Estimating maximum friction coefficient based on knowledge of the loads and self-alignment torque generated in a tire contact zone

• US 20050065699A1
• Filed: 08/04/2004
• Published: 03/24/2005
• Est. Priority Date: 02/08/2002
• Status: Active Grant

• Current Assignee: Compagnie Generale DES Etablissements Michelin SCA
• Original Assignee: Michelin Recherche et Technique SA (Compagnie Generale DES Etablissements Michelin SCA)
• Inventors: Bertrand, David
• Granted Patent: US 7,069,135 B2
• Time in Patent Office: Days
• US Class Current: 701/80
• CPC Class Codes: B60T 2210/12 Friction
• CPC Class Codes: B60T 8/1725 Using tyre sensors, e.g. Sidewall Torsion sensors [SWT] for tyre pressure and temperature detection B60C23/00

Abstract

A method for estimation of the maximum grip coefficient on the basis of knowledge of the forces and the self-alignment torque which are generated in the contact area of a tire, includes the steps of: selecting a plurality of fixed points in space, which lie at different azimuths along the circumference in at least one sidewall of the tire, carrying out a corresponding number of measurements of circumferential distance variation (extension or contraction) at these fixed points when the tire is rolling on the road, processing the measurement signals so as to extract the three components of a resultant of forces which are exerted by the road on the contact area of a tire and the self-alignment torque generated by the tire from them, processing the evaluation signals of the three components of a resultant of forces which are exerted by the road on the contact area of a tire and of the self-alignment torque generated by the tire so as to extract the said grip coefficient μ from them.

Claims

1. A method of determining the grip coefficient μ

in the contact area of a tire on a road, comprising the steps of;

determining the three components of a resultant of forces which are exerted by the road on the contact area of a tire and the self-alignment torque generated by the tire, and

processing the signals of the three components of a resultant of forces which are exerted by the road on the contact area of a tire and of the self-alignment torque generated by the tire to extract the grip coefficient μ

therefrom.

2. The method according to claim 1, wherein the step of determining the three components of a resultant of forces which are exerted by the road on the contact area of a tire and of the self-alignment torque generated by the tire comprises the steps of:

selecting a plurality of fixed points in space, which points each lie at different azimuths along the circumference in at least one sidewall of the tire,

measuring a corresponding number of circumferential distance variations at these fixed points when the tire is rolling on the road, and,

processing the measurement signals to extract the three components of the resultant of forces which are exerted by the road on the contact area of a tire and the self-alignment torque generated by the tire from them.

3. The method according to claim 2, in which the determination of the three components of a resultant of forces which are exerted by the road on the contact area of a tire and of the self-alignment torque generated by the tire further comprises the steps of obtaining at least five measurements of circumferential distance variation in at least one sidewall of the tire at five fixed points in space which lie at different azimuths along the circumference, and deriving the grip coefficient μ

from the determination of the three components of a resultant of forces which are exerted by the road on the contact area of a tire and of the self-alignment torque generated by the tire.

4. The method according to claim 2, wherein the step of measuring the circumferential variation is carried out by at least one sensor integrated in a sidewall of the tire, a signal delivered by the sensor being processed to read a value of the signal at a plurality of azimuths corresponding to the fixed points.

5. The method according to claim 2, wherein the step of measuring the circumferential variation is carried out by a plurality of sensors equal in quantity to a quantity of azimuths being observed, the sensors being external to the tire and arranged in fixed space.

6. The method according to claim 2, further comprising the step of establishing a transfer function between the measurements of circumferential variation, the three components of a resultant of forces which are exerted by the road on the contact area of a tire, and the self-alignment torque generated by the tire, and wherein the maximum grip coefficient is determined by the steps of:

compiling a data base containing Fx, Fy, Fz, N and p for a set of constraints of the tire on ground surfaces having different grip coefficients, the parameters Fx, Fy, Fz, and μ

being imposed on the tire, the value of the self-alignment torque N being measured,

training a neural network for determining the weightings of a neural network to reconstruct μ

on the basis of knowledge of Fx, Fy, Fz and N.

7. The method according to claim 2, further comprising the step of establishing a transfer function between the measurements of circumferential variation, the three components of a resultant of forces which are exerted by the road on the contact area of a tire, and the self-alignment torque generated by the tire, and a percentage of grip potential being used, wherein the percentage of grip being used is defined by

$p_u=\frac{{\mu }_{}}{}$

used μ

= F x 2 + F y 2 μ

·

F z ,

and comprising the steps of;

compiling a data base containing Fx, Fy, Fz, N and p_u for a set of constraints of the tire on ground surfaces having different grip coefficients, the parameters Fx, Fy, Fz, and μ

being imposed on the tire, the value of the self-alignment torque N being measured,

training a neural network for determining the weightings of the neural network to reconstruct p_u directly on the basis of knowledge of Fx, Fy, Fz and N.

8. The method according to claim 2, wherein a transfer function is established between the measurements, the three components of a resultant of forces which are exerted by the road on the contact area of a tire and the self-alignment torque generated by the tire, and the maximum grip coefficient by the steps of:

compiling a data base containing Fx, Fy, Fz, N and μ

for a set of constraints of the tire on ground surfaces having different grip coefficients, the parameters Fx, Fy, Fz, and μ

being imposed on the tire, the value of the self-alignment torque N being measured,

calculating the reduced values

$\frac{\mathrm{Fx}}{\mathrm{Fz}},\frac{\mathrm{Fy}}{\mathrm{Fz}},\frac{N}{\mathrm{Fz}},$

training a neural network for determining the weightings of a neural network to reconstruct μ

on the basis of knowledge of

$\frac{\mathrm{Fx}}{\mathrm{Fz}},\frac{\mathrm{Fy}}{\mathrm{Fz}},\frac{N}{\mathrm{Fz}}.$

9. The method according to claim 2 in which, setting

N=F _y .dx−

F _x .dy=N _y −

N _x

where dx is the distance from the point of application of the force Fy to the vertical plane which passes through the center of the contact area and is orthogonal to the X axis, and

where dy is the distance from the point of application of the force Fx to the vertical plane which passes through the center of the contact area and is orthogonal to the Y axis,

the method includes the step of obtaining values of N_x and N_y from the measurements of the three torques L, M, N, L being the moment about the X axis, M being the moment about the Y axis and N being the moment about Z.

10. The method according to claim 6, using a neural network of the perceptron type.

11. The method according to claim 2, wherein the circumferential variation is estimated by measuring the distance between the threads of a carcass ply in the sidewalls of the tire.

12. The method according to claim 2, wherein the circumferential variation is estimated by measuring the distance between wires forming a sensor which measures a variation in capacitance linked with the distance separating two electrodes.

13. The method according to claim 2, wherein at least three fixed points in space are used, which points are defined such that:

one point corresponds to one of the azimuth of the center of the contact area or the azimuth of the point opposite to the contact area, and,

two other points are symmetrical with respect to a vertical plane passing through the center of the contact area.

14. The method according to claim 2, wherein measurement azimuths are selected symmetrically with respect to the azimuth of the center of the contact area (180°

+α

° and

180°

−

α

°

), with cc different from α

₀, where α

₀ is the azimuth at the entry of the contact area, V₁¹ and V₂¹ being the values measured at these azimuths on the first sidewall and V₁² and V₂² being the values measured at these azimuths on the second sidewall, the method comprising the steps of;

estimating a component Fz by f_z(a₁V₁¹+a₂V₂¹+b₁V₁²+b₂V₂²), where a₁, a₂, b₁ and b₂ are positive real coefficients and f_z is a continuous function,

estimating a component Fx by f_x(c₁V₁¹−

c₂V₂¹+d₁V₁²−

d₂V₂²), where c₁, c₂, d₁ and d₂ are positive real coefficients and f_x is a continuous function,

estimating a component Fy by f_y(e₁V₁¹+e₂V₂¹−

f₁V₁²−

f₂V₂²), where e₁, e₂, f₁ and f₂ are positive real coefficients and f_y is a continuous function, and

estimating the self-alignment torque N by f_n(g₁V₁¹−

g₂V₂¹−

h₁V₁²+h₂V₂²) where g₁, g₂, h₁ and h₂ are positive real coefficients and f_n is a continuous function.

15. The method according to claim 2, wherein the measurement azimuths being selected symmetrically with respect to the azimuth of the center of the contact area (180°

+α

° and

180°

−

α

°

), with α

not equal to α

₀, where α

₀ is the azimuth at the entry of the contact area, and V₁ and V₂ being the values measured at these other azimuths, the method comprises the steps of;

estimating a value Fx by f_x(V₂−

rV₁), where r is a positive real coefficient and f_x is a continuous function,

estimating a value Fz by f_z(V_c−

(s₁V₁+s₂V₂)), where s₁ and s₂ are positive real coefficients and f_z is a continuous function, and,

estimating a value Fy by f_y(V_c+u₁V₁+u₂V₂), where u₁ and u₂ are positive real coefficients and f_y is a continuous function.

16. The method according to claim 2, wherein a confidence index is determined by the steps of:

determining in a training phase a model μ

=g(Fx, Fy, Fz, N, . . . ) and a model N_grip=f(Fx, Fy, Fz, . . . ) representing the value of N when the tire is gripping perfectly;

estimating in a use phase Fx, Fy, Fz and N with the aid of the circumferential extension measurement, and determining Ngrip;

responsive to |N|−

[N_grip| being less than a threshold value, determining that the contact of the tire on the ground is gripping, and that the determination of μ

is unreliable; and

,

responsive to |N|−

[N_grip| being greater than the threshold, determining that a slip zone in the contact is sufficient and the estimate of μ

is reliable.