation equation K = g (Y ″, Z) is a multiple regression equation in which the coefficients a1, a2, and a3 are obtained by multiple regression analysis.

Images