Method for a vehicle including determining an off-road environment complexity and control device

Abstract

The invention relates to a method for determining (20) at least one functional specification (18) for a driving operation function of a vehicle (26), comprising the steps of environmental detection (22) of a vehicle environment (24) of the vehicle (26) by means of at least one environmental sensor device (28) of the vehicle (26), determining (34) an external roadway environmental complexity (36) of external roadway environmental factors (38) in the vehicle environment (24) depending on the environmental detection (22), calculating (60) the functional specification (18) at least depending on the determined external roadway environmental complexity (36).

Description

The invention relates to a method for determining at least one function specification for a driving operation function of a vehicle according to Claim 1. Furthermore, the invention relates to a method for calculation, a method for control and a control device.Prior ArtDE 10 2012 216 875 A1 describes a method for ascertaining a driving strategy of a vehicle, in which a speed range of a setpoint speed of the vehicle is calculated as a function of a traffic density.DE 10 2018 005 261 A1 describes a method for reducing a speed of a vehicle for a cutting-in process of another vehicle as a function of a traffic density in the lane behind the vehicle.Disclosure of the InventionAccording to the present invention, a method for determining at least one function specification for a driving operation function of a vehicle having the features according to claim 1 is proposed. As a result, the driving strategy of the vehicle can be more predictable and more comprehensible for other road users in the vehicle environment. The vehicle can be operated more safely and comfortably. The driving operation of the vehicle may be closer to the usual human behavior of the vehicle guidance, in which a driver of the vehicle automatically adjusts his speed to the ambient density in order to have sufficient time for the processing of the information and making correct decisions. Because the human ability to evaluate concurrent information is limited, drivers normally reduce vehicle speed as the ambient density increases. The driver and other road users may estimate and experience the vehicle at high ambient density as appropriate driving in the situation.The vehicle may be an engine driven vehicle. The vehicle may be a motor vehicle, a truck or a two-wheel vehicle. The vehicle may be a mobile robot.The function specification can be a predefined driving operation variable, in particular at least one setpoint value of the driving operation variable, for the driving operation function. The driving operation variable can be a speed, acceleration, pose, position, rotational speed and / or rotational acceleration of the vehicle.The driving operation function can be a function for setting, maintaining and / or changing a speed, acceleration, pose, position, rotational speed and / or rotational acceleration of the vehicle. The driving operation function may be adjustable by a driver assistance system of the vehicle and / or a vehicle system performing a semi-autonomous or autonomous driving operation of the vehicle. The driver assistance system can carry out a lane change, a distance control to another vehicle, in particular a vehicle driving ahead, a speed control and / or a parking process in an automated manner.The driving operation function can regulate or control a vehicle function, a driving operation and / or an operating state of the vehicle as a function of the function specification.The vehicle environment may be the immediate environment around the vehicle.The environment detection may involve a detection of environment sensor data of the environment sensor device. The environment detection can involve processing of the environment sensor data. The environment detection can involve an environment perception as an interpretation of the vehicle environment. The environment sensor data can contain information about environment objects, in particular other vehicles, living beings, devices, stationary devices, mobile devices and / or road conditions.The environment sensor device may include at least one radar sensor, a LIDAR sensor, a camera and / or an ultrasonic sensor.Environmental complexity is defined as the difficulty and variety of environmental factors. Off-road environmental complexity is defined as the difficulty and variety of off-road environmental factors. The external-roadway environmental complexity can be an intensity, dynamics, spatial proximity, a number and / or a density of the external-roadway environmental factors and / or their interaction or influence on the vehicle and / or the driver.The environmental factors may be environmental objects, environmental events, environmental conditions, and / or relationships to be considered by the vehicle or driver to which the vehicle is responding, needs to react, may be misresponsive, and / or with which the vehicle or driver is interacting, needs to interact, and / or needs to interact.The roadway is the spatial area intended for vehicle traffic to travel the vehicle and other vehicles. The roadway may have one or more lanes.The spatial area outside the roadway is considered to be external to the roadway. In this case, the part of the road which is temporarily or permanently unusable for the traffic as a driving lane can also be considered external to the roadway, for example because of other vehicles parking there or because of construction site devices.The environmental factors which are present or take place in the spatial region outside the roadway are circumscribed as environment factors external to the roadway.The determination of the external-roadway environment complexity may include access to the environment sensor data and / or retrieval of stored information, in particular stored environment sensor data. The off-road environmental complexity can be determined in real time. The determination may involve a calculation of the off-road environmental complexity. The determination of the roadway-external environmental complexity can be carried out supported by information from external data sources, in particular from other vehicles or devices external to the vehicle, in particular from the cloud and / or from map data.The determination of the external-roadway environmental complexity may comprise a calculation of the external-roadway environmental complexity by a calculation model. The calculation model can calculate a relationship between the environment detection, for example between the environment sensor data of the environment sensor device, and the external environment complexity. The relationship may be calculated by machine learning. The computation model may employ artificial intelligence techniques. The calculation model can be learned. The relationship may be adjustable during vehicle operation. The context may be individualized, in particular with respect to the vehicle user. For example, a function specification can be output with the adapted relationship, which corresponds to the expectation of the vehicle user with regard to the roadway-external environmental complexity.The calculation model may include a neural network. The calculation model may be a deep learning model. A deep learning model is a neural network which is constructed from a plurality of planes and contains in each of the planes a plurality of neurons which are connected to one another and describe their connection by trainable parameters.A training goal in training the neural network can be estimating the environmental complexity as a number, vector or tensor. The neural network may be trained to learn a context that maps the environment sensor data to an estimate of the out-of-roadway environmental complexity. The mean square error that measures the difference between the estimated and actual environmental complexity may be reduced during training.The neural network can be trained to assign the environment sensor data to a plurality of complexity classes of the external environment complexity. The classes of complexity may be characterized by predefined thresholds or intervals. The cross entropy indicating the information content between the predicted and actual classes of complexity may be used as a loss function.The calculation model may subdivide the out-of-roadway environmental complexity into regions (clusters). The computation model may include support vector machines (SVM). SVMs as a supervised learning measure attempt to find a decision boundary (hyperplane) that separates two or more classes of complexity from maximum distance data points. The data points closest to the hyperplane are referred to as support vectors. These support vectors are decisive for defining the separation limit. SVM can be used for both linear and non-linear classification problems. In non-linear problems, the data is transformed into a higher dimensional space to find a separable hyperplane.The computation model may include Bayesian networks. Bayesian networks are graphical models that represent the probabilistic relationships between a series of variables. Each node in the network represents a variable and the edges represent conditional dependencies. These networks make it possible to calculate the common probability distribution over all variables. They are particularly useful for modeling uncertainties and making inferences about unknown variables. The Bayesian networks can be used to calculate the probability of different external-roadway environmental complexes based on the environment sensor data and to take into account the uncertainty of this environment sensor data.The determination of the off-road environment complexity may apply multiple computation models. For example, a deep learning model may first be used as a preprocessing step in order to project the environment sensor data into a low-dimensional feature space which contains the external environment complexity on the road surface. Then, a computation model with SVM can be used as a post-processing step to estimate or classify the off-road environmental complexity. Subsequently, a Bayesian network computation model may be used to evaluate and update the uncertainty or information content of the estimation or classification of the off-road environmental complexity.The calculation of the function specification can be carried out by a further calculation model. The further calculation model can calculate a relationship between the external environment complexity and the function specification. The calculation model can be implemented analogously to the calculation model for calculating the external-roadway complexity. The calculation model may be integrated in the calculation model for calculating the external environment complexity.The function specification can be displayed to the vehicle user of the vehicle in an optically, acoustically and / or vibrating manner, for example via at least one human-machine interface.In a preferred embodiment of the invention, it is advantageous if the function specification is furthermore calculated as a function of a determined roadway-related environmental complexity of roadway-related environmental factors. The roadway-related environmental complexity may involve a roadway-related environmental density of the roadway-related environmental factors. The roadway-related environmental factors can be environmental objects located on the roadway, in particular road users, in particular other vehicles. The roadway-related ambient density may be a density of the roadway-related ambient factors, in particular a traffic density. The roadway-related ambient density may include a roadway-related signaling density, for example, a density, a frequency, and / or a number of switched-on brake lights of the other vehicles. The roadway-related ambient density may include a density, a frequency and / or a number of traffic dynamics, for example acceleration events, braking events, changes of direction or lane of the other vehicles, and / or a traffic intensity, for example a number of road users.The spatial area of the roadway used by the vehicle or usable for the movement is considered to be roadway-related. In multi-lane lanes, the area includes all lanes of the lane.In a preferred embodiment of the invention, it is provided that the external environment complexity involves an external environment density of the external environment factors. The off-road ambient complexity may be proportional to the off-road ambient density. The roadway-external ambient density may directly indicate the density of the roadway-external ambient factors and / or the number of roadway-external ambient factors due to the spatially limited extent of the vehicle environment. The greater the number and / or the density of environment factors external to the roadway, the greater the environment density external to the roadway.In a specific embodiment of the invention, it is advantageous if the environment density outside the roadway includes a signaling density of signaling environment factors acting on the vehicle outside the roadway. Signaling is understood to mean the environmental factors which emit or convey information with respect to the vehicle and / or the vehicle user. The signaling environmental factors may include traffic signs, advertisements, advertisements, display panels, or the like.A preferred embodiment of the invention is advantageous in which the external environment complexity involves spatial distances between the vehicle and the external environment factors. A narrow roadway or a narrow roadway intersection may have a greater environmental factor density of the off-roadway environmental factors and / or a smaller distance of the off-roadway environmental factors from the roadway than a wide roadway or a generous roadway intersection.In a preferred embodiment of the invention, it is advantageous if the function specification is calculated as a function of a comparison between the ascertained external-roadway environmental complexity and a threshold value of the external-roadway environmental complexity. The threshold value can be changed depending on the traffic situation, the operating state of the vehicle, the driving mode of the vehicle and / or the environment detection.In an advantageous embodiment of the invention, it is provided that the calculated function specification is restrictive with respect to a function specification that was previously effective for the driving operation function if the ascertained road-external environmental complexity reaches or exceeds the threshold value. The calculated function specification can be restrictive with respect to the previous function specification, in that the calculated function specification is a smaller and / or larger driving operation variable than the driving operation variable of the previously effective function specification. The greater the external environmental complexity on the road surface, the more restrictive the function specification can be.If the external environment complexity, in particular the external environment density, increases, the predefined driving operation variable, for example the predefined speed, can be smaller. As the roadway-external environmental complexity decreases, in particular the roadway-external environmental density, the predefined driving operation variable may be greater.According to the present invention, a method for calculating a driving strategy having the features according to claim 8 is furthermore proposed. According to the present invention, a method for controlling a driving operation function having the features according to claim 9 is also proposed. According to the present invention, there is further provided a control device having the features of claim 10.Further advantages and advantageous embodiments of the invention result from the description of the figures and the figures.DESCRIPTION OF THE FIGURESThe invention will be described in detail below with reference to the drawings. They show in detail: FIG. 1 : shows a method for controlling with a method for calculation and a method for determining a function specification in each case in a specific embodiment of the invention. FIG. 2 : a schematic comparison of low and high external-roadway complexities.FIG. 1 shows a method for controlling with a method for calculating and a method for determining a function specification in each case in a specific embodiment of the invention. The method for controlling 10 a driving operation function 11 of a vehicle controls the driving operation function 11, for example an adaptive cruise control, as a function of a driving strategy 12, which is calculated by a method for calculating 14 a driving strategy 12 of the vehicle. The method for controlling 10 may be executed by a control device 16 of the vehicle.The driving strategy 12 is in turn determined as a function of a function specification 18, for example a predefined driving operation variable, in particular a setpoint speed of the vehicle, by a method for determining 20 a function specification 18 for the driving operation function 11 of the vehicle.Method for ascertaining 20 includes surroundings detection 22 of a vehicle surroundings 24 of vehicle 26 by at least one surroundings sensor device 28 of vehicle 26. surroundings sensor device 28 has, for example, at least one radar sensor, LIDAR sensor, or a camera. Environment detection 22 includes, in particular, processing 30 of environment sensor data 32 of environment sensor device 28.Subsequently, an ascertainment 34 of an external-roadway environmental complexity 36 of external-roadway environmental factors 38 in the vehicle environment 24 takes place as a function of the environment detection 22, in particular as a function of the processed environment sensor data 32. The environmental factors can also comprise signaling environmental factors 42, in particular traffic signs 44 and advertisements 46.The off-road environmental complexity 36 may involve an off-road environmental density 48 of the off-road environmental factors 38. The off-road environment density 48 may be a signaling density 50 of the off-road signaling environment factors 42, such as the traffic signs 44 and advertisements 46. The greater the number or density of off-road signaling environmental factors 42 in the off-road vehicle environment 24 of the vehicle 26, the greater the signaling density 50.The off-road environmental factors 38 may also include other vehicles 54 parking outside the roadway 52 being traveled by the vehicle 26, or buildings 56 standing outside the roadway 52. The off-road environment complexity 36 may also incorporate spatial distances 58 between the vehicle 26 and these off-road environment factors 38. A narrow roadway 52 having smaller spatial distances 58 between the vehicle 26 and the off-roadway environmental factors 38 may have a greater off-roadway environmental complexity 36 than a wide roadway 52 having larger spatial distances 58 between the vehicle 26 and the off-roadway environmental factors 38.A calculation 60 of the function specification 18 is carried out at least as a function of the ascertained roadway-external environmental complexity 36. Function specification 18 may also be calculated as a function of a ascertained roadway-related environmental complexity 62 of roadway-related environmental factors 64, for example other road users on roadway 52. The roadway-related environmental complexity 62 may include a roadway-related environmental density 66 of the roadway-related environmental factors 64, for example, a traffic density 68.Function specification 18 is preferably calculated as a function of a comparison 70 between ascertained external-roadway environmental complexity 36 and a threshold value 72 of external-roadway environmental complexity 36. The calculated function specification 18 is restrictive with respect to a function specification 18 that was previously effective for the driving operation function, if the ascertained road-external environmental complexity 36 reaches or exceeds the threshold value 72. For example, when threshold value 72 is reached or exceeded, the driving operation variable, in particular the setpoint speed, of the adaptive cruise control is reduced.FIG. 2 shows a schematic comparison of low and high ambient factor densities. FIG. 2 a) shows a vehicle environment 24 of the vehicle 26 with a greater road-external environmental complexity than FIG. 2 b). In FIG. 2 a), the road-external environmental factors 38 are formed by other vehicles 54 which park outside the road 52 and which have smaller spatial distances 58 than in the case of the wider road 52 in FIG. 2 b) with respect to the vehicle 26. As a result, the external-roadway environmental complexity in FIG. 2 a) is increased compared to the external-roadway environmental complexity in FIG. 2 b). Also, the ambient complexity would be even further increased if the other vehicles 54 were to be even closer together, corresponding to a greater ambient density.FIG. 2 c) shows a vehicle environment 24 of the vehicle 26 with a greater road-external environmental complexity than FIG. 2 d). In FIG. 2 c), the roadway-external environmental factors 38 are formed by traffic signs 44 and advertisements 46 arranged outside the roadway 52, which have a greater signaling density 50 than in FIG. 2 d) and thereby increase the roadway-external environmental complexity compared to the roadway-external environmental complexity in FIG. 2 d).References included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Patent Literature citedDE 10 2012 216 875 A1

[0002] DE 10 2018 005 261 A1

[0003]

Claims

Method for ascertaining (20) at least one function specification (18) for a driving operation function of a vehicle (26), having the steps of environment detection (22) of a vehicle environment (24) of the vehicle (26) by at least one environment sensor device (28) of the vehicle (26), ascertaining (34) an external-roadway environment complexity (36) of external-roadway environment factors (38) in the vehicle environment (24) as a function of the environment detection (22), calculating (60) the function specification (18) at least as a function of the ascertained external-roadway environment complexity (36).Method for ascertaining (20) according to Claim 1, characterized in that the function specification (18) is furthermore calculated as a function of a ascertained roadway-related environmental complexity (62) of roadway-related environmental factors (64).Method for determination (20) according to Claim 1 or 2, characterized in that the external-roadway environmental complexity (36) involves an external-roadway environmental density (48) of the external-roadway environmental factors (38).Method for ascertaining (20) according to Claim 3, characterized in that the road-external ambient density (48) uses a signaling density (50) of road-external signaling ambient factors (42) acting on the vehicle (26).Method for determination (20) according to one of the preceding claims, characterized in that the external environment complexity (36) uses spatial distances (58) between the vehicle (26) and the external environment factors (38).Method for ascertaining (20) according to one of the preceding claims, characterized in that the function specification (18) is calculated as a function of a comparison (70) between the ascertained external-roadway environmental complexity (36) and a threshold value (72) of the external-roadway environmental complexity (36).Method for ascertaining (20) according to Claim 6, characterized in that the calculated function specification (18) is restrictive with respect to a function specification which was previously effective for the driving operation function if the ascertained road-external environmental complexity (36) reaches or exceeds the threshold value (72).Method for calculating (14) a driving strategy (12) of a vehicle (26) as a function of a function specification (18) of a driving operation function of the vehicle (26) calculated by a method for determining (20) a function specification (18) according to one of Claims 1 to 7.Method for controlling (10) a driving operation function of a vehicle (26) as a function of a driving strategy (12) of the vehicle (26) calculated by a method for calculating (14) a driving strategy (12) according to Claim 8.A control device (16) for a vehicle (26), configured to execute the method for controlling (10) a driving operation function of the vehicle (26) according to claim 9.

Images