DE102019205892B4 - Method for operating a motor vehicle and motor vehicle designed to carry out such a method - Google Patents

Abstract

Method for operating a motor vehicle (10), which comprises at least one vehicle system (12) which is designed to control the motor vehicle (10) at least partially autonomously, with a predefined driving behavior mode (16) being activated in the vehicle system (12), which has at least one Control parameter (17) which defines a driving behavior in a predetermined geographical area (20) and which is taken into account when the motor vehicle (10) is controlled at least partially autonomously, comprising the steps of:- detecting a current geographical position (11c) of the motor vehicle (10) (S9);- Determination of the predetermined geographical area (20) in which the motor vehicle (10) is currently located, based on the detected geographical position (11c) (S10);- Determining which predetermined driving behavior mode (16) of several predetermined position-dependent Driving behavior modes (16) is assigned to the determined geographic area (20) (S11); - checking whether the determined Driving behavior mode (16) is already activated in the vehicle system (12) (S12); - If the determined driving behavior mode (16) is currently not activated in the vehicle system (12), deactivating the already activated driving behavior mode (16) and activating the determined driving behavior mode (16) (S14); and- at least partially autonomous control of the motor vehicle (10) taking into account the at least one control parameter (17) of the currently activated driving behavior mode (16) (S15); wherein the geographic area (20) for the respective driving behavior mode (16) is defined by the following steps:- providing vehicle data (30) from at least one vehicle that is operated by the respective driver and the vehicle data (30) with at least one acquisition device of the vehicle, the respective vehicle data (30) describing at least one driving maneuver of the at least one vehicle and additionally comprising location data (31) characterizing a location where the vehicle data (30) was recorded (S1);- determining the driving behavior data (40) and/or the road user behavior data ( 42) by applying a machine learning method to the provided vehicle data (30) (S2);- Recognizing at least one data pattern (50) in the determined driving behavior data (40) and/or the determined road user behavior data (42) by applying the method of machine learning on the determined driving behavior data (40) and / or di e determined road user behavior data (42) (S3);- defining an extension of the respective geographical area (20) taking into account the at least one recognized data pattern (50) and the vehicle data (30) included location data (31) (S4).

Description

The invention relates to a method for operating such a motor vehicle that includes at least one vehicle system that is designed to control the motor vehicle at least partially autonomously. The invention also relates to a motor vehicle with a vehicle system, the motor vehicle being designed to carry out said method.

A prerequisite for a fully autonomous driving function for a motor vehicle that can be used worldwide is that the driving function imitates the driving behavior of an average human driver during a fully autonomous trip. It is therefore necessary for the driving behavior of the fully autonomous motor vehicle to be based on local conditions and driving styles. It may be the case, for example, that a speed limit indicated on a road sign before a bend in the road is more serious in one country than it is in another country. This means that if a motor vehicle negotiates the curve at a speed corresponding to the specified maximum speed, the motor vehicle will experience higher lateral forces in the first-mentioned state than in the second-mentioned state. A human driver learns such regional peculiarities quickly and can adapt to them. During a next stay in the first-mentioned state, the human driver will therefore, for example, drive more carefully into a curve and drive there, for example, more slowly than the maximum speed specified there allows. In addition, the human driver will in particular avoid driving there at a higher speed than the displayed maximum speed allows. There are also various differences in driving styles and in an interpretation of traffic signs and traffic rules, which a driver usually already learns as part of his training in a driving school in the corresponding state or district. Because driving behavior in large urban areas, such as Paris or Beijing, differs significantly from driving behavior in more rural regions, such as in Rwanda or on a serpentine road in the Alps. It can therefore be assumed that there is no globally standardized model that would make fully autonomous driving possible everywhere.

Furthermore, it is assumed that a motor vehicle designed for fully autonomous driving no longer has a steering wheel or operating elements for other intervention options for a human during fully autonomous driving. This motor vehicle must therefore be able to drive the motor vehicle completely independently, that is to say fully autonomously. However, motor vehicles that have, for example, driver assistance systems that are designed at least for semi-autonomous driving must also be able to control the motor vehicle in such a way that there is no atypical behavior of the semi-autonomous motor vehicle, which, for example, can lead to accidents with other motor vehicles that are driven by manually controlled by human drivers. Such atypical behavior is particularly common in situations in which a manual driver would not brake his motor vehicle, but the at least partially autonomously driven motor vehicle would perform an emergency stop, for example due to traffic regulations or conflicting decisions.

In the DE 10 2012 016 802 A1 describes a method for controlling a vehicle system designed for autonomous operation of a motor vehicle. In this case, a current position of the motor vehicle is recorded and, depending on this position information, the permissibility of using a driver assistance system is checked. Depending on the position of the motor vehicle, for example, a maximum speed up to which autonomous driving is permitted, a spatial area in which autonomous driving is permitted, and limit values with regard to other parameters can be set.

the DE 10 2014 217 389 A1 shows a system for autonomous driving of a vehicle. Here, a message is transmitted to the vehicle, whereupon a driving state of the vehicle is determined based on whether the message was received and/or a content of the message was identified. A driving instruction is determined based at least in part on the driving condition.

the DE 10 2016 120 508 A1 FIG. 1 shows a system, method, and apparatus for predicting a driver's intention and future movement of a proximal vehicle, which is either an automated vehicle or a human-driven vehicle. The goal here is for autonomously driving vehicles to be able to predict the movement of surrounding vehicles, even if they are being driven by a human. This is based on a detected movement of the other vehicles, body language from drivers of the other vehicles, a direction indicator of the vehicles or their absence and wireless communications, for example between individual vehicles.

the DE 10 2016 219 121 A1 shows a method for selecting autonomous driving modes in an at least partially autonomously driving vehicle. The current location of the vehicle and information on area-specific driving behavior are determined and the driving mode is automatically adapted to the area-specific driving behavior.

the U.S. 2017/0 106 876 A1 shows a method in which a manually controllable vehicle automatically switches from a manual driving mode to an automatic driving mode if it is determined that a driver of the vehicle is unqualified to continue to manually control the vehicle.

the U.S. 2015/0 166 062 A1 shows a control system for a vehicle, which is designed to adapt a control of the vehicle to a driving behavior of other vehicles in an environment of the vehicle and a driving behavior in a specific geographical area.

It is the object of the invention to provide a solution by means of which, in the case of at least partially autonomous driving of a motor vehicle, regional driving behavior peculiarities are taken into account.

This object is solved by the subject matter of the independent patent claims. Advantageous configurations with expedient and non-trivial developments of the invention are specified in the dependent claims, the following description and the figures.

A motor vehicle is operated using the method according to the invention. The motor vehicle comprises at least one vehicle system, this vehicle system being, for example, a central vehicle controller or a control device of a driver assistance system of the motor vehicle, each of which is designed to control the motor vehicle with regard to longitudinal guidance and/or lateral guidance. The at least one vehicle system is designed to control the motor vehicle at least partially autonomously, with a predefined driving behavior mode being activated in the vehicle system. The predefined driving behavior mode includes at least one control parameter that defines a driving behavior in a predefined geographical area and is taken into account in the at least partially autonomous driving of the motor vehicle. The geographic area is, for example, a particular state or city. Within the meaning of the invention, driving behavior is understood to mean behavior of a driver that is imitated by the vehicle system when the motor vehicle is driven manually in road traffic in the geographical area. A total of technical properties of the motor vehicle that occur while the driver controls the vehicle are understood as driving behavior. The driving behavior can specify, for example, whether speed limits, for example a typical underspeed and/or overspeed of the driver, on an expressway, country road and/or or in inner-city traffic, are taken into account. The driving behavior can thus be quantified using the control parameter, which includes, for example, that the average driver typically always drives 10 kilometers per hour slower on the expressway than the maximum speed specified on this expressway in, for example, the specific area allows, for example there the expressway has more bends than a typical expressway, which is the basis for the posted speed limit. In this example, the control parameter therefore includes the information on the underspeed of 10 kilometers per hour on expressways, which reflects the typical driving behavior. This information contained in the control parameter is then taken into account when the motor vehicle is actually controlled during its at least partially autonomous driving, which is carried out, for example, by an autopilot. The driving behavior mode thus includes the autopilot, which represents the vehicle system of the motor vehicle in this example, always falling below the locally specified maximum speed by 10 kilometers per hour and thereby adapting to the typical driving behavior in the specified geographical area.

The method according to the invention comprises the following steps: First, a current geographic position of the motor vehicle is recorded. This detection takes place, for example, by means of a detection device in the motor vehicle, which is provided with data from a global navigation satellite system (GNSS), for example. The predetermined geographical area in which the motor vehicle is currently located is then determined on the basis of the recorded geographical position. This determination can take place, for example, in an evaluation device of the vehicle system or in a control device of the motor vehicle that is coupled to the vehicle system. In this method step it can be determined, for example, that the motor vehicle is located in a specific country, for example Portugal, based on its current geographic position. Then, for example also by the evaluation unit, it is now determined which predefined driving behavior mode from a plurality of predefined position-dependent driving behavior modes is assigned to the determined geographical area. It is therefore determined which predefined driving behavior mode with its respective at least one control parameter defines the driving behavior in the ascertained geographic area in which the motor vehicle is located. For example, a respective driving behavior mode can be differentiated for individual states, such as Portugal, Germany and France, in which case, in addition to the state-based assignment of the geographic area, the area of the state can also be subdivided into individual urban areas, for example. For example, a driving behavior mode for large cities, such as Paris or Berlin, can be specified separately from the driving behavior modes for France or Germany.

It is then checked whether the determined driving behavior mode is already activated in the vehicle system. If the determined driving behavior mode is not currently activated in the vehicle system, the already activated driving behavior mode is deactivated and the determined driving behavior mode is activated. In the vehicle system, therefore, only one driving behavior mode is activated at any one time. An interaction of several driving behavior modes, such as the driving behavior mode for France and the driving behavior mode for Paris, for example, within the city of Paris is therefore not provided, but a separate driving behavior mode is specified for Paris. For example, if the motor vehicle is driving in a border area between two predetermined geographical areas, for example because it has entered the urban area of Paris from a highway in rural France, it may be that the driving behavior mode for France is still activated in the motor vehicle and not the predetermined driving behavior mode for the city of Paris. If this is the case, after it has been determined that the motor vehicle is actually located in Paris, the previously activated driving behavior mode for France is deactivated and the determined driving behavior mode, ie the driving behavior mode for the city of Paris, is activated.

In a next step, the motor vehicle is controlled at least partially autonomously, taking into account the at least one control parameter of the currently activated driving behavior mode. In the example mentioned, the at least partially autonomous control of the motor vehicle will take place after the activation of the determined driving behavior mode for the city of Paris, taking into account the control parameter of the driving behavior mode for Paris. The method described is based on the knowledge that, for example, within a city, the typical driving behavior of drivers of motor vehicles is often different than is observed, for example, in a rural area and thus in other parts of the state in which the city is located. So that a motor vehicle that is driving fully autonomously, for example, does not represent an obstacle or annoyance for other road users who are still manually controlling their motor vehicle, for example, the driving behavior mode for the urban area can include at least one control parameter other than the driving behavior mode for the rural area of the same state.

With the help of the driving behavior mode assigned to the current geographical area, it can thus be achieved that the motor vehicle behaves in its at least partially autonomous driving in exactly the same way as an average human driver actually behaves in the current geographical area. This reduces, for example, the probability and frequency of traffic complications due to an unexpected and/or atypical behavior of the partially or fully autonomously driving motor vehicle, for example, since it adapts to the local driving conditions. The method according to the invention thus allows the local driving behavior to be taken into account and also ensures that the appropriate driving behavior mode is always activated at a specific point in time and at a specific location. The method according to the invention therefore has the advantage that regional characteristics of driving behavior are taken into account when driving the motor vehicle at least partially autonomously.

The invention deals with the question of which subsequent steps are used to define the geographic area for the respective driving behavior mode. First of all, vehicle data is provided by at least one vehicle that is operated by the respective driver and collects the vehicle data with at least one detection device of the vehicle. The at least one detection device is, for example, a sensor device of the vehicle, such as a camera system, a radar device, a laser scanner and/or an infrared camera system. Alternatively or additionally, the detection device can be, for example, a temperature sensor, a speed sensor and/or a navigation system of the motor vehicle, which provides position data of the vehicle. The vehicle data is stored, for example, in a memory unit of the respective vehicle and, for example, is transmitted to an external computing device via a communication link after a trip. This computing device is designed, for example, as a server device. The respective vehicle data ten describe at least one driving maneuver of the at least one vehicle and also include location data that characterize a location where the vehicle data was recorded. A respective acquisition location in the form of location data is also stored for the respective vehicle data, so that conclusions can be drawn as to where the respective vehicle data was acquired by the acquisition device. If the vehicle data is, for example, speed data that indicates the speed at which the vehicle was driving, the corresponding data contain the respective location data in addition to the respective speed values, which indicate where the respective speed value was recorded. The computing device can have a total of vehicle data from a number of vehicles, which were recorded and stored, for example, during a number of driving maneuvers during which the respective vehicle was controlled by at least one respective driver.

In a next step, the driving behavior data and/or the road user behavior data are determined by the computing device. The driving behavior data and/or road user behavior data is determined by applying a machine learning method to the vehicle data provided. This driving behavior data then contains, for example, information about by what percentage or by how many kilometers per hour the driver drove too slowly and/or too fast on a road of a certain type of road, such as an expressway, which can be determined on the basis of the vehicle data and the location data assigned to the vehicle data can be determined. This is because the vehicle data include, for example, the recorded speed of the vehicle and image data recorded by the vehicle's camera system from a street sign with the specified maximum speed on the street. Data can thus be obtained that characterize how the at least one driver of the vehicle behaves while driving. In addition or as an alternative to this, data can be obtained which describe how at least one road user behaves. The information about the other road users can be determined, for example, using data recorded with the camera system, the laser scanner, the radar device and/or the infrared camera system, which are also provided as vehicle data by the at least one vehicle.

In a next step, at least one data pattern is recognized in the determined driving behavior data and/or the determined road user behavior data by applying a machine learning method to the determined driving behavior data and/or the determined road user behavior data. A method of machine learning is therefore used both to determine the driving behavior data and/or the road user behavior data and to recognize at least one data pattern in the driving behavior data determined in this way and/or the determined road user behavior data. This is a method of artificial intelligence, which is carried out, for example, by using a respective artificial neural network. It is therefore provided that, for example, an artificial neural network is applied to the vehicle data provided to the computing device.

An artificial neural network is a network of artificial neurons. In neuroinformatics, artificial neural networks are used as artificial intelligence, since they are suitable, for example, for evaluating this data without detailed knowledge of how it works or of a communication relationship between individual subsets of data. An artificial neural network is an independently learning computer program that can recognize complicated regularities and relationships in large amounts of data in a relatively short time. A neural network is based on an abstract model of interconnected neutrons, so that an artificial neural network is modeled on the structure of a biological brain, at least to a certain extent. To put it simply, the structure and functioning of an artificial neural network can be described with a network of neurons, which each receive information from the outside or from other neurons and pass it on to other neurons in a modified form or output it as the end result. Here, for example, positive or negative weightings can be made, which represent an exciting or inhibitory influence of the neurons on one another. The knowledge and thus the artificial intelligence of an artificial neural network is ultimately stored in the connections between the individual neurons and their weightings. During training of the neural network, these weights of the connections change depending on the applied learning rules and the results achieved.

In other words, by applying the artificial neural network to the vehicle data, the driving behavior data and/or the road user behavior data can be determined from this vehicle data within a short period of time. Except conspicuous data packets, the so-called data patterns, can be identified in the determined driving behavior data and/or the road user behavior data. In a simplified example, the differences between the permitted driving speed and the actual maximum speed determined when determining the driving behavior data can be plotted as a function of a number of vehicles for which such a difference was observed. A function graph resulting from this has, for example, several maxima and minima. The drivers whose data are assigned to a respective area of the maxima of such a function graph consequently show a similar behavior with regard to a specific underspeed, for example due to a winding road. Such a maximum occurs, for example, when the speed falls below the permissible maximum speed of 10 kilometers per hour.

In a next step, an extension of the respective geographic area is defined, taking into account the at least one recognized data pattern and the location data included in the vehicle data. If, for example, it is now determined that all the vehicle data that indicate that the speed limit is 10 kilometers per hour below the permitted speed limit was found in a specific, limited area, for example on various roads in a mountainous region, the extent of this mountainous region can be defined as a geographical area . By using artificial intelligence methods, conclusions can be drawn about relationships or similar behavior of individual vehicle data subsets, driving behavior data subsets and/or road user behavior data subsets, so-called clusters. These similarities can ultimately be used to identify a geographic area where drivers behave similarly. Ultimately, for example, coordinates that describe the geographical area are inferred. This makes it possible, on the basis of vehicle data provided by, for example, several vehicles, to ultimately use machine learning methods to differentiate, in particular in a short period of time and thoroughly, individual geographical areas that differ in the driving behavior of drivers of vehicles and/or drivers that are usual there distinguish between other road users. Ultimately, this helps to enable a suitable geographical subdivision of the world traveled by vehicles, so that ultimately, based on the recorded extents of the respective geographical areas, it is recognized how many and which specific driving behavior modes must be determined and made available for motor vehicles in order to enable that with at least semi-autonomous driving, local peculiarities in driving behavior can be taken into account.

The invention also includes configurations that result in additional advantages.

In an advantageous embodiment of the invention, it is provided that the at least one control parameter is determined from data of at least one of the following types of data: Traffic rule data that characterizes at least one traffic rule currently in force in the geographical area, driving behavior data that describes the situation for a driver of a vehicle in the geographical Characterize typical driving behavior in the area, road user behavior data that characterizes a behavior that is typical for at least one road user in the geographical area, interpretation data that characterizes an interpretation rule for at least one traffic symbol in the geographical area and / or vehicle characteristics that characterize the motor vehicle.

The at least one control parameter on which the driving behavior mode is based is thus determined, for example, from data that describes the applicable traffic regulations in the geographical area. This traffic regulation data includes, for example, information on left-hand or right-hand traffic that is prescribed in the geographic area. In addition, this data includes legal regulations regarding traffic routing in the geographic area. The driving behavior data, in turn, includes information about how the driver of any vehicle in the geographic area typically behaves. In this case, the driving behavior data can characterize the respective typical driving behavior of several drivers of a total of several respective vehicles in the geographical area. The driving behavior data therefore include, for example, how likely it is that the driver will stop in the geographic area for a pedestrian in front of a traffic light that is green for the vehicle and/or a zebra crossing and/or by what percentage he will be slower, for example, within an urban area and/or drives faster than the specified maximum speed. Stopping or at least reducing the speed of the vehicle by the driver for a pedestrian despite a green traffic light for the vehicle can occur in certain geographical areas as a reaction to impatiently waiting pedestrians, especially in urban areas, and can be taken into account by the at least one control parameter will. The Road User Association On the other hand, stopping data describe typical behavior of other road users in the geographical area, such as how far these other road users keep from the vehicle of the driver providing the driving behavior data, or how other road users are in a dangerous situation, for example a child running onto the road , react. The interpretation data includes information on how traffic signs, lane markings and thus different traffic symbols are recorded in the geographic area and affect driving behavior. Here, for example, interpretations of traffic rules that are not written down in a set of rules but that are already learned in the driving school in the geographical area, for example, are taken into account, which are reflected in the actual driving behavior of drivers in the geographical area. The vehicle characteristic data include, for example, a model type of the motor vehicle, such as its length and width, and characteristic data with regard to the vehicle system and/or, for example, a drive of the motor vehicle. This information is particularly relevant, for example, when driving on a narrow street or in a parking lot.

The at least one control parameter is thus based on the information that ultimately contributes to defining the driving behavior in the geographical area. The at least one control parameter can be determined by means of a suitable calculation and weighting of the information that can be taken from the individually named data. All available data regarding the behavior of manual drivers, other road users, the existing traffic regulations, the interpretation of traffic symbols and information about the motor vehicle itself are therefore available to determine the control parameter, so that a reliable control parameter that describes the driving behavior that is typical for the geographical area can be determined.

A particularly advantageous embodiment of the invention also provides that defining the at least one control parameter of the predefined driving behavior mode for the respective geographical area includes the following steps: First, a real behavior model is defined for the respective geographical area, taking into account the determined driving behavior data and/or road user behavior data, assigned to each specified geographical area. Ultimately, a model is determined that describes how a typical driver of a vehicle behaves in the geographical area. This model can be used, for example, to simulate the average driving behavior of a human driver for the respective geographic area. The real behavior model is also defined, for example, in the external computing device and preferably not within the motor vehicle itself.

The computing device is also provided with the traffic regulation data, interpretation data and/or vehicle characteristic data that are associated with the respective specified geographical area or the respective motor vehicle. An ideal behavior model is then defined for the at least one motor vehicle in the defined respective geographical area, taking into account the traffic control data, interpretation data or vehicle characteristic data provided. The information about the traffic regulations in force in the geographical area, the interpretation rules that prevail in the geographical area, and the vehicle characteristics that describe the motor vehicle in which the respective driving behavior mode is to be activated are taken into account in order to achieve an ideal and thus to determine the driving behavior model that corresponds to the legally prescribed behavior, which is referred to as the ideal behavior model.

The at least one control parameter for the respective geographic area is then determined from the real behavior model and ideal behavior model defined for the respective defined geographic area by applying a predetermined weighting criterion. All the data or at least parts of the data that have already been mentioned above as data from which the at least one control parameter is determined are therefore used to determine the control parameter. With the help of an evaluation system, which is provided by the weighting criterion, a comparison is made between the real behavior model and the ideal behavior model, from which the driving behavior mode is ultimately derived. For example, no arbitrarily serious driving maneuvers that do not comply with the applicable traffic rules, which are stored in the real behavior model, for example, are transferred to the driving behavior mode, but the actually observed behavior is compared with the legally prescribed behavior contained in the ideal behavior model. This makes it possible for a driving behavior that is dependent on the specified geographical area to be quantified in such a way that the at least one control parameter is provided, based on which the vehicle system can ultimately control the motor vehicle at least partially autonomously. By choosing the weight The criterion consequently makes it possible for a driving behavior that is as realistic as possible, but still sufficiently compliant with the rules, to be implemented in the at least partially autonomously driving motor vehicle using the driving behavior mode. An irritating and impeding of manually driving road users by a motor vehicle that is controlled according to the driving behavior mode is therefore particularly small, since the driving behavior mode is based directly on the actual driving behavior in the specific geographical area.

A further embodiment provides that the weighting criterion is determined by a simulation and/or a shadow mode function. The evaluation criterion is thus determined using a learning system. For example, when simulating a detected driving maneuver based on the vehicle data by simulating the driving behavior in the corresponding situation according to the real behavior model and the ideal behavior model, it can be determined which of the models is weighted and how heavily in order to achieve a desired functional quality, i.e. a desired behavior during the driving maneuver , to reach. When determining the weighting criterion using the shadow mode function, a first version of the weighting criterion is provided in the motor vehicle, the motor vehicle itself driving without taking into account the weighting criterion and any driving behavior mode. However, an evaluation unit of the motor vehicle determines, for example, an expected behavior of the motor vehicle for each manually driven driving maneuver when the driving behavior mode is activated. The driving behavior mode is therefore only activated in the background, so that the weighting criterion can be adjusted by comparing the driving behavior results of this driving behavior mode, which works more or less in the shadows, with the actual driving behavior of the driver. The weighting criterion is thereby adjusted in such a way that an optimum correspondence can be achieved between the simulated activation using the driving behavior mode and the actual driving behavior of the manual driver. The mode referred to here as the shadow mode function is also often referred to with the English term shadow mode. The suitable weighting criterion is thus established in a particularly simple manner by the simulation and/or the shadow mode function.

A further embodiment of the invention provides that when determining a driving route from the current geographic position to a specified target position of the motor vehicle, it is checked whether the motor vehicle drives through several specified geographical areas when traveling along the determined driving route. For example, as part of determining the driving route, it is already determined whether the motor vehicle will be in more than one predetermined geographical area on this driving route. If several predetermined geographical areas are driven through, the driving behavior modes that are associated with the respective predetermined geographical area are downloaded from the external device via a communication link of the motor vehicle. It is therefore ensured that all driving behavior modes required for this driving route are already available in the motor vehicle when the journey starts, so that, depending on the current position of the motor vehicle, which is recorded, for example, at predetermined time intervals, it is possible to switch between the individual driving behavior modes that have already been downloaded. It is possible, for example, that when driving from one city to another, both the respective driving behavior modes for the starting city and the destination city as well as corresponding driving behavior modes for a cross-country trip between the two cities are downloaded. It is therefore not necessary for all driving behavior modes available for, for example, an entire country, continent or even the entire world to always be stored in the vehicle system, in a control device coupled to the vehicle system or in a memory device of the motor vehicle coupled to the vehicle system. Rather, only the driving behavior modes can always be stored in the motor vehicle, for example by temporarily storing them in the memory device that the motor vehicle actually needs during the planned route. Especially with regard to fully autonomous motor vehicles, it is advantageous that the complete route is already known when starting the journey, so that it is ensured, for example, independently of a regionally available or accessible mobile data network, that the vehicle system can access all driving behavior modes required for this route as soon as they are activated are. As a result, the use of the method for controlling the motor vehicle is also reliably possible in remote regions which, for example, do not have nationwide coverage with a mobile data network. In this context, one advantageous embodiment provides for a transition from the previous driving behavior mode to the determined driving behavior mode to take place when the motor vehicle is traveling over a predefined transition route. There is therefore not necessarily an abrupt transition from the previous driving behavior mode to the current driving behavior mode predetermined on the basis of the current position of the motor vehicle. Rather, it can be provided that, for example, a transitional area of, for example, 5 kilometers in length between two geographically adjacent areas areas is specified. If it is detected that the motor vehicle is in this transition area, a continuous transition from the previous driving behavior mode to the new driving behavior mode in the adjacent and soon to be reached geographical area is activated for the 5 km transition route through the transition area. This avoids the occupants of the motor vehicle, which is driving fully autonomously, for example, being surprised by a driving style that suddenly changes abruptly. Rather, the occupants of the motor vehicle can continuously get used to the now changed driving behavior of the motor vehicle and adjust to it.

Alternatively, an advantageous embodiment of the invention provides for the determined driving behavior mode to be downloaded from an external device via a communication link to the motor vehicle, with it being determined that the motor vehicle is located in the geographical area assigned to the determined driving behavior mode. In this alternative embodiment, the driving behavior mode is always first downloaded into the motor vehicle when it is determined that the motor vehicle is located in the associated geographic area. The prerequisite for this is that a communication connection with the external device in which the respective driving behavior modes are stored can be established and maintained. This enables the advantage that the large amounts of data for the respective driving behavior modes do not have to be stored in the motor vehicle at all times, since the corresponding data are only downloaded when they are actually required by the vehicle system. As a result, the requirement of the method for a storage capacity of the motor vehicle is kept as low as possible.

In a particularly advantageous embodiment of the invention, it is provided that the geographic area includes at least one of the following areas: a state, a federal state, a district and/or a settlement. The geographic area thus encompasses a relatively large area, particularly when it encompasses a single state, i.e. a single country of the world. As an alternative or in addition to this, a separate driving behavior mode can be provided for different driving behaviors for the individual member states, such as federal states, federal states, at least partially autonomous regions or other political sub-states. It may also be necessary for a specific district, such as a rural district, or a metropolitan area, such as a large city, a small town or a village, i.e. for a settlement, to have its own driving behavior mode, since, for example, the typical Driving behavior in these small geographic areas is strikingly different from surrounding areas. Depending on how locally differentiated the driving behavior is, different types of geographical areas are possible, so that a sufficiently differentiated division into geographical areas can always take place.

According to the invention, a motor vehicle with a vehicle system is also provided, the vehicle system being designed to control the motor vehicle at least partially autonomously. The vehicle system or a control device coupled to the vehicle system is designed to carry out a method as described above. The preferred embodiments presented in connection with the method according to the invention and their advantages apply correspondingly, to the extent applicable, to the motor vehicle according to the invention. For this reason, the corresponding developments of the motor vehicle according to the invention are not described again here. In this case, the vehicle system can be a central control device of the motor vehicle, for example a central vehicle controller. However, the vehicle system can also be a control device of a driver assistance system. In this case, the method is partially carried out by the control device coupled to the vehicle system. The control device is then designed to control, for example, a longitudinal control and/or lateral control of the motor vehicle, taking into account the control commands of the control device of the driver assistance system.

The invention also includes the vehicle system and the control device for the motor vehicle coupled to the vehicle system. The vehicle system or the control device each have a processor device that is set up to carry out an embodiment of the method according to the invention. For this purpose, the processor device can have at least one microprocessor and/or one microcontroller. Furthermore, the processor device can have program code which is set up to carry out an embodiment of the method according to the invention when executed by the processor device. The program code can be stored in a data memory of the processor device.

The invention also includes the combinations of features of the described embodiments.

• 1 eine schematische Darstellung von drei Kraftfahrzeugen in verschiedenen geografischen Gebieten,
• 2 in schematischer Darstellung einen Signalflussgraph für ein Festlegen eines geografischen Gebiets;
• 3 in schematischer Darstellung einen Signalflussgraph für ein Festlegen von zumindest einem Ansteuerparameter für einen Fahrverhaltensmodus für ein bestimmtes geografisches Gebiet; und
• 4 in schematischer Darstellung einen Signalflussgraph für ein Verfahren zum Betreiben eines Kraftfahrzeugs.

Exemplary embodiments of the invention are described below. For this shows:

• 1 a schematic representation of three motor vehicles in different geographic areas,
• 2 a schematic representation of a signal flow graph for defining a geographic area;
• 3 a schematic representation of a signal flow graph for specifying at least one control parameter for a driving behavior mode for a specific geographic area; and
• 4 a schematic representation of a signal flow graph for a method for operating a motor vehicle.

The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiment each represent individual features of the invention to be considered independently of one another, which also develop the invention independently of one another and are therefore also to be regarded as part of the invention individually or in a combination other than that shown. Furthermore, the described embodiments can also be supplemented by further features of the invention that have already been described.

Elements with the same function are each provided with the same reference symbols in the figures.

In 1 A plan view of a landscape is sketched in which three automobiles 10, referred to as automobiles 10a, 10b and 10c, are outlined. The motor vehicles 10a, 10b and 10c each include a vehicle system 12 which is designed at least for partially autonomous control of the respective motor vehicle 10a, 10b, 10c. The motor vehicles 10a, 10b and 10c are driving on a road 14, the road 14 leading through two states which are separated from one another by a state border 22. The motor vehicle 10a is located in a first state, i.e. in a geographical area 20, which is referred to as geographical area 20a, whereas the motor vehicle 10b is located in the second state beyond the state border 22 in the direction of travel and in a geographical area associated with the second state Area 20b drives. The third motor vehicle 10c is located in a city referred to as geographic area 20c. A city boundary 24 defines the boundary between this geographic area 20c and the second state geographic area 20b surrounding the city.

In the respective vehicle system 12, a predefined driving behavior mode 16 (shown with the reference number 16 in 3 ) activated, which activates at least one control parameter 17 (shown with the reference number 17 in 3 ) includes. The control parameter 17 defines a driving behavior in the respective predefined geographical area 20a, 20b, 20c and is taken into account when the motor vehicle 10a, 10b, 10c is controlled at least partially autonomously. A change between such driving behavior modes 16 can take place, for example, when the state border 22 is reached, i.e. for example whenever the motor vehicle 10a reaches the state border 22 and then, as outlined here by the motor vehicle 10b, to the second state, which corresponds to the geographical area 20b is assigned, drives. As an alternative to this, the transition between the two driving behavior modes 16 mentioned can take place smoothly, namely over a predetermined transition route 15, as shown in 1 is sketched using dashed lines encompassing a portion of the road 14 . Within this transition route 15, it is possible, for example, to switch continuously from the at least one control parameter 17 of the driving behavior mode 16 assigned to the geographical area 20a to the driving behavior mode 16 assigned to the geographical area 20b.

In 2 individual method steps are outlined that take place in order to define the geographical area 20c. The other geographic areas 20a and 20b are defined analogously. In a first step S1, vehicle data 30 from at least one vehicle is made available to a computing device. This vehicle is operated by a driver and the vehicle data 30 is recorded with at least one recording device of the vehicle, such as a camera system, a radar device, a speed measuring sensor and/or a navigation system. The respective vehicle data 30 describe at least one driving maneuver of the at least one vehicle. For example, vehicle data 30 describes the speed at which the vehicle was traveling on road 14 . In addition to this, the vehicle data 30 includes location data 31 which characterize a location where the vehicle data 30 was recorded. Consequently, for each speed value, it is known where this speed value was detected by the corresponding detection device of the vehicle, that is to say in this example by the speed measuring sensor of the vehicle. The vehicle data 30 can be provided, for example, by a fleet of test vehicles, but also by any vehicles controlled by manual drivers whose drivers have agreed to provide their data for evaluation and optimization purposes.

In a next step S2, driving behavior data 40 and/or road user behavior data 42 are determined by applying a machine learning method, for example by applying an artificial neural network, to the vehicle data 30 provided. Driving behavior data 40 is understood to be data that characterizes the driving behavior that is typical for the driver of the vehicle. Traffic user behavior data 42 refers to data that characterizes a behavior that is typical for at least one traffic user. For example, driving behavior data 40 quantifies how often at least one vehicle stops when a person waiting is detected at a traffic light that is green for the vehicle and/or at a zebra crossing, what distance it maintains from cyclists riding at the side of the road, what speed it is driving on road 14 with a speed limit of 30 kilometers per hour and/or which braking acceleration is selected before a red traffic light in order to brake the vehicle to a standstill. In general, driving behavior data 40 can therefore express a frequency or probability for a specific driving behavior. The road user behavior data 42 is, for example, data that specifies a frequency of emergency stops due to pedestrians who, for example, still walk across the road 14 when a pedestrian traffic light is red, and/or the speeds of other vehicles in an area surrounding the vehicle data 30 providing Vehicles that are also driving on road 14, for example. In general, road user behavior data 42 can therefore express a frequency or probability for a specific driving behavior.

It can therefore be determined, for example, at what speed reduction and/or speed increase the driver of the at least one vehicle is driving on the road 14 . For this purpose, image data from a road sign, which shows the maximum permissible speed on the road 14 and/or information about the maximum permissible speed stored in a navigation system of the vehicle, provided by the vehicle's camera system, for example, may be taken into account.

In a step S3, at least one data pattern 50 is recognized in the determined driving behavior data 40 or the determined road user behavior data 42 by applying the machine learning method to the determined driving behavior data 40 or the determined road user behavior data 42 . In step S3 it is thus determined, for example, which number 41 of vehicles has which driving behavior data 40, that is to say it is entered, for example, which number 41 of vehicles drove at which reduced speed. In this example, for example, there are two maxima in the in 2 function graphs shown under step S3, which can be clearly interpreted as a recognizable data pattern 50. In particular when using artificial neural networks on the vehicle data 30 and the driving behavior data 40 and/or road user behavior data 42 determined from the vehicle data 30, regularities in this large amount of data can ultimately become known. The pattern recognition thus relates to the most frequent behaviors, which are recognized as a local maximum in the respective frequency distribution.

In a next step S4, taking into account the at least one recognized data pattern 50 and the location data 31 recorded from the vehicle data 30, an extent of the respective geographic area 20c is defined. If, for example, the driving behavior data 40 associated with the respective data patterns 50 are plotted for different geographic coordinates, which in this example are each assigned to an x-value and a y-value, then individual areas appear, the two maxima of the function graph determined in step S3 can be assigned. Consequently, an actual geographic extent of two cities, each referred to as geographic area 20c, can be ascertained. In these, the driving behavior behaves significantly differently or noticeably differently than in the surrounding area of the second state, ie geographical area 20b. Such individual geographic regions 20a, 20b, 20c determined as a function of various parameters can thus be used to identify the regions in which the driver of the at least one vehicle typically behaves at least similarly, and their respective extent can be determined. These geographic areas 20a, 20b, 20c may be respective states, as is the case for geographic areas 20a and 20b in this example, a state, district and/or settlement, as is in this example for the geographic area 20c is the case.

In 3 the next procedural steps are described that are required to set the at least one control parameter 17 of the driving behavior mode 16 for the geographic area 20c. In this case, in a step S5, a real behavior model 60 is first defined for the specific geographic area 20c. This takes place taking into account the determined driving behavior data 40 and/or road users behavioral data 42 associated with the geographic area 20c determined as described above. Analogously to the steps now described below, the respective at least one control parameter 17 for the other geographical areas 20a and 20b can also be determined. In a next step S6, the computing device, by means of which, for example, the in 3 described method steps were carried out, traffic regulation data 44, interpretation data 46 and / or vehicle characteristic data 48 are provided. The traffic regulation data 44 characterizes at least one traffic regulation currently in force in the geographical area 20c. The interpretation data 46 include an interpretation rule for at least one traffic symbol in the geographic area 20c. The vehicle identification data 48 characterize the motor vehicle 10a, 10b, 10c, for example its respective length and width and a respective vehicle model type. In a next step S7, an ideal behavior model 70 for the at least one motor vehicle 10a, 10b, 10c in the specified geographical area 20c is defined from the provided traffic regulation data 44, interpretation data 46 and/or vehicle characteristic data 48. It is thus determined how motor vehicle 10a, 10b, 10c should behave, taking into account, for example, the legally prescribed behavior in geographical area 20c.

In contrast, the real behavior model 60 determined in step S5 reflects how a real driver in the geographical area 20c typically behaves. In a next method step S8, the real behavior model 60 and the ideal behavior model 70 are now calculated with one another by applying a predetermined weighting criterion in such a way that the at least one control parameter 17 is determined for the geographic area 20c. Consequently, the driving behavior mode 16 for the geographical area 20c is hereby defined at least with regard to the control parameter 17 for the speed of the motor vehicle 10a, 10b, 10c. Taking into account data of the various vehicle data types mentioned above and taking into account the as in 2 Described defined geographical areas 20a, 20b, 20c can thus be determined several control parameters 17, so that overall a comprehensive driving behavior mode 16 is provided that the driving behavior in the geographical area 20a, 20b, 20c so close to a real driving behavior of drivers in this geographical area 20a, 20b, 20c as possible. In this case, the weighting criterion can be determined by a simulation and/or a shadow mode function. Ultimately, the weighting criterion reflects how compliant or in what way individual traffic rules are interpreted in the respective geographical area 20a, 20b, 20c, so that ultimately the driving behavior mode 16 for a partially autonomously or fully autonomously driving motor vehicle 10a, 10b, 10c always leads to control of the motor vehicle 10a, 10b, 10c, which largely corresponds to a manual driver in the geographical area 20a, 20b, 20c.

In 4 the individual method steps are shown with which the motor vehicle 10a, 10b, 10c is operated. In 4 this is the example of the motor vehicle 10c from 1 sketched. In a method step S9, a current geographic position 11c of the motor vehicle 10c is first detected by means of a detection device of the motor vehicle 10c. A corresponding position request can, for example, be transmitted from the vehicle system 12 to a navigation system of the motor vehicle 10c, which then provides the position data for the vehicle system 12 . The motor vehicle 10c is here on the road 14 on the outskirts of the city that is assigned to the geographic area 10c. The motor vehicle 10c is therefore currently in the vicinity of the city limits 24. In a next step S10 it is determined in which predetermined geographical area 20c the motor vehicle 10c is currently located, which is determined using the detected geographical position 11c. In this example, the motor vehicle 10c is located in the geographical area 20c, as already described above. Shortly before that, however, it was still outside the city and consequently in geographical area 20b. In a next method step S11, it is determined which predefined driving behavior mode 16 of a plurality of predefined position-dependent driving behavior modes 16 is assigned to the determined geographic area 20c. In this example, the behavioral mode 16 associated with geographic area 20c is referred to as behavioral mode 16c. In a next method step S12, it is checked whether the determined driving behavior mode 16c is already activated in the vehicle system 12. It is determined here that the previous driving behavior mode 16b, which is assigned to the geographical area 20b, is currently still activated in the vehicle system 12.

In a next step 13a, the driving behavior mode 16c required in the geographical area 20c is downloaded into the motor vehicle 10c. For this purpose, a communication link 19 is set up with an external device 18, such as an external server device. This communication connection is set up as soon as it is determined that the motor vehicle 10c is located in the geographical area 20c associated with the determined driving behavior mode 16c de. This is the case here just behind the city limits 24 for the motor vehicle 10c at the recorded geographic position 11c. The driving behavior mode 16b previously stored in the vehicle system 12 can now be deleted, for example, as soon as the driving behavior mode 16c assigned to the current position 11c has been downloaded. Alternatively, it can be provided in a step S13b that already as part of a determination of a driving route of the motor vehicle 10c, which is driving, for example, from a location outside the geographical area 20c to a destination within the city, i.e. within the geographical area 20c, is checked was whether motor vehicle 10c travels through a plurality of predetermined geographic areas 20a, 20b, 20c when driving along the specific route. If several of the predefined geographical areas 20a, 20b, 20c are driven through, the driving behavior modes 16, which are associated with the respective predefined geographical areas 20a, 20b, 20c, are downloaded from the external device 18 via the communication link 19 into the motor vehicle 10c. If the starting position of motor vehicle 10c is, for example, in geographical area 20b and the target position is in geographical area 20c, the respective driving behavior modes 16b and 16c are therefore already stored in vehicle system 12 and/or in a storage device of motor vehicle 10c coupled to vehicle system 12 at the start of the journey.

In a next step S14, if the determined driving behavior mode 16c is not currently activated in the vehicle system 12, the already activated driving behavior mode 16b is deactivated and the determined driving behavior mode 16c is activated. Then, in a step S15, the motor vehicle 10c is controlled at least partially autonomously. This takes place taking into account the at least one control parameter 17 of the currently activated driving behavior mode 16c. This at least partially autonomous control of motor vehicle 10c is outlined here with an arrow in a direction of travel 13 of motor vehicle 10c on road 14 . The related to 1 The transition route 15 described here is only possible as a transition area between the previous driving behavior mode 16b and the determined driving behavior mode 16c if both respective driving behavior modes 16b, 16c have already been downloaded into the motor vehicle 10c.

Overall, the examples show a localized driver model and a location-based driver model selection for highly automated driving. In a first part, the invention relates to the finding that control, regulation and guidance of an autonomous vehicle of the highest level, ie a vehicle designed for fully autonomous driving, should be carried out depending on a locality. It is assumed here that there is an active model in the vehicle that controls the vehicle. This model includes knowledge about the specific vehicle behavior and the applicable traffic regulations. This model is referred to as driving behavior mode 16 in the context of this patent application. The driving behavior mode 16 thus knows which actions have to take place when the vehicle system 12 receives corresponding sensor signals from sensor units of the vehicle. It is also able to carry out planning, for example for the next several 100 meters in front of the vehicle, ie it carries out path planning for a driving route of the vehicle. The driving route itself from the starting position to the target position is specified by an external module. This external module is, for example, a control device of a navigation system that transmits the corresponding data to vehicle system 12 . Since fully autonomous vehicles and vehicles are expected to coexist with human drivers for at least a predetermined period of time, it is necessary for the vehicle system 12 to correctly interpret and predict the human behavior of the driver and/or other road users. However, every driver who has driven a vehicle in a previously foreign country experiences personally that people behave differently in different parts of the world. For example, China therefore requires a new driving test and does not recognize foreign driving licenses.

If the at least semi-autonomous vehicle is to behave like a typical human, it must know, understand and be able to imitate the behavior of a typical human. Behavioral patterns that a person learns on site in a driving school, but which may not be derived from the respective traffic regulations, i.e. from the relevant legal texts, are also relevant here. There is therefore not one driving behavior mode, but several driving behavior modes. These can be implemented in the classic way with the help of a software program or generated by methods of artificial intelligence, i.e. by methods of machine learning.

In a second part, the invention relates to the fact that the individual driving behavior modes 16 are activated in the vehicle depending on the locality, ie the place where the vehicle is or will soon be. For example, driving behavior mode 16 for Portugal will be activated shortly before crossing a border into Portugal, ie shortly before the border crossing the typical portuguese driver is loaded. The same happens, for example, when the vehicle enters the metropolitan area of Paris. Then the driving behavior mode 16 for the settlement of Paris is activated. For example, the change between the previous driving behavior mode 16 and the driving behavior mode 16 now provided for Portugal can take place directly at the border between Spain and Portugal, for example with a corresponding state sign. The change can be made smooth, for example, so that there is a continuous transition from the driving behavior mode 16 specified for Spain to the driving behavior mode 16 specified in Portugal via the transition route 15 .

The driving behavior mode 16 or, strictly speaking, the at least one control parameter 17, which is covered by the driving behavior mode 16, contains at least: The applicable traffic rules in the current area, such as left-hand and right-hand traffic, ie traffic rule data 44; the unwritten but valid interpretations of this rule, information how to interpret local signs and lane markings, that is, the interpretation data 46; the typical behavior of other road users at this location, ie the road user behavior data 42, as well as data on the type of vehicle, ie possibly the exact vehicle model, as can be seen from the vehicle identification data 48. A size, i.e. an extension, of the respective locations, i.e. a state, a settlement or a district, as well as the assignment of the individual driving behavior modes 16 is arbitrary and must be used as part of the generation of the driving behavior modes 16, which can also be referred to as learning , to be determined. Theoretically, the driving behavior mode 16 can already be loaded in the vehicle, but it is more likely that the driving behavior mode 16 will be loaded into the vehicle in good time via an online connection, ie via the communication link 19 . It is therefore possible to update, for example, when there are new traffic regulations. In addition, the driving behavior modes 16 require a not inconsiderable amount of space in a memory device of the vehicle. It can also generally be assumed that an autonomous vehicle always knows where it wants to go. The route ahead is therefore already known to the vehicle when the journey begins. The timely downloading of the appropriate driving behavior mode 16 can therefore be planned well. This can be done, for example, when planning the route or, for example, only when it is determined that the vehicle is now in a new geographical area 20 .

If it is now assumed that the driving behavior mode 16 for the vehicle system 12 is generated in the computing device using artificial intelligence, the learning as such would certainly be located in an external computing device, where all relevant vehicle data 30 converge and joint learning from all this data takes place. In this context, learning means applying the methods of machine learning to data in order to determine the at least one control parameter 17 of the specified driving behavior mode 16 for the respective geographic area 20 . At this point, a decision must also be made as to which locations, ie which geographic areas 20, are assigned their own driving behavior mode 16 in order to achieve a desired functional quality, ie performance, of the method. For this purpose, the individual driving behavior modes 16 or their respective functional quality for the respective place of use, ie for the respective geographical area 20, must be evaluated. This can start with a uniform driving behavior mode 16, which is taught the known traffic rules, ie the ideal behavior model 70 is determined. In addition, an evaluation system must be introduced in the form of the weighting criterion, which makes it possible to test the functionality of what has been learned. This can be a simulation system or a shadow mode functionality, in which real vehicles read and use the previously specified driving behavior mode 16 . A constant evaluation of the functional quality for different locations will inevitably show that an improvement in location A is bought at the cost of a deterioration in location B. This is something that is already being observed in navigation and assistance systems in vehicles. In this case, the learning system must recognize the conflict of objectives and generate a location-specific expression of the driving behavior mode 16 so that both location A and location B can be driven with their own optimal driving behavior mode 16 . This is because the vehicle is then always loaded with the just optimal driving behavior mode 16 for the current location, ie for the current position 11c, of the vehicle. It is assumed here that the optimal driving behavior mode 16 for Berlin, for example in Paris or Shanghai, does not have sufficient functionality. Therefore, location-specific driving behavior modes 16 are ultimately created and used for controlling autonomous vehicles that are designed for partially autonomous and/or fully autonomous driving. Here, local conditions are taken into account based on the driving behavior data 40, the road user behavior data 42, the interpretation data 46 and the traffic control data 44. In addition, a typical behavior is taken into account, which can be inferred from the driving behavior data 40 and the road user behavior data 42, that is Real behavior model 60. The assessment and decision about the localities for the individual driving behavior modes 16 can be created in the way that relates to the above 2 and 3 was described. As in 4 described, at a suitable point in time at which the change of location took place, ie at which the geographic area 20 was changed, the predetermined driving behavior mode 16 assigned to the geographic area 20 now entered can be activated. In addition, the respective driving behavior modes 16 are loaded into the vehicle from the external device 18 at exactly the right time, for example via an online service, for example before the upcoming change of location, which is known from the vehicle-external navigation systems.

Learning, characterization and localization of the individual driving behavior modes 16 takes place in the external device and is made available to the vehicles online. There is also a simulation environment that enables the functional quality to be checked, i.e. within the framework of which the evaluation criterion is determined. Changes to the location and its traffic rules are automatically updated for a new driving behavior mode 16 since, for example, renewed calculations with regard to the predefined driving behavior modes 16 can always be carried out. Ideally, it is possible that after a first vehicle has been in Portugal, the next vehicle will also know how to drive there.

Learning, ie defining the geographical area 20 and the at least one control parameter 17 of the predefined driving behavior mode 16 can theoretically also take place in the vehicle itself. In this case, this learning takes place purely locally and without exchange with other vehicles. To this end, everything related to learning, as described in the 2 and 3 was described, located in an evaluation device of the vehicle. The evaluation, ie the application of the weighting criterion to the real behavior model 60 and the ideal behavior model 70, also takes place in the vehicle, for example, and would be generated taking into account the current position 11c. The functional quality for a location where the vehicle is not located is again possible via simulation, although the simulation would also have to take place in the vehicle.

Individual vehicles can also exchange information directly and also pass on driving behavior modes 16 to one another without using an external device 18, for example. In this case, a standardization for the driving behavior mode 16 and an associated exchange protocol would be necessary. The standardized driving behavior mode 16 does not, for example, take into account the vehicle characteristic data 48. A vehicle that comes to Portugal, for example, could then receive information from a local vehicle on how to drive there. This would be done by providing the default driving behavior mode 16 for Portugal to the vehicle newly arriving in Portugal. However, the other method steps described remain in principle identical.

Theoretically, a global driving behavior mode 16 can be generated gradually, which has a complexity that knows every site-specific characteristic and can therefore actually be used worldwide. Furthermore, it must be taken into account here that this global driving behavior mode 16 includes individual driving behavior modes 16 assigned to the geographic regions 20 .

There is also the consideration of offering a fully autonomous vehicle in such a way that it can be controlled remotely and therefore in a situation that the vehicle-activated driving behavior mode 16 cannot handle, by a human working for example in a call center and having access to the vehicle system 12 of the vehicle is controlled remotely. The corresponding call center employee then has the degree of local knowledge or knows the local traffic rules in order to control the vehicle remotely as an alternative to a driving behavior mode 16 that is automatically determined and made available.

All in all, it is a self-taught system that enables fully autonomous driving in a mix with manually driven vehicles and can be used worldwide. A typical driver or the driving behavior of a typical driver is thus made directly available to the vehicles using the respective predefined driving behavior mode 16 when the vehicle drives into the geographical area 20 associated with the predefined driving behavior mode 16 .

The method described is particularly suitable for fully autonomous vehicles. However, a driver assistance system configured for semi-autonomous driving as a vehicle system 12 can also benefit from such localized driving behavior modes 16 . For example, a navigation system, a multimedia device and other precursors to fully autonomous driving, such as a lane departure warning system, light control, a switching device and so on. It therefore seems probable that the use of several local driving behavior modes 16 is only possible efficiently if methods of artificial intelligence are used, which means that a individual driving behavior modes 16 can be learned. In general, the method described is suitable for all systems that work with models that benefit from local and location-related knowledge. This can not only be vehicles. At the moment, market-specific software variants are usually produced, but precisely if the methods of artificial intelligence should prevail, a more precise division and optimization depending on the geographical area 20 can take place.

Reference List

10, 10a, 10b, 10c

motor vehicle

11c

position

12

vehicle system

13

driving direction

14

road

15

transition route

16, 16b, 16c

driving behavior model

17

control parameters

18

external facility

19

communication link

20, 20a, 20b, 20c

geographic area

22

state border

24

city limits

30

vehicle data

31

location data

40

driving behavior data

41

number

42

road user behavior data

44

traffic regulation data

46

interpretation data

48

vehicle identification data

50

data pattern

60

real-world behavior model

70

ideal behavior model

S1 to S15

steps

Claims (9)

Method for operating a motor vehicle (10), which comprises at least one vehicle system (12) which is designed to control the motor vehicle (10) at least partially autonomously, with a predefined driving behavior mode (16) being activated in the vehicle system (12), which has at least one Control parameter (17) which defines a driving behavior in a predetermined geographical area (20) and which is taken into account when the motor vehicle (10) is controlled at least partially autonomously, comprising the steps of: - detecting a current geographical position (11c) of the motor vehicle (10) (S9); - Determining the predetermined geographical area (20) in which the motor vehicle (10) is currently located, based on the detected geographical position (11c) (S10); - Determining which predetermined driving behavior mode (16) of a plurality of predetermined position-dependent driving behavior modes (16) is assigned to the determined geographical area (20) (S11); - Check whether the determined driving behavior mode (16) is already activated in the vehicle system (12) (S12); - If the determined driving behavior mode (16) is currently not activated in the vehicle system (12), deactivating the already activated driving behavior mode (16) and activating the determined driving behavior mode (16) (S14); and - at least partially autonomous control of the motor vehicle (10) taking into account the at least one control parameter (17) of the currently activated driving behavior mode (16) (S15); wherein the geographic area (20) for the respective driving behavior mode (16) is defined by the following steps: - providing vehicle data (30) from at least one vehicle that is operated by the respective driver and the vehicle data (30) with at least one acquisition device of the Vehicle detected, wherein the respective vehicle data (30) describe at least one driving maneuver of the at least one vehicle and additionally include location data (31) characterizing a detection location of the vehicle data (30) (S1); - Determining the driving behavior data (40) and/or the road user behavior data (42) by applying a machine learning method to the provided vehicle data (30) (S2); - Recognition of at least one data pattern (50) in the determined driving behavior data (40) and/or the determined road user behavior data (42) by applying the machine learning method to the determined driving behavior data (40) and/or the determined road user behavior data (42) (S3 ); - Defining an extension of the respective geographic area (20) taking into account the at least one recognized data pattern (50) and the vehicle data (30) included location data (31) (S4). Method according to the preceding claim, wherein the at least one control parameter (17) is determined from data of at least one of the following data types: - traffic rule data (44) characterizing at least one traffic rule currently in force in the geographic area (20); - driving behavior data (40) which characterize the driving behavior typical of a driver of a vehicle in the geographical area (20); - Road user behavior data (42) which characterize a behavior that is typical for at least one road user in the geographical area (20); - interpretation data (46) which characterize an interpretation rule for at least one traffic symbol in the geographical area (20); - Vehicle identification data (48) which characterize the motor vehicle (10). procedure after claim 2 , wherein defining the at least one control parameter (17) of the specified driving behavior mode (16) for the respective geographical area (20) comprises the following steps: - defining a real behavior model (60) for the specific respective geographical area (20) taking into account the determined driving behavior data (40) and/or road user behavior data (42) which are assigned (S5) to the respective specified geographical area (20); - Providing the traffic regulation data (44), interpretation data (46) and/or vehicle identification data (48) which are associated with the defined geographic area (20) (S6); - Defining an ideal behavior model (70) for the at least one motor vehicle (10) in the defined respective geographical area (20), taking into account the provided traffic control data (44), interpretation data (46) and/or vehicle characteristic data (48) that correspond to the respective defined geographical area region (20) are assigned (S7); - Determining the at least one control parameter (17) for the respective geographic area (20) from the real behavior model (60) and ideal behavior model (70) specified for the respective specified geographic area (20) by applying a predetermined weighting criterion (S8). Method according to the preceding claim, wherein the weighting criterion is determined by a simulation and/or a shadow mode function. Method according to one of the preceding claims, wherein when determining a route from the current geographical position (11b) to a predetermined target position of the motor vehicle (10), it is checked whether the motor vehicle (10) when traveling along the specific route has several predetermined geographical Areas (20) travels through and, if several predetermined geographical areas (20) are traveled through, the driving behavior modes (16), which are assigned to the respective predetermined geographical areas (20), from an external device (18) via a communication link (19) downloaded into the motor vehicle (10). Method according to the preceding claim, wherein a transition from the previous driving behavior mode (16) to the determined driving behavior mode (16) takes place while the motor vehicle (10) is traveling over a predetermined transition route (15). Procedure according to one of Claims 1 until 4 , wherein the determined driving behavior mode (16) is downloaded from an external device (18) via a communication link (19) into the motor vehicle (10) as soon as it is determined that the motor vehicle (10) is in the determined driving behavior mode (16). geographical area (20). Method according to any one of the preceding claims, wherein the geographic area (20) comprises at least one of the following areas: - a state; - a constituent state; - a district; - a settlement. Motor vehicle (10) having a vehicle system (12) which is designed to actuate the motor vehicle (10) at least partially autonomously, the vehicle system (12) or a control device coupled to the vehicle system (12) being designed to carry out a method according to one of the to carry out the preceding claims.

Images