EP1907226A2

EP1907226A2 - System for determining inflating pressure of tyres mounted on a motor vehicle front and rear wheels

Info

Publication number: EP1907226A2
Application number: EP06779008A
Authority: EP
Inventors: Zahir Djama
Original assignee: Peugeot Citroen Automobiles SA
Current assignee: PSA Automobiles SA
Priority date: 2005-07-26
Filing date: 2006-07-04
Publication date: 2008-04-09
Also published as: FR2889111B1; WO2007012771A2; WO2007012771A3; US20080208515A1; FR2889111A1; US7823443B2

Abstract

The invention concerns a system for determining inflating pressure of tyres mounted on a motor vehicle front and rear wheels. Said system comprises means (12, 14) for acquiring inflating pressures of the front and rear wheels, means (16, 18) for acquiring vertical accelerations of the front and rear wheels, and means (32) for merging the acquired pressures and the acquired accelerations to estimate the inflating pressures of the front and rear tyres.

Description

System for determining the inflation pressure of tires mounted on front and rear wheels of a motor vehicle.

The present invention relates to a system for determining the inflation pressure of tires mounted on front and rear wheels of a motor vehicle.

In the state of the art, such systems are known comprising accelerometers equipping a front wheel and a rear wheel of the vehicle to measure their vertical accelerations, means for estimating the stiffness coefficients of the tires thereof as a function of the accelerations. measured and means for calculating the inflation pressures of these tires as a function of the estimated stiffness coefficients.

In such systems, when one of the accelerometers is faulty, the estimate of the tire pressure is no longer available. Moreover, when the measurements of the accelerometers are defective, for example when they are stained with a not insignificant noise or that the measures of the front acceleration and the rear acceleration are not more coherent, these systems return estimates of the false pressure.

Similarly, these systems estimate unsatisfactorily tire inflation pressure when they experience rapid drops in pressure. Moreover, the accuracy of the estimate depends strongly on the type of trajectories taken by the vehicle.

Also known are tire pressure monitoring systems of a vehicle which comprise pressure sensors directly located in the tires.

The pressure sensors of such systems comprise information transmission means operating in association with corresponding receiving means embedded in the vehicle.

However, communication problems between the transmitting and receiving means are often present, for example because of too much noise, so that the function of monitoring the pressure of the tires is not satisfactorily provided. . In addition, when a pressure sensor is defective, no measurement of the tire pressure of the tire associated with this sensor is then available.

The present invention aims to solve these problems by providing a system based on redundancy of information and allowing increased reliability and accuracy.

To this end, the subject of the invention is a system for determining the inflation pressure of tires mounted on front and rear wheels of a motor vehicle, characterized in that it comprises: means for acquiring pressure inflating the tires of the front and rear wheels;

means for acquiring vertical accelerations of the front and rear wheels; and

means for melting the acquired pressures and the accelerations acquired in order to estimate the inflation pressures of the tires of the front and rear wheels.

According to particular embodiments, the system according to the invention may comprise one or more of the following characteristics:

the melting means comprise: means for estimating the stiffness coefficients of the tires of the front and rear wheels as a function of the vertical accelerations acquired therefrom; and

means for estimating the inflation pressures of the tires of the front and rear wheels as a function of the inflation pressures acquired and the estimated stiffness coefficients of the tires of the front and rear wheels,

the means for estimating the stiffness coefficients comprise means for temporally resetting one of the accelerations acquired on the other of the accelerations acquired and means for calculating the stiffness coefficients as a function of the accelerations thus time-adjusted;

the means for estimating the stiffness coefficients comprise bandpass filtering means of the accelerations acquired arranged between means for acquiring acceleration and time resetting means;

the band-pass filtering means are adapted to implement a filtering in a frequency range substantially equal to [8, 20] Hz;

the means of time registration comprise means for calculating the intercorrelation of the accelerations acquired and means for applying a delay corresponding to the maximum of the calculated correlation to the acceleration gained from the front wheel; the means for calculating the stiffness coefficients are adapted to implement a recursive least squares algorithm in real time based on a predetermined mechanical model of the wheel;

the estimation means are adapted to estimate said stiffness coefficients from a single-wheel mechanical model of the front and rear wheels;

the estimation means are capable of estimating said stiffness coefficients based on discrete time modeling of the corrected accelerations of the front and rear wheels according to the relation:

Avr (k) = where k is the k ^th time of sampling, ^Avr and Ava are the vertical accelerations of the rear and front wheels respectively, Zvr and Zva are the altitudes of the centers of the rear and front wheels respectively, Kpr and Kpa are the stiffness coefficients of the tires front and rear wheels respectively, and n is a sampling instant corresponding to a time shift between the rear and front wheels undergoing the same portion of roadway;

Ava (k) = - (mrrxAvr (k + n) Zvr (k + n) - Zva (k)) ^{Kpa () Kpr (} 'I mra I Kpa (k) J where k is the k ^th time of sampling, ^Avr and Ava are the vertical accelerations of the rear and front wheels respectively, Zvr and Zva are the altitudes of the centers of the rear and front wheels respectively, Kpr and Kpa are the stiffness coefficients of the tires front and rear wheels respectively, and n is a sampling instant corresponding to a time shift between the rear and front wheels undergoing the same portion of roadway;

the estimation means are adapted to estimate said stiffness coefficients from a bicycle mechanical model thereof;

where k is the k ^ι J è ^e ^m e ^th sampling instant, Avr and Ava are the vertical accelerations of the rear and front wheels, respectively, and Zvr Zva are the altitudes of the rear wheels and front centers, respectively, and Kpr Kpa are the stiffness coefficients of the tires of the front and rear wheels respectively, n is a sampling instant corresponding to a time shift between the rear and front wheels undergoing the same portion of roadway, Ra and Rr are stiffness coefficients of the suspensions of the front and rear wheels respectively, and Zva and Zvr are the first derivatives of the altitudes of the centers of the front and rear wheels respectively;

the means for estimating the inflation pressures are adapted to implement, for each of the front and rear wheels, a Kalman estimator based on a model linking the inflation pressure and the stiffness coefficient of the tire of the tire. wheel; the melting means are adapted to implement a Kalman estimator based on a model linking the tire inflation pressures and the vertical accelerations of the front and rear wheels;

the system further comprises means for diagnosing the operating state of the means for acquiring the inflation pressures and the mounting state of the tires as a function of the estimated and acquired inflation pressures and the acquired vertical accelerations of the wheels. front and rear ;

if an acquired inflation pressure and its associated estimated pressure differ by more than X%, where X is a predetermined number, the diagnostic means are adapted to calculate the tire pressures of the front and rear wheels as a function of the vertical accelerations acquired from them, to compare the pressure acquired with the calculated pressures and to determine that, if the pressure acquired corresponds to one of the calculated pressures, the tire associated with the calculated pressure corresponding to the pressure acquired is subject to an inverted montage; and

if none of the tires has an inversion of assembly, then the diagnostic means are able to diagnose that the acquisition means associated with the acquired pressure are faulty. The invention will be better understood on reading the following description, given solely by way of example, and with reference to the accompanying drawings, in which:

FIG. 1 is a schematic view of a mechanical model used by a system according to the invention; FIG. 2 is a diagram illustrating a calculation hypothesis used by a system according to the invention;

FIG. 3 is a schematic view of a system according to the invention; and

- Figure 4 is a schematic view of a second mechanical model used by another embodiment of the system according to the invention. The system according to the invention is based on a mechanical model of the interactions between the body C of a vehicle, of mass Mc, the wheels R of it and the ground S.

A first example of mechanical modeling of these interactions is illustrated in FIG. 1, which is a schematic view of a "single-wheel" type of model, of the interactions between a vehicle wheel Ro, the vehicle body C and the vehicle. soil S.

As this figure shows, in this model with two degrees of freedom, the body C of the vehicle is assimilated to a mass Mc suspended from the wheel Ro mass Mr by a suspension Su assimilated to a spring / damper stiffness Kc and damping coefficient

R.

The wheel Ro and the body C move in a vertical direction and occupy respective altitudes Zr and Zc with respect to a reference level, for example the altitude of the ground when the vehicle is started.

The wheel Ro carries a pneumatic tire Pn resting on the ground S and similar to a stiffness spring K composed of a spring modeling the tire envelope Pn of structural stiffness Ks in parallel with a spring modeling the gas contained in the tire of stiffness Pneumatic Kp, all in series with a spring modeling rubber gum tire eraser Kg.

The behavior of this mechanical system is controlled by the evolution over time of the altitude Zs of the ground relative to the reference level, that is to say the profile of the roadway. It is also known that the inflation pressure Pg of the tire is directly related to the pneumatic stiffness Kp thereof, this dependence being for example modeled according to the relation:

where c and α are predetermined constants, for example substantially equal to 6.7 and 0.85 respectively for a given tire. The system according to the invention is also based on the following observation presented in FIG. 2 which illustrates the progress of a vehicle on a roadway between two instants t and t + Δt.

As illustrated in this figure, the front and rear wheels of the vehicle undergo, most of the time, with a time shift Δt depending on the speed V and the wheelbase of the vehicle, the same road profile.

The system according to the invention is then advantageously based on the following relationship to determine the stiffness coefficients of the tires and therefore the inflation pressure thereof, as will appear in more detail below:

Z _sa (t) = Z _sr (t + Δt) (2) where t is the time, Δt is the time between the passage of a front wheel on a point of the road, the passage of a rear wheel on this same point, Z _sa is the ground altitude at the front wheel and Z _sr is the ground altitude at the rear wheel.

FIG. 3 schematically illustrates, under the general reference 10, a system for determining the inflation pressure of the tires mounted on a front wheel and a rear wheel of a motor vehicle, arranged on the same side of the vehicle. this.

This system includes, associated with each of the front and rear wheels:

a pressure sensor 12, 14 for measuring and delivering the inflation pressure of the tire Pa, Pr of the wheel, for example mounted inside the tire;

an accelerometer 16, 18 for measuring and delivering the vertical acceleration Avr, Ava of the wheel at its center, for example a single-axis or tri-axis accelerometer mounted in the center of the wheel; and

a temperature sensor 20, 22 for measuring and delivering the temperature Ta, Tr of the gas contained in the tire of the wheel, for example a temperature sensor mounted inside thereof. Each of these sensors 12-22 comprises means 12a, 14a, 16a, 18a, 20a, 22a forming a transmitting antenna for the delivery of a low frequency electromagnetic signal representative of the quantity it measures. Receiving antenna means 24 are provided in the system 10 for receiving the signals emitted by the sensors 12-22 and extracting therefrom the corresponding measured quantities Ava, Avr, Pa, Pr, Ta, Tr.

The reception antenna means 24 are connected to an analog / digital converter 26 (CAN), for example a 0-order blocker. The CAN 26 is capable of sampling all the measurements Ava, Avr, Pa, Pr, Ta , Tr at a predetermined sampling frequency fe, for example 50 Hz, and delivering sampled measurements Ava (k), Avr (k), Pa (k), Pr (k), Ta (k), Tr (k), where k is the k ^th sampling instant.

Compensation means 28, 30 are connected to the CAN 26 and are adapted to correct the measurement of the inflation pressure Pa (k), Pr (k) of each tire by the corresponding temperature measurement Ta (k), Tr (k ) in order to compensate for the temperature drifts of the pressure sensors 12, 14 with respect to a nominal operating state, as is known per se in the state of the art. The system 10 further comprises means 32 for melting tire inflation pressures and vertical accelerations of the front and rear wheels. These means 32 of fusion are particularly suitable for estimating the tire inflation pressure using the redundancy of information contained in the measurements. The estimation of the inflation pressures of the tires is thus more reliable and more accurate.

This estimate therefore largely eliminates the aberrant measurements of sensors and their corruption due to the wireless transmission of information between the sensors and the rest of the system.

The merging means 32 comprise means 34 for estimating the stiffness coefficients Kpa, Kpr of the tires of the front and rear wheels as a function of the measurements of the vertical accelerations thereof.

These estimation means 34 comprise a bandpass filter 36 connected to the CAN 26 and adapted to process the accelerations Ava (k), Avr (k) sampled by applying band pass filtering. This filtering is implemented in a frequency range in which the power of the modes of the front and rear wheels is concentrated. This frequency range corresponds to the rolling resistance range and is for example substantially equal to the range [8; 20] Hz.

The estimation means 34 also comprise time-shift means 38 connected to the band-pass filter 36 and adapted to temporally reset the sampled acceleration Ava (k) of the front wheel to the sampled acceleration Avr (k) of the wheel. rearward to output the accelerated accelerations Avr (k), Ava (kn) front and rear wheels, corresponding to the same altitude of the ground to apply the hypothesis according to the relationship (2) described above.

These recalibration means 38 comprise, for this purpose, calculation means 40 adapted to estimate the digital intercorrelation IC (N) of the accelerations Avr (k), Ava (k) delivered by the filter 36 according to the relation:

+ OO

IC (N) = ΣAvr (k) xAva (N -k) (3) k = -oo

The means 40 for calculating the cross-correlation are adapted to implement an estimator of this cross-correlation, as is known per se in the field of signal processing. The resetting means 38 also comprise, connected to the calculation means 40, means 42 for determining the maximum of the correlation IC (N) and the sampling instant n corresponding to this

maximum. This instant n corresponds to the time difference between the front and rear wheels undergoing the same portion of roadway. Time delay means 44 are connected to the means

42 and the filter 36, and are adapted to apply a delay of n samples to the sampled acceleration Ava (k) of the front wheel and thus to deliver a sampled acceleration Ava (kn) time-corrected on the sampled acceleration Avr (k) of the rear wheel. The means 34 for estimating the stiffness coefficients of the tires of the front and rear wheels further comprise means

46 for calculating the latter, connected to the filter 36 to receive the accelerations Avr (k), Ava (k) of the rear and front wheels and the means 44 for resetting to receive the acceleration Ava (k-n) recalibrated the front wheel.

These calculation means 46 are adapted to calculate the stiffness coefficients of the tires of the front and rear wheels as a function of the accelerations they receive based on the mechanical model of Figure 1 to model the dynamic behavior of the front and rear wheels. More particularly, by using the fundamental principle of the dynamics applied to this model in relation to the hypothesis according to relation (2), it can be shown that the vertical accelerations Avr (k), Ava (k) of the wheel centers can to be modeled in discrete time according to the relations:

Avr (k) = (4)

Ava (k) (5) where mrr and mra are the masses of the rear and front wheels respectively, and Zvr and Zva are the altitudes of the centers of the rear and front wheels respectively with respect to the reference level. The calculation means 46 are adapted to implement a recursive algorithm of the least squares in real time based on the relation (4), according to the relations: θ (k + 1) = θ (k) + K (k + 1) (Avr (k + 1) - A (k + 1) θ (k)) (6)

K (k + 1) = GT ¹ S (k) X ^τ (k + 1) (σ ² (k) + ϋ5 ^{~ l} A (k + 1) s (k) A ^τ (k + 1)) ¹ (7) s (k + l) = 05 ^"1 (s (k) -K (k + l) (k + l) s (k)) (8)

X (k + 1) = E (A ^τ (k + I) A + 1)) ¹ (9) σ (k) = Var (e (k)) (10) where (») ^τ is the symbol of the transpose, θ (k) is the vector estimate

at the instant k, A (k) is the regression vector at the instant k, E (A ^τ (k) A (k)) is the Variance of the vector A ^τ at time k, Var (e (k)) is the variance of the estimation error e (k) = Avr (k) -A (k) θ (k) at the moment k, 05 is a predetermined forgetting factor and κ (k), x (k) and s (k) are vectors or intermediate matrices used in estimating the vector θ.

Preferably, the means 46 are suitable for calculating the altitudes Zvr (k), Zva (kn) of the centers of the rear and front wheels at each sampling instant as a function of the vertical accelerations Avr (k) and Ava (kn), by example by performing a double integration thereof after filtering between 8 Hz and 20 Hz. Another example of a calculation of the altitude of a wheel according to its vertical acceleration is described in the French patent application FR 2,858,267 in the name of the applicant. Alternatively, the estimation means 34 are adapted to implement a recursive least squares algorithm in real time based on the relation (5) in a manner similar to that described above.

In a variant, the means 34 are suitable for implementing an inversion or deconvolution algorithm based on the relation (4) or (5) for estimating the stiffness coefficients.

The calculation means 46 are thus adapted to deliver, at each sampling instant, estimated values Kpa (k) and Kpr (k) of the pneumatic stiffness coefficients of the front and rear wheels.

The merging means 32 also comprise, connected to the calculation means 46 and the compensation means 28, 30, means 48 for estimating the inflation pressures of the tires of the front and rear wheels. These estimating means 48 are adapted to implement, for each of the front and rear wheels, an algorithm merging the compensated inflation pressure Pac (k), Prc (k) with the estimate of the coefficient of stiffness. Kpa (k), Kpr (k) of the tire of the wheel to estimate the inflation pressure thereof.

For the sake of brevity, only the merging of the data associated with the front wheel will be described, the merging of the data associated with the rear wheel being obtained in the same way as that of the front wheel.

For example, the estimating means 48 implement an extended Kalman algorithm for calculating an estimate Kpa (k) of the stiffness coefficient of the front wheel as a function of the compensated inflation pressure Pac (k) of its pneumatic and stiffness coefficient Kpa (k) estimated by the calculation means 46, from an observation model according to the relation:

Y (k + 1) = h (Pac (k)) + w (k) (11) where Y is the magnitude observed at time k + 1, here the coefficient of stiffness, h (Pac (k)) = cPac (k) ^α and w (k) is a measurement noise on the predetermined variance stiffness coefficient σ _Kpa .

For example, the estimation means 48 are adapted to implement a Kalman algorithm according to the relationships:

Kpa (k + 1) = Kpa (k) + K (k + 1) (Kpa (k + 1) -h (Pac (k)) (12)

Q (k + 1 / k) = E (Kpa (k) Kpa (k)) (13)

H (k + 1) = - ^ - (Pac (k + 1)) = cαPac (k + 1) ^α-1 (14) dPac

Q (k + 1 / k + 1) = Q (k + 1 / k) - K (k + 1) H (k + 1) Q (k + 1 / k) (15)

K (k + 1) = Q (k + 1 / k) H (k + 1) (σκ _pa + H (k + 1) Q (k + 1 / k) H (k + 1)) (16) where K (k + 1) is the Kalman gain matrix at time k + 1, Q (k + 1 / k) is the prediction of the covariance matrix of the estimation error at time k + 1, and Q (k + 1 / k + 1) is the correction of the covariance matrix of the estimation error at time k + 1.

As can be noted, the calculation at time k + 1 of the estimate

Kpa (k + 1) of the stiffness coefficient of the tire of the front wheel is produced as a function of an estimation error formed by the difference in the coefficient of stiffness Kpa (k + 1) of the tire of the front wheel estimated by the calculation means 46 and a coefficient of stiffness h (Pac (k)) of this tire calculated according to the relation h (Pac (k)) = cPac (k) ^α as a function of the compensated inflation pressure Pac (k) at the moment k of the tire of the front wheel. Thus, the estimation means 48 are able to estimate the stiffness coefficient of the tire of a wheel, and therefore the inflation pressure thereof via the relation (1), with increased accuracy because two sources of information. Additional information regarding this pressure is used. This makes it possible in particular to reject outliers in the measurement of the inflation pressure by the pressure sensor 12, 14 and the information transmission errors between means 12a, 14a transmitting antenna of this sensor 12, 14 and the means 24 forming receiving antenna.

In a variant, the calculation means 48 implement an extended Kalman algorithm for calculating an estimate Pac (k) of the stiffness coefficient of the front wheel as a function of the compensated inflation pressure Pac (k) of its tire. and the coefficient of stiffness Kpa (k) estimated by the calculation means 46 from an observation model according to the equation: Y (k + 1) = g (Kpa (k)) + v (k) (17) where Y is the magnitude observed at time k + 1, here the pressure of

and v (k) is a measurement noise on the pressure compensated inflation of predetermined variance σp _a .

For example, in this variant, the estimation means 48 are suitable for implementing an algorithm according to the relationships:

Pac (k + 1) = Pac (k) + K (k + 1) (Pac (k + 1) - g (Kpa (k)) (18)

S (k + 1 / k) = E (Pac (k) Pac (k)) (19)

G (k ₊ 1) (20)

S (k + 1 / k + 1) = S (k + 1 / k) - K (k + 1) G (k + 1) S (k + 1 / k) (21) K (k + 1) = S (k + 1 / k) G (k + 1) (σκ _pa + G (k + 1) S (k + 1 / k) G (k + I) J ¹ (22) where S (k + 1 / k) is the prediction of the covariance matrix of the estimation error at time k + 1, and S (k + 1 / k + 1) is the correction of the matrix of covariance of the estimation error at time k + 1. As can be noted, the computation at time k + 1 of estimation

Pac (k) of the inflation pressure of the tire of the front wheel is made as a function of an estimation error formed by the difference of the compensated inflation pressure Pa (k + 1) of the tire of the front wheel and of the an inflation pressure g (Kpa (K)) of this tire calculated according to the relation

according to the coefficient of stiffness Kpa (k) of this pneumatic at the moment k.

Thus, in this variant also, the calculation means 46 are able to estimate the tire inflation pressure of a wheel with increased precision because two sources of additional information concerning this pressure are used. Like the previous variant, this makes it possible in particular to reject outliers in the measurement of the inflation pressure by the pressure sensor 12, 14 and the information transmission errors between means 12a, 14a forming the transmission antenna of this device. sensor 12, 14 and means 24 forming receiving antenna. In another embodiment of the system according to the invention, the melting means 32 are adapted to directly merge the inflation pressures of the tires with the vertical accelerations of the wheels.

In this embodiment, the merging means 32 are structurally similar to those described in relation to FIG. 3. They comprise the bandpass filter 36 as well as the time resetting means 38.

The means 48 for estimating the inflation pressures of the tires are connected to the filter 36 to receive the accelerations Avr (k), Ava (k) of the rear and front wheels, to the resetting means 44 for receiving the acceleration Ava (kn) recalée of the front wheel, and means 28, 30 of compensation for receiving the compensated inflation pressures Pac (k), Prc (k).

In this embodiment, the estimating means 48 implement an extended Kalman algorithm for calculating an estimate Kpa (k) of the stiffness coefficient of the front wheel as a function of the compensated inflation pressure Pac (k ) of his tire and the acceleration they receive from an observation model according to the relation:

Y (k + 1) = d (Pac (k), Pr c (k)) + x (k) (23) where Y is the magnitude observed at time k + 1, here the vertical acceleration of the wheel before,

d (Pac (k), Prc (k)) = (mraxAva (k-n) (24) and x (k) is a measurement noise on the vertical acceleration of the rear wheel of predetermined variance σ _vr .

For example, in this embodiment, the estimation means 48 are suitable for implementing an algorithm according to the relationships:

Pac (k + 1) = Pac (k) + K (k + 1) (Avr (k + 1) -d (Pac (k), Prc (k)) (25)

R (k + 1 / k) = E (Pac (k) Pac (k)) (26)

R (k + 1 / k + 1) = R (k + 1 / k) - K (k + 1) D (k + 1) R (k + 1 / k) (28)

K (k + 1) = R (k + 1 / k) D (k + 1) (σ _vr + D ^τ (k + 1) R (k + 1 / k) D (k + 1)) ^{~ 1} ( 29) where R (k + 1 / k) is the prediction of the covariance matrix of the estimation error at time k + 1, and R (k + 1 / k + 1) is the correction of the covariance matrix of the estimation error at time k + 1.

The ground altitudes of the front and rear wheels are for example calculated in the manner described above.

As can be noted, the calculation at time k + 1 of the estimate

Pac (k) of the tire inflation pressure of the front wheel is made according to an estimation error formed by the difference of the acceleration vertical acquired Avr (k + 1) of the tire of the front wheel and a vertical acceleration d (Pac (k), Prc (k)) thereof calculated according to relation (24) as a function of the compensated inflation pressures tires at the moment k. Thus, in this embodiment, the means 46 this calculation are able to estimate the inflation pressure of the tire of a wheel with increased precision because two sources of additional information concerning this pressure are used. As above, this makes it possible in particular to reject outliers in the measurement of the inflation pressure by the pressure sensor 12, 14 and the transmission errors of information between means 12a, 14a forming the transmitting antenna of this sensor 12 , 14 and means 24 forming receiving antenna.

The system 10 according to the invention finally comprises diagnostic means 50 connected to the compensation means 28 and 30 and to the means 48 for estimating the inflation pressures.

These diagnostic means 50 are particularly adapted to diagnose the malfunctions at each tire, that is to say a failure of the pressure sensors mounted therein or an inversion of the mounting of the tire (for example the mounting of the tire on the left front wheel instead of the left rear wheel as would normally be the case) by comparing the compensated inflation pressures to the respective estimated inflation pressures.

For example, the means 50 diagnose that the pressure sensor has failed or that the tire comprising this sensor has been mounted on a wrong wheel, if the compensated and estimated pressures associated with this tire differ by more than X%, where X is a predetermined number, for at least a predetermined duration.

The measurement of the pressure sensor 12, 14 is included in a communication frame comprising a sensor identification field and the frame is transmitted by the means forming a transmission antenna 12a, 14a to the receiving antenna means 24. These then extract the measurement of the received frame and associate it with the sensor identified by the identification field. Thus, if the tire is mounted on the left front wheel instead of the left rear wheel, the means 24 combine the measurement extracted from the communication frame to the left rear wheel and not to the left front wheel. It then follows errors in the calculations using such a measurement and therefore an estimate of the tire inflation pressure invalid.

However, the accelerometers 16, 18 are not mounted in the tires but on the wheels so that they can not be subject to an inversion mounting. If the estimated pressure and the compensated pressure associated by the system according to the invention to the same tire differ by more than X%, the means 50 then calculate the inflation pressures of the tires of the front and rear wheels from the relation (1 ) as a function of the stiffness coefficients calculated by the means 46 In a variant, the means 50 calculate these stiffness coefficients as a function of the vertical accelerations as previously described if they have not already been calculated and then calculate the inflation pressures. according to these.

The means 50 then compare the compensated pressure at each of the pressures thus calculated.

If the compensated pressure corresponds to one of the pressures calculated as a function of the stiffness coefficients, then the diagnostic means 50 locate the tire from which the failure originates as being that associated with the pressure calculated as a function of the stiffness coefficients, c that is to say the tire associated with the accelerometer whose measurements were used to calculate the coefficient of stiffness of the tire of the wheel on which it is mounted.

If the compensated pressure does not correspond to any of the pressures calculated as a function of the stiffness coefficients, then the means 50 diagnose a defective state of the pressure sensor associated with this compensated pressure.

In a variant, the means 50 diagnose that a pressure sensor is faulty or an inversion of the mounting of a tire if the matrix Q (k / k), R (k / k), S (k / k) covariance of the associated estimation error calculated by the estimation means 48 does not tend to 0.

Finally, if the compensated pressures and the predicted pressures correspond, then the means 50 diagnoses the inflation state of the tires according to the estimated inflation pressures.

For example, the means 50 are adapted to compare each of these estimated pressures to a predetermined set of pressure intervals each representative of a tire inflation state (puncture, under-inflation, normal inflation, over-inflation). The means 50 thus determine the state of inflation of the tire associated with this estimated pressure as a function of the membership of the latter at one of the pressure intervals.

Specific embodiments of the invention have been described.

In a variant, the means 28, 30 for compensating the inflation pressures and the temperature sensors 20, 22 are omitted, the algorithms described above being executed as a function of the inflation pressures acquired delivered by the pressure sensors 12, 14.

It has been described a system based on a mechanical model of the wheel illustrated in Figure 1. Alternatively, the system is based on the mechanical model shown in Figure 4. Figure 4 is a schematic view of a mechanical model generally referred to as "bicycle model". This type of model allows in particular to take into account the case of active suspensions equipping the vehicle and applies to front and rear wheels arranged on the same side of the vehicle.

The difference with the model of Figure 1 consists in the fact that the body C of the vehicle is likened to a mass Mc suspended both on the front wheel Roa and the rear wheel Ror.

Based on the fundamental principle of the dynamics applied to this bicycle model as well as the hypothesis according to relation (2), it can be shown that the vertical accelerations Ava (k), Avr (k) of the front and rear wheels are modeled according to the relation: mra

Ava (k -n) mrr Kpr (k) / Kpa (k)

- (Zva (k -n) - Zvr (k)) Kpr (k)

Avr (k) = mrr (30)

Zva (k -n) (Kpr (k) / Kpa (k)) xRa (k) mnr Rr (k)

1

Zvr (k) mrr where Ra and Rr are stiffness coefficients of the suspensions of the front and rear wheels respectively, and Zva and Zv are the first derivatives of the altitudes of the centers of the front and rear wheels respectively, i.e. the speeds vertical displacement thereof.

Previous relationships are then modified to account for the introduction of these first derivatives.

Likewise, fusion means 32 have been described based on an estimation of inflation pressures by Kalman filtering. Other types of estimates are possible. For example, in a variant, the estimating means 48 implement an estimate of inflation pressures based on Baysian filtering using one or other of the observation models according to the relationships (11), (17) and (23).

It has been described a system according to the invention applied to a pair of front and rear wheels of a motor vehicle arranged on the same side thereof. Of course, it will be understood that this system can also be applied to each pair of front and rear wheels arranged on the same side of the vehicle.

Claims

1. System for determining the inflation pressure of tires mounted on front and rear wheels of a motor vehicle, characterized in that it comprises: means (12, 14) for acquiring tire inflation pressures front and rear wheels;

means (16, 18) for acquiring the vertical accelerations of the front and rear wheels; and

means (32) for melting the acquired pressures and the accelerations acquired in order to estimate the inflation pressures of the tires of the front and rear wheels.

2. System according to claim 1, characterized in that the means (32) for melting comprise:

means (34) for estimating the stiffness coefficients of the tires of the front and rear wheels as a function of the vertical accelerations acquired therefrom; and

means (48) for estimating the inflation pressures of the tires of the front and rear wheels as a function of the inflation pressures acquired and the estimated stiffness coefficients of the tires of the front and rear wheels.

3. System according to claim 2, characterized in that the means (34) for estimating the stiffness coefficients comprise means (38) of time registration of one of the accelerations acquired on the other accelerations acquired and means calculation of the stiffness coefficients as a function of the accelerations thus recalibrated temporally.

4. System according to claim 2 or 3, characterized in that the means (34) for estimating stiffness coefficients comprise means (36) for bandpass filtering acquired accelerations arranged between means of acquisition accelerations and time resetting means.

5. System according to claim 4, characterized in that the means (36) of band-pass filtering are adapted to implement a filtering in a frequency range substantially equal to [8, 20] Hz.

6. System according to claim 3, 4 or 5, characterized in that the means (38) of time registration comprise means (40) for calculating the intercorrelation of the accelerations acquired and means (42) of application of a delay corresponding to the maximum of the intercorrelation calculated at the acceleration gained from the front wheel.

7. System according to any one of claims 3 to 6, characterized in that the means (46) for calculating the stiffness coefficients are adapted to implement a recursive least squares algorithm in real time based on a mechanical model predetermined wheel.

8. System according to any one of claims 2 to 7, characterized in that the estimating means (34) are adapted to estimate said stiffness coefficients from a single-wheel mechanical model of the front and rear wheels.

9. System according to claim 8, characterized in that the estimating means (34) are adapted to estimate said stiffness coefficients based on discrete time modeling of the recalibrated accelerations of the front and rear wheels according to the relation:

Avr (k) = where k is the k ^th time of sampling, ^Avr and Ava are the vertical accelerations of the rear and front wheels respectively, Zvr and Zva are the altitudes of the centers of the rear wheels and respectively before, Kpr and

Kpa are the stiffness coefficients of the tires of the front and rear wheels respectively, and n is a sampling instant corresponding to a time difference between the rear and front wheels undergoing the same portion of roadway.

10. System according to claim 8, characterized in that the estimating means (34) are adapted to estimate said stiffness coefficients based on a discrete time modeling of the recalibrated accelerations of the front and rear wheels according to the relation:

Ava (k) = - (mrrxAvr (k + n) Zvr (k + n) - Zva (k)) ^{Kpa () Kpr (} 'I mra I Kpa (k) J where k is the k ^th time of sampling, ^Avr and Ava are the vertical accelerations of the rear and front wheels respectively, Zvr and Zva are the altitudes of the centers of the rear and front wheels respectively, Kpr and Kpa are the stiffness coefficients of the tires front and rear wheels respectively, and n is a sampling time corresponding to a time shift between the rear and front wheels undergoing the same portion of roadway.

11. System according to any one of claims 2 to 7, characterized in that the means (34) of estimation are adapted to estimate said stiffness coefficients from a bicycle mechanical model thereof.

12. System according to claim 11, characterized in that the estimating means (34) are adapted to estimate said stiffness coefficients based on a discrete time modeling of the corrected accelerations of the front and rear wheels according to the relation:

where k is the k ^th time of sampling, ^Avr and Ava are the vertical accelerations of the rear and front wheels respectively, Zvr and Zva are the altitudes of the centers of the rear and front wheels respectively, Kpr and Kpa are the stiffness coefficients of the tires front and rear wheels respectively, n is a sampling instant corresponding to a time shift between the rear and front wheels undergoing the same portion of roadway, Ra and Rr are stiffness coefficients of the suspensions of the front and rear wheels respectively, and Zva and Zvr are the first derivatives of the altitudes of the centers of the front and rear wheels respectively.

13. System according to any one of claims 2 to 7, characterized in that the means (48) for estimating the inflation pressures. are adapted are adapted to implement, for each of the front and rear wheels, a Kalman estimator based on a model binding the inflation pressure and the coefficient of stiffness of the tire of the wheel.

14. System according to claim 1, characterized in that the means (32) of fusion are adapted to implement an estimator of

Kalman relying on model linking tire inflation pressures and vertical accelerations of the front and rear wheels.

15. System according to any one of the preceding claims, characterized in that it comprises means (50) for diagnosing the operating state of the means for acquiring the inflation pressures and the mounting state of the tires. depending on the estimated and acquired inflation pressures and the acquired vertical accelerations of the front and rear wheels.

System according to claim 15, characterized in that, if an acquired inflation pressure and its associated estimated pressure differ by more than X%, where X is a predetermined number, the diagnostic means (50) are adapted to calculate the tire pressures of the front and rear wheels as a function of the vertical accelerations acquired therefrom, to compare the pressure acquired with the calculated pressures and to determine that, if the acquired pressure corresponds to one of the calculated pressures, the associated tire at the calculated pressure corresponding to the acquired pressure is the subject of an inverted assembly.

17. System according to claim 16, characterized in that, if none of the tires has a mounting inversion, then the diagnostic means (50) are able to diagnose that the acquisition means associated with the acquired pressure are faulty. .