Method for warning of under-inflation on a twin wheel tyre

Abstract

The present invention relates to a method for detecting under-inflation of a tyre on a twin wheel of a vehicle, the method comprising the steps of: - moving the twin wheel at an identified speed V over the same ground provided with sensors capable of detecting contact between the tyre and the ground, wherein two neighbouring sensors are spaced apart by a distance Di; - determining, for each sensor, a time at which the contact between the tyre and the sensor changes state; - determining a discrete metric line of contact for each tyre of the twin wheel; - determining a parabolic arc that minimises the deviation from the points of the discrete line for each tyre; - algebraically comparing the second-order coefficients of the parabolic arcs; - if the algebraic comparison exceeds a threshold S, issuing a warning that one of the tyres of the twin wheel is under-inflated.

Description

DESCRIPTION TITLE: METHOD FOR WARNING UNDER-INFLATION OF A TWIN WHEEL TIRE Field of invention

[0001] The present invention relates to the field of vehicles transporting significant loads, which requires the use of twin wheels on certain axles of the vehicle in order to reduce ground pressure. Technological background

[0002] In the freight transport sector in particular, twin-wheel vehicles are frequently used for long-distance journeys between warehouses. The leitmotif of this industry is speed of delivery, which forces the vehicles to be used intensively. However, under-inflation of a tire, caused for example by a puncture, can lead to immobilization of the vehicle to repair the tires. However, if the puncture is not detected early, the damage caused to the tire can lead to the scrapping of the tire or even the destruction of the wheel assembly. In the first case, this constitutes a financial loss for the carrier with controlled immobilization of the vehicle.In the second case, this can cause an interruption in the use of the truck to best change the tire, or even repair the damage caused by the destruction of the tire on the vehicle, not to mention the possible damage generated on the road. It is common to retread worn tires in the road transport industry. A new tire that has been driven under pressure can thus lose its "second life" and the corresponding value.

[0003] The difficulty in inspecting twin wheel tires is the accessibility of the control on the tire located on the inside of the vehicle for instrumented checks such as inflation pressure but also visual checks on the entire tire. In fact, these checks are generally carried out while the vehicle is stationary, the workshop check, for which accessibility is facilitated, only takes place at the place of vehicle maintenance. As a result, the most efficient checks are scheduled and are generally too far apart to detect under-inflation before it becomes too serious for the integrity of the tire.

[0004] The introduction of electronic pressure sensors such as TPMS (Tyre Pressure Monitoring System) in tires is a solution to provide early warning of under-inflation. However, this technological solution is expensive since it requires equipping each tire with a pressure sensor. These sensors are generally battery-powered, which requires specific maintenance. As a result, this solution is too expensive and intrusive compared to the current standard approach used in the transportation industry.

[0005] Furthermore, patent FR3030742A1 discloses the principle of estimating the contact area of ​​a tire using proximity or force sensors. When a tire passes over it, the time signal from such a sensor successively shows an approach phase, a plateau phase corresponding to the unfolding of the flattened tread in the contact area, and a moving away phase. Appropriate signal processing makes it possible to estimate the duration of the plateau, knowing the speed, and its longitudinal extent, i.e. the length of the contact area. Assuming that the contact width is constant and known, it is possible to estimate the surface of the contact area.

[0006] However, twin wheels have, by geometric construction, an axis of rotation positioned at the same distance from the ground. As a result, the contact area is almost identical between the twin wheels, regardless of the inflation pressure. In other words, the sensitivity of the contact surface is low to the variation in inflation pressure. As a result, the contact surface between the ground and the tire is not a sensitive indicator of a drop in inflation pressure between the tires of the twin wheel.

[0007] The object of the invention which follows is to propose a solution for early and reliable detection of under-inflation of the tires of a twin wheel of a vehicle with low operating costs, even for small variations in inflation pressure. Description of the invention

[0008] The invention relates to a method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle comprising the following steps: • Move the two tires of the twin wheel of the vehicle at an identified speed V on the same flat ground equipped with sensors, two neighboring sensors being distant from each other transversely to the direction of movement of the vehicle by a distance Di, so that each tire is located vertically above at least three sensors during the movement, said sensors being capable of detecting contact between the tire and the ground; • Determine for each sensor at least one moment when the contact between the tire and the sensor changes state. • Determine a discrete metric line of contact of the tire with the ground for each tire of the twin wheel by means of the identified speed V, the transverse distances Di and the instants of change of state of the at least three sensors located vertically of the tire corresponding to the entry and / or exit of the contact area; • Determine a parabola arc that minimizes the deviation from the points of the discrete line for each tire; • Algebraically compare the second-order coefficients of the arcs of parabolas, preferably perform an algebraic difference; and • If the algebraic comparison exceeds a threshold S, alert that one of the tires on the twin wheel is under-inflated.

[0009] The method is based on obtaining the entry and / or exit front of the contact patch. Necessarily, several sensors are required per tire. Knowing that the sensitive parameter according to the invention is the curvature of this entry and / or exit front of The contact patch requires a minimum of 3 sensors per tire to identify the equation of the parabola and which will minimize the difference between the points of this front and the theoretical parabola. By nature, tires have entry or exit fronts that are similar to parabolas. Of course, the greater the number of measurement points, the greater the robustness of the result by minimizing measurement noise and measurement errors.

[0010] To obtain this theoretical curvature, it is first necessary to identify the instants of change of state of the sensor. The sensor signal is similar to a square wave consisting of a rising edge at the entrance to the contact area, a plateau when crossing the contact area and a falling edge when leaving the contact area.

[0011] The rising and falling edges reflect several phenomena simultaneously. First of all, the edge is an illustration of the spatial approach of the tire following its trajectory relative to the sensor. Indeed, whatever the sensor technology used, the tire gradually approaches the measuring sensor, which quickly passes from a non-contact state to a progressive state of contact. The rising or falling edge reflects this evolution of the distance between the sensor and the tire. Then, the edge is also the translation of the length of the sensitive zone of the sensor following the direction of the tire's trajectory. The sensitive zone of the sensor has a certain length, the contact is progressive along this length. A part of the sensitive zone is in contact while the other is still out of effective contact and the resultant of the sensor signal translates these two phenomena by averaging them.Therefore, it is impossible to identify the exact moment of contact on the sensor signal since it is primarily spatial. The method therefore proposes a morphological approach to signal processing on the sensor signal to, on the one hand, identify the plateau which corresponds to complete contact over the entire active area of ​​the sensor and, on the other hand, identify the rising and falling edges which correspond to partial contact over the entire active area of ​​the sensor and to a distance between the sensor and the tire which varies greatly depending on the sensitivity of the sensor.

[0012] Therefore, it is necessary to define a unique method for determining the start or end of the contact period, which is called a sensor state change. One approach by thresholds relative to a maxima that resembles the sensor response at the plateau is a technical solution. More complex techniques can also be used. As long as the same technique is used on all sensors for the entry and / or exit of the contact area, the under-inflation detection method is robust since the shape of the state change line remains identical. This is a strong point of the method since the length of the contact area and consequently the surface area evolve according to the detection method used.

[0013] Then, the time signals must be transformed into metric signals. For this, the determination of the speed V and the transverse distances Di between the ground sensors is essential. This provides a set of measurement points in a two-dimensional metric frame of reference. This makes it possible to obtain a method independent of the speed of passage of the tires on the ground equipped with sensors and to determine a threshold S that is independent of the passage speed V, which makes the method reliable.

[0014] Speed ​​V corresponds to an instantaneous average speed at the moment the tire passes over the sensors. Since the time spent passing the sensors is short, the speed changes little during this period. This can be obtained by a specific measurement system located upstream or downstream of the ground sensors following the trajectory of the vehicle. This determination of the signals can also take place on the signal itself from the measurement sensor in connection with the sampling frequency of the measurement sensor and the shape of the output signal.

[0015] Once the discrete metric definition of the contact between the ground and the tire is established, we can identify the equation of the parabola which minimizes the deviation, generally the quadratic deviation, between the discrete measurement points and the equation of the parabola.

[0016] For each tire of the twin wheel, we obtain the equation of the parabola, this translates the convexity or the concavity of the contact line by positioning itself from the point of view of the contact area, respectively from the point of view of the exterior of the contact area, such as for example, the geometric barycenter of the contact area, respectively any point of the trajectory of the tire outside the contact area. The coefficient of order 2 of the equation of the parabola is the descriptor of the level of convexity or concavity of the parabola. If the coefficient of order 2 is positive, the parabola, seen for example from the contact area, is convex. If the coefficient of order 2 is negative, the parabola, for example for an observation from the contact area, is concave.

[0017] We then perform the algebraic comparison of the second order coefficients, whether it is the ratio of the coefficients or their difference. In the case of a twin wheel having the same distance between the wheel center and the ground for both tires, the two mounted assemblies being inflated to the same inflation pressure and the tires being identical, as well as the wheels, the two coefficients are identical, the comparison tends towards a target value, 0 for the difference and 1 for the ratio. As soon as one tire is in a situation of under-inflation compared to the other, the two coefficients differ. For normal pressure and imposed load conditions, the parabolas are generally convex from the point of view of the contact patch.The more the pressure of one of the mounted assemblies differs from the other, the more the second order coefficients move away from each other to the point where the under-inflated tire ends up with a concave parabola seen from the inside of the contact patch, respectively convex seen from the outside of the contact patch. Thus, the comparison of the coefficients moves away from the target value. As initially, we do not know which tire is under-inflated compared to the other, the difference in the coefficients is insensitive to the order of the coefficients, unlike the ratio which will have to tend from the target value 1 towards the value 0 or infinity depending on whether the coefficient of the deflated tire is in the numerator or denominator.

[0018] When the distance from this target value is greater than a threshold S, under-inflation of one of the sets of twin wheels is confirmed and an alert is sent to the operator operating the ground measuring device.

[0019] The method is more accurate than that used using the surface of the contact patch which is sensitive for twin wheels only when the inflation differences are significant between the two mounted sets, approximately a loss of inflation pressure of half, or about 4 bars for tires generally inflated to 7 or 9 bars. Here, a pressure difference of 2 bars, or about 20 percent, will be identified by the method which allows to be earlier on the appearance of under-inflation and to act before a significant structural defect occurs in the structure of the tire itself.

[0020] Advantageously, the method comprises the complementary steps: • Identify the discrete line with the smallest second-order coefficient; • Identify the tire that generated the discrete line with the smallest second-order coefficient; the identified tire is in an under-inflated condition.

[0021] While detecting under-inflation is a minimum for vehicle safety, detecting an under-inflated individual can speed up the repair or restoration of the vehicle, which saves time and therefore money for the goods and passenger transport industry, which generally uses twin wheels.

[0022] Having previously identified the two second-order coefficients of the parabolas, it is sufficient to evaluate the coefficients between them to identify the under-inflated tire compared to the other.

[0023] Preferably, each tire has a trajectory supported by at least five sensors during its movement on the ground.

[0024] A minimum of three sensors is required to identify the equation of the parabola. However, increasing the number of measurement points increases the accuracy of the method by minimizing measurement errors. Furthermore, an odd number of measurement points is preferable compared to a theoretical parabola, whether convex or concave.

[0025] Specifically, the comparison of the second order coefficients being carried out by an algebraic difference, the threshold S corresponds to a curvature of value 5 m' 1 .

[0026] The inventor has found that for heavy goods vehicle tires, having generally rectangular contact areas with contact lines at the entrance or exit of the contact area that are rather rectilinear, a threshold value S of 5 makes it possible to state that one of the mounted assemblies of the vehicle is in an under-inflated condition compared to the other when the tires comply with the ETRTO standards (acronym in English for European Tyres and Rim Technical Organization).

[0027] According to a first embodiment, the contact between the sensor and the ground being previously identified on the sensor signal by a constant amplitude A of the signal, the change of state of the sensor signal corresponds to the crossing of a threshold T by the signal, the threshold T corresponding to a percentage of the amplitude A of the signal.

[0028] If the detection of the plateau on the measurement signal which corresponds to the complete contact between the active zone of the sensor and the tire is easy by a morphological approach or threshold or optimum detection, a first definition of the contact line between the tire and the ground is given by the exceeding of a threshold T which is a function of the amplitude level A of the signal which corresponds to the plateau. This threshold T is a percentage of the amplitude A of the plateau. The percentage is a function of the sensor technology used to take into account the non-linearity of the sensor and the measurement noise associated with the technology. Thus, the contact line, at the entrance or exit of the contact area, corresponds to all the measurement points having crossed the threshold T by rising on the one hand or falling on the other hand.

[0029] According to a second embodiment, the contact being previously identified on the sensor signal by a constant amplitude A of the signal, the instants of change of state of the sensor correspond to the intersection between the tangent of the sensor signal at the abscissa point corresponding to the ordinate at mid-amplitude A / 2 and a straight line whose ordinate is equal to the value A.

[0030] If the detection of the plateau on the measurement signal which corresponds to the complete contact between the active zone of the sensor and the tire is easy by a morphological approach or threshold or optimum detection, a second definition of the contact zone is given by the set of points included between two particular points.

[0031] These particular points, which are not necessarily measurement points, are defined by the intersection between two lines. The first line corresponds to the one defining the plateau of amplitude A previously identified. The second line corresponds to the tangent of the sensor signal at the point whose abscissa is half the amplitude A of the plateau.

[0032] Regardless of the method used on the measurement signals to identify the contact area, as long as it is the same between the various measurement sensors, the method remains robust. Indeed, the parabolic shape of the contact line is similar regardless of the definition of the identified state change points. As a result, the comparison gives an equivalent result by possibly adjusting the threshold S to the method used.

[0033] According to a specific embodiment, the identification of the speed of passage of the tire over the sensor is obtained by the estimated duration of the rising edge and / or the falling edge of the sensor signal as the tire passes, knowing that the distance of these edges is constant depending on the geometric characteristics of the tire, preferably its diameter.

[0034] It is entirely possible to use the sensor signal to evaluate a speed of passage of the mounted assembly when crossing the measurement sensor. Indeed, the inventor noted that the shape of the sensor signal, whatever the technology used, is always related to a trapezoid and that the inclined edges of the trapezoid have a constant spatial length according to the identity of the tire, whatever the inflation pressure, the static load applied to the mounted assembly. This length can be pitched according to the method of definition of the contact line used and then serve as a reference distance. Knowing the sampling frequency of the sensor and the number of measurement points included on the inclined edge of the trapezoid, it is possible to define an instantaneous speed of passage of the tire on the measurement sensor. In fact, two measurements can be carried out during the same passage, at the entry and exit of the contact area.By taking either of the values ​​or the median value, it is entirely possible to identify a realistic passage speed which is only a function of the discretization of the measurement signal.

[0035] Preferably, having previously identified the two moments when the contact between the tire and the sensor changes state when the tire passes over the sensors, the signals from the sensors are realigned with each other by making the moments corresponding to the middle of the interval defining the contact state of each sensor coincide.

[0036] If the sensor signals are not phased with each other in time, it is important to temporarily re-time the signals from each ground sensor to identify the equation of each parabola associated with each mounted assembly. The method consists of taking advantage of a particularity of the assemblies mounted in free rolling conditions. Indeed, the geometric barycenter of the contact area is located at an equal distance from the contact lines located at the entrance to the contact area and at the exit of the contact area. Each sensor making it possible to identify a so-called contact state between the tire and the sensor on the signal, the middle of this time interval corresponds to this geometric barycenter of the contact area. Thus, we can define a point, potentially imaginary, time per measurement signal corresponding to this middle. It is then sufficient to make these different potentially imaginary points coincide between the various measurement signals to phase all the sensors together. Brief description of the drawings

[0037] The invention will be better understood on reading the following description, given solely as a non-limiting example and made with reference to the appended figures in which the same reference numbers designate identical parts throughout and in which: • Fig. 1 illustrates an example of application of the method according to the invention for a vehicle with twin wheels; • Figs. 2a and 2b show two views of an evaluation system according to the invention; • Fig. 3 shows a signal measured at the output of a sensor of the system and a processing for determining the instants of change of state of the sensor according to the invention; • Fig. 4 shows the same signal measured at the output of a sensor of the system and another processing for determining the instants of change of state of the sensor according to the invention; • Fig. 5 shows a representation of the contact areas of the axles of a vehicle on a measuring system, with one of the tires of a twin wheel under-inflated; Fig. 6 is a time-domain representation of the second-order coefficients of the parabolas of the tires of a twin wheel, one of the tires being deflated from a given instant. Detailed description of the embodiments

[0038] Fig. 1 shows a convoy consisting of a tractor 12 and a semi-trailer 13 passing over a measuring system 200 as described in Figs. 2a and 2b.

[0039] In this example, the tractor 12 consists of a front axle 401 and a rear axle 501 and the semi-trailer 13 consists of a group of three axles 601. In addition, the rear axle 501 of the tractor 12 is provided with two “twin” tires at each of its ends.

[0040] Only the twin wheel 510 located on the right side of the axle 501 of the tractor 12 is in an under-pressure situation. In the case of twin tires, the increase in length or contact patch surface area resulting from under-pressure on one of the twin tires will be distributed between the two tires on the side affected by the under-pressure. Thus, even if a single tire of the twin wheel 510 is in an under-pressure situation, it is indeed the length and the contact patch surface area of ​​the two tires that will increase. Consequently, it is not possible to distinguish which tire of the twin wheel 510 is under-inflated.

[0041] The sensitivity to possible underpressure of a tire is therefore much lower than when the axle is fitted with a single tire.

[0042] The measuring system 200, given here as an example, appears in figures 2a and 2b and consists of: • A housing 10 consisting of two access ramps 15 and a measuring zone located between the two access ramps 15. • Two tire presence detection devices each consisting of three piezoelectric sensors 110, positioned along a line transverse to the direction of travel of a vehicle arriving at the housing. In this example, the piezoelectric sensors are buzzers glued to the structure of the housing 10. • A line of measuring sensors 100 positioned along a line transverse to the direction of travel of the vehicle arriving at the housing 10. These measuring sensors can be either variable reluctance sensors or eddy current sensors. Alternatively, this line of electromagnetic sensors can also be replaced by an optical measuring system applying the principle of laser triangulation. • Processing electronics 140 to which the measurement sensors 100 and the tire presence detection sensors 110 are connected. In this example, the processing electronics 140 also contains an RFID reader allowing the reading of RFID chips integrated into the tires or stuck on the tires whose under-inflation is evaluated.

[0043] When a tire 20 passes over the housing 10 of the measuring system, the presence of the tire is first detected by a first line of tire presence detection sensors, then, when the tire leaves the housing 10 of the measuring system, its presence is detected by a second line of tire presence detection sensors. The distance between the two tire presence detection devices being known, it is then possible to calculate the speed of the tire by a very simple formula: Average speed = d / tO.

[0044] In this formula, the distance d is the distance separating the two transverse lines of tire presence detection sensors 110, and the time t0 is the time elapsed between the detection of the presence of the tire by any one of the sensors belonging to the first line of tire presence detection sensors 110 and the detection of the presence of the tire by any one of the sensors belonging to the second line of tire presence detection sensors 110. The speed of the tire being known, it is then very easy to estimate a local length of the contact area of ​​the tire, vertical to the sensor considered by taking into account the sampling frequency of the sensor as we will see in Fig 3.

[0045] Fig. 3 is the time representation of the output signal of a sensor of a measuring system over which a tire has passed and a first method of detecting the instants of change of state of the sensor.

[0046] Curve 1001 is the mathematical representation of the various measurement points recorded by the sensor. This curve 1001 appears as a trapezoid as it passes through the tire. From a low level, which reflects the absence of contact between the sensor and the tire, we observe a first rising edge up to a plateau that lasts a few moments before a falling edge that returns to the initial level.

[0047] Here, the upper plate is almost constant and tangent to a straight line 1002 whose ordinate is equal to the value A. This value A is constant since when the sensor is under the tire, the detected value is constant during the crossing of the tire on the sensor.

[0048] The rising and falling edges reflect, on the one hand, the progressive approach of the tire in rotary motion towards the sensor and, on the other hand, the progressive covering of the tire on the active part of the sensor. As these phenomena occur both during the attack movement on the sensor and during the escape movement made by the tire on the sensor, it is normal for the rising and falling edges to be symmetrical.

[0049] According to a first embodiment, the detection of the change of state of the sensor is carried out using a threshold value T which is proportional to the value A of the plateau. In this case, we are at mid-amplitude of A for the value of T. We can therefore define a straight line 1003 corresponding to this value which intercepts the curve 1001 at the rising edge by a point 1011 and at the falling edge by a second point 1012.

[0050] The interval of the sensor measurement points between these two reference points defines the contact zone of the tire with the sensor. Points outside this zone are called out of contact. Thus, points 1011 and 1012 define the instants of change of state of the sensor. The duration of the zone between 1011 and 1012 represents the temporal length of the contact between the two structures, which will determine the metric distance of this contact by multiplying this duration by the instantaneous speed of passage of the tire over the sensor.

[0051] Finally, the symmetry of this signal 1001 makes it possible to determine an axis of symmetry passing through the point 1015 which therefore corresponds to a given instant which is defined by the abscissa of this point 1015. By construction, this abscissa of the point 1015 corresponds to the median plane of the contact area of ​​the tire which is common to the entire transverse width of the tire, which makes it possible to phase the signals 1001 from the various sensors measuring the passage of the tire on the measuring system.

[0052] Fig. 4 is the same time representation of the output signal of a sensor of a measuring system over which a tire has passed as well as a second method of detecting the instants of change of state of the sensor.

[0053] Curve 1001 is the mathematical representation of the various measurement points recorded by the sensor. This curve 1001 appears as a trapezoid as it passes through the tire. From a low level, which reflects the absence of contact between the sensor and the tire, we observe a first rising edge up to a plateau that lasts a few moments before a falling edge that returns to the initial level.

[0054] Here, the upper plate is almost constant and tangent to a straight line 1002 whose ordinate is equal to the value A. This value A is constant since when the sensor is under the tire, the value detected is constant during the crossing of the tire on the sensor. Of course, depending on the sensor technology and the tire dimensions, the value can change.

[0055] The rising and falling edges reflect, on the one hand, the gradual approach of the tire in rotary motion towards the sensor and, on the other hand, the gradual covering of the tire on the active part of the sensor. As these phenomena occur both during the attack movement on the sensor and during the escape movement made by the tire on the sensor, it is normal for the rising and falling edges to be symmetrical.

[0056] According to a second embodiment, the detection of the change of state of the sensor is carried out using a property of the curve 1001. Generally at mid-amplitude A / 2 of the plateau, the rising and falling edges have points inflection. It is then possible to pass a tangent 1005 on the rising edge of signal 1001 passing through the inflection point of the rising edge. Similarly, we can construct tangent 1006 on the falling edge. Here, the inflection points were imposed by the intersection of line 1003 with signal 1001. Line 1003 has the equation that the abscissa is equal to half the amplitude A of the plateau of signal 1001.

[0057] The intersection of the tangent 1005, respectively 1006, with the curve 1002 representing the signal plateau, defines a single point which is called the sensor state change point whose time abscissa is referenced 1011, respectively 1012.

[0058] The interval of the sensor measurement points between these two reference points defines the contact zone of the tire with the sensor. The points outside this zone are called out of contact. Thus, points 1011 and 1012 define the instants of change of state of the sensor. The duration of the zone between 1011 and 1012 represents the temporal length of the contact between the two structures, which will allow the metric distance of this contact to be determined by multiplying this duration by the instantaneous speed of passage of the tire over the sensor.

[0059] It is noted here that the contact times are not identical depending on the method used. However, this does not affect the metric parabola which will be based on the instant of change of state 1011 of the sensor, respectively the instant 1012.

[0060] Finally, the symmetry of this signal 1001 makes it possible to determine an axis of symmetry passing through the point 1015 which therefore corresponds to a given instant which is defined by the abscissa of this point 1015. By construction, this abscissa of the point 1015 corresponds to the median plane of the contact area of ​​the tire which is common to the entire transverse width of the tire, which makes it possible to phase the signals 1001 from the various sensors measuring the passage of the tire on the measuring system.

[0061] Fig. 5 is a representation of the contact patches of vehicle tires across their width for various axles of a vehicle.

[0062] Here, at the top of Fig. 5 is the right side of the vehicle when you are the driver of the vehicle. Knowing that the convoy moves in a straight line according to the solid arrow shown in the middle of Fig. 5.

[0063] The footprints on the left of the figure represent the front of the convoy. The footprints on the right of the figure correspond to the rear of the convoy. The first axle of the vehicle consists of a single wheel for which the method does not apply. However, the implementation of the method on this steering axle shows a symmetry of behavior between the right and left sides of the vehicle, which is expected by the method due to the trajectory of the vehicle expressed by the solid arrow.

[0064] The second and third axles of the vehicle are made up of twin wheels of different dimensions. The second axle, which traditionally represents the drive axle, only carries the weight of the engine, which rests vertically on this axle. The third axle, which corresponds to the loading area of ​​the vehicle, is equipped with twin wheels of larger dimensions to support a load much greater than that carried by the first two axles. Here, a tire on a twin wheel of this third axle has been deliberately deflated. This is the twin wheel on the right side of the vehicle and the tire located inside the vehicle. Generally, inspection of this tire is difficult due to its location on the vehicle.

[0065] The vehicle passed through the straight-line measuring system at a speed of 30 km / h. The measurement signals were recorded by each sensor. Here, we see that for the third axle, each tire passed through approximately 7 sensors during its passage through the system.

[0066] After processing the signals as illustrated in Fig.3 or Fig.4, the contact time between the tire and the sensor was transformed through the measurement of the vehicle's passing speed into a metric length. In addition, the system, including the sensors, being placed on the ground, the transverse distances Di between two neighboring sensors of the system is known whether the sensors are aligned transversely to the direction of passage of the vehicle or staggered. All of this information makes it possible to construct a metric contact length for each sensor and to phase the sensors between them spatially and temporally thanks to the imaginary time points representing the center of contact according to the direction of movement of the tire.

[0067] The set of contact lines from each sensor and the spacing of these lines using the distances Di make it possible to obtain a representation of the contact patch of each tire. Each contact patch measurement line begins and ends using the previously identified instants of change of state of the sensor.

[0068] It is then possible to determine for the entry line, respectively the exit line, of the contact patch of each tire, a parabola which minimizes the deviation from the points of change of state at the entry of the contact patch, respectively the points of change of state at the exit of the contact patch.

[0069] Here in the figure, we have identified a parabola for the input line and the output line. Then we have extracted from the equation of the parabola, the second order coefficient of the parabola which is similar to the curvature of the parabola. And we have taken the average "a" of these coefficients which are displayed for each tire in Fig. 5.

[0070] We note that the average “a” between the right and left tires of the vehicle are almost identical for the first two axles, which is expected.

[0071] On the other hand, for the third axle, on the right side, there is a significant disparity in curvature between the two tires of the twin wheel, unlike what happens on the left side. This is normal since the tire inside the vehicle of the right twin wheel of the third axle is deliberately under-inflated. Here, there is a difference of 7 units on the right side for a difference of only one unit on the left side.

[0072] Fig. 6 is a time-domain representation of the second-order coefficients of the parabolas of the tires of a twin wheel, one of the tires being deflated from a given instant.

[0073] Fig. 6 is a graph whose abscissa is time with periodic measurements of the second order coefficient of the parabola best describing the line of contact of the tire with the ground at the entrance of the contact point following the direction of passage of the vehicle on the measuring system.

[0074] The 2002 series corresponds to the measurements of the outer casing of the twin wheel vehicle and the 2003 series corresponds to the measurements of the inner casing of the vehicle of the same twin wheel. At the beginning, although a gap remains between the two values ​​of the second order coefficient of the parabola, it is stable over time and of low level. Thus, we can define a 2001 master curve of theoretical average evolution of the coefficient. From a moment, the inner tire began to gradually deflate. Looking at the 2003 series, we observe that the coefficient began to drift compared to the 2001 master curve. The same level of drift is observable on the 2002 series corresponding to the measurements on the outer envelope of the vehicle of the twin wheel. Thus, the gap between the two values ​​Aa, materialized by the double vertical arrow, increased until it exceeded a threshold S signifying that one of the two envelopes of the twin wheel is under-inflated. To identify the incriminated wheel, it is necessary to identify the one whose coefficient decreases instead of increasing.

[0075] The proposed method thus makes it possible to detect under-inflation of a tire of a twin wheel in an early and robust manner. In addition, the method makes it possible to identify which of the two tires is under-inflated relative to the other.

Claims

CLAIMS 1. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle comprising the following steps: - Move the two tires of the twin wheel of the vehicle at an identified speed V on the same flat ground equipped with sensors, two neighboring sensors being distant from each other transversely to the direction of movement of the vehicle by a distance Di, so that each tire is located vertically above at least three sensors during the movement, said sensors being capable of detecting contact between the tire and the ground; - Determine for each sensor at least one moment when the contact between the tire and the sensor changes state. - Determine a discrete metric line of contact of the tire with the ground for each tire of the twin wheel by means of the identified speed V, the transverse distances Di and the instants of change of state of the at least three sensors located vertically of the tire corresponding to the entry or exit of the contact area; - Determine a parabola arc that minimizes the deviation from the points of the discrete line for each tire; Algebraically compare the second-order coefficients of the arcs of parabolas, preferably perform an algebraic difference; and If the algebraic comparison exceeds a threshold S, alert that one of the tires on the twin wheel is under-inflated.

2. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to claim 1 in which the method comprises the additional steps: Identify the discrete line with the smallest second-order coefficient; Identify the tire that generated the discrete line with the smallest second-order coefficient; the identified tire is in an under-inflation condition.

3. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 2 in which each tire has a trajectory bearing on at least five sensors simultaneously during its movement on the ground.

4. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 3 in which the comparison of the second order coefficients is carried out by an algebraic difference, the threshold S corresponds to a curvature of value 5 m' 1 .

5. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 4 in which, the contact between the sensor and the ground being previously identified on the sensor signal by a constant amplitude A of the signal, the change in state of the sensor signal corresponds to the signal crossing a threshold T, the threshold T corresponding to a percentage of the amplitude A of the signal.

6. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 4 in which, the contact being previously identified on the sensor signal by a constant amplitude A of the signal, the instants of change of state of the sensor correspond to the intersection between the tangent of the sensor signal at the abscissa point corresponding to the ordinate at mid-amplitude A / 2 and a straight line whose ordinate is equal to the value A.

7. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 6 in which the identification of the speed of passage of the tire over the sensor is obtained by the estimated duration of the rising edge and / or falling edge of the signal from the sensor as the tire passes, knowing that the distance between these edges is constant depending on the geometric characteristics of the tire, preferably its diameter.

8. Method for detecting under-inflation of a tire forming part of a twin wheel of a vehicle according to one of claims 1 to 7 in which, having previously identified the two instants at which the contact between the tire and the sensor changes state when the tire passes over the sensor, the signals from the sensors are time-recalibrated with each other by making the instants corresponding to the middle of the interval defining the contact state of each sensor coincide.

Images