Method and system for determining the traction utilization of a motor vehicle

Abstract

The present invention relates to a method for determining the traction utilization of a motor vehicle. The method comprises a step of acquiring (50) sound pressure data within a tire cavity (16) of a wheel (10) of the motor vehicle. Furthermore, the method comprises a step of determining (70) the current traction utilization as a function of the sound pressure data. The invention also relates to a system for determining the traction utilization of a motor vehicle.

Description

Technical area

[0001] The present invention relates to a method and system for determining traction utilization by a motor vehicle. State of the art

[0002] Motor vehicles have systems that intervene in the vehicle's control system. These are designed to prevent uncontrolled driving situations and / or transition them to controlled driving situations. Such situations can occur, for example, when one or more of the vehicle's tires change from a state of static friction to a state of sliding friction. An example of such a system is an ESP system. To control the system, a current coefficient of adhesion of the vehicle can be determined. This is determined based on sensor data, for example, the respective wheel speeds and / or a current acceleration vector of the vehicle. On this basis, a current coefficient of adhesion can then be estimated. However, such a determination is based on many assumptions and can therefore be inaccurate.For example, a sufficiently accurate estimate can only be made in a borderline area or when the vehicle has already entered a driving situation in which mainly sliding friction rather than static friction is present.

[0003] The accuracy of the above estimate can be improved by classifying a surface or an intermediate layer between the road and the tires. For example, it can be visually or acoustically detected whether the road is dry, wet, or covered with snow. However, such a classification only allows for a rough correction of the adhesion coefficient determination and is also very complex.

[0004] DE 10 2015 217 482 A1, DE 10 2015 217 474 A1, and DE 10 2015 217 472 A1 each describe a vehicle tire with a tire module, one of which is equipped with studs. Studded vehicle tires cause specific rolling noises and vibrations in the vehicle tire, which are recorded and evaluated by a sensor in the tire module. The sensor in the tire module can be, for example, a special microphone or special acceleration sensors. The evaluated measurement data can be used to draw conclusions about, for example, the condition of the studs in the tread or the friction values ​​with the road surface. However, these documents do not describe whether and how this can be used to determine current traction utilization.By determining the traction utilization, it is possible to determine how far a vehicle is from a critical situation in terms of driving dynamics, in which loss of control is imminent. Furthermore, the sensors must be disposed of or recycled along with the tires when a tire is changed, which is not very sustainable.

[0005] EP 1 337 404 A1 describes a method and device for determining the wear status of a tire. For this purpose, a microphone records the sound emission from the tire as it rolls.

[0006] DE 10 001 272 A1 describes a device and method for determining forces in the tire of a vehicle wheel. For this purpose, a rim deformation is recorded.

[0007] DE 19 807 004 A1 describes a sensor system and a method for monitoring the adhesion of a vehicle tire to the road surface and other physical tire data. This document also does not describe how to determine the current adhesion utilization.

[0008] DE 101 58 906 C1 describes a method for determining the maximum transmittable force on a tire and road surface during braking or acceleration of a vehicle. During braking or acceleration, the sound level in the vicinity of the tire is measured, and based on a previously determined mu-slip curve, the amplitude of the measured sound level is used as a control variable for an ABS system or traction control system.

[0009] DE 10 2017 207 825 A1 describes a method and a device for detecting vehicle speed. First, a dynamic pressure sensor signal is detected in a tire of a vehicle. An analysis of this dynamic pressure sensor signal allows the tire speed to be determined.

[0010] DE 10 2019 132 245 A1 describes a method for acoustic tire type recognition.

[0011] DE 10 2019 209 787 A1 describes an acoustic determination of static friction and / or sliding friction between at least one tire of a road vehicle and a road surface. An acoustic sensor positioned close to the tire is used to obtain an audio signal comprising vibrations of profile elements of a tire contact patch during travel. Description of the invention

[0012] A first aspect of the invention relates to a method for determining a traction utilization by a motor vehicle, for example during a journey. A motor vehicle can be designed, for example, as a passenger car, truck, or motorcycle. A motor vehicle has, for example, at least one wheel. A motor vehicle has, for example, a drive engine. A traction utilization can be a ratio of the current friction to a maximum possible friction with a road surface. The traction utilization can be a one-dimensional value or have multiple key figures. For example, the traction utilization can have a value in the longitudinal direction, transverse direction, and / or direction of rotation about a vertical axis of the motor vehicle and / or the wheel. A maximum traction utilization can be defined by a maximum traction potential.The maximum utilization of adhesion can be a maximum friction force between the road surface and the motor vehicle, individual wheels, and / or all wheels. The maximum utilization of adhesion can also have only one friction characteristic or multiple friction characteristics, for example, for the longitudinal and transverse directions of the motor vehicle and / or the wheel. The maximum utilization of adhesion can, for example, be achieved at a certain percentage of slip. The maximum utilization of adhesion can, for example, be achieved in a driving condition in which a specific wheel and / or one of the wheels of the motor vehicle is about to change from a state of static friction to a state of sliding friction.

[0013] The method comprises a step of acquiring alternating sound pressure data within a tire cavity of a wheel of the motor vehicle. The acquiring can be done by sensor. Alternating sound pressure data can comprise one or more acquired alternating sound pressures. For example, alternating sound pressure data can result from a pressure profile in the tire cavity and / or deviations from a static pressure in the tire cavity. The acquiring can be done by means of a sensing device that has, for example, a sound pressure sensor. The sound pressure sensor can be designed as a microphone, in particular as a MEMS microphone. The alternating sound pressure can correspond to pressure fluctuations in the air in the wheel, which are typically many orders of magnitude smaller than a static pressure in the wheel.

[0014] A wheel, for example, may comprise a rim and a tire. The tire may be attached to the rim. The essentially airtight cavity between the rim and tire may be the tire cavity. The sound pressure sensor, for example, may be attached to the rim. This means that any imbalance caused by the sound pressure sensor can be minimal and easily compensated for. The rim may be made of steel, aluminum, magnesium, and / or carbon fiber, for example. The tire may have rubber as a main component, for example. The wheel enables the motor vehicle to roll along the road.

[0015] By recording the alternating sound pressure data in the tire cavity, many disturbances can be avoided. For example, wind or noise sources in the vehicle's surroundings, such as other vehicles, have little or no influence on the alternating sound pressure data. Furthermore, the sensors can be protected from environmental influences and do not require additional installation space. This results in a high signal-to-noise ratio. The alternating sound pressure data can contain information about both airborne sound and structural sound. For example, oscillations, such as vibrations, in the tire and / or rim can also cause alternating sound pressure. This allows additional information regarding the utilization of traction to be taken into account.

[0016] The method comprises a step of determining a current adhesion utilization as a function of the alternating sound pressure data. The adhesion utilization can be determined one-dimensionally or multi-dimensionally, for example, separately for the longitudinal and transverse directions of the motor vehicle and / or the wheel. The adhesion utilization can also be determined as a vector, which extends, for example, in a plane spanned by the vehicle's longitudinal and / or transverse directions. For a horizontal road surface, this vector can also be horizontal. There is usually a force between the tire and the road surface that results from shear stress at the tire contact patch. This usually involves a combined slip condition. For processing, this force can be divided into longitudinal and transverse vectors.Accordingly, both adhesion and adhesion utilization can be present, for example, as a vector or split into two values ​​for the transverse direction and the longitudinal direction. By determining the adhesion utilization, it is possible to know which driving dynamics reserves are still available. This can enable preventative driving control intervention. This can also reliably prevent an autonomous driving system from steering the vehicle into adhesion limit zones. The current adhesion utilization can take into account the current possible friction with the road surface. No intermediate layer needs to be classified for this. Rather, the recorded alternating sound pressure data can directly allow conclusions to be drawn about the possible friction between the tires and the road surface, since the sound pressure change in the tire cavity varies depending on the road surface and existing intermediate layers in otherwise identical driving situations.The sound pressure change data can also allow tire wear and / or wheel imbalance to be taken into account, since the sound pressure change in the tire cavity can vary depending on the tire wear and / or wheel imbalance in otherwise identical driving situations and on otherwise identical road surfaces.

[0017] For example, the current traction utilization can be determined using an evaluation device. The evaluation device can, for example, be an on-board computer connected to the detection device. The traction utilization can be determined continuously, for example at a high frequency. The evaluation device can also be an integral part of the sound pressure sensor and / or arranged in the tire cavity. For example, the evaluation device can also be attached to the rim. For example, the traction utilization can also be determined in normal driving situations, such as straight-ahead driving, and be sufficiently accurate. For example, the traction utilization can be determined at half the frequency of a measuring frequency of the respective sound pressure sensors.For example, the sound pressure can be determined with any detection frequency in a range of at least 0.5 kHz, 1 kHz, 5 kHz, 10 kHz, 20 kHz or more up to 40 kHz, 50 kHz, 60 kHz, 80 kHz, 100 kHz or more. Correspondingly, alternating sound pressures can be detected with a frequency within the detection range. In contrast to a normal tire pressure sensor for detecting pressure losses due to a leak, the detection frequency can therefore be high. In said normal tire pressure sensor, for example, a pressure in the tire cavity is compared with a target pressure only every 30 seconds. The sensitivity of the sound pressure sensor can be high because, for example, only small pressure fluctuations need to be detected. The sound pressure sensor can, for example, be calibrated to a target pressure of the wheel.

[0018] For example, a current adhesion for the wheel can be determined depending on the alternating sound pressure data. The adhesion can be determined one-dimensionally or multi-dimensionally, for example, separately for the longitudinal direction and the transverse direction of the motor vehicle and / or the wheel. The current adhesion can correspond to the respective circumferential forces transmitted to the wheel. A current adhesion coefficient can also be determined, which results from the normal force acting on the wheel. In addition, a current adhesion potential for the wheel can be determined depending on the alternating sound pressure data. The adhesion potential can be determined one-dimensionally or multi-dimensionally, for example, separately for the longitudinal direction and the transverse direction of the motor vehicle and / or the wheel.The current adhesion potential can, for example, be the currently maximum possible adhesion, which can result from the friction between the tire and the road surface as well as the normal force on the wheel. The current adhesion potential can be influenced by the condition of the tire, the road surface, an intermediate layer between the tire and the road surface, and the current driving dynamics. The current adhesion utilization can be determined depending on the determined current adhesion and the determined current adhesion potential, for example, by division.

[0019] Overall, the quality of determining adhesion utilization can be high because only a small number of chained estimators are used, as adhesions and adhesion potentials can be deduced directly from the alternating sound pressure data. The use of alternating sound pressure sensors, which enable measurement of the alternating sound pressure, enables the estimation of performance parameters that are physically directly traceable to the adhesion coefficient. This avoids the chaining of different estimation algorithms. Furthermore, an air column in the tire cavity is excited to different high-frequency vibrations depending on the road surface, which can be detected by the alternating sound pressure sensors. The signals thus contain information about the road surface, which can be incorporated into the estimation of the adhesion potential.It is also possible to estimate the adhesion potential during non-critical driving maneuvers, such as constant straight-line driving, because the quality of the input data does not depend on the utilization of the adhesion potential. In addition, the effects on the alternating acoustic pressure resulting from bumps or tire rolling are minimal compared to the acceleration of the wheel or tire, so the measuring range of the alternating acoustic pressure sensors can be significantly smaller than that of acceleration sensors. This allows the sensors to be designed with significantly greater sensitivity and better resolve even smaller events or influences.

[0020] The method may include a step of generating a control signal. The control signal may control a vehicle function, such as a driver assistance system or an autonomous vehicle control system. For example, the driver assistance system may change its intervention behavior depending on the control signal or the current traction utilization. The method may be designed as a method for operating the driver assistance system and / or the autonomous vehicle control system. Likewise, the determined current traction utilization may be output, for example, to the driver via a display.

[0021] In one embodiment of the method, at least a first alternating sound pressure is recorded within a tire cavity of a first wheel of the motor vehicle and at least a second alternating sound pressure is recorded within a tire cavity of a second wheel of the motor vehicle. The alternating sound pressure data can therefore, for example, comprise recorded pressures from two different wheels. The alternating sound pressure data can comprise the respective recorded alternating sound pressures or are at least generated depending thereon. As a result, information from several or even all wheels of the motor vehicle can be taken into account when determining the traction utilization. For example, an associated alternating sound pressure is recorded for each wheel of the motor vehicle.

[0022] The current adhesion utilization can be determined for each individual wheel. For example, a first current adhesion utilization is determined for the first wheel and a second current adhesion utilization for the second wheel. This makes it possible to determine separately for each of these wheels how far this wheel is from a loss of adhesion. This allows better control to avoid the loss of adhesion. For example, individual wheels can be steered and / or braked or accelerated in a targeted manner in order to avoid the loss of adhesion. A driving maneuver can also be adapted, for example to transfer less power to certain wheels that are already close to a grip limit. For example, a wheel-specific current adhesion and / or a wheel-specific current adhesion potential can be determined for this purpose, for example depending on the associated recorded alternating sound pressure.For example, a first current adhesion and / or a first current adhesion potential can be determined for the first wheel, each as a function of the first alternating sound pressure. For example, a second current adhesion and / or a second current adhesion potential can be determined for the second wheel, each as a function of the second alternating sound pressure.

[0023] Alternatively or additionally, the current adhesion utilization can be determined overall for the vehicle, for example as a function of the first alternating sound pressure and the second alternating sound pressure. This can be done, for example, by merging the sensor data. However, the current adhesion and / or the current adhesion potential can also be determined first for each individual wheel and then the current adhesion utilization can be determined for the entire vehicle based on these values. However, an overall current adhesion and / or an overall current adhesion potential can also be determined directly, for example as a function of all recorded alternating sound pressures. Determining the overall adhesion utilization for the vehicle can require little computing power and be very accurate.

[0024] The following explanations are primarily based on a single wheel and a single recorded acoustic pressure. However, the respective explanations also apply to embodiments in which the acoustic pressures of two or more tires are recorded and / or in which the total current traction utilization and / or the current traction utilization for each individual wheel are determined, if applicable.

[0025] In one embodiment of the method, the method comprises a step of determining a current adhesion as a function of the alternating sound pressure data. The determination can be made for each individual wheel or for the entire motor vehicle. The current adhesion can be determined analytically as a function of the alternating sound pressure data. For example, an empirically determined relationship between the alternating sound pressures in wheels and the current adhesion can form the basis for a calculation. Since no assumptions need to be made regarding a driving condition, the determination can be simple and very precise. For example, such a correlation can be 97% accurate. In the analytical determination, additional data can be taken into account in addition to the alternating sound pressure data, such as the temperature of the air in the tire cavity and / or the static air pressure in the tire cavity.The current traction utilization can then be determined depending on the specific current traction. Alternatively, the current traction can be determined, for example, using a neural network. In this case, no empirical determination is necessary. For example, the neural network can be trained using simulated or experimental data.

[0026] A neural network can, for example, be a mathematical model. The neural network can be created using a computer and / or designed as an artificial neural network. The neural network can have input nodes, output nodes, and a plurality of intermediate nodes arranged between the input nodes and the output nodes. Respective connections between the nodes can, for example, have a weighting. The input nodes can, for example, be designed as data interfaces via which input data can be fed into the neural network. The output nodes can, for example, be designed as data interfaces via which output data can be output from the neural network. The input nodes can be connected to the intermediate nodes, and the intermediate nodes can be connected to one another. The intermediate nodes can be connected to the output nodes.The neural network can, for example, be designed as a convolutional neural network.

[0027] In one embodiment of the method, the method includes a step of determining a current adhesion potential as a function of the alternating sound pressure data. The determination can be made for each individual wheel or for the entire motor vehicle. The current adhesion potential can be determined using a neural network as a function of input data. The input data for the neural network can include at least the alternating sound pressure data. With sufficient training, the neural network can easily consider intermediate layers and / or different road surfaces without having to be separately identified and / or classified, since the alternating sound pressure data contain information on this. Furthermore, a complex analytical determination of such relationships can be dispensed with.For example, a traction potential can be determined experimentally on a test bench, thus providing corresponding training data for the neural network. It has been shown that a neural network based on alternating sound pressure data can determine the current traction potential with high accuracy, for example, without using additional data on the road surface and / or possible intermediate layers. During the determination, the input data can also include additional data in addition to the alternating sound pressure data, such as the air temperature in the tire cavity and / or the static air pressure in the tire cavity. Alternatively, the current traction potential can also be determined analytically.

[0028] The neural network can also be configured to determine both the current adhesion potential and the current adhesion and output them as output data. In this case, separate determination is unnecessary. The neural network can also be configured to directly determine the adhesion potential without first determining the current adhesion and the current adhesion potential. Determining both the current adhesion potential and the current adhesion potential first can enable a particularly precise determination of the adhesion utilization. It has been shown that the analytical determination of the current adhesion and the AI-based determination of the adhesion potential enable a particularly precise determination of the adhesion utilization and are also very easy to implement. The AI ​​can be implemented here by the neural network.As output data, the neural network can, for example, output the current adhesion potential and / or current adhesion utilization, each for the entire vehicle and / or for each wheel individually, for which the alternating acoustic pressure in the tire cavity is recorded.

[0029] In one embodiment of the method, the method includes a step of filtering the alternating sound pressure data. For example, respective alternating sound pressures can be filtered according to their frequency and / or amplitude. Filtering can occur before determining the current adhesion utilization, before determining the current adhesion, and / or before determining the current adhesion potential. Filtering can occur before data is transmitted to an evaluation device and / or before a data reduction step. Filtering can filter out interference signals and / or extract useful signals. Filtering can be performed, for example, using a bandpass filter, high-pass filter, or low-pass filter. Filtering can increase the accuracy of respective determinations and / or reduce the amount of data.The filter's corner frequencies can be adapted to the expected frequency bandwidth of the desired signal. Filtering can be performed, for example, by a data preprocessing device. The data preprocessing device can be arranged on the wheel, for example, together with the associated sound pressure sensor on the rim. The data preprocessing device can comprise a microprocessor. The data preprocessing device can also be an integral part of the sound pressure sensor.

[0030] In one embodiment of the method, the method comprises a step of transmitting the alternating sound pressure data to an evaluation device. The data transmission can be wired or wireless, for example. The data transmission can be wireless, for example, via the IEEE 802.11 standard, Bluetooth, or wired via a slip ring in a wheel hub. The alternating sound pressure data can be filtered, reduced in quantity, and / or transformed before transmission. The data transmission can enable a computationally intensive evaluation to determine the current traction utilization to be carried out centrally in the motor vehicle, for example, also for multiple wheels and / or taking into account the alternating sound pressure data from multiple wheels.

[0031] The method comprises a step of data reduction of the alternating sound pressure data, which, for example, occurs before the alternating sound pressure data is transmitted to the evaluation device. The data reduction can reduce the amount of data. The data reduction can compress and / or transform the alternating sound pressure data. This can make data transmission fast. Furthermore, evaluation can be simplified, particularly through a neural network. The data reduction can comprise feature extraction. This extracts only data relevant for the determination, such as patterns, from the acquired data. For example, data reduction from the field of speech recognition can be used, such as the determination of Mel Frequency Cepstral Coefficients (MFCC). Only these coefficients are then transmitted as alternating sound pressure data.For example, data reduction can be achieved through transformations such as wavelet or fast Fourier transformations, image-level transformations, and / or time-frequency transformations. Subsequently, data conversion can be performed, for example, into a format that can be processed by the evaluation device.

[0032] In one possible variant of the method, Mel Frequency Cepstral Coefficients (MFCC) are determined based on the recorded alternating sound pressures in order to generate the data-reduced alternating sound pressure data. This method is established in the field of speech recognition and has been shown to be very suitable for the acoustic determination of vehicle tire adhesions. In a first step, the useful signal data is divided into time periods, each of which is multiplied by a window function. An overlap is created between the time windows. The windowed data is then transformed into the frequency domain using a discrete Fourier transform (DFT), then logarithmized and multiplied by triangular filters whose center frequencies are selected to map the frequency spectrum to the Mel scale.The resulting amplitudes are integrated for each triangular filter, resulting in a scalar value for each mel filter. Since the number of triangular filters is significantly lower than the frequency sampling points resulting from the DFT, the data volume is significantly reduced. For example, with a time window width of 512 sampling points, 257 frequency sampling points are generated, which can be reduced to 13 single values ​​with a typical selection of 13 MFCCs. In the next step, the data is converted for transmission to the evaluation device. In addition to the actual data vector with the MFCCs for each time step, the data stream can contain a timestamp so that the data can be chronologically classified for further processing.

[0033] In one embodiment of the method, it is provided that the method additionally comprises a step of detecting a vehicle state. The current adhesion utilization can additionally be determined depending on the vehicle state. The vehicle state can, for example, be used to determine the current adhesion and / or the current adhesion potential. The vehicle state can, for example, be detected by sensors or through an input. The vehicle state can, for example, include an outside temperature, a vehicle location, or information relating to vehicle movement, such as acceleration, driving speed, direction of travel, and / or a second derivative of the vehicle speed. This vehicle movement information can be available for each individual wheel or for the entire vehicle. For example, appropriate sensors can be provided for this purpose, such as a 9D inertial measuring unit.The vehicle state may also include a wheel state. The vehicle state may, for example, also include information about the type of tires currently installed and the usage of these tires. The wheel state may, for example, be detected by sensors or an input. For example, the wheel state may include an inflation pressure, an air temperature in the wheel, and / or a wheel speed. An associated wheel state may be recorded for each wheel state.

[0034] A second aspect relates to a system for determining traction utilization by a motor vehicle. The system can be configured to carry out the method according to the first aspect. Respective advantages and further features can be found in the description of the first aspect, with embodiments of the first aspect also forming embodiments of the second aspect, and vice versa.

[0035] The system comprises a detection device with at least one sound pressure sensor. The detection device can be configured to generate alternating sound pressure data for each wheel in whose tire cavity a sound pressure sensor is arranged.

[0036] This generation can occur depending on the detected alternating sound pressures. The sound pressure sensor can be attached to a rim of a wheel of the motor vehicle within a tire cavity. Alternatively, the sound pressure sensor can also be attached to the tire. The detection device can have an associated sound pressure sensor for each of two or more wheels, which is arranged in the tire cavity of the associated wheel. The sound pressure sensor can be designed to detect alternating sound pressure data within the tire cavity. For example, the sound pressure sensor can be designed as a microphone. The frequency range and / or a pressure range in which the sound pressure sensor can detect data can be adapted to a useful signal range. The useful signal range can correlate particularly well with the adhesion of the tire to the road surface and / or the adhesion potential of the friction pairing of tire and road surface.

[0037] The system can, for example, have a power source that is also attached to the rim and supplies the sound pressure sensor with power. The power supply can also come from outside the tire. For example, the power supply can be provided by a sliding contact or induction.

[0038] The energy source can be designed as an energy generating device. The energy generating device can be designed to generate energy for at least supplying the sound pressure sensor with energy. The energy generating device can also be designed to supply energy to the data preprocessing device, the entire detection device, and / or the evaluation device. The energy generating device can be designed to obtain energy from a wheel movement. For example, the energy generating device can obtain energy from deformations and / or movements of the wheel or parts thereof. For example, the energy generating device can have a piezo element that is fastened in or on the wheel, for example in or on the tire or rim, and generates electrical current to supply energy to parts of the system through deformation and / or vibrations acting on the piezo element.For example, permanent magnets can be arranged on a stationary vehicle component near the wheel, inducing currents in coils located within the rotating wheel. For example, the energy generation device can be configured to supply at least the sound pressure sensor and / or other parts of the system with energy independently of the on-board power supply of the rest of the motor vehicle.

[0039] Compared to vehicle-mounted alternating sound pressure sensors, alternating sound pressure sensors inside the tire cavities offer insulation from noise outside the tire. This significantly reduces the level of background noise. Furthermore, the alternating sound pressure sensors inside the tire cavity are not exposed to moving air, meaning there is no wind noise. These conditions lead to a very high signal-to-noise ratio. The resulting high level of useful signal simplifies data preprocessing, which separates useful signals from noise.

[0040] The system may include an evaluation device configured to determine a current traction utilization based on the alternating sound pressure data. For example, the evaluation device may be configured to analytically calculate a current traction for each wheel from which alternating sound pressure data is acquired. Alternatively or additionally, the evaluation device may be configured to determine a current traction potential for each wheel from which alternating sound pressure data is acquired using a neural network.

[0041] In one embodiment of the system, the system comprises at least one data preprocessing device. The system may comprise an associated data preprocessing device for each sound pressure sensor. The data preprocessing device may be arranged together with the sound pressure sensor in the tire cavity and attached to the rim. For example, each sound pressure sensor and each data preprocessing device may be formed as a common component. The data preprocessing device may be supplied with power together with the associated sound pressure sensor.

[0042] The data preprocessing device can be configured to filter the alternating sound pressure data. Alternatively or additionally, the data preprocessing device can be configured to perform data reduction on the alternating sound pressure data, for example, by feature extraction. The feature extraction can be feature extraction. This data reduction can occur before data is transmitted to the evaluation device. The data preprocessing device can be configured for this transmission. The data transmission can be wireless or wired, for example. The data preprocessing device can comprise a microprocessor. Short description of the characters Fig. 1 schematically illustrates a system for determining traction utilization by a motor vehicle. Fig. 2 schematically illustrates a method for determining traction utilization by a motor vehicle. Detailed description of embodiments

[0043] Fig. 1 schematically illustrates a system for determining traction utilization by a motor vehicle. A wheel 10 of the motor vehicle is shown, partially in a sectional view. The wheel 10 has a tire 12 and a rim 14. The tire 12 is attached to the rim 14 so that together they define an air-filled tire cavity 16. A sound pressure sensor 18 and an associated data preprocessing device 20 are arranged in the tire cavity. The sound pressure sensor 18 is designed as a micro-electro-mechanical system (MEMS). The data preprocessing device 20 and the sound pressure sensor 18 are attached to the rim 14 so that the tire 12 can be replaced without replacing these components. A power supply device 22 is also shown, which is designed to supply power to the sound pressure sensor 18 and the data preprocessing device 20.In the example shown, the energy supply device 22 comprises a battery. In other embodiments, the energy supply device 20 is designed for connection to an on-board power system, for example via sliding contacts. The energy supply device 20 can also be designed to generate energy upon rotation of the wheel 10, for example, through a coil into which magnets fixed to the vehicle induce a voltage. In the example shown, the energy supply device 22 is arranged outside the tire cavity 16 on the rim 14. The energy supply device 22 can also be arranged in a partially or fully protected manner within the tire cavity 16.

[0044] The data preprocessing device 20 is designed to transmit data wirelessly or by cable, for example via sliding contacts, to an evaluation device 24 of the system. The evaluation device 24 is designed as an on-board computer of the motor vehicle and is, for example, permanently attached to a vehicle body. The system can have additional pairs of sound pressure sensors 18 and data preprocessing devices 20, each arranged on an associated wheel 10 of the motor vehicle. Each of these pairs then has an associated energy supply device 22. The data acquired and preprocessed there is also transmitted to the evaluation device 24 and evaluated there together.

[0045] Fig. Figure 2 illustrates a method for determining the traction utilization by a motor vehicle, which is determined by the system according to Fig.1 is carried out. In a first step 50, alternating sound pressure data is recorded within each tire cavity of each wheel 10 of the motor vehicle, which is equipped with a sound pressure sensor 18. This raw data is recorded, for example, at a frequency of at least 1 kHz and has, for example, pressure fluctuations of up to 10 Pa. For example, a frequency relevant for determining the adhesion utilization lies between 0.5 kHz and 10 kHz or 20 kHz. The sound pressure sensor 18 outputs a voltage as a measurement signal, which is first converted into alternating sound pressures by the data preprocessing device 20 using a stored sensitivity curve and a calibration curve of the sound pressure sensor 18. In addition, an initial useful signal extraction already takes place by filtering using a bandpass filter, which only allows a specific frequency range and / or a specific amplitude range to pass.This filtering and generation of the alternating sound pressures takes place in a step 52. In a further step 54, the already filtered alternating sound pressure data is reduced by the data pre-processing device 20. Feature extraction takes place using the Mel Frequency Cepstral Coefficients (MFCC). This reduces the amount of data to be transmitted to the evaluation device 24 and the computing power required there for evaluation. In a step 56, the amount of alternating sound pressure data reduced by the data reduction is converted by the data pre-processing device 20 for transmission to the evaluation device 24. During this conversion, a timestamp is assigned to the actual data vector with the MFCC for each time period. This data is then transmitted to the evaluation device 24, in this case, for example, via Bluetooth or WiFi.Individual ones of these steps 52 to 56 can be omitted in other embodiments and / or can only be carried out after the transmission of corresponding data to the evaluation device 24 by the evaluation device 24.

[0046] The evaluation of the received data by the evaluation device 24 then takes place in two paths. In a first path, the received data is made available in a step 58 to a neural network implemented in the evaluation device 24 in order to determine a current adhesion potential as a function of the alternating sound pressure data. Optionally, the input data can also include recorded data on a vehicle condition. The neural network was previously trained with training data, which was previously determined experimentally on a test bench, for example, for various pairings of tires and road surfaces with any intermediate layers present, for a relationship between the preprocessed input data and a current adhesion potential. In one embodiment, the neural network determines the adhesion potential individually for each wheel 10.Alternatively or additionally, the traction potential for the entire vehicle is determined.

[0047] As output data, the neural network can, for example, output a classification vector with various adhesion coefficient classes and an associated probability. Alternatively, it can output only one adhesion coefficient, which could, for example, be an estimated adhesion potential. It can also output one adhesion coefficient for a vehicle's longitudinal direction and one adhesion coefficient for a vehicle's transverse direction.

[0048] Optionally, the system is configured to estimate the adhesion potential based on additional sensor values ​​and / or other methods. For example, the adhesion potential can also be estimated based on wheel speeds and a detected vehicle movement. For example, a determined wheel slip can be used as a basis for determining the adhesion potential. The wheel slip can, for example, enable the adhesion potential to be determined relatively reliably, at least in a limit range of adhesion utilization. In an optional step 60, the evaluation device 24 merges the various determined adhesion potentials, for example, by averaging the classification vectors.

[0049] In the second path, the received data from the evaluation device 24 is used to determine the current adhesion as a function of the alternating sound pressure data through an analytical calculation. For this purpose, in a step 62, the evaluation device 24 first calculates a power calculation for all wheels jointly or for each individual wheel. In the process, a rolling power and a thermal power in the tire contact area are determined. The conversion of the Mel coefficients into the power variables is based on empirically determined relationships that show a high correlation between the variables. In a further step 64, the evaluation device 24 calculates a current adhesion coefficient based on the calculated power variables. Multiple adhesions and / or adhesion coefficients can also be determined per wheel 10, for example, one for the vehicle's longitudinal direction and one for the vehicle's transverse direction.In the case of sound pressure sensors 18 in several wheels 10, the adhesion coefficients can be further processed individually for each wheel or fused in an optional step 66, for example by averaging.

[0050] In a step 70, the evaluation device 24 determines the current adhesion utilization based on the current adhesion potential determined by the neural network and the analytically determined current adhesion, and thus based on the acoustic pressure data. The current adhesion utilization can be determined individually for each wheel 10 and / or for the entire vehicle. This allows the reserve of longitudinal and lateral forces to be determined, enabling a more precise estimation of the driving condition and the derivation of possible interventions by vehicle dynamics control systems. Reference symbol 10 wheels 12 tires 14 rim 16 tire cavity 18 Sound pressure sensor 20 Data preprocessing device 22 Power supply device 24 Evaluation device 50 Step / Acquisition of alternating sound pressure data 52 Step / Filtering the sound pressure data 54 Step / Data reduction of the sound pressure data 56 Step / Conversion of sound pressure data 58 Step / Determine a current adhesion potential 60 Step / Fusion of the determined adhesion potential with another determined adhesion potential 62 Step / Power calculation of the wheels 64 Step / Determine a current adhesion 66 Step / Fusion of the specific power connections 70 Step / Determine a current traction utilization

Claims

[1] Method for determining the use of traction by a motor vehicle, the method comprising at least the following steps: - detecting (50) alternating sound pressure data within a tire cavity (16) of a wheel (10) of the motor vehicle; - Determining (70) a current traction utilization as a function of the alternating sound pressure data. [2] Method according to claim 1, wherein at least one first alternating sound pressure is detected within a tire cavity (16) of a first wheel (10) of the motor vehicle and at least one second alternating sound pressure is detected within a tire cavity (16) of a second wheel (10) of the motor vehicle, where the current traction utilization is determined individually for each wheel or whereby the current traction utilization is determined overall for the motor vehicle. [3] Method according to claim 1 or 2, wherein the method comprises a step (62, 64) of determining a current frictional connection as a function of the alternating sound pressure data, where the current adhesion is determined analytically depending on the sound pressure data and where the current traction utilization is determined depending on the determined current traction. [4] Method according to one of the preceding claims, wherein the method comprises a step (58) of determining a current adhesion potential as a function of the alternating sound pressure data, where the current adhesion potential is determined by means of a neural network depending on input data, wherein the input data for the neural network comprises at least the alternating sound pressure data. [5] Method according to one of the preceding claims, wherein the method comprises a step (52) of filtering the alternating sound pressure data. [6] Method according to one of the preceding claims, wherein the method comprises a step of transmitting the alternating sound pressure data to an evaluation device (24), wherein the method comprises a step (54) of data reduction of the alternating sound pressure data before transmission of the alternating sound pressure data to the evaluation device (24). [7] Method according to one of the preceding claims, wherein the method additionally comprises a step of detecting a vehicle state, wherein the determination (70) of the current traction utilization is additionally carried out as a function of the vehicle state. [8] Method according to claim 7, the vehicle state may comprise at least one of the following information: - Information about a vehicle movement; - Information about an outside temperature; - Information about a vehicle location; and - Information about a wheel condition, such as a currently mounted tire, a usage of the tire, an inflation pressure in the wheel and / or a wheel speed. [9] System for determining the use of traction by a motor vehicle, wherein the system comprises a detection device with at least one sound pressure sensor (18), wherein the sound pressure sensor is mounted within a tire cavity (16) on a rim (14) of a wheel (10) of the motor vehicle and wherein the sound pressure sensor (18) is designed to detect alternating sound pressure data within the tire cavity (16), wherein the system comprises an evaluation device (24) which is designed to determine a current traction utilization as a function of the alternating sound pressure data. [10] System according to claim 9, wherein the system comprises at least one data preprocessing device (20) which is designed to the sound pressure data for filters and / or to carry out a data reduction of the alternating sound pressure data before transmitting them to the evaluation device (24). [11] System according to claim 9 or 10, wherein the system comprises at least one energy supply device (20) designed as an energy generating device, which is designed to supply at least the sound pressure sensor with energy.

Images