Method for determining the compensated dynamic radius of a vehicle wheel, method for estimating the tread depth of a tire, and motor vehicle for implementing said methods

Abstract

Method for determining a compensated dynamic radius of a vehicle wheel, method for estimating the tread depth of a tire, and motor vehicle for implementing said methods. The invention relates to a method for determining a compensated dynamic radius of a vehicle wheel, said compensated dynamic radius being a function of: - a raw dynamic radius, - momentary values ​​of variables, - compensation factors specific to said variables. According to the invention, the compensation factors are obtained by a learning phase where: - values ​​of the raw dynamic radius are acquired for each variable while said vehicle is in operation, as a function of the evolution of said variables; - the compensation factors are calculated from each of said values.The invention also relates to a method for estimating the tread depth of a tire and a motor vehicle implementing said methods. Figure for the abstract: Figure 7.

Description

Title of the invention: Method for determining a compensated dynamic radius of a wheel of a vehicle, method for estimating the depth of a tread of a tire and motor vehicle for implementing said methods Technical field

[0001] The field of the present invention is that of wheel tires of a motor vehicle.

[0002] More specifically, the invention relates to a method for determining a compensated dynamic radius of a wheel of a vehicle, a method for estimating the depth of a tread of a tire and a motor vehicle for implementing said methods. Prior art

[0003] Conventionally, the tire has at its outer surface a zone called “tread”, corresponding to the outer surface of the tire which is in contact with the road.

[0004] The tread includes a relief, also called a "sculpture", which allows in particular the evacuation of rainwater, snow, dust and heat, in order to limit the loss of grip of the tire or to avoid aquaplaning.

[0005] As the vehicle travels, the tire's tread wears down and becomes smooth, increasing the risk of loss of grip. Therefore, once the vehicle reaches a certain level of wear, it is necessary to replace the worn tire with a new one.

[0006] In order to detect the wear of a tire, it is known to use wear indicators in the form of a colored label integrated into the tread. As the tread comes into contact with the road, the tire wears and its thickness decreases until the tire wear indicator is revealed.

[0007] This solution is unsatisfactory insofar as the vehicle owner must visually check his tire to determine whether it needs to be replaced. He must also remember to check his tires himself. However, a person with average attention does not check his tires systematically and risks, after a certain time, traveling on board a vehicle equipped with worn tires offering little grip, which presents a danger manifest.

[0008] Also, there is a growing need to autonomously monitor the wear of a tire, particularly in the context of vehicle fleet management, a model which is part of the growing development of new modes of mobility for which the driver is not the owner and where monitoring and maintenance are carried out by specific organizations on large sets of vehicles.

[0009] Also, autonomous monitoring of tire wear applies even more to autonomously driven vehicles.

[0010] In order to overcome these drawbacks, it is known to monitor the state of wear of the tire tread from the dynamic radius of the wheel.

[0011] We refer to [Fig. 1] to define what is meant by dynamic radius.

[0012] A tire 1, mounted in a vehicle wheel (not shown) rests on a ground 3. The loaded tire 1 is deformed at the contact zone between the tire 1 and the ground 3, at its tread. At this contact zone, the radius of the tire 1 is defined by the loaded radius Rc, which corresponds to the distance between the axis of rotation 5 of the wheel and the ground 3.

[0013] The loaded radius Rc is less than the nominal radius Rn, defined by the radius of the tire when the wheel is not loaded, i.e. not mounted in the vehicle. It is therefore understood that the distance traveled by the tire 1 differs depending on whether the nominal radius or the loaded radius is considered. The length developed by the loaded radius Rc over one wheel revolution is thus less than that obtained by the nominal radius Rn.

[0014] The purpose of determining the dynamic radius is to give a radius value as close as possible to the actual distance traveled by the tire 1.

[0015] More precisely, the dynamic radius Rdyn of a wheel is the radius of the virtual wheel which would have a point contact with the ground, with the same linear speed and the same rotational speed as the real wheel.

[0016] The prior art has identified several physical variables which influence the value of the dynamic radius. Among these variables, we particularly remember pressure, speed, speed squared, load.

[0017] The way in which these variables influence the value of the dynamic radius depends on so-called “compensation” parameters, which are representative of the state of wear of the tire and which thus change over time.

[0018] Document EP2867036B1 discloses a solution aimed at quantifying the momentary dynamic radius by taking into account in part the state of wear of the tire.

[0019] To do this, the method proposes to first measure a so-called “raw” dynamic radius, equal to the speed of the vehicle divided by the rotation speed of the wheel, then to compensate this dynamic radius measurement according to the current variables (i.e. measured at a given time) retained as influencing the value of the dynamic radius (in particular pressure, speed, square speed, load).

[0020] We then obtain a so-called “compensated” dynamic radius which is a function of the raw dynamic radius and the current variables.

[0021] The method described in this document also proposes to integrate into the calculation of the dynamic radius compensation parameters representative of the state of wear of a tire.

[0022] These parameters are learned through specific preliminary tests carried out in the factory and then integrated into the dynamic radius calculation algorithm in the form of charts giving compensation parameter values ​​to be applied for a type of tire considered.

[0023] A disadvantage of this solution is that the compensation parameters applied in the calculation algorithm apply systematically regardless of the tire in a family considered, and this independently of any changes that may occur during the life of the tire.

[0024] Also known is prior art document WO2021247036Al which describes a system using a tire monitoring sensor to determine data on dimension, pressure, identification, load parameters, deformation, etc. based on the current circumference of the tire. After determining these parameters, an algorithm determines the tire wear value. A controller adjusts the tire parameters, such as the tire radius based on different variables such as temperature, stiffness, acceleration, load, etc.

[0025] As with document EP2867036B1, the system and method described in document WO2021247036Al do not make it possible to take into account the behavior of the tire when it is used on a vehicle, i.e. during the life of the tire on the vehicle.

[0026] Also known is document US2021188017A1, which describes a system and method for monitoring tread wear measuring various parameters such as tread depth, tire acceleration, inflation pressure, tire loads, etc. The system determines a calibrated tire wear pattern based on the measured tire parameters. This system operates using an electronic control unit or a cloud computing system installed on the vehicle to execute the system.

[0027] However, as for documents EP2867036B1 and WO2021247036A1, the system and method described in document US2021188017A1 proposes a characterization upstream of the use of the tire in a vehicle, and this for each type of tire.

[0028] Thus, the systems and methods of document US2021188017A1 also do not make it possible to take into account changes in the behavior of the tire when it is used on a vehicle. Statement of the invention

[0029] The present invention aims to overcome the drawbacks of the prior art, and to do this relates to a method for determining a compensated dynamic radius of a wheel of a vehicle, said wheel comprising a tire and said vehicle comprising a set of sensors capable of acquiring signals representative of variables chosen from a group comprising at least: the speed of the vehicle, the rotational speed of said wheel, the pressure of said tire, the load of said wheel, said compensated dynamic radius being a function of: - a raw dynamic radius calculated from the momentary speed of said vehicle and the momentary rotational speed of said wheel, - of momentary values ​​of the variables considered, - of reference values ​​of the variables considered, said method being remarkable in that the compensated dynamic radius is furthermore a function of compensation factors, specific to said variables considered and obtained by the execution of a learning phase according to which: - when said vehicle is in operation, for each of said variables, sets of values ​​of the raw dynamic radius are acquired as a function of the evolution of said variables, - the said compensation factors are calculated from each of the said sets of acquired values.

[0030] Consequently, unlike the prior art where the compensation factors were learned through specific preliminary tests carried out in the factory, the present invention provides for acquiring online, that is to say during the life of the tire mounted on the vehicle in operation, sets of values ​​of the raw dynamic radius as a function of the evolution of the variables identified as influencing the value of the dynamic radius.

[0031] Based on these sets of acquired values, the compensation factors specific to each variable are calculated.

[0032] The compensation factors are thus updated throughout the life of the tire.

[0033] The integration of these updated compensation factors into the algorithm for calculating the compensated dynamic radius makes it possible to take into account the changes in behavior that occur during the life of the tire, regardless of the type of tire.

[0034] According to optional characteristics of the method according to the invention: - in a first implementation, the learning phase is executed continuously; - in a second embodiment, the acquisition step is carried out over a limited period; - the learning phase is initiated following a step of detecting one or more predefined trigger events; - the triggering events are chosen from a group comprising at least: a change in the location of the wheel, a change of wheel or tire on the same axle, a detection of a change in the driving habits of the vehicle, aging of the tires, an overload situation detected when a predetermined overload threshold has been exceeded over a predetermined period, an overspeed situation detected when a predetermined overspeed threshold has been exceeded over a predetermined period, a predetermined minimum or maximum pressure threshold value reached by a tire, expiration of a predetermined period of use; - the method comprises a monitoring phase initiated when no triggering event has been previously detected and according to which, when said vehicle is in operation, over a limited period and for each of said variables, sets of values ​​of the raw dynamic radius are acquired as a function of the evolution of said variables; - according to one provision, the tracking phase is subsequent to the learning phase and, to determine the compensated dynamic radius, the compensation factors acquired during said learning phase are used; - according to one provision, the step in which the compensation factors are calculated from each of the sets of acquired values ​​is obtained by applying a multilinear regression to each of said sets of acquired values; - in one embodiment, the learning phase includes, prior to the step of executing the multilinear regression, a severe filtering step, applied to the sets of values ​​of the acquired raw dynamic radius; - in one embodiment, the tracking phase includes a “low severity” to “moderate severity” filtering step, applied to the sets of values ​​of the acquired raw dynamic radius.

[0035] The invention also relates to a method for estimating the depth of a tread of a tire, remarkable in that the depth of said tread is estimated from a temporal variation of the compensated dynamic radius determined according to the invention.

[0036] The invention also relates to a motor vehicle comprising hardware and / or software means for implementing the methods according to the invention. Brief description of the drawings

[0037] Other characteristics, aims and advantages of the invention will appear on reading the detailed description which follows, for the understanding of which reference will be made to the appended drawings in which:

[0038] [Fig-1] is a schematic view of a tire of a vehicle wheel.

[0039] [Fig.2] details the steps of the process for determining a dynamic radius compensated for a wheel of a vehicle according to the invention.

[0040] [Fig.3] is a graph representing the dynamic radius values ​​acquired in function of the pressure measured for one of the vehicle's tires.

[0041] [Fig.4] is a graph representing the dynamic radius values ​​acquired in function of the vehicle speed measured for one of the vehicle's tires.

[0042] [Fig.5] is a graph representing the dynamic radius values ​​acquired in function of the pressure and speed measured for one of the vehicle's tires.

[0043] [Fig.6] is a graph showing an example of linear regression applied to dynamic radius values ​​acquired as a function of measured pressure values.

[0044] [Fig.7] is a flowchart showing an example of an algorithm that can be implemented work to carry out the method for determining a compensated dynamic radius of a wheel of a vehicle according to the invention. Description of the embodiments

[0045] In the remainder of the description, elements having an identical structure or similar functions are designated by the same reference.

[0046] The method for determining a compensated dynamic radius of a wheel of a vehicle according to the invention is implemented in a vehicle (not shown in the figures) equipped with hardware and software means adapted to implement said method.

[0047] For this purpose, the software means comprise a computer program code means, comprising in particular the algorithm implemented to execute the method of the invention.

[0048] The hardware means comprise a set of dedicated sensors capable of acquiring signals representative of physical variables identified as influencing the value of the dynamic radius.

[0049] These variables are chosen from a group comprising at least: the speed of the vehicle, the rotation speed of said wheel, the pressure of said tire, the load of said wheel.

[0050] In a known manner, the speed of the vehicle can be acquired by a positioning system, known as “GPS” (English acronym for “Global Positioning System”). The rotation speed of the wheel can be provided by a sensor wheel speed sensor, known as "WSS" (an acronym for "Wheel Speed ​​Sensor"). The pressure is provided by a tire pressure monitoring system, known as "TPMS", an acronym for "Tire Pressure Monitoring System". The load can be provided by calculating data from the TPMS system.

[0051] Other variables can also be monitored using sensors known to those skilled in the art, such as temperature, square speed, wheel torque, and the wheel footprint quotient on the ground.

[0052] According to the invention, the compensated dynamic radius of the wheel is a function of: - the gross dynamic radius, calculated from the momentary speed of said vehicle and the momentary rotational speed of said wheel, - of momentary values ​​of the variables considered, - of reference values ​​of the variables considered, and - of compensation factors specific to the variables considered updated throughout the life of the tire.

[0053] To do this, the method which is the subject of the invention provides for obtaining the compensation factors by carrying out learning phases initiated throughout the life of the tire.

[0054] [Fig.2] details the steps of a learning phase of the method of the invention.

[0055] The learning phase comprises a first step E1 of signal acquisition, followed by a second step E2 of calculating the compensation factors from the acquired signals.

[0056] The signal acquisition step E1 aims to acquire, when said vehicle is in operation, for each variable identified as influencing the value of the dynamic radius, sets of values ​​of the raw dynamic radius as a function of the evolution of these variables.

[0057] For example, when the variables monitored are the vehicle speed, the wheel rotation speed, the wheel tire pressure and the wheel load, the evolution of the dynamic radius is monitored as a function of each of these variables.

[0058] Thus, using the set of sensors previously described, the values ​​of the dynamic radius are acquired as a function of the pressure values ​​measured for each tire over a given period, the values ​​of the dynamic radius are acquired as a function of the vehicle speed values, for each wheel of the vehicle and the values ​​of the dynamic radius are acquired as a function of the load values ​​measured for each wheel.

[0059] We thus obtain two- or three-dimensional point clouds, as illustrated in Figures 3 to 5.

[0060] As shown in [Fig.3], we obtain a two-dimensional point cloud, re showing the dynamic radius values ​​as a function of the pressure values ​​measured over a given period. These values ​​are acquired for all the vehicle's tires (the values ​​relating to a single tire are illustrated in [Fig.3]).

[0061] As shown in [Fig.4], a two-dimensional point cloud is obtained, representing the values ​​of the dynamic radius as a function of the vehicle speed values ​​over a given period. These values ​​are acquired for all the vehicle's tires (the values ​​relating to a single tire are illustrated in [Fig.4]).

[0062] As shown in [Fig.5], a three-dimensional point cloud is obtained, representing the values ​​of the dynamic radius as a function of the pressure values ​​measured over a given period and as a function of the vehicle speed values ​​over a given period. These values ​​are acquired for all the vehicle's tires (the values ​​relating to a single tire are illustrated in [Fig.5]).

[0063] When the values ​​are acquired, step E2 is initiated, according to which the compensation factors specific to each variable are calculated from each of the sets of acquired values.

[0064] This calculation step is carried out by means of mathematical functions including in particular a multilinear regression, applied to each set of acquired values.

[0065] For reasons of memory resource limitations, the algorithm that determines the value of the compensation factor for the variable considered (step E2) is recursive, which means that the calculation is updated after each new measurement. More precisely, step E2 is an intermediate update iteration of the compensation factor considered. Step E2 is applied as many times as necessary until reaching end-of-learning conditions. The value of the compensation factor of the last iteration then becomes the value retained for the calculation of the compensated dynamic radius.

[0066] [Fig.6] shows the straight line obtained by linear regression applied to the values ​​of the dynamic radius for one of the vehicle's tires, as a function of the measured pressure values.

[0067] Here, the slope of the line is equal to 0.18 mm / 10 kPa, which means that if the pressure of a tire has increased by 10 kPa (0.1 bar), the dynamic radius increases by 0.18 mm.

[0068] The compensation factor retained for the pressure and for the tire considered is then equal to 0.18 mm / 10 kPa.

[0069] The same procedure is followed for all the vehicle's tires and for all the variables monitored.

[0070] The compensated dynamic radius is then calculated by the following formula: Rdync„mp ~ Rdynbru! + E ^aimmp Rdynmtnri is the dynamically compensated radius of the wheel, Rdvn is a raw dynamic radius of the wheel, defined by the following formula: • raw n _ V^fuath , where Vvehicle is the momentary speed of the vehicle and œ is the speed Rdynhnl, ~ « of momentary rotation of said wheel, where ®ieanp is the compensation factor specific to the variable i considered, where Ki is a momentary value of the variable i considered, where K; r is a reference value of the variable i considered. 'ref

[0071] Thus, when the variables monitored are the vehicle speed, the wheel rotation speed, the wheel tire pressure and the wheel load, the compensated dynamic radius is calculated by the following formula: S + V-Vref) +Peamp>^P-Pref) + Lmmp* ( L - Lref)™ : Rdynamt„ is the compensated dynamic radius of the wheel, V is the momentary speed of the vehicle, a) is the momentary rotational speed of said wheel, Vcomp is the compensation parameter of the vehicle speed, Vref is a reference value of the vehicle speed, P is the momentary pressure of the wheel tire, PComP is the compensation parameter of the wheel tire pressure, Pref is a reference value of the wheel tire, L is the momentary load of the wheel, Lcomp is the compensation parameter of the wheel load, Lref is a reference value of the wheel load.

[0072] According to one arrangement of the invention, the step of calculating the compensation factors during the learning phase can be supplemented by a filtering step initiated prior to the step of executing the multilinear regression.

[0073] The type of filtering can be so-called “severe” filtering and is applied to the sets of values ​​of the raw dynamic radius acquired as a function of the variables considered, this in order to guarantee learning with input signals which are as clean as possible.

[0074] Filtering is obtained by using a statistical criterion related to the average standard deviation of the dynamic radius. To do this, we choose to reject an instantaneous value of the dynamic radius when it deviates from a predetermined value depending on this standard deviation.

[0075] For example, if the current value of the dynamic radius is greater than the absolute value of three times the standard deviation, then it is rejected.

[0076] For example, the test results demonstrate that the standard deviation for “severe” filtering is equal to 0.5 millimeters. Thus, so-called “severe” filtering rejects dynamic radius values ​​deviating by + / - 1.5 millimeters from the current average.

[0077] According to one execution of the method of the invention, the acquisition step (step E1) can be executed over a limited period.

[0078] According to one arrangement of the invention, the learning phase can be initiated following the detection of triggering events.

[0079] As non-limiting examples, these triggering events are notably chosen from a group comprising at least: - a change in wheel location, - a change of wheel or tire on the same axle, - detection of a change in vehicle driving habits, such as detection of a change in average speed due to a change in vehicle ownership, - aging of tires, - overload situation that can be detected when a predetermined overload threshold has been exceeded over a predetermined period, - overspeed situation that can be detected when a predetermined overspeed threshold has been exceeded over a predetermined period, - predetermined minimum or maximum threshold value of pressure reached by a tire, - expiration of a predetermined period of use.

[0080] According to one arrangement of the invention, when no triggering event has been detected, a so-called monitoring phase is initiated rather than a learning phase.

[0081] According to this monitoring phase, sets of values ​​of the raw dynamic radius are acquired as a function of the evolution of the retained variables but the compensation factors are not calculated.

[0082] The compensation factors used for calculating the compensated dynamic radius are then those acquired and stored during the learning phase. The tracking phase is therefore preferably subsequent to the learning phase.

[0083] In the case where the monitoring phase is not subsequent to the learning phase and the compensation factors have not yet been calculated, compensation factors provided by the manufacturer for a type of tire considered can be used.

[0084] When a tracking phase is carried out, a less severe filtering than that applied for the learning phase is applied to the sets of values ​​of the ray raw dynamics acquired. This is a so-called "mildly severe" to "moderately severe" filtering. As with "severe" filtering, "mildly severe" filtering “moderately severe” is applied using a statistical criterion related to the average standard deviation of the dynamic radius, rejecting an instantaneous value of the dynamic radius when it deviates from a predetermined value depending on this standard deviation.

[0085] If the current value of the dynamic radius is greater than the absolute value of three times the standard deviation, then it is rejected. For example, test results demonstrate that the standard deviation from "mild" to "medium" filtering is equal to 1 millimeter.

[0086] Thus, a filtering called “not very severe” to “moderately severe” rejects dynamic radius values ​​deviating by + / - 3 millimeters from the current average.

[0087] According to an alternative embodiment of the method of the invention, the acquisition step (step E1) can be executed continuously. In this embodiment, there is no longer any need for a trigger element to initiate the learning phase. The learning is then continuous, which means that there is no distinction between the learning phase and the tracking phase. In this way, the compensation parameters are updated at each iteration and are directly taken into consideration for the calculation of the compensated dynamic radius.

[0088] Reference is made to [Fig.7] showing a flowchart describing an example of an algorithm that can be implemented to execute the method for determining a compensated dynamic radius of a wheel of a vehicle according to the invention.

[0089] The algorithm runs through successive iterations. When the cycle reaches a new iteration, the events that may have occurred since the last iteration are evaluated (step E10).

[0090] If a triggering event has occurred, then we move into a learning phase.

[0091] We then reset all the previously calculated compensation factors, that is to say we erase all the previously acquired compensation factors (Eli).

[0092] According to step E12, the signals representative of physical variables identified as influencing the value of the dynamic radius (corresponding to step E1 previously described) are acquired using the dedicated sensors.

[0093] The question of whether one is still in the learning phase is again asked (step El3). A priori, if an event that triggered a learning phase was detected previously, then the probability that one is still in the learning phase is high.

[0094] If the answer is yes, then the step in which the com- thinking is initiated (step E2 of the method of the invention).

[0095] Severe filtering can be applied to the sets of values ​​of the raw dynamic radius acquired as a function of the variables considered (step El4).

[0096] Multilinear regression is then applied to the values ​​of the dynamic radius acquired as a function of the variables considered, in order to deduce the compensation factors (step E15).

[0097] The algorithm then asks the question of whether end-of-learning conditions are reached (step El6).

[0098] If the answer is no, we return to the initial step and a new iteration of the algorithm is initiated without memorizing the calculated compensation factors.

[0099] If the answer is yes, the compensation factors are stored (step E17) and a monitoring phase is moved on (step El8).

[0100] The algorithm then returns to the initial step and a new iteration is initiated with the acquired compensation factors stored in step E17.

[0101] If no triggering event is detected (step E10), which is probably the case since an end of learning step has previously been detected, we move on to step E12 of signal acquisition, according to which we acquire the signals representative of physical variables identified as influencing the value of the dynamic radius. The compensation factors used for calculating the compensated dynamic radius are then those acquired and stored during the learning phase.

[0102] The question of whether one is in the learning phase is asked again (step El3) and the answer will probably be negative.

[0103] A “low severity” to “moderate severity” filtering can be applied to the sets of values ​​of the raw dynamic radius acquired as a function of the variables considered (step E19).

[0104] We then move on to step E20 according to which we apply the compensation factors stored during the learning phase to calculate the compensated dynamic radius.

[0105] According to a particular application of the method of the invention, the temporal variation of the compensated dynamic radius measured by the method which has just been described is used in order to estimate the depth of the tread of a tire.

[0106] Indeed, if we observe temporal variations in the compensated dynamic radius when we have already compensated the raw dynamic radius by the compensation factors applied to the monitored variables, then the variations in the compensated dynamic radius reflect a variation in the depth of the tire tread.

[0107] The vehicle (not shown in the figures) comprises hardware and / or software means making it possible to implement the method according to the invention for estimating the depth of a tread of a tire which is the subject of the invention. The software means may in particular comprise a computer program code means, comprising in particular the algorithm adapted to carrying out the steps of the method for estimating the depth of a tread of a tire.

[0108] It should be noted that the determination of the compensated dynamic radius may have advantages other than that of estimating the depth of a tire tread, such as in particular that of allowing an estimation of the longitudinal speed of the vehicle or that of an application linked to odometry, that is to say the precise estimation of the distance traveled.

[0109] As goes without saying, the present invention is not limited to the sole embodiments of this method for determining a compensated dynamic radius of a wheel of a vehicle, of this method for estimating the depth of a tread of a tire and of this vehicle for the implementation of said methods, described above solely as illustrative examples, but on the contrary it embraces all the variants involving the technical equivalents of the means.

Claims

Claims

1. Method for determining a compensated dynamic radius (o ) vn I -■ compj of a wheel of a vehicle, said wheel comprising a tire and said vehicle comprising a set of sensors capable of acquiring signals representative of variables considered in a group comprising at least: the speed of the vehicle, the rotational speed of said wheel, the pressure of said tire, the load (L) of said wheel, said compensated dynamic radius ( p ) being a function of: vn 1 1 compj - a raw dynamic radius (Rdw ) calculated from the momentary speed (V) ' noise of said vehicle and the momentary rotational speed (œ) of said wheel, - momentary values (K^ of the variables considered - reference values (^ \ of the variables considered, h-ef / said method being characterized in that the compensated dynamic radius ( P \ is furthermore a function of compensation factors («.), specific to said variables considered and obtained by the execution of a learning phase according to which: - we acquire (step El), when said vehicle is in operation for each of said variables, sets of values of the raw dynamic radius (Rdw ) as a function of the evolution of said variables, " raw - we calculate (step E2), from each of said sets of acquired values, said compensation factors 1. ^icomp / .

2. Method according to claim 1, characterized in that the learning phase is executed continuously.

3. Method according to claim 1, characterized in that the acquisition step (step E1) is carried out over a limited period.

4. Method according to claim 2, characterized in that the learning phase is initiated following a step of detecting one or more predefined trigger events (step E10).

5. Method according to claim 4, characterized in that the triggering events are chosen from a group comprising at least: - a change in location of the wheel, - a change of wheel or tire on the same axle, - a detection of a change in driving habits of the vehicle, - tire aging, - overload situation detected when a predetermined overload threshold has been exceeded over a predetermined period, - overspeed situation detected when a predetermined overspeed threshold has been exceeded over a predetermined period, - predetermined minimum or maximum pressure threshold value reached by a tire, - expiry of a predetermined period of use.

6. Method according to one of claims 4 or 5, characterized in that it comprises a monitoring phase initiated when no triggering event has been previously detected and according to which, when said vehicle is in operation, over a limited period and for each of said variables, sets of values of the raw dynamic radius (Rdvn) are acquired as a function of the evolution of said raw variables (step El2).

7. Method according to claim 6, characterized in that the tracking phase is subsequent to the learning phase and in that the compensation (a) acquired during said learning phase is used to determine the compensated dynamic radius (d Y the factors of VH 1 - comp). ic&mp

8. Method according to any one of claims 1 to 7, characterized in that the step in which the compensation factors ( n ] are calculated from each of the sets of acquired values is obtained (uicomp) by applying a multilinear regression to each of said sets of acquired values (step El5).

9. Method according to claim 8, characterized in that the learning phase comprises, prior to the step of executing the multilinear regression, a “severe” filtering step (step E14), applied to the sets of values of the raw dynamic radius (Rdvn) acquired.

10. Method according to any one of claims 6 to 9, characterized in that the tracking phase comprises a “low severity” to “moderate severity” filtering step (step El9), applied to the sets of values of the raw dynamic radius (RdVn) acquired. ' raw

11. Method for estimating the depth of a tread of a tire, characterized in that the depth of said tread from a temporal variation of the compensated dynamic radius (o \ determined according to any one of dyncompj claims 1 to 10.

12. Motor vehicle comprising hardware and / or software means for implementing the method according to any one of claims 1 to 11.

Images