DE102018217193A1 - Method and system for providing a coefficient of friction to a vehicle - Google Patents

Abstract

The invention relates to an improved possibility of providing a coefficient of friction to a vehicle on the basis of a static coefficient of friction classification. A method for providing a coefficient of friction (.mu.) For a traffic area section (RS) to a vehicle (V) provides for the first data to be recorded, which includes vehicle coefficient of friction data of the traffic area section (RS). In addition, second data are recorded, which include environmental condition data of the traffic area section (RS). Weather data are obtained which can be assigned to the traffic area section at a determination time. The coefficient of friction (.mu.) For the determination time is then determined on the basis of the first data and the second data, the condition being fulfilled that the coefficient of friction (.mu.) From the first data in the case of environmental condition data from the second data, with the weather data obtained correlate with respect to a certain time parameter lies within a range of values.

Description

The present invention relates to a method for providing a coefficient of friction to a vehicle. The invention further relates to a system for providing a coefficient of friction to a vehicle.

State of the art

The coefficient of friction between a vehicle, in particular its tire, and a traffic area section, such as a roadway, influences the dynamic driving limits of a vehicle with regard to longitudinal and transverse guidance. In a conventional vehicle driven by a human driver, the driver may e.g. Using empirical values, brief braking, visual perception or the like, at least to some extent, determine whether the current coefficient of friction is rather high or rather low. For example, the current coefficient of friction on a well-developed and clean road surface and when it is dry can be rather high and thus promote good static friction between the vehicle and the road surface, whereas wetness, black ice, dirt on the road surface or the like can adversely affect the coefficient of friction.

In the case of a partially or fully autonomous driving operation of a vehicle, it is more difficult to estimate the current coefficient of friction of a traffic area section, since the information of the human vehicle driver is at least partially eliminated. To counter this problem, for example, the WO 2016/120092 A1 a database-based coefficient of friction map, in which information sent by sending vehicles is received and stored in the database, the information comprising at least the data describing the coefficient of friction of a road segment, the location data describing the geometric position of this segment of the road, and the time data describing the time of determination of the coefficient of friction data, and in data stored in the database can be called up by receiving vehicles. Although the vehicle can thus be provided with useful information on the current coefficient of friction of a traffic area section, there is a desire to be able to provide a vehicle with even more precise coefficient of friction data.

Disclosure of the invention

Embodiments of the invention provide an improved way of providing friction data to a vehicle according to the independent claims. Advantageous further developments of the invention result from the dependent claims, the description and the accompanying figures.

A proposed method for providing e.g. predicted coefficient of friction for a traffic area section on a vehicle can in particular be carried out with a data processing device which e.g. can have a determination module, in particular an artificial intelligence module, KL module, which can be implemented by software, possibly also by hardware. The data processing device can have a memory for storing acquired and / or received data and a processor etc. A data interface can be provided for acquiring and / or receiving data and for exchanging data with the vehicle. The data processing device can be a server system or a part thereof and allow access to the coefficient of friction for a large number of vehicles. The traffic area section can in principle be any geographically determinable, e.g. mapped, roadway, but especially a street. The vehicle can be an at least partially automated vehicle that can use the coefficient of friction provided, for example, for its route planning, trajectory planning, etc.

• - Es werden erste Daten erfasst, die fahrzeugseitig bestimmte Reibwertdaten des Verkehrsflächenabschnitts umfassen.

The process involves the following steps:

• - First data are recorded, which include vehicle coefficient of friction data of the traffic area section.

• - Es werden zweite Daten erfasst, die Umgebungsbedingungsdaten des Verkehrsflächenabschnitts umfassen.

The coefficient of friction data can be obtained from a driving dynamics control of the vehicle and other sensor data, such as a rain sensor, a camera and the like, and can be carried out either on the vehicle side or on the server side, for example by the data processing device. From a functional point of view, one or more vehicles can drive on the traffic area section and in doing so acquire the first data and make it available to the data processing device, for example. The coefficient of friction data can be present, for example, as a feature vector or the like.

• - Second data are recorded, which include environmental condition data of the traffic area section.

• - Es werden Wetterdaten erhalten, die dem Verkehrsflächenabschnitt zu einer Bestimmungszeit zuordbar sind.

The environmental condition data can include, for example, weather data that can be assigned to the coefficient of friction data of the first data, for example, was recorded at a similar or the same time. The environmental data can include, for example, an ambient temperature, a precipitation amount, a wind strength, a wind direction, a sun intensity, street temperature, traffic volume and the like. The environmental condition data can be obtained, for example, at least in part from the vehicles that also record the coefficient of friction data of the first data. As an alternative or in addition to this, the environmental condition data can also be acquired or obtained by a weather service, a local weather station, road sensors, etc.

• - Weather data are obtained that can be assigned to the traffic area section at a determined time.

• - Es wird der bereitzustellende Reibwert für die Bestimmungszeit auf Basis der ersten Daten und der zweiten Daten bestimmt, wobei zumindest die Bedingung erfüllt ist, dass der Reibwert aus den ersten Daten bei Umgebungsbedingungsdaten aus den zweiten Daten, und optional auch ersten Daten, die mit den erhaltenen Wetterdaten korrelieren, bezogen auf einen bestimmten Zeitparameter innerhalb eines Wertebereichs liegt.

The weather data can include, for example, an ambient temperature, a rainfall, a wind strength, a wind direction, a sun intensity, road temperature, traffic volume and the like. The weather data can be acquired or obtained from a weather service, a local weather station, road sensors, etc.

• The coefficient of friction to be provided for the determination time is determined on the basis of the first data and the second data, at least the condition being fulfilled that the coefficient of friction from the first data in the case of environmental condition data from the second data, and optionally also first data associated with the correlate weather data obtained, based on a specific time parameter lies within a range of values.

In this context, correlating data can be understood to mean that in the first and / or second data, which are stored, for example, in a history table or in some other way, similar, approximately identical or identical environmental condition data have been recorded, which have a relationship between the coefficient of friction that can be assigned to them and allow the received weather data at the destination time. This is based on the idea that the traffic area section has at least approximately the same coefficients of friction under at least approximately the same environmental conditions, which are determined on the one hand from the first and / or second data and on the other hand from the weather data obtained, and this does not change over a limited period of time. Expressed in graphic terms, measurement points of the first and / or second data that can be assigned can be filtered at the determination time, which can include a point in time, a time segment or the like, the traffic area segment under consideration here being able to be spatially and temporally limited.

With this method, a static road coefficient of friction classification can be carried out for different intermediate layers, the coefficient of friction to be provided being determined on the basis of the weather data, which form the exclusive input variable for the determination time, from the first and / or second data, which represent a type of database . This means that the coefficient of friction of the traffic area section can only be determined using known weather data, without further real-time measurement data from e.g. Vehicles are needed. This allows a precise coefficient of friction to be determined or predicted and made available to the vehicle with simple means.

In a further development, the weather data obtained can be defined by features from which a classifier determines the coefficient of friction from the first and / or second data. The characteristics can e.g. exist as a feature vector. A classifier, which may previously have been trained with training data, can therefore be used at the determination time in order to determine the coefficient of friction from the weather data as the only input variable.

According to a further development, the determination of the coefficient of friction can include a next-neighbor classification, which in particular can comprise a k-nearest-neighbor algorithm, on the basis of the first and / or second data.

In another development, the determination of the coefficient of friction can include determining an intermediate layer of the traffic area section from the first and / or second data. The intermediate layer can be present on a road surface of the traffic area section, which can be dry, moist, wet, snow-covered, icy, covered with chippings, etc. The intermediate layer can be determined by a model, a classifier or the like, which can be implemented in the AI module.

, wobei n die Anzahl der Messungen in den ersten und/oder zweiten Daten und µ_i einzelne Messwerte darstellen. Dadurch kann der bereitzustellende Reibwert noch genauer bestimmt werden.According to a further development, the determination of the coefficient of friction can include a determination of an average error of the coefficient of friction, or possibly several possible coefficients of friction, from the first and / or second data. For example A mean value can be determined, in particular calculated, for, in particular filtered, measuring points of the first and / or second data. An error of the mean value can then be determined for this. This can be determined according to, for example $$\mathrm{\Delta }{\mathrm{\mu }}_{\text{err}}=\frac{\sqrt{\frac{1}{n}{\displaystyle {\sum }_{i1}^n{\left(\overline{\mu }-{\mathrm{\mu }}_i\right)}^{\mathrm{2nd}}}}}{\sqrt{n}}$$

, where n represents the number of measurements in the first and / or second data and µ _i individual measured values. As a result, the coefficient of friction to be provided can be determined even more precisely.

In einer anderen Weiterbildung kann aus der Gewichtung ein gewichteter Mittelwert des Reibwerts bestimmt werden. Dieser kann beispielsweise den Reibwert des Verkehrsflächenabschnitts für eine Zwischenschicht wiedergeben. Dieser Reibwert kann beispielsweise bestimmt werden nach $$\mathrm{\mu =}{\displaystyle \sum _{i=1}^n{w}_i\ast {\mathrm{\mu }}_i.}$$

In another development, the determination of the coefficient of friction can include a weighting of possible coefficients of friction, the coefficient of friction data included in the first data being weighted more if these are less in time from the determination time. In other words, a time distance from the determination time can be determined for, in particular filtered, measurement points of the first and / or second data. A weight w _i can then be assigned to each measuring point based on the distance. The sum of the weights can be equal to 1. More recent or younger friction values of the first and / or second data can be weighted higher than older ones. A temporal correlation can also be mapped to derive these weights. The weight can be determined according to, for example $$w_i=\frac{\frac{1}{t_{diff}}}{{\displaystyle \sum \frac{1}{t_{diff}}}}.$$

In another development, a weighted average of the coefficient of friction can be determined from the weighting. This can, for example, reflect the coefficient of friction of the traffic area section for an intermediate layer. This coefficient of friction can be determined, for example, according to $$\mathrm{\mu \; =}{\displaystyle \sum _{i=1}^n{w}_i\ast {\mathrm{\mu }}_i.}$$

According to a further development, a one-sigma error of the mean can be determined over a one-sigma confidence interval of the measuring points, assuming a standard distribution $$\mathrm{\Delta }{\mathrm{\mu }}_{err}=\frac{{\displaystyle {\sum }_{i=1}^n{w}_i\ast \mathrm{\Delta }{\mathrm{\mu }}_i}}{\sqrt{n}}$$

In another development, the time parameter can be determined using an optimization method. As mentioned above, the time parameter in particular gives a time limit, e.g. a period of time over which the coefficient of friction changes little or not at all in correlating ambient conditions. This time parameter can be determined by an optimization method, for example a hyper parameter optimization. As a result, the coefficient of friction or its change over time can be determined even more precisely.

In a further development, the optimization method can be selected from: grid search, random search, Bayesian optimization and the like. These optimization methods, in particular hyper parameter optimization methods, have proven to be particularly suitable for the non-linear system here.

According to a further development, a hypothesis test of the determined coefficient of friction can be carried out before the coefficient of friction is provided. The hypothesis test can comprise, for example, a three-sigma test, a specification of a level of significance or the like. In particular, it can be used to check whether the determined mean value with Δµ _{err is} above a limit value. For example, it is conceivable that the vehicle for which the coefficient of friction is to be provided has a driving function, for example a driver assistance system, for which a minimum coefficient of friction should be given. This can be checked in the hypothesis test.

In a further development, the traffic area section can be delimited locally or spatially by a polygon. The first and / or second data can thus allow an assignment via points of the polygon.

• - erste Daten zu erfassen, die fahrzeugseitig bestimmte Reibwertdaten des Verkehrsflächenabschnitts umfassen,
• - zweite Daten zu erfassen, die Umgebungsbedingungsdaten des Verkehrsflächenabschnitts umfassen,
• - Wetterdaten zu erhalten, die dem Verkehrsflächenabschnitt zu einer Bestimmungszeit zuordbar sind, und
• - den bereitzustellenden Reibwert für die Bestimmungszeit auf Basis der ersten Daten und der zweiten Daten zu bestimmen, wobei die Bedingung erfüllt ist, dass der Reibwert aus den ersten Daten bei Umgebungsbedingungsdaten aus den zweiten Daten, die mit den erhaltenen Wetterdaten korrelieren, bezogen auf einen bestimmten Zeitparameter innerhalb eines Wertebereichs liegt.

The invention also relates to a system for providing a coefficient of friction for a traffic area section to a vehicle, having a data interface and a data processing device, which can be implemented, for example, as described above. The data processing device is set up to

• to acquire first data, which includes vehicle coefficient of friction data of the traffic area section,
• to acquire second data, which include environmental condition data of the traffic area section,
• Obtaining weather data which can be assigned to the traffic area section at a determination time, and
• - Determine the coefficient of friction to be provided for the determination time on the basis of the first data and the second data, the condition being fulfilled that the coefficient of friction from the first data in the case of environmental condition data from the second data, which correlate with the weather data obtained, based on a specific one Time parameter lies within a range of values.

The first and / or second data can be collected and / or stored in the system or the data processing device. The advantages explained for the method described above can thus be achieved. The data processing device can have the above-mentioned AI module.

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

Figure list

• 1 eine schematische Übersicht eines Systems zum Bereitstellen eines Reibwerts eines Verkehrsflächenabschnitts an ein Fahrzeug und
• 2 ein Flussdiagramm eines Verfahrens zum Bereitstellen eines Reibwerts eines Verkehrsflächenabschnitts an ein Fahrzeug.

In the following, advantageous exemplary embodiments of the invention are described in detail with reference to the accompanying figures. Show it:

• 1 a schematic overview of a system for providing a coefficient of friction of a traffic area section to a vehicle and
• 2nd a flowchart of a method for providing a coefficient of friction of a traffic area section to a vehicle.

The figures are only schematic and are not to scale. In the figures, the same, equivalent or similar elements are provided with the same reference numerals throughout.

Embodiments of the invention

1 shows a schematic overview of a system 100 , which is used to provide a coefficient of friction µ of a traffic area section RS to a vehicle V is suitable. The traffic area section RS can be determined geographically, for example via a street map, and, as in 1 indicated below, spatially delimited by a polygon. So is the traffic area section RS representative for in principle any mapped lane or similar.

The system 100 is designed as a, for example server-based, computer system and here includes, by way of example, a data processing device 110 comprising a processor, a memory for storing program instructions executable by the processor, a memory for storing acquisition data, etc. In the system 100 or the data processing device 110 is a classifier, which is an example of a k-nearest-neighbor algorithm, implemented by program instructions and / or hardware and can be executed by the processor. Furthermore, in the system 100 or the data processing device 110 an optimization method, which is an example of a grid search method, Bayesian optimization method or the like, is implemented by program instructions and / or hardware and can be carried out by the processor.

The system 100 or the data processing device 100 also has a (not specified) data interface, via which it can be connected to one or more data acquisition devices 120 , 130 , 140 is connectable or connected for data reception. The data acquisition device 120 is here, for example as a sensor system, electronic control unit or the like, in a plurality of vehicles MV of arranged, two of which are shown as examples. The vehicles MV usually drive through the traffic area section at different (day) times, weather conditions etc. RS . The data acquisition device 120 is, for example, part of a vehicle dynamics control system of the respective vehicle MV or accesses data from the same, and possibly also a rain sensor, operating parameters or environmental data of the respective vehicle MV to. The data acquisition device 120 represents the system 100 or the data processing device 100 First data are available via the data interface, which include at least friction data of the traffic area section. The data acquisition device 130 is designed here as an example as a road sensor and is set up to detect, for example, a road temperature, passing vehicles or the like. The data acquisition device 140 As an example, a weather service, a local weather station or the like, which may also be further away from the traffic area section RS1, is set up here and is set up to record, for example, weather data, such as the amount of precipitation, sun intensity or the like. From a functional point of view, the data acquisition devices 120 , 130 , 140 set up to deliver measurement and / or estimation data with a time reference. With this, the data acquisition devices 130 , 140 the system 100 or the data processing device 100 Second data are available via the data interface, which include at least environmental condition data of the traffic area section. However, the environmental condition data can also be generated by the additional data of the data acquisition device 120 be supported.

The system 100 or the data processing device 100 also has a (not specified) data interface, which you can use with a large number of vehicles V is connectable or connected for data reception. So it can through the system 100 or the data processing device 100 Certain coefficient of friction μ are made available to the vehicle for processing there, for example for route calculation, trajectory calculation, etc.

Further referring to 1 will now be an example operation of the system 100 or the data processing device 100 explained.

As mentioned above, to form a database serve as input data for the system 100 that of the vehicles MV , i.e. the data acquisition devices 120 recorded, measured and / or determined coefficient of friction data in the form of the first data. In addition, the second data is acquired, that of, for example, the data acquisition devices 130 , 140 and or 120 captured environmental condition data of the traffic area section, in which the coefficient of friction data from the vehicle MV were recorded. The environmental condition data either directly contain information on an intermediate layer that can be assigned to the detected friction values, for example dry, wet, moist, etc., of the traffic area section RS or the data processing device 100 is set up to determine this, for example using a model.

At a designated time, for example at the request of the vehicle V For example, a route calculation is made available to the system as input data, current or current weather data, which may also be predicted for a future point in time. This can be done in particular by the data acquisition device 140 to provide.

At the destination time, then through the system 100 or the data processing device 110 Associated measurement points from the data memory, for example a history table in which the first and second data are stored, are filtered using the classifier, for example the k-nearest-neighbor algorithm. The considered traffic area section is in the first and second data RS spatially or locally limited by the polygon and limited in time by a time parameter. The time parameter is determined using the optimization method mentioned above.

The coefficient of friction μ to be provided is determined by the k-nearest-neighbor algorithm in such a way that the condition is fulfilled that the coefficient of friction recorded in the first data relates to an environmental data recorded in the second data, which correlate with the weather data obtained certain time parameters lies within a range of values. The range of values is limited, for example, in such a way that slight deviations in the collected first and / or second data can be taken into account and a static value can be assumed as a whole.

In some embodiments, it is assumed that it is for the traffic area section RS static friction values, so that a mean value of the filtered measuring points is determined. An error of the mean value can be determined from this, for example $$\mathrm{\Delta }{\mathrm{\mu }}_{\text{err}}=\frac{\sqrt{\frac{1}{n}{\displaystyle {\sum }_{i1}^n{\left(\overline{\mu }-{\mathrm{\mu }}_i\right)}^{\mathrm{2nd}}}}}{\sqrt{n}}.$$

In some embodiments, static coefficients of friction are not assumed. A time distance to the current determination time is determined. A weighting w _i can then be assigned to each measuring point on the basis of the respective distance, with more recent coefficients of friction being weighted higher than older coefficients of friction in the first data. A time correlation can be formed, for example, in order to derive these weightings. For example: $$w_i=\frac{\frac{1}{t_{diff}}}{{\displaystyle \sum \frac{1}{t_{diff}}}}.$$

A weighted average value can then be determined via the weightings wi, which is the coefficient of friction of the traffic area section RS for a respective intermediate layer. This coefficient of friction applies to the time parameter, which is optimized as described above. For example, the following applies to the weighted average of the coefficient of friction: $$\mathrm{\mu }={\displaystyle \sum _{i=1}^n{w}_i\ast {\mathrm{\mu }}_i.}$$

A one-sigma error of the mean value can also be determined, for example according to, using one-sigma confidence intervals of the measurement points $$\mathrm{\Delta }{\mathrm{\mu }}_{err}=\frac{{\displaystyle {\sum }_{i=1}^n{w}_i\ast \mathrm{\Delta }{\mathrm{\mu }}_i}}{\sqrt{n}}.$$

The determined mean value with the determined error can be checked by means of a hypothesis test, for example a three-sigma test, a specification of a significance level, etc., to determine whether these are with sufficient certainty over a limit value around the vehicle V to provide the most reliable coefficient of friction µ.

2nd summarizes a method for providing the coefficient of friction µ for the traffic area section RS to the vehicle V together. In one step S1 the first data, the vehicle's coefficient of friction data of the traffic area section, is recorded RS include. In one step S2 the second data, the environmental condition data of the traffic area section, is recorded RS include. In one step S3 weather data corresponding to the traffic area section is obtained RS are assignable to a determination time. In one step S4 the coefficient of friction µ is determined for the determination time on the basis of the first data and the second data, the condition being fulfilled that the coefficient of friction µ from the first data in the case of environmental condition data from the second data, which correlate with the weather data obtained, based on one certain time parameters lies within a range of values.

QUOTES INCLUDE IN THE DESCRIPTION

This list of documents listed by the applicant has been generated automatically and is only included 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

• WO 2016/120092 A1 [0003]

Claims (10)

Method for providing a coefficient of friction (µ) for a traffic area section (RS) to a vehicle (V), comprising the steps: - Acquisition of first data, which includes vehicle coefficient of friction data of the traffic area section (RS), - Acquisition of second data, which include environmental condition data of the traffic area section (RS), Obtaining weather data which can be assigned to the traffic area section at a determination time, and Determining the coefficient of friction (µ) for the determination time on the basis of the first data and the second data, the condition being fulfilled that the coefficient of friction (µ) from the first data with environmental condition data from the second data, which correlate with the weather data obtained, based on a certain time parameter lies within a range of values. Procedure according to Claim 1 , characterized in that the determination of the coefficient of friction comprises a nearest neighbor classification based on the first and / or second data. Procedure according to Claim 1 or 2nd , characterized in that determining the coefficient of friction (µ) comprises determining an intermediate layer of the traffic area section (RS) from the first and / or second data. Method according to one of the preceding claims, characterized in that the determination of the coefficient of friction (µ) comprises a determination of an average error (Δµ _err ) of the coefficient of friction (µ) from the first and / or second data. Method according to one of the preceding claims, characterized in that the determination of the coefficient of friction (µ) comprises a weighting of possible coefficients of friction, the coefficient of friction data included in the first data being weighted more strongly if these are less in time from the determination time. Procedure according to Claim 5 , characterized in that a weighted mean value of the coefficient of friction (µ) is determined from the weighting. Method according to one of the preceding claims, characterized in that the time parameter is determined by an optimization method. Procedure according to Claim 8 , characterized in that the optimization method is selected from: grid search, random search, Bayesian optimization and the like. Method according to one of the preceding claims, characterized in that a hypothesis test of the determined coefficient of friction is carried out before providing the coefficient of friction (µ). System for providing a coefficient of friction (µ) for a traffic area section (RS) to a vehicle, with a data interface and a data processing device (110), which is set up to - to acquire first data which include the vehicle's friction coefficient data of the traffic area section (RS), to acquire second data, which include environmental condition data of the traffic area section (RS), Obtaining weather data which can be assigned to the traffic area section at a determination time, and - Determine the coefficient of friction (µ) to be provided for the determination time on the basis of the first data and the second data, the condition being fulfilled that the coefficient of friction is obtained from the first data in the case of environmental condition data from the second data, which correlate with the weather data obtained to a certain time parameter lies within a range of values.

Images