DE102019212487A1 - Method and computer system for determining a road friction coefficient - Google Patents

Abstract

Disclosed is a method for determining a road surface coefficient of friction (16) of a road surface (8), the method comprising the steps: selecting (15) at least one algorithm (10) of a plurality of algorithms (10) for determining a coefficient of friction (12) of the road surface ( 8) based on input data (9) which are representative of at least one measured friction value at a measurement position of the roadway (8) and at a measurement time, the algorithms (10) for determining the coefficient of friction (12) with regard to a requirement for an input quality ( 13) differentiate between the input data (9) and / or with regard to a result quality (14) of the coefficient of friction (12) determined using the respective algorithm (10); wherein the at least one algorithm (10) is selected based on the input quality (13) of the input data (9) and / or based on the result quality (14) of the algorithms (10); Determining (11) a coefficient of friction (12) for the roadway (8) using the at least one selected algorithm (10) of the plurality of algorithms (10); and providing (17) the coefficient of friction (12) determined with the at least one selected algorithm (10) as a coefficient of friction for the roadway (16).

Description

The present invention relates to a method for determining a roadway coefficient of friction of a roadway. The invention also relates to a computer system, a computer program and a computer-readable medium.

State of the art

Based on measured values from sensors present in a vehicle, a coefficient of friction of a roadway or another roadway property can be estimated.

For example, the DE 10 2008 047 750 A1 a method for estimating the coefficient of friction, where up to four estimated values of a coefficient of friction are estimated using different algorithms from sensor signals from sensors present in a vehicle and then merged. It is disclosed that a first coefficient of friction is estimated based on torsional vibrations of a wheel of the vehicle, that a second coefficient of friction is estimated based on a slip of the wheel, that a third coefficient of friction is estimated based on driving dynamics of the vehicle, that based on a statistical evaluation of others Sensor signals and the first three estimates, a fourth coefficient of friction is estimated, and that the four estimated coefficients of friction are merged together in a coefficient of friction classification. A driver assistance system can be adapted using the friction coefficient classification.

Disclosure of the invention

With embodiments of the invention, an improved method and an improved computer system for determining a roadway coefficient of friction can be provided in an advantageous manner.

The invention is defined in the independent claims. Advantageous developments of the invention emerge from the dependent claims and the following description.

One aspect of the invention relates to a method for determining a roadway coefficient of friction of a roadway, the method at least the steps of selecting at least one algorithm of a plurality of algorithms for determining a coefficient of friction of the roadway, determining a coefficient of friction for the roadway using the at least one selected algorithm and providing of the determined coefficient of friction as the road surface coefficient of friction. The road coefficient of friction is therefore determined in several steps.

The roadway can designate a surface that can be driven and / or driven on by a vehicle, such as a land vehicle, an amphibious vehicle and / or an aircraft, examples of the roadway being a road, a path, a runway, a terrain and / or a railroad track include. The roadway can designate a permanent, such as stationary, subsurface and / or surface, the friction properties of which vary due to, for example, an application, a material property and / or an environmental condition; In this regard, an asphalt pavement of a road can be assumed as the roadway, the coefficient of friction of which can vary in terms of amount due to a snow application, aging behavior and / or partial melting of the asphalt as a result of high solar radiation.

The coefficient of friction can also be referred to as a force-fit coefficient or coefficient. In particular, it characterizes a ratio of a given contact force of a tire or wheel to an achievable longitudinal and / or lateral force in tire-road contact. It can be determined and made available as a single value in order to facilitate simple processing, for example. It can also be determined and provided for the longitudinal force and the lateral force together or separately. It can also be determined and provided as a function of a slip between the tire and the road surface in order to promote precise processing. In addition, a coefficient of friction distribution and / or another statistical statement about the coefficient of friction or the coefficient of friction function can be determined and provided in order to increase precision. It can thus be provided to provide a determined minimum, average and / or maximum road friction coefficient of the roadway. In this disclosure, the “coefficient of friction” preferably denotes a coefficient of friction determined using an algorithm, “measured coefficient of friction” preferably denotes a real coefficient of friction at a determinable measurement position of the roadway, and “coefficient of friction” preferably denotes a coefficient of friction determined with a selected algorithm.

At least one algorithm from a plurality of algorithms for determining a coefficient of friction of the road surface is selected based on input data.

An algorithm can be an abstract instruction or regulation or, in particular, an instruction or regulation implemented for use in a specific environment, such as a specific computer system and / or operating system, for deterministic, partially-terministic, non-deterministic or apparently non-deterministic for a person Determining a coefficient of friction of a roadway can be understood. Examples of an algorithm include a sub-method for determining, estimating and / or calculating a determined, estimated and / or calculated Coefficient of friction of a roadway and / or an AI module for it. Each algorithm preferably determines the coefficient of friction of a roadway using an estimate, so that the coefficient of friction determined as the result is an estimated result. The estimation result can additionally contain an assessment of the estimation, such as a number of data used, a tolerance of the determined coefficient of friction and / or the like. Estimating can in particular mean ascertaining with a probability, wherein the probability can mean, for example, a measure of a ratio or a tolerance of a specifically estimated coefficient of friction determined from specific input data with respect to a generally unknown measured coefficient of friction or real coefficient of friction. Thus, an estimated coefficient of friction can be a coefficient of friction which is above, around and / or below the real coefficient of friction with a predeterminable / determined and / or specifiable probability within a predefined / determined and / or specifiable tolerance.

The input data represent the measured friction value at a measurement position on the roadway and at a measurement time. The input data can, for example, be measurement points obtained in at least one measurement and / or values derived therefrom and / or aggregated. In addition to a measured value or measurement information as an indication of a measured variable, a measurement point can contain at least one additional item of information such as the measurement position or a measurement location, the measurement time or a measurement time, measurement method information and / or measurement instrument information as an indication of location, time, method, instrument and / or configuration of the instrument of the measurement. A speed calculated from two measuring positions and measuring times is an example of a value derived from measuring points. A measurement series, such as a spatial and / or temporal sequence of measurement values, in particular with at least one associated measurement information item, is an example of values aggregated from measurement points. A measurement position can be a section of the road, such as a lane of a multi-lane road and / or a segment of a road network. A measurement time can be a measurement time or a measurement period.

The algorithms for determining the coefficient of friction differ with regard to a requirement for an input quality of the input data and / or with regard to the quality of the result of the coefficient of friction determined using the respective algorithm. An input quality of the input data can be a measure of usability and / or suitability for determining or estimating a coefficient of friction, in particular by means of a respective one of the plurality of algorithms. Likewise, the quality of the result of the determined coefficient of friction can be a measure of the usability and / or suitability for the reliable operation of a driver assistance application. Examples of criteria for determining or measuring the respective input quality and examples of criteria for determining or measuring a respective result quality are described later.

In the step of selecting, the at least one algorithm is selected based on the input quality of the input data and / or based on the result qualities of the algorithms. By using the claimed method to select an algorithm for determining a coefficient of friction based on the input quality of the input data of each algorithm and / or the respective result quality of each algorithm, the at least one algorithm is selected based on a suitability measure or usability measure or, conversely, is a determined one Coefficient of friction selected based on the suitability level. Thus, the claimed method enables the road surface's coefficient of friction to be determined as precisely as possible. The same advantage follows for a computer system according to the invention, which is set up or prepared for executing steps of the method according to the invention, for a computer program according to the invention which can guide execution of steps of the method according to the invention, and for a computer-readable medium according to the invention, on which a computer program according to the invention is stored is.

The coefficient of friction determined with the at least one selected algorithm is provided as the coefficient of friction for the roadway. The provision can be an output from the method directly to at least one receiver or indirectly to a downstream method, such as collecting road friction coefficients and transmitting a collection of road friction coefficients, for example as part of an update (in technical terms: an update). The provision can additionally or alternatively be a transmission to the at least one recipient with or without an acknowledgment of receipt by the recipient. The provision can additionally or alternatively be a transfer to a memory that can be called up by the receiver, such as a database, for example a road map-based database and / or a database set up to keep data available for a navigation application and / or a driver assistance application. The road surface coefficient of friction is thus uncomplicated to use for a recipient. The receiver is preferably a vehicle. More precisely, the receiver can be a control device, such as a driver assistance device, or a part thereof, such as a receiving part or a transmitting / receiving part, in the vehicle.

Based on the road surface friction coefficient and / or a signal representing the road surface friction coefficient, a control signal can be used for the control of a vehicle can be determined, for example by a control unit of the vehicle. For example, a speed of the vehicle can be changed and / or adapted based on the control signal. Alternatively or additionally, a direction of travel of the vehicle can be changed based on the control signal.

According to a further aspect, the result qualities of the estimation results of the algorithms can be divided or classified or graded into classes or levels based on the respective requirement of the input quality. If a classification is provided, more than one result quality is preferably provided for each class. If a grading is provided, only one quality of results is preferably provided per level. Classes and levels can be staggered in terms of amount, for example with increasing quality of results. By means of this aspect it can be avoided that an estimated result of only a relatively low result quality is determined from input data of a relatively high input quality, so that an overall efficient method is provided.

The input quality can be determined based on a criterion of a number or quantity of the input data so that, for example, the input quality increases, the more input data, in particular the more measured values per measured variable, are available. The number can be assessed depending on the measured variable. This enables a more meaningful and precise determination of the coefficient of friction.

Additionally or alternatively, the input quality can be determined based on a criterion of a density of the input data, in particular a measurement position-related or local and / or measurement time-related or temporal density of the input data and / or a density of measured values of the input data. For example, for a high input quality it can be required that, for example, when measuring temperature over a measuring range between, for example, -40 ° C and +60 ° C, a minimum density is not undershot, or that, for example, according to a standard distribution, a min. Area containing 65% of all measured values, i.e. an area of high density, can be identified. The same applies to a measurement position or a measurement location or a measurement position density and a measurement time or measurement time density. This increases the reliability of the method.

Additionally or alternatively, the input quality can be determined based on a criterion of a type of input data. The type of input data can, for example, be a described variable, such as the presence of a road surface temperature or the presence of an ambient temperature, an origin of a variable such as a forecast or a measurement, a degree of processing of a variable, such as raw data or processed raw data or derived measured values, and / or the like. This enables, for example, the algorithms that can be used due to the presence of all variables to be determined on the basis of the presence of the measured variables in the input data, and increases the flexibility of the method.

Additionally or alternatively, the input quality can be determined based on a criterion of an accuracy of the input data. The accuracy can, for example, follow from a data source or source of the respective data and / or a measurement method of the respective data, and / or it can be transmitted from the respective source, for example as variable metadata or measurement information. This makes it easier to assess the input quality.

Additionally or alternatively, the input quality can be determined based on a criterion of regularity of the input data. The regularity can in particular relate to a measurement position-related or local regularity of the input data, a measurement time-related or temporal regularity of the input data, and / or a regularity of measured values of the input data. For example, input data that are available at relatively regular intervals, such as weather information from a weather service that can be accessed via the Internet, can be assigned a higher input quality than input data that are available at irregular intervals, such as measured values from a vehicle fleet. This increases the reliability of the method.

Additionally or alternatively, the input quality can be determined based on a criterion of a deviation in the measured value range of the input data. This can in particular include a deviation of measurement points or measurement values and / or measurement information of the input data relative to one another. For example, a few measured values that are relatively close together can be assigned a higher input quality than many, but widely spread, measured values. The deviation can in particular be a statistically describable deviation. This increases the reliability of the method.

Additionally or alternatively, the input quality can be determined based on a criterion of a quality of the input data. The quality of the input data can in particular relate to an overall consideration of measuring points and / or measured values of a measured variable and / or a data source, such as a weighing or assessment based on at least two of the criteria mentioned. This determination can be predetermined very efficiently.

Additionally or alternatively, the input quality can be determined based on a criterion of the availability of at least one data source for the input data. For example, input data from a permanently communicable measuring station can be assigned a higher input quality than input data from a vehicle that can only be communicated during operation. This increases a minimum security of the procedure.

Additionally or alternatively, the input quality can be determined based on a criterion of a reliability of at least one data source for the input data. For example, it can be defined that input data from vehicle-mounted sensors have a higher input quality than input data from a forecast service. This increases the reliability of the method.

The at least one criterion for determining the input quality of the input data is preferably established or determined or defined separately for each algorithm. The at least one criterion for determining the input quality of the input data is preferably established or determined or defined separately for each measured variable and / or for each data source. The at least one criterion for determining the input quality of the input data is furthermore preferably determined or determined or defined separately for each measured variable for each algorithm and / or for each data source for each algorithm. In other words, the input quality can be a plurality of input qualities. The input quality can be established or determined or defined once, at irregular intervals, at regular intervals, event-related and / or continuously for each measured variable and / or for each data source and / or for all input data.

In a further development, the input quality of the input data can be determined with regard to all algorithms in order to make a complete comparison. The quality of the results of the coefficients of friction determined using all algorithms can also be determined in order to make a complete comparison.

If the step of selecting has to determine the input quality of the input data, the input quality can be made directly related to the selection so that, for example, criteria that are not unconsidered are evaluated so that the method becomes more efficient.

If the at least one algorithm is selected before the coefficient of friction is determined with the at least one selected algorithm, this enables an early decision or selection if only the input quality is considered.

That algorithm of the plurality of algorithms for determining the coefficient of friction can be selected whose requirements for the input quality best match the determined input quality of the input data. In this context, provision can be made, in a further development, to select an algorithm from the plurality of algorithms based on the greatest possible degree of correspondence between the determined input quality and the respective requirement of each algorithm for the input quality. The agreement can be a comparison in terms of amount. The matching can include or take into account or evaluate the fulfillment of a predetermined minimum requirement for the input quality. The possibilities mentioned allow a relatively simple automation of the selection step, such as by means of a computer.

In order to save effort, such as computational effort, provision can be made to determine the coefficient of friction of the roadway exclusively with the selected algorithm. An available capacity, such as a computation capacity of a computer system, can thus be concentrated on only one algorithm, and an efficient method is thus achieved.

Furthermore, using each algorithm of the plurality of algorithms, a coefficient of friction can be determined based on the input data, so that the same number of coefficients of friction as algorithms are obtained in order to form a broad decision-making basis for the selection step.

In a further development, the step of selecting can include determining the quality of the results of the friction values determined by applying each algorithm of the plurality of algorithms. As a result, the result qualities actually achieved are available as a basis for decision-making for the selection step directly related to the selection.

If the at least one algorithm is selected after the coefficients of friction have been determined using each of the algorithms, the presence or knowledge of the coefficients of friction actually determined facilitates selection of the most precise coefficient of friction.

If the result quality is determined based on at least one characteristic value, an easily applicable determination is given. Examples of the characteristic value (technical language: key performance indicator or KPI) contain a convergence requirement, a stability requirement and a minimum number of reference values.

If the result quality is determined based on at least one comparison of the determined friction values with at least one control measurement, the determined coefficient of friction that comes closest to the measured coefficient of friction can be determined. The control measurement is preferably not included or taken into account in the determined coefficient of friction; it is therefore, for example, not used for training and / or parameterizing the algorithm or the algorithms. The comparison is preferably a field comparison (in technical terms: ground truth), such as a comparison with a coefficient of friction measurement, for example from an ESP sensor.

If that algorithm of the plurality of algorithms for determining the coefficient of friction is selected, the application of which is used to determine the coefficient of friction with the best result quality, the measured coefficient of friction can be approximated very precisely.

If, on the other hand, the input quality of the input data is also determined, and that algorithm of the plurality of algorithms for determining the coefficient of friction is selected which is the best combination of the result quality determined for the coefficient of friction determined using this algorithm and a match between the requirement of this Algorithm to the input quality and the determined input quality of the input data can be achieved in two ways: On the one hand, by combining the input quality and the result quality, a selection based on a randomly unrepresentative control measurement can be identified and prevented. On the other hand, taking the input quality into account can be more efficient than comparing it with a plurality of control measurements.

If it is provided that all algorithms are used in parallel in the determination step, the coefficients of friction can be determined quickly. If it is provided that the algorithms are at least partially applied serially in the determination step, the coefficients of friction can, for example, be determined in a prioritized manner. These possibilities allow a wide selection based on all determined or determinable coefficients of friction.

The input data can contain data from a data provision service, such as a weather service and / or an Internet source or other Internet source, from a road sensor system, from a stationary measuring station, from a vehicle and / or from a vehicle fleet. A data provision service can be a service that provides data that can be called up for at least one measured friction value at a measurement position on the roadway and at a measurement time. A weather service can be a service for the forecast-based and / or measurement-based provision of environmental data. An Internet source is generally available regularly and can be accessed using simple means; The internet source can provide, for example, road map data, road surface data, weather data, vehicle type data, tire type data and / or traffic data. A lane sensor system, such as a road sensor system, can have a device for determining a parameter representing a condition of a road, such as a device for determining a road temperature, a device for determining a parameter representing a condition of the surroundings of a road, such as a device for determining air humidity, and / or a device for determining a parameter representing a state of a road user, such as a device for speed control or a device for measuring a wheel contact force. A stationary measuring station can be a lane sensor system and / or a device for determining environmental data. A vehicle fleet can contain vehicles of the same type and / or of different types with regard to a coefficient of friction, such as, for example, vehicles of the same type or model or model year with the same tires or the same tire type or vehicles of different types with different tires. The different data sources enable the method to be used with a high degree of estimation quality with different availability of the data sources.

An algorithm used in this method can be a data-based algorithm, such as an algorithm derived from a data analysis or an algorithm containing a parameter derived from a data analysis. Additionally or alternatively, an algorithm used in this method can be an algorithm that is suitable and / or designed for processing in a time sequence, such as an algorithm that processes measured values in a time sequence with or without time values. Additionally or alternatively, an algorithm used in this method can be an algorithm learned or trained or parameterized in a preceding training or training step. Additionally or alternatively, an algorithm used in this method can be a learnable algorithm, such as an algorithm that can be learned online or in a communicating exchange and / or an adaptively learnable algorithm. Additionally or alternatively, an algorithm used in this method can be an algorithm that is independent of at least one further algorithm used, with all algorithms used preferably being independent of one another. "Independent" can mean that the respective algorithms are developed, parameterized, trained and / or validated based on independent, i.e. different, measuring points or measured values or input data.

The road friction coefficient is preferably provided with reference to the measurement position of the road, for which the input data are representative, on the basis of which the selected algorithm has determined the coefficient of friction. The step of providing with reference to the measurement position can include the fact that the road surface coefficient of friction is provided in response to a request for the road surface coefficient of friction at this position. Additionally or alternatively, providing with reference to the measurement position can include the fact that the road surface friction coefficient is provided together with an indication of this measurement position. A combination of these possibilities increases the security of the process.

The provided coefficient of friction of the roadway can be input into a driver assistance system on the part of a receiver, such as a vehicle, in order to adapt a driving behavior that can be influenced by the driver assistance system to the route. For example, a maximum vehicle speed can be adapted or adapted to a required braking distance. For example, an adjustable maximum braking force can be adapted or adapted to an achievable traction, in particular to avoid wheel locking. For example, a maximum vehicle speed can be adapted and / or adapted to an achievable lateral force before and / or during an expected cornering, for example. For example, a maximum steering angle can be adapted and / or adapted to an achievable lateral force while driving through a curve. For example, a maximum vehicle acceleration can be adapted or adapted to an achievable traction, in particular to avoid wheelspin. A combination of the above options increases vehicle safety. The road surface coefficient of friction provided, in particular provided to a vehicle, can therefore be input into a control device in order to influence, in particular limit, an output or a signal of the control device relating, for example, to a vehicle speed, a braking force, a steering angle and / or a vehicle acceleration. Furthermore, the road friction coefficient provided, in particular for a vehicle, can be input into a control device in order to transmit an output or a signal from the control device relating to a warning and / or information signal to a driver relating, for example, to a vehicle speed, a braking force, a steering angle and / or to generate a vehicle acceleration.

In a further development, the method can have at least one measuring point or measured value representative of a measured friction value at a measuring position of the roadway at a measuring time and / or data representative of a measured friction value at a measuring position of the roadway at a measuring time in a database containing the for at least one measured friction value input data representative of the same or a different measurement position of the roadway at the same or a different measurement time, so that the method can be expanded and / or updated with respect to the input data. In a further development, the method can include acquisition of input data, input data for the algorithms, such as derived and / or aggregated, being obtained from the database containing measuring points or measured values.

In a further development, the determination of the coefficient of friction for the roadway using the at least one selected algorithm can be a server-side determination or a determination that can be carried out by means of a server or a communicable determination. Thus, a result of a friction coefficient determination carried out only once can ultimately be provided to a plurality of vehicles, so that the same estimation is not carried out more than once in an efficient manner. In a further development, the selection can be a server-side selection so that the same selection is only carried out once in an efficient manner. Alternatively, the selection can be a vehicle-specific selection after the server-side determination, for example a vehicle-specific and / or road-specific selection based on a vehicle-specific and / or road-specific comparison with a vehicle-specific and / or road-specific control measurement. The selection in the vehicle can contain and / or follow up on the selection, for example to the database.

Another aspect of the invention relates to a computer system that is set up to carry out steps of the method as described above and below. The computer system can, for example, have at least one determining means or estimating means which is set up to carry out the determining or estimating. Such a determination means can contain a computer processor, preferably a plurality of computer processors and / or preferably at least one computer processor set up for parallel computing. In addition, the computer system can have at least one selection means, such as a computer processor, and a provision means, such as a communication interface. The computer system can in particular be a single computer, such as a desktop computer, a server computer or a computer emulated by software, or a computer network, such as a computer cluster or a computer cluster emulated by software. In addition, the computer system can also be selected from one of the classes of personal portable computer systems and / or embedded computer systems.

Another aspect of the invention relates to a computer program that, when it is executed at least one processor of a computer system instructs the computer system to carry out steps of the method as described above and below. The computer program can in particular be modularized in order to be expandable, for example by adding additional algorithms. The computer program can in particular be set up for determining or estimating distributed over a plurality of processors and / or for determining or estimating in parallel.

Another aspect of the invention relates to a computer-readable medium on which a computer program is stored which, when it is executed on at least one processor of a computer system, instructs the computer system to carry out steps of the method as described above and below.

Further measures improving the invention are shown in more detail below together with the description of the preferred exemplary embodiments of the invention with reference to figures.

Figure list

• 1 schematisch einen Aufbau eines Systems, in dem ein Verfahren gemäß einer ersten Ausführungsform der Erfindung angewendet wird;
• 2 schematisch einen Ablauf des Verfahrens gemäß der ersten Ausführungsform der Erfindung;
• 3 schematisch einen Ablauf eines Verfahrens gemäß einer zweiten Ausführungsform der Erfindung; und
• 4 schematisch einen Ablauf eines Verfahrens gemäß einer dritten Ausführungsform der Erfindung.

It shows:

• 1 schematically a structure of a system in which a method according to a first embodiment of the invention is applied;
• 2 schematically a sequence of the method according to the first embodiment of the invention;
• 3 schematically a sequence of a method according to a second embodiment of the invention; and
• 4th schematically a sequence of a method according to a third embodiment of the invention.

The figures are only schematic and not true to scale. Identical, identically acting or similar elements with the same reference symbols can be used in the figures.

The 1 shows schematically a system with a computer system 1 , one as a whole with 2 designated vehicle fleet of a plurality of structurally different vehicles, a lane sensor system 3 , a weather service provider 4th , an internet source 5 and a vehicle 6th . The computer system 1 In the present case, the architecture of the server backend is not specified. The vehicle fleet 2 , the lane sensors 3 , the weather service provider 4th and the internet source 5 are common as data sources 7th designated. The vehicle 6th drives on a roadway 8th , in this case an asphalt road.

The data sources 7th put the computer system 1 Measurement points ready, which in the present case each contain a measurement value, a measurement position and a measurement time. For example, determine the vehicles in the vehicle fleet 2 the measuring points by means of vehicle-based sensors, and they transmit them to the computer system by means of a cellular network (not shown) 1 .

In the computer system 1 does it run in the 2 outlined, later described procedure, which a road friction coefficient 16 the roadway 8th determined. This coefficient of friction 16 is then the vehicle 6th provided as a receiver or target vehicle by using cellular to the vehicle 6th is transmitted. Thus the vehicle receives 6th information about the measured friction value of the road currently being driven on 8th . From this coefficient of friction 16 This results in a gain in information for the vehicle with regard to the roadway 8th whereby driver assistance systems of the vehicle 6th be able to control assisted and / or partially automated driving more reliably. For example, the security of a predictive driver assistance system is increased in the present case. In other words, the vehicle 6th receives information about the roadway 8th which it cannot generate with this level of accuracy with on-board resources. As a result, the provided road friction coefficient 16 the roadway 8th be entered in a coefficient of friction map, which is a driver assistance system of the vehicle 6th For example, it is used to automatically set or limit a vehicle speed before cornering, so that a dangerous situation, such as a departure from a lane, especially in road conditions that are perceived as difficult, such as wetness or snow on the roadway 8th , can be avoided.

The 2 shows the method according to the invention for determining the road surface coefficient of friction 16 the roadway 8th according to the first embodiment. Are shown generically with 9 designated input data 9a , 9b , 9c , which in each case in algorithms designated generically with 10 10a , 10b , 10c enter. The input data 9 can be partially the same, that means, for example, that the input data 9a in the algorithm 10a and the input data 9b in the algorithm 10b partly the same and partly different input data 9 are.

From the input data 9 an input quality is generically denoted by 13 13a , 13b , 13 determined. The algorithms 10 are in the present case estimation algorithms which are used during a step of determining 11 of friction coefficients generically designated by 12 12a , 12b , 12c for the roadway 8th be applied. From the determined coefficients of friction 12 is based on a comparison with a control measurement in each case a result quality generically designated by 14 14a , 14b , 14C determined.

Here is the algorithm 10a for example, set up and designed to include all those input data 9a for estimating the coefficient of friction 12a use which is defined to have the highest input quality 13a have, for example, because they are of the highest quality and in large quantities. The algorithm 10b is set up and designed for this, based on data from the weather service provider 4th , which are assumed to be always available, and on data from the vehicle fleet 2 , which are assumed to be present in a relatively dense location and / or time, the coefficient of friction 12b to determine with an estimate. The algorithm 10c is set up and designed, for example, based on the data from the weather service provider 4th and the internet source 5 the coefficient of friction 12c to determine with an estimate. The algorithms 10 are in the present case independent, data-based and teachable or parameterizable algorithms, which the input data 9 process in time series.

In the first embodiment, the step of determining follows 11 a step of choosing 15th . In the step of choosing 15th is based on matching the input qualities 13 with a respective requirement for each algorithm 10 to the input quality 13 and based on the results 14th one of the determined coefficients of friction 12 Goodness-based selected to the vehicle 6th as road friction coefficient 16 to be provided. For example, the determined coefficient of friction is used here 12a selected. The determined coefficient of friction 12a thus becomes the road surface friction coefficient 16 the roadway 8th , which in a step marked 17 the vehicle 6th For example, is provided for use in a driver assistance system.

The 3 shows the method for determining a road surface coefficient of friction 16 the roadway 8th according to a second embodiment of the invention. In contrast to the first embodiment, there are no input quality and only the result quality 14th determined, and is selected 15th one of the determined coefficients of friction 12 only based on the quality of results 14th selected. For the rest, reference is made to the description of the above embodiment.

The 4th shows the method for determining a roadway coefficient of friction of the roadway 8th according to a third embodiment of the invention. In contrast to the first embodiments, only the input qualities are used 13 and no quality of result is determined, and is in selecting 15th only one determinable coefficient of friction 12 or only one of the determinable coefficients of friction 12 only based on a respective agreement of the input quality 13 with a respective requirement for each algorithm 10 to the input quality 13 selected. In contrast to the first embodiments, this is the step of selecting 15th the step of determining 11 upstream, so only one algorithm 10 on its input data 9 , e.g. the algorithm 10a on the input data 9a , is applied. In the third embodiment, the coefficient of friction is thus determined 12a even before its determination as the road friction coefficient 16 preselected which in the step of providing 17th the vehicle 6th For example, is provided for use in a driver assistance system. For the rest, reference is made to the descriptions of the above embodiments.

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• DE 102008047750 A1 [0003]

Claims (17)

Method for determining a roadway coefficient of friction (16) of a roadway (8), the method comprising the steps: Selecting (15) at least one algorithm (10) of a plurality of algorithms (10) for determining a coefficient of friction (12) of the roadway (8) based on input data (9) which are representative of at least one measured friction value at a measurement position of the roadway (8) and at a measurement time are wherein the algorithms (10) for determining the coefficient of friction (12) with regard to a requirement for an input quality (13) of the input data (9) and / or with regard to a result quality (14) of the coefficient of friction (12) determined using the respective algorithm (10) ) distinguish; wherein the at least one algorithm (10) is selected based on the input quality (13) of the input data (9) and / or based on the result quality (14) of the algorithms (10); Determining (11) a coefficient of friction (12) for the roadway (8) using the at least one selected algorithm (10) of the plurality of algorithms (10); and Providing (17) the coefficient of friction (12) determined with the at least one selected algorithm (10) as a coefficient of friction for the roadway (16). Procedure according to Claim 1 , further comprising: determining the input quality (13) based on at least one criterion selected from the group consisting of: a number of the input data (9), a density of the input data (9), a type of the input data (9), an accuracy of the input data (9), a regularity of the input data (9), a deviation from measured values of the input data (9), a quality of the input data (9), an availability of at least one data source (7) for the input data (9) and a reliability of at least one data source (7) for the input data (9). A method according to any preceding claim, wherein the step of selecting (15) comprises: Determining the input quality (13) of the input data (9). Procedure according to Claim 3 wherein the at least one algorithm (10) is selected before the coefficient of friction (12) is determined with the at least one selected algorithm (10). Procedure according to Claim 3 or 4th , wherein that algorithm (10) of the plurality of algorithms (10) for determining the coefficient of friction (12) is selected whose requirements for the input quality (13) best match the determined input quality (13) of the input data (9). Procedure according to Claim 5 , wherein the coefficient of friction (12) of the roadway (8) is determined exclusively with the selected algorithm (10). Method according to one of the Claims 1 to 2 wherein, using each algorithm (10) of the plurality of algorithms (10), a coefficient of friction (12) is determined based on the input data (9); and wherein the step of selecting (15) further comprises: determining the result qualities (14) of the coefficients of friction (12) determined using each algorithm (10) of the plurality of algorithms (10). Procedure according to Claim 7 , wherein the at least one algorithm (10) is selected after the determination (11) of the friction values (12) using each of the algorithms (10). Method according to one of the Claims 7 to 8th , the result qualities (14) being determined based on at least one characteristic value and / or based on at least one comparison of the determined friction values (12) with a control measurement. Method according to one of the Claims 7 to 9 , wherein that algorithm (10) of the plurality of algorithms (10) for determining the coefficient of friction (12) is selected under whose application the coefficient of friction (12) with the best result quality (14) is determined. Method according to one of the Claims 7 to 9 wherein the step of selecting (15) comprises: determining the input quality (13) of the input data (9); wherein that algorithm (10) of the plurality of algorithms (10) for determining the coefficient of friction (12) is selected which is the best combination from the quality of the result (14) that determines the coefficient of friction (12) determined using this algorithm (10) and a correspondence between the requirement of this algorithm (10) for the input quality (13) and the determined input quality (13) of the input data (9). Method according to one of the Claims 7 to 11 , all algorithms (10) being applied in parallel in the estimating (11). Method according to one of the preceding claims, wherein the input data (9) contain data from a data provision service (4, 5), from a lane sensor system (3), from a stationary measuring station, from a vehicle and / or from a vehicle fleet (2). Method according to one of the preceding claims, wherein the road surface coefficient of friction (16) with reference to the measurement position of the road surface (8) is provided, for which the input data (9) are representative, on the basis of which the selected algorithm (10) has determined the coefficient of friction (12). Computer system (1) which is set up to carry out steps of the method according to one of the preceding claims. Computer program which, when executed on at least one processor of a computer system (1), instructs the computer system (1) to do so, steps of the method according to one of the Claims 1 to 11 execute. Computer-readable medium on which a computer program according to the Claim 16 is stored.

Images