SENSOR

Technical field

[0001] The invention relates to a method to improve the measurement accuracy of a Weigh in Motion (WIM) sensor. The invention also relates to a WIM sensor with an improved meas uring accuracy arranged in a roadway.

Background art

[0002] WIM sensors measure forces that a vehicle exerts on a surface of a section of a roadway by the wheels of the ve hicle while driving on the section of the roadway. For the sake of simplicity, section of the roadway will be referred to just as roadway. Such WIM sensors are often used in a WIM system in combination with other sensors for vehicle identi fication, speed measurement or measurement of environmental factors. A WIM sensor is arranged in the roadway in such a way that at least one wheel of a vehicle moves across the WIM sensor .

[0003] WIM systems with at least one WIM sensor are often used to measure a wheel force or an axle force or the total weight of a vehicle. The wheel force corresponds mainly to a proportion of the total weight of a vehicle that exerts on the roadway surface by the wheel. The axle force is the sum of all wheel forces of the wheels of an axle. The total weight is the sum of all axle forces of all axles of the ve hicle. If a force exerts on a WIM sensor arranged in the roadway, the force is provided in form of a force signal. [0004] An excessively high wheel force, for example when an overloaded vehicle is driving on a roadway, has a strong effect on the roadway and can lead to greater wear or even damage to the roadway.

[0005] Applications such as the direct enforcement of an overloaded vehicle or a toll payment dependent on the wheel force, axle force or total weight place high demands on the accuracy of a WIM system and the force measurement of the WIM sensor .

[0006] A WIM sensor is known from EP0491655B1. The WIM sensor is designed as a hollow-profile, with several force measuring elements spaced at a distance from each other with in the profile. The force measuring elements provide a force signal. The calibration of the WIM sensor is done by driving a vehicle with a known weight over the WIM sensor. The weight of the vehicle can be determined beforehand on a static weighing station, where the vehicle is weighted while sta tionary. A disadvantage of a static weighing station is that dynamic effects that have an influence on the force exerted by a wheel on the roadway are not taken into account. The term "dynamic" refers to a "time dependent change". A dynamic effect is, for example, a swinging of the body mass, which is connected to the wheel via a wheel suspension. The body mass is the mass of the body of a vehicle. The body of the vehicle comprises all parts of the vehicle excluding wheels and wheel suspension. For example the engine, the chassis, one or more persons if applicable, and freight if applicable, are part of the body mass . [0007] A swinging of the body mass can be caused by a pre vious unevenness in the roadway surface. A swinging is a tem poral change of a distance, for example the distance of the body to the roadway. For each temporal position, the acceler ation of the body mass can be determined. An acceleration of the body mass causes a change of the wheel force a wheel of the vehicle is exerting to the roadway. Another example for a dynamic effect is an acceleration or deceleration of the ve hicle, causing an acceleration of the body mass as well. Due to dynamic effects, a force acting on the roadway surface though the wheel may not only be caused by the proportion of the total weight of the vehicle but also by the acceleration of a moving body mass.

[0008] Providing a signal refers to the signal being available for further use. Providing a signal also includes storing the signal in an electronic memory and loading the signal from said memory. Providing a signal also includes displaying the signal on a display unit.

[0009] Dynamic effects are considered in W02011120176A1 , disclosing a WIM sensor which is calibrated by a calibration vehicle passing over the WIM sensor. The calibration vehicle is equipped with at least one measuring wheel, a so-called wheel force dynamometer, which directly determines the dynam ic wheel force, which the wheel exerts on the roadway sur face, when passing over the WIM sensor. A calibration func tion is determined by comparing the wheel force measured by the calibration vehicle with the force measured by the WIM sensor for that wheel. Dynamic effects of vehicles passing the calibrated WIM sensor after calibration procedure is con- eluded cannot be eliminated by using the calibration func tion .

[0010] Dynamic effects are caused by unevenness in the roadway surface. The term unevenness refers to all elevations and depressions within a surface of the roadway that deviate more than 0.01 mm from a plane defining the average surface of the roadway.

[0011] The average surface defines two directions being parallel to this surface: the longitudinal direction and the lateral direction. The longitudinal direction is the direc tion vehicles are moving on a roadway. The lateral direction is perpendicular to the longitudinal direction. The lateral direction defines a lateral axis being parallel to the lat eral direction. The longitudinal direction defines a longitu dinal axis being parallel to the longitudinal direction. A position on the roadway is defined by its coordinate on the lateral axis and its coordinate on its longitudinal axis.

[0012] A third direction is given by the vertical direc tion. The vertical direction is perpendicular to the lateral direction. The vertical direction is perpendicular to the longitudinal direction. The vertical direction defines a ver tical axis being parallel to the vertical direction. The height, defined by a coordinate on the vertical axis is a measure for a deviation of an unevenness from the average surface at a certain position. A height of zero defines a po sition on the average surface of the roadway.

[0013] The average surface is the surface for which the root mean square deviation considering all positions on the roadway for which a height of an unevenness is available is minimal .

[0014] A three dimensional representation of the surface of a roadway consists of at least three positions on the roadway defined by a lateral coordinate and a longitudinal coordinate having a height. Such a three dimensional repre sentation of the surface of a roadway is referred to as "road profile" and can be determined with high precision, for exam ple with a roadway surveying vehicle as disclosed in US2017350698AA.

[0015] For all positions on a surface of a roadway, a height can be recorded. If a road profile does not contain a recorded height for a certain position, a height for this po sition is determined using an interpolation of the heights of the closest positions the height are available for.

[0016] To increase the accuracy of a WIM sensor arranged in a roadway, the force measuring elements of the WIM sensor are calibrated at predetermined intervals, often in accord ance with regulations of the jurisdictional district. These calibrations are expensive and time consuming.

[0017] The task of the invention is to increase the meas uring accuracy of a WIM sensor.

Summary of the invention

[0018] At least one of the tasks is solved by the features of the independent claim.

[0019] The invention relates to a method for generating a calibration function of a WIM sensor; which WIM sensor is ar- ranged in a roadway ; which WIM sensor measures a wheel force exerted on the surface of the roadway; wherein the following steps are performed: a) The road profile of a roadway is recorded, the road profile of the roadway being a three dimensional rep resentation of the surface of the roadway; b) The wheel force is determined by a simulation, which wheel force is exerted on the surface of the roadway by a wheel; which wheel is moving with a speed across the road profile of step a) in a longitudinal direc tion; which wheel is pressed onto the surface of the roadway; c) The dependency of the wheel force determined in step b) on the road profile determined in step a) is de termined by simulation for at least one position of the road profile recorded in step a) ; d) The dependency of the position determined in step c) is used to minimize the influence of the road profile on the wheel force measured by the WIM sensor.

[ 0020 ] Minimizing the dependence of the wheel force meas ured by the WIM sensor from the road profile increases the measuring accuracy of the WIM sensor.

[ 0021 ] A vehicle comprises at least two wheels, which wheels are connected to a body mass by means of a wheel sus pension. The body mass comprises the combined masses of all parts of the vehicle excluding wheels and wheel suspension. For example the masses of the engine, the chassis, one or more persons if applicable, and freight if applicable, are part of the body mass. The suspension comprises a spring ele ment and a damping element. The suspension minimizes the in fluence of an unevenness of the surface of a roadway on the movement of the body mass. For example the suspension mini mizes a swinging of the body mass upon a wheel connected to the body mass encounters an unevenness while driving on the surface of the roadway. The wheel comprises a running surface which is in contact with the surface of the roadway. Further more, the wheel comprises a wheel hub, the wheel hub being the centre of rotation of the wheel while driving across the surface of the roadway and the wheel hub being connected to the wheel suspension. The running surface is part of a usual ly elastic element of the wheel, for example a gas-filled el ement in case the wheel is a gas-filled tire, or a rubber el ement in case the wheel is a full-cushion tire. The wheel therefore shows a spring like behaviour and a damping behav iour .

[ 0022 ] Despite suspension of the wheel, the body mass shows a swinging when the wheel is driving across an uneven surface. The swinging of the body mass influences the force pressing the wheel onto the surface of the roadway. If the body mass is swinging while driving across a WIM sensor being arranged in the roadway, the wheel force or axle force is not determined correctly by the WIM sensor. The wheel force shows a dynamic behaviour over time. This dynamic behaviour which can be translated into a position dependent wheel force for each a position on the roadway, if the speed of the vehicle is known. The wheel force of a vehicle influenced by the swinging of the body is denominated dynamic wheel force. [0023] The method according to the invention is used to simulate the dynamic wheel force of a vehicle and to use this simulation to improve the accuracy of a WIM sensor.

[0024] The simulation is preferably based on a so-called quarter car model, which is simulating one wheel of a vehicle on a surface with a corresponding partial body mass. The par tial body mass is the mass pressing the wheel against the surface of the roadway. The partial body mass is for conven ience's sake denominated body mass. Of course, also the so- called half-car model can be used, where one side of the car, i.e. one wheel per axle is simulated. The simulation may also be based on a so-called full track model, also known as full car model, where the whole vehicle is simulated. While the last two mentioned models promise a slightly more realistic model as movement of one wheel also influences the other wheels of the simulation, the increase in computing power is immense. For the application described hereafter, although the models may be used, the so-called quarter car model is deemed sufficient. Therefore, in the following the descrip tion is given for the quarter car model. However, it is clear that a person known in the art may use the half car model or the full car model instead.

[0025] The body mass and the mass of the wheel are con nected by a first spring and a first damping element, first damping element and first spring being arranged in parallel between the body mass and the mass of the wheel. The position of the surface the wheel is pressed onto is connected to the mass of the wheel by a second spring and a second damping el ement, second damping element and second spring being ar ranged in parallel between the mass of the wheel and the po- sition on the surface the wheel is pressed against. When resting, the first spring and the second spring are both in a state of static preloaded equilibrium.

[0026] The wheel in motion will encounter the unevenness of the surface of the roadway, resulting in a change of the distance of the mass of the wheel and the position on the surface the wheel is pressed against. This change of distance changes the loading of the second spring. The movement of the surface is thus exciting a movement of the wheel mass depend ing on the spring constant and the damping constant of the second spring and second damping element. Movement of the wheel mass is not free, as the wheel mass is connected to the body mass by the first spring and first damping element. Movement of the wheel mass changes the distance of the wheel mass to the body mass. This change of distance changes the loading of the second spring. The change of the between wheel mass and body mass is thus exciting a movement of the body mass depending on the spring constant and the damping con stant of the first spring and first damping element. In fact, a change of the position on the surface the wheel is pressed against thus results in a complex movement of body mass and wheel mass coupled by a damped two spring system. The move ment of body mass and wheel mass coupled by a damped two spring system can be described by system of differential equations solving the equation of movement of the wheel mass and the body mass. The movement of the wheel mass and body mass determines the force pressing the wheel onto the surface of the roadway.

[0027] Thus, the dynamic wheel force can be determined with knowledge of the unevenness of the roadway in the wheel track. The wheel track is the section of the roadway the wheel is covering during its movement across the surface of the roadway.

[ 0028 ] Preferably the simulation is performed for differ ent speeds, wheel masses, body masses, first spring con stants, first damping constants, second spring constants, second damping constants, one combination of which defining a unique quarter car travelling with a unique speed on the sur face of the roadway. The dynamic wheel force and its devia tion from the static wheel force is determined for each posi tion of the wheel track on the surface of the roadway. The static wheel force is the force a wheel of a resting vehicle is exerting onto the surface of the roadway.

[ 0029 ] With the result of the simulation it is possible to identify locations best suited for installation of a WIM sen sor, by evaluating which position on the roadway shows the least deviation for a predetermined quarter-car. It has been seen in recent tests that spring and damping constants of ve hicles are usually within a small range for vehicles of a certain range of body mass. For a predetermined parameter, like the desired travelling speed and/or one or more body mass of vehicles, the best location can be chosen that shows the least deviation of the dynamic weight from the static weight. Usually, this location should extend in a longitudi nal direction for a distance of at least 1 m. More, a loca tions best suited for installation of a WIM sensor can be chosen as a best compromise when considering multiple parame ters, like different speeds of vehicles, the location showing the least deviation of the dynamic weight from the static weight for all parameters considered. This is important for example, if the WIM installation is arranged in a zone where a variable-message sign changes the speed limit of the sec tion of the roadway with respect to weather conditions, time or traffic conditions. The selection of parameters for speed and weight can be done with respect to the desired criteria for vehicle mass measurement. A WIM sensor arranged in a lo cation selected by the method described above shows a higher accuracy than a similar WIM sensor arranged at a random posi tion in the roadway. Thus the method improves the accuracy of the WIM sensor.

[0030] Furthermore, the simulation determines a calibra tion function for a WIM sensor. An evaluation element for the WIM sensor evaluates at least one force signal determined and provided by the measuring element of the WIM sensor. The evaluation element applies the calibration function to the force signal and provides the result as a calibrated force signal. The calibrated force signal shows a smaller deviation from the static wheel force than a wheel force measured with out use of the calibration function, which calibration func tion is determined by the simulation described above. Thus the method improves the accuracy of the WIM sensor measuring the wheel force.

Brief description of the drawings

[0031] The drawings used to explain the embodiments show:

Fig. 1 A illustration of the method for improving the pre cision of a WIM sensor,

Fig. 2 A representation of a sectional view of a vehicle on a roadway, Fig. 3 A representation of a sectional view of a wheel of a vehicle on a roadway,

Fig. 4 A representation of a mass-spring model represent ing a vehicle on a roadway,

Fig. 5 A representation of the mass-spring model repre senting a vehicle on a roadway of Fig. 4 illustrating the forces involved,

Fig. 6 A sketch of a sectional view of a wheel of a vehi cle on a roadway, with a WIM sensor arranged in the roadway, and

Fig. 7 A sketch of a top view of a section of a roadway,

Preferred Embodiments

[0032] Fig. 1 shows an illustration of the method for im proving the precision of a WIM sensor arranged in a roadway, comprised of step a) 101, step b) 102, step c) 103 and step d) 104. The steps are described in detail below.

[0033] Step a) comprises recording a profile of a roadway 1, the profile of a roadway being a three dimensional repre sentation of the surface of the roadway. The profile of the roadway determined is stored electronically in an electronic memory, for example in form of a three dimensional object. An electronic memory can be a random access memory (RAM) and/or a memory, for example an EEPROM (electrically erasable pro grammable read-only memory) . A three dimensional object can be represented in form of a 3 dimensional wire frame model known from computer assisted design software. The electroni cally stored road profile 2 is provided for further use in the simulation of step b) 102.

[0034] Step b) 102 comprises determining the wheel force F9 by a simulation 102; which wheel force F9 is exerted on the road profile 2 by a wheel 8; which wheel 8 is moving with a speed across the road profile 2 of step a) in longitudinal direction X; which wheel 8 is pressed onto the surface of the roadway 1. The simulation is run on a processing element of a computer system. The simulation is a computer program, which defines certain operations the processing element should per form on an input. The input of the simulation are at least the road profile 2 and a description of the vehicle and vehi cle speed. The description includes all data the simulation needs to calculate a wheel force F9 as an output as described below .

[0035] Fig. 2 shows a sketch of a sectional view of a ve hicle on a roadway in a plane given by the lateral axis Y'Y and vertical axis Z'Z. The wheels 8 of the vehicle 10 are connected to the body mass 3 by the suspension 4. The vehicle 10 moving on the surface of the roadway 1 in longitudinal di rection X (not shown) .

[0036] Fig 3 shows a sectional view of the wheel 8 of a vehicle 10 on a roadway 1 in a plane given by the longitudi nal axis Y'Y and the vertical Z'Z axis simulated in one em bodiment of the simulation 102. In step b) the wheel is con nected to at least one body mass m3. The wheel 8 in step b) 102 is pressed onto the road profile 2 by the body mass m3. [0037] In one embodiment, in step b) the wheel 8 is con nected to the body mass m3 via resilient connection with a predefined spring element constant c42,c84 and damper element constant k41,k83, for example a resilient connection, in par ticular a wheel suspension 4 with spring element 42 and damp er element 41 or shock absorber 41 of a vehicle 10. In the simulation, the wheel moves across the surface of the roadway 1 in longitudinal direction X. The wheel 8 comprises a wheel hub 82, which is the centre of rotation of the wheel 8 and the theoretical location of the wheel mass 8 in the simula tion .

[0038] In one embodiment of the simulation 102, in step b) the wheel 8 features a spring element 84 and a damper element 83, for example an elastic running surface with a predeter mined spring element 84 and damper element 83. Therefore, the wheel 8 with the spring element 84 and damper element 83 is a damped mass-spring system 200.

[0039] In one embodiment, in step b) the simulation 102 comprises at least one damped mass-spring system 200; the damped mass-spring system 200 comprising at least one spring element 42,84, at least one damper element 41,83 and at least one mass m3 , m8.

[0040] Fig. 4 shows the damped mass-spring system 200 sim ulated in one embodiment of step b) 102. The mathematical de scription of the model of a vehicle 10 on a roadway 1 can be represented as a one-dimensional juxtaposition of a first spring 84, a wheel mass 8, a second spring 42 and a body mass 3. Parallel to the first spring 84, a first damper element 83 is arranged. Each spring 42, 84 and damper element 41,83 can be moved in one dimension, which is parallel to the vertical axis Z Z .

[0041] The simulation determines the wheel force F9 a wheel 8 exerts onto a contact point 9 between wheel 8 and roadway 1 on the surface of the roadway 1 by the mass m3,m8 connected to the contact point 9 via at least one damper ele ment 41,83 and at least one spring element 42,84.

[0042] The first spring 84 and the first damper element 83 are fixed to the wheel mass 8. The second end of the first spring mass 84 and the first damper element 83 are commonly connected to a theoretical contact point 9. The contact point 9 is in contact with the surface of the roadway 2. The con tact point 9 is the point in the middle of a contact surface where the wheel 8 touches the surface of the roadway 2.

[0043] Parallel to the second spring 42, a second damper element 41 is arranged. The second spring 42 and the second damper element 41 are connected to the body mass 3 and con nected to the wheel mass 8.

[0044] Body mass 3 and wheel mass 8 can move freely within the vertical axis Z'Z and within the constraints given by the forces the springs 42,84 and damper elements 41,83 exert on the body mass 3 and wheel mass 8.

[0045] For a resting vehicle 10, forces acting on the body mass 3 and the wheel mass 8 are in equilibrium. For a moving vehicle 10, only forces have to be considered which are dif fering from the equilibrium forces. The forces involved in the simulation 102 of the vehicle 10 moving on the surface of the roadway 1 are shown in Fig. 5. [ 0046 ] The forces acting on the body mass are given by differential equations

with d/dt being the first time derivate, d ² /dt ² being the second time derivate, h3 the distance between body mass m3 and wheel mass m8, F3 is a force due to an acceleration or deceleration of the vehicle 10 and/or due to the vehicle 10 driving along curve on the roadway 1, c42 the spring constant of the second spring 42, and k41 the damping constant of the second damper 41.

[ 0047 ] The forces acting on the wheel mass m8 are given by differential equations

with d/dt being the first time derivate, d ² /dt ² being the second time derivate, h8 the distance between wheel mass m8 and contact point 9, F8 is a force due to an potential unbal ance of the wheel 8 or a non-uniformity of the wheel 8 or suchlike, c84 the spring constant of the first spring 84, and k83 the damping constant of the second damper 83. [ 0048 ] The wheel force F9 acting on the road profile 2 is given by differential equation

F9 = F84 + F83.

The simulation 102 determines the movement of the wheel mass m3 and the body mass m8 by introducing an unevenness of the roadway 1 in the road profile 2 as an excitation of the damped spring mass system 200. A movement of the wheel 8 across the road profile 2 is simulated by moving the contact point 9 in a longitudinal direction X. This changes the dis tance h8. The temporal rate the distance h8 changes depends on the speed of the vehicle 10. Thus, for calculating the wheel force F9 at a given position on the roadway 1, the sim ulation has to get the speed of the vehicle 10 and the road profile 2 of the roadway 1 the wheel 8 encounters for all po sitions on the roadway as an input data. As this is not pos sible in practise, especially since the road profile 2 of the roadway 1 recorded in step a) is limited in a longitudinal direction, the simulation is started with starting conditions of the damped spring systems involved in equilibrium. In or der to determine the wheel force F9 at a certain position on the roadway, the starting position of the simulation is cho sen to be at least 1 m distance, preferably at least 100 m distance, optimal 500 m distance, from the position the wheel force should be calculated for by the simulation.

[ 0049 ] In a preferred embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the damped mass spring system 200 given above. [0050] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein some terms of the system of differential equations are neglected. A term is neglected if it does not contribute to the result ing wheel force F9 more than 10 percent, preferably less than one percent.

[0051] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein first spring 84 and second spring 42 are combined in one spring el ement (not shown) , first damper element 83 and second damper element 41 are combined into one damper element (not shown) and wheel mass 8 is neglected and forces F8 acting on the wheel 8 are neglected. The differential equations and the system of differential equations are adapted respectively. In this embodiment, the calculation is performed faster compared to calculating the first and second damper elements 41,83 separately .

[0052] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein wheel mass 8 and contact point 9 are connected with an element of constant length, comprising distance h8 constant. First damp er element 83 and first spring element 84 are omitted. The differential equations and the system of differential equa tions are adapted respectively. [0053] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein first damper element 83 and second damper element 41 are omitted. The differential equations and the system of differential equations are adapted respectively.

[0054] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein first damper element 83 is omitted. The differential equations and the system of differential equations are adapted respective ly.

[0055] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein the force F8 acting on the wheel mass 8 due to a potential unbalance of the wheel 8 or a non-uniformity of the wheel 8 or suchlike is omitted. The differential equations and the system of differ ential equations are adapted respectively.

[0056] In another embodiment of the invention, the wheel force F9 is determined by establishing and solving a system of differential equations given by the differential equations for the mass spring system 200 given above, wherein the force F3 acting on the body mass 3 due to an acceleration or decel eration of the vehicle 10 and/or due to the vehicle 10 driv ing along curve on the roadway 1 is omitted. The differential equations and the system of differential equations are adapted respectively.

[0057] In another embodiment of the invention, the force F3 includes forces exerted on the body mass m3 by wheel forc es F9 of other wheels 8 of the vehicle 10. A wheel force F9 can be determined with the simulation 102 and each wheel 8 has its own body mass in its simulation. As body masses m3 of each wheel 8 are connected in a real vehicle 10, movement of the body mass m3 of one wheel 8 influences the movement of the other body masses m3 of the other wheels 8 of a vehicle 10. The differential equations and the system of differential equations are adapted to consider the movement of other body masses m3. This simulation 102 thus uses a full car model or a half car model mentioned above.

[0058] In the different embodiments described above, the calculation is faster due to the omitting of terms, combina tion of terms, or rendering terms constant. This way, calcu lation power can be saved and a more cost effective computing device for the simulation can be used.

[0059] The different embodiments may be combined where possible and embodiments resulting from such a combination of embodiments described above are part of the invention as well .

[0060] In step c) of Fig. 1, the dependence of the wheel force F9 determined in step b) 102 on the road profile 2 rec orded in step a) is determined by simulation 102 for at least one position P of the road profile 2 recorded in step a) . [0061] The dependency of the position P determined in step c) is used to minimize the influence of the road profile 2 on the wheel force W measured by the WIM sensor 7. This increas es the measurement accuracy of the WIM sensor 7, as perturba tions of the measurement by the road profile 2 are minimized.

[0062] It has been found that in order to obtain realistic results in the simulation 102 of a dynamic wheel force F9, the road profile 2 recorded in step a) 101 is recorded with an accuracy of at least 1 mm, preferred 0.2 mm, along a ver tical axis Z'Z of the roadway 1; and the road profile 2 rec orded in step a) 101 is recorded with an accuracy of at least 100 mm, preferred 2 mm, along a longitudinal axis X'X of the roadway 1; and the road profile 2 recorded in step a) is rec- ordeded with an accuracy of at least 100 mm, preferred 2 mm, along a lateral axis Y'Y of the roadway 1.

[0063] In one embodiment, in step b) the simulation deter mines an excitation of the damped mass-spring system 200, ex cited by the wheel 8 moving with a speed across an unevenness of the surface of the roadway 1. In terms of the simulation 102, the unevenness is being represented within the road pro file 2 of step a) . This ensures that results of the simula tion 102 are showing the same wheel force F9 as a vehicle 10 that would be travelling on the roadway 1.

[0064] In one embodiment, for a position P in step c) the wheel force F9 is depending on the spring element 42,84 de fined by a spring constant and damper element 41,83 defined by a damping constant of the at least one damped mass-spring system 200. The wheel force F9 is depending on the speed of the wheel 8 moving across the surface of the roadway 1. The wheel force F9 is depending on at least one mass m3,m8 of the mass spring system 200 for a position in step c) . The speed of the wheel 8 changes the excitation of the mass spring sys tem 200 and has therefore an influence on the dynamic motion of the excited damped mass spring system 200 as well as the damped mass spring systems dynamic motion is depending on the mass m3,m8 being part of the damped mass spring system 200.

[0065] In one embodiment, step c) is performed for a num ber of adjacent positions P on the longitudinal axis X'X and/or adjacent positions P on the lateral axis Y'Y; wherein an adjacent position has a distance between 2 mm and 500 mm to each other. Furthermore, step c) is performed for at least one damped mass-spring system 200 comprising at least one mass m3,m8 and at least one spring element 42,84 and at least one damper element 41,83 and for at least one speed of the wheel 8 on the road profile 2; the speed in combination with the road profile 2 of the roadway 1 determining the excita tion of the damped mass-spring system 200.

[0066] In one embodiment the road profile (2) in step a) 101 is determined by a recording vehicle (not shown recording the unevenness of the roadway (1) while driving on the road way (1) or that the road profile (2) of step a) 101 is rec orded by an aircraft (not shown) recording the unevenness of the roadway (1) while flying over or next to the roadway (1) .

[0067] Minimizing the dependence of the wheel force F9 measurement of the WIM sensor 7 from the road profile 2 is achieved by generation of a calibration function (C) for a WIM sensor 7. For generation of the calibration function (C) the following steps are performed: dla) The road profile 2 in step a) is recorded for at least one lane of a section of the roadway 1, starting at a dis tance in front of the position P the WIM sensor 7 is arranged at with respect to the longitudinal direction X and ending at least 100 mm behind the position P the WIM sensor 7 is ar ranged at; with the distance being at least 25 m, preferred 100 m, optimal 500 m; d2a) the wheel force F9 according to step b) is determined by the simulation 102; d3a) the dependency of the wheel force F9 on the road profile 2 at the positions P the WIM sensor 7 is arranged in the roadway 1 is determined according to step c) 103 for differ ent speeds ranges within an overall speed range between and including 5 km/h and 250km/h; and step c) 103 is performed for different wheel force ranges within an overall wheel force range between and including 1000 N and 50000 N per wheel 8; and spring elements 42,84 and damper elements 41,83 are predetermined for a respective wheel force range; d4a) the deviation of the wheel force F9 according to d3a) depending on the speed range and mass range and on the road profile 2 from a measured wheel force W of a stationary vehi cle 10 with a mass within the mass range is determined; and that the calibration function (C) generated is able to sub tract the deviation from the measured wheel force W of the WIM sensor 7 according to the wheel force range and according to the speed range of a speed determined for the wheel 8.

[0068] It was found that the length of the road profile 2 given in step dla) is enough to obtain realistic results of the wheel force F9 from the simulation 102. With the result of the simulation 102 of Step d4a) a calibration function (C) can be generated which is not only calibrating a measured wheel force W depending on the measured wheel force W but al so calibrating the measured wheel force W depending on the speed of the vehicle 10.

[0069] Thus, the calibration function (C) calibrates the measured wheel force W. The calibration function (C) is a function of measured wheel force W and speed of the vehicle 10.

[0070] The calibration function (C) can be used readily in a WIM System 77. One embodiment of a WIM system 77 is shown as an example in fig. 6. The WIM system 77 is having an eval uation element 6. The WIM system 77 is adapted to determine the speed of a wheel driving across the WIM sensor 7. The WIM sensor 7 is part of the WIM system 77 and the evaluation ele ment 6 uses the calibration function (C) to subtract the de viation from a measured wheel force of the WIM sensor 7 ac cording to the wheel force range the measured wheel force is within and according to the speed range the speed determined for the wheel 8 is within; and that the evaluation element 6 provides the result as a calibrated wheel force CW.

[0071] The wheel force F9 determined by the WIM sensor 7 of a WIM system 77 is calibrated with the calibration func tion (C) ; that the calibrated wheel force CW reduces the in fluence of the road profile 2 on the measured wheel force W by at least 75% with respect to the measured wheel force W of a stationary vehicle 10.

[0072] In one embodiment, the simulation is used to select a location for a WIM sensor 7. This is advantageous, as devi- ations of a measured wheel force W depending on the road pro file 2 can be minimized before the WIM sensor 7 is arranged in the roadway 1.

[ 0073 ] For selection of a location for a WIM sensor 7 to be arranged in the roadway 1 the following steps are per formed : a2) The road profile 2 of a section of the roadway 1 is rec orded according to step a) for at least one lane of the road way 1, in which lane the WIM sensor 7 is to be arranged; which section of the roadway 1 comprises a length in longitu dinal direction X of at least 25 m, preferred 1000 m, optimal 5000 m; b2) The wheel force (F9) is determined by a simulation (102) according to step b) (102); c2) The dependence of the wheel force F9 exerted on the road profile 2 determined in step a) 101 is determined by simula tion 102 for a number of adjacent positions P in the longitu dinal direction X, wherein adjacent positions P have a dis tance between 250 mm and 5000 mm to each other; which posi tion lie within the road profile (2) recorded in step a) 101; c3) Step c2) is performed for different speed ranges within an overall speed range between and including 5 km/h and 250 km/h; and step c2) is performed for different wheel force ranges within an overall wheel force range between and in cluding 1000 N and 50000 N per wheel; and spring elements (42,84) and damper elements 41,83 are predetermined for a re spective wheel force range; d2 ) From the results of step c3) a position P of the roadway 1 is selected, where the dependence of wheel force F9 is min imal for at least one preselected wheel force range and at least one preselected speed range.

[0074] Of course, the preselected wheel force range has to be chosen according to the demands on the wheel force meas urement W. Such demands are different for wheel force meas urements W at a toll booth and wheel force measurements W at a mining site of at an airport. The same is true for prese lected speed ranges. The selection of the wheel force range and the speed range is therefore at the discretion of the owner of the WIM sensor 7.

[0075] In one embodiment, the location a WIM sensor 7 is arranged in the roadway is selected for all WIM sensors 7 of the WIM System 77. The WIM system 77 comprising at least one WIM sensor 7. When all WIM sensors 7 of a WIM System are in respective positions P where deviations of a measured wheel force W depending on the road profile 2 are minimized, the overall accuracy of the WIM system 77 is increased.

[0076] It is understood that the different aspects and em bodiments described above may be combined where possible and embodiments resulting from such a combination of embodiments described above are part of the invention as well. List of reference symbols

1 roadway

10 vehicle

11 track of wheel

101 step a)

102 step b) / simulation

103 step c)

104 step d)

2 road profile

3 body

4 suspension

41 damper element

42 spring element

6 evaluation element

7 WIM sensor

77 WIM system

8 wheel

81 running surface

82 hub

83 damper element

84 spring element

9 contact point

c42 spring constant c84 spring constant

CW calibrated wheel force

F9 determined wheel force h3 distance

h8 distance

k41 damping constant k83 damping constant m3 body mass m8 wheel mass

P position

W measured wheel force

X longitudinal direction

Y lateral direction

Z vertical direction

X'X longitudinal axis

Y'Y lateral axis

Z'Z vertical axis