Driving support system, vehicle, method for controlling vehicle, and computer program product

Abstract

The present invention provides a driving support system 100 for a vehicle. The driving support system 100 includes: a control module 30 for generating control signals; an environment sensor module 10 for providing environmental data about the environment of the vehicle on the basis of detection results of at least one optical sensor 11; and a support model module 20 for providing a driving support model to the control module 30. The support model module 20 includes: a model storage device 22; a visibility determination device 21; and a model selection device 23 for selecting a main driving support model from at least two driving support models, stored in the model storage device 22, according to a characteristic visibility parameter determined by the visibility determination device 21. The control module 30 generates the control signals on the basis of a combination of the data provided by the environment sensor module 10 and the selected main driving support model provided by the support model module 20.

Description

Driving assistance system, vehicle, method for controlling a vehicle, and computer program product

[0001] The present invention relates to a driver assistance system for use with a vehicle, a vehicle equipped with such a driver assistance system, a method for controlling a vehicle, and a computer program product.

[0002] Recent developments in the field of the automotive industry have increased the importance of driver assistance systems in vehicles in order to increase comfort and safety during vehicle operation.

[0003] Driver assistance systems typically provide a variety of support and safety measures, such as collision prevention, lane keeping, or lane change assistance, and are generally characterized by significantly faster reaction times compared to "manual" operation alone.

[0004] As a further development, autonomous vehicles, which require minimal interaction between the operator (driver) and the vehicle, are also gaining attention as a comfortable means of personal transportation.

[0005] In either case, the traditional driver is no longer the "master sensor and master controller" of the vehicle, and the respective driver assistance system takes over the various tasks required for the aforementioned assisted or autonomous driving functions, namely vehicle localization, environment detection, and vehicle control. This requires a reliable and fast way to collect and process environmental data and initiate respective actions based on it.

[0006] Thus, a driver assistance system includes several sensors that provide data for various tasks such as localization and recognition, and these sensors can typically be used individually or in combination with each other.

[0007] The quality of the data acquired by these sensors and the quality of its processing is crucial for making safe, fast and reproducible decisions under all conditions, which is particularly challenging in today's rapidly changing traffic conditions.

[0008] Therefore, Patent Document 1 proposes a driving assistance system that uses an overall illumination map to select a route or path to be taken by a vehicle based on the illumination conditions described by the illumination map. The system is configured to select a specific route from multiple available routes that is likely to have the best illumination conditions. This maximizes the accuracy and reliability of the detection results of the optical sensor of the driving assistance system throughout the entire route, thereby reducing the risk of malfunction of the driving assistance system due to a decrease in the recognition ability of the optical sensor.

[0009] However, this strategy can result in a longer and potentially inefficient route and generally requires processing of vast amounts of data, which in turn requires time and increases computational costs due to the need for models with more parameters.

[0010] US Patent Application Publication No. 2022 / 0107203

[0011] Therefore, the technical problem that the present invention aims to solve is to provide fast and efficient vehicle driving assistance that enables real-time processing under a variety of different environmental conditions.

[0012] This problem is advantageously solved by a driving assistance system according to claim 1, a vehicle according to claim 20, a method according to claim 21 and a computer program product according to claim 23.

[0013] Each independent claim in this specification refers to a preferred aspect / embodiment, each of which may be provided individually or in combination with each other.

[0014] According to a first aspect of the present invention, there is provided a driving assistance system for a vehicle, the driving assistance system comprising: a control module configured to generate a control signal; an environmental sensor module configured to provide environmental data of an environment of the vehicle based on detection results of at least one optical sensor; and an assistance model module that provides a driving assistance model to the control module, the assistance model module comprising: a model storage device in which at least two driving assistance models corresponding to different visibility conditions in the environment of the vehicle are stored; a visibility determination device configured to determine a characteristic visibility parameter that describes a current visibility condition in the environment of the vehicle; and a model selection device configured to select a primary driving assistance model from the at least two driving assistance models stored in the model storage device in accordance with the characteristic visibility parameter determined by the visibility determination device, and the control module is coupled to at least the assistance model module and the environmental sensor module and is configured to generate a control signal based on the data provided by the environmental sensor module in combination with the selected primary driving assistance model provided by the assistance model module.

[0015] The quality of the optically acquired detection results of the environmental sensor module naturally depends strongly on the visibility conditions in the vehicle's environment, which may vary, inter alia, depending on the day, lighting, i.e. illumination, weather, shading, etc. in the environment. Thus, detection results during clear, cloudless days are expected to be of higher quality than, for example, during foggy nights.

[0016] In order to encounter this large variety of different visibility conditions, the proposed driver assistance system is able to select from the various assistance models the primary assistance model that is most appropriate for processing the data provided by the environmental sensor module.

[0017] Therefore, instead of providing one advanced assistance model that takes into account various visibility scenarios, the proposed system comprises various "smaller" assistance models, from which one is selected prior to processing the data provided by the environmental sensor module (=environmental data).

[0018] This reduces latency across the entire driver assistance system compared to large-scale "synthetic models," allowing control signals to be provided in near real time, thereby increasing safety levels and enabling energy-efficient driving.

[0019] Furthermore, less computing power, e.g., in terms of CPU and / or RAM, is required to evaluate the provided data by the smaller model compared to a large "synthetic model", which reduces the manufacturing costs of the driver assistance system and also reduces power consumption during operation.

[0020] An assistance model is to be understood as a computer-based model that includes at least one or more algorithms each set up to process the data provided by the environmental sensor modules and to issue one or more control signals based on the output data, each of which results in the feeding of environmental data into the model (algorithm).

[0021] The term "primary" support model refers to the selected model that will be used to process environmental data until another model is selected. Preferably, the model storage device contains 2-5 different models, and more preferably 2-10 different models, which may be selected as the primary support model.

[0022] Preferably, the assistance model module is configured to determine the characteristic visibility parameter and to select a primary assistance model based on the characteristic visibility parameter at predetermined time intervals, in other words, an update function is provided to ensure that the most suitable model is always used to process the environmental data throughout the entire operation interval of the driver assistance system.

[0023] The characteristic visibility parameter is to be understood as a measure of the expected quality of the optically detected data of the environmental sensor module. Since optical sensors may encounter various distorting effects, it is not possible to expect a constant quality, i.e. a constant level, of optical interference throughout the entire operation interval of the driver assistance system.

[0024] Thus, the detected data may be used directly as is, although different signal refinement or error compensation methods, or different levels of these methods, may be required depending on the conditions in the environment.

[0025] In other words, the characteristic visibility parameters essentially correlate with the predictable perceived quality and / or accuracy of the detection results transmitted by the optical sensor.

[0026] Thereby, at least two different values ​​can be assigned to the characteristic visibility parameter, such as a value corresponding to "good visibility" and a value corresponding to "poor visibility", which for simplicity reasons may be represented numerically as "0" and "1". The term "value" as used in this context is not necessarily limited to the form of a number, but can also be a string or any other data type.

[0027] Preferably, the characteristic visibility parameter can be assigned three or more values ​​corresponding to different levels of visibility in the vehicle's environment. Particularly preferably, the values ​​of the characteristic visibility parameter are continuously distributed over a particular range.

[0028] The visibility determining device may be any device capable of receiving inputs and determining characteristic visibility parameters.

[0029] In one reduced complexity embodiment, an operator (driver) may manually select the characteristic visibility parameter via a user interface of the determination device such that the parameter is determined based on operator input.

[0030] Preferably, the characteristic visibility parameter is a characteristic illumination, which describes the illumination level in the vehicle's environment. Illumination level has been found to be an easy-to-determine quality that correlates strongly with the quality of perception of optical sensors, especially cameras.

[0031] In a preferred embodiment, the visibility determining device is coupled to the environmental sensor module and configured to determine a characteristic visibility parameter in response to detection results of the at least one optical sensor.

[0032] In this manner, the environmental sensor module is used to obtain data regarding visibility conditions, and based on this data, characteristic visibility parameters are determined.

[0033] Preferably, the detection results used to determine the characteristic visibility parameters are different from, i.e., not part of, the environmental data used for control.

[0034] In this manner, characteristic visibility parameters are determined in advance of an operating condition where the control module uses subsequently acquired environmental data to generate control signals.

[0035] In a preferred embodiment, the at least one optical sensor is a camera, and the environmental data comprises at least in part image and / or video data from the camera.

[0036] Cameras represent a reliable and relatively inexpensive optical sensor, potentially capable of detecting many visibility conditions in the vehicle's environment, such as brightness and shading.

[0037] In a preferred embodiment, the visibility determination device comprises a data interface connectable, in particular wirelessly connectable, to an external data source containing data relating to environmental conditions in the vehicle's environment, and is configured to process data received from the external data source via the data interface and to determine the characteristic visibility parameter accordingly.

[0038] In this way, the driver assistance system is given access to a large amount of data, in particular real-time data, from which data can be extracted regarding the visibility conditions in the vehicle's environment.

[0039] Such data received from external data sources preferably includes real-time data on the sun position and / or sun angle, particularly in the area in which the vehicle is located, and / or geographical 3D and / or 2D ground data, particularly in the area in which the vehicle is located, and / or real-time weather data.

[0040] Advantageously, all of the aforementioned data can be used to extract data regarding current visibility conditions. For example, 3D ground data can be used to estimate the likelihood of shading, e.g., due to buildings or terrain. Also, meteorological data can be used to estimate the likelihood of shading, e.g., due to clouds, or even to estimate the average visibility, which depends, e.g., on fog, rain, etc.

[0041] In a preferred embodiment, the driver assistance system further comprises a global positioning device configured to determine the current global position of the vehicle, in particular by communicating with a satellite-based positioning system such as GPS or GNSS.

[0042] In a preferred embodiment using a global positioning device, the visibility determination device is configured to access data in an external data source to receive location-specific data based on the global position determined by the global positioning device.

[0043] In this manner, the exemplary sun, geographical surface, and weather data mentioned above can be accessed and received for the vehicle's current location, enhancing the accuracy of such data for purposes of determining current visibility conditions in the vehicle's environment.

[0044] In a preferred embodiment, the visibility determination device is configured to determine the characteristic visibility parameter in response to the detection results of the at least one optical sensor and in response to data received from an external data source, in particular in response to a weighted average value of the received data.

[0045] In this way, the visibility determination device advantageously combines inputs from the on-board environmental sensor module with inputs received from external sources to enable an even more accurate estimation of current visibility conditions.

[0046] The characteristic visibility parameters may be extracted, for example, by a predetermined function that uses data from the optical sensors of the environmental sensor module and data received from external sources as arguments of the function.

[0047] In a preferred embodiment, the visibility conditions described by the characteristic visibility parameters are light level, field of view, light contrast, or illumination level, or a combination of two or more of the foregoing, in the vehicle's environment.

[0048] All of the aforementioned conditions affect the visibility conditions in the vehicle's environment, which in turn affect the perceived accuracy of the optical sensors used by the driver assistance systems.

[0049] In a preferred embodiment, the model selection device is configured to select a first model of the at least two models as the primary support model if the determined characteristic visibility parameter is within a first predetermined range, and to select a second model of the at least two models as the primary support model if the determined characteristic visibility parameter is within a second predetermined range.

[0050] In this way, a relatively simple implementation of the selection process is provided, based on rules criteria.

[0051] In a preferred embodiment, the model selection device is configured to select the main support model by means of a neural network and / or by means of machine learning methods.

[0052] The method is based on training data provided to the model selection device during a training phase, from which the device sets up a decision-making model as the basis for the subsequent operational phase. In this way, a programmer or operator does not need to predefine the rules of the selection process, and the model selection device learns on its own.

[0053] In a preferred embodiment, the assistance model module is further configured to switch the driver assistance system into a standby mode when the determined characteristic visibility parameter exceeds a critical threshold, during which the environmental sensor module is specifically switched off.

[0054] When visibility conditions become so poor that safe and reliable operation of the driving assistance system based on optically detected environmental data can no longer be guaranteed, i.e., when the determined characteristic visibility parameter exceeds the critical threshold, the driving assistance system switches to standby mode and, in particular, hands over partial control of the vehicle to the operator (driver) and switches off energy consumers that are not necessary for this operating mode.

[0055] Preferably, the control module is configured to issue a notification to the operator (driver) in this case, in particular by means of an optical or acoustic notification device.

[0056] In a preferred embodiment, the control module comprises a path planning device configured to determine path data for a path to be followed by the vehicle based on data provided by the environmental sensor module in conjunction with a selected primary assistance model provided by the assistance model module, and configured to generate control signals based on the determined path data, in particular for controllable actuators of the drive system and / or steering system of the vehicle.

[0057] In a preferred embodiment, the at least two driving assistance models stored in the model storage device are vehicle localization models, each of which includes an algorithm for determining the position of the vehicle relative to its environment, and in particular the algorithm is a visual odometry algorithm.

[0058] In this way, driver assistance systems can provide location-based services such as lane keeping.

[0059] In a preferred embodiment, the at least two driving assistance models stored in the model storage device are perceptual models, each of which includes an algorithm for determining the position of objects in the vehicle's environment, in particular an algorithm for obstacle detection.

[0060] In this way, the driver assistance system may provide services such as collision warning and / or lane keeping.

[0061] Preferably, the model storage device includes at least four models, two of which are perception models and two of which are localization models, in which the model selection device is configured to select a primary perception model and a primary localization model from the models, and the control module is configured to generate a control signal based on data provided by the environmental sensor module in combination with the selected primary localization model and the selected primary perception model.

[0062] In this way, the driver assistance system combines the advantages of the previous examples that utilize either a localization model or a perception model.

[0063] In a preferred embodiment, the environmental sensor module comprises at least one further sensor, in particular an ultrasonic sensor or a LIDAR sensor, configured to detect environmental conditions in the vehicle's environment.

[0064] In this way, the environmental sensor module's field of perception can be expanded to provide more accurate and more reliable data of the vehicle's environment.

[0065] Preferably, the control module is configured to process the respective data received from all sensors of the environmental sensor module.

[0066] In a preferred embodiment, the driver assistance system further comprises at least one actuator controllable by a control signal generated by the control module, in particular the at least one actuator being an optical and / or acoustic notification device configured to notify the driver of the vehicle.

[0067] Alternatively or additionally, the control module may be coupled to at least one actuator of the vehicle that is controllable by a control signal generated by the control module, in particular the at least one actuator of the vehicle is part of the drive system of the vehicle or of the steering system of the vehicle.

[0068] All of the above modules, devices and units of the driver assistance system are preferably to be understood as electronic components, and the determination steps performed by the respective modules, devices and units, such as determining the characteristic visibility parameters, are typically performed by a processor-based computing unit of the respective module, device or unit.

[0069] According to a second aspect of the present invention, there is provided a vehicle comprising a driving assistance system according to the first aspect or one of the preferred embodiments of the first aspect.

[0070] Preferably, the driver assistance system is coupled to or is part of a control device of the vehicle, configured to control one of the vehicle's actuators, in particular based on control signals received from the control module of the driver assistance system.

[0071] According to a third aspect of the present invention, there is provided a method for controlling a vehicle by a driving assistance system of the vehicle, the method comprising the steps of: providing at least two driving assistance models corresponding to different visibility conditions in an environment of the vehicle; providing environmental data of the vehicle's environment including a sub-step of optically detecting environmental conditions of the vehicle's environment; determining a characteristic visibility parameter describing the current visibility conditions in the vehicle's environment; selecting a primary driving assistance model from the at least two provided driving assistance models in response to the determined characteristic visibility parameter; and generating a control signal based on the provided environmental data in conjunction with the selected primary assistance model.

[0072] The aforementioned advantages of the driver assistance system according to the invention apply analogously to this subordinate method, so that a repetition of them will be omitted at this point.

[0073] In a preferred embodiment, the method further comprises the step of outputting the control signal to an actuator of the vehicle that is controllable by the control signal, in particular the actuator being part of the drive and / or steering system of the vehicle.

[0074] Preferably, the step of determining the characteristic visibility parameter comprises the sub-steps of optically detecting an environmental condition of the vehicle's environment and determining the characteristic visibility parameter depending on the detection results of said sub-step.

[0075] The sub-step of detecting environmental conditions within the step of determining the characteristic visibility parameters is preferably separate from the step of detecting environmental data of the vehicle's environment, as it occurs at an earlier point in time. In this way, the data used from determining the characteristic visibility parameters is "older" than the data used to generate the control signals, i.e., the data fed to the selected primary support model.

[0076] Preferably, the step of determining the characteristic visibility parameter comprises the sub-steps of receiving, from an external data source, data relating to environmental conditions in the vehicle's environment, and determining the characteristic visibility parameter in response to the received data.

[0077] Preferably, the step of determining the characteristic visibility parameter includes the sub-steps of receiving data relating to environmental conditions of the vehicle's environment from an external data source, and determining the characteristic visibility parameter in response to the received data and to the detection results of the sub-step of detecting the environmental conditions.

[0078] Preferably, the method includes the step of determining a global position of the vehicle, wherein the sub-step of receiving data from an external data source is performed in response to the determined global position.

[0079] Preferably, the step of selecting the primary driving assistance model includes the sub-steps of selecting a first model of the at least two models as the primary assistance model if the determined characteristic visibility parameter is within a first predetermined range, and selecting a second model of the at least two models as the primary assistance model if the determined characteristic visibility parameter is within a second predetermined range.

[0080] Preferably, the method includes a route planning step in which route data for a route to be followed by the vehicle is determined based on the provided environmental data together with the selected primary assistance model, and the step of generating control signals includes a sub-step of generating control signals, in particular for controllable actuators of the vehicle's drive system and / or steering system, in response to the determined route data.

[0081] According to a fourth aspect of the present invention there is provided a computer program product for use in a driver assistance system according to the first aspect or one of the preferred embodiments of the first aspect, the computer program product comprising instructions which, when the computer program product is executed, cause the driver assistance system to perform a method according to the third aspect or the preferred embodiments of the third aspect.

[0082] It is emphasized that the present invention is in no way limited to the embodiments and their features described above: the invention also includes variants of these embodiments, in particular those resulting from variants and / or combinations of individual or several features of the described embodiments, within the scope of protection of the independent claims.

[0083] Further aspects and advantages of the present invention, as well as particular examples of the aforementioned embodiments, are described below with reference to the drawings illustrated in the accompanying figures.

[0084] 1 shows a schematic layout of a first embodiment of a driving assistance system according to the invention, 2 shows a schematic layout of a second embodiment of a driving assistance system according to the invention, and 3 shows a flow chart of an embodiment of a method for controlling a vehicle according to the invention.

[0085] FIG. 1 shows a schematic layout of a first embodiment of a driver assistance system 100 according to the invention.

[0086] The driving assistance system 100 comprises a control module 30 configured to generate a control signal, an environmental sensor module 10 configured to provide environmental data of the vehicle's environment, and an assistance model module 20 that provides a driving assistance model to the control module 30.

[0087] The environmental sensor module 10 comprises at least one optical sensor 11, which is in particular a camera, and in this embodiment comprises another sensor 12, which may be another optical or non-optical sensor, for example a LIDAR sensor or an ultrasonic sensor.

[0088] The environmental sensor module 10 is coupled to at least the control module 30 for providing the control module 30 with environmental data based on the detection results of the sensors 11, 12.

[0089] The assistance model module 20 that provides the driving assistance model to the control module 30 includes a visibility determination device 21, a model storage device 22, and a model selection device 23.

[0090] The visibility determining device 21 is configured to determine characteristic visibility parameters describing the current visibility conditions, in particular the illumination level, in the vehicle's environment.

[0091] For that purpose, the visibility determination device 21 comprises a data interface 21a, via which it is coupled to the optical sensor 11 of the environmental sensor module 10 and to an external data storage 200. The coupling to the external data storage 200 is generally realized in a wireless manner.

[0092] In this embodiment, both the data received from the environmental sensor module 10 and the data from the external data storage 200 may be utilized by the visibility determination device 21. However, it should be noted that in alternative embodiments, the data interface 21 a may be coupled only to the environmental sensor module 10, or only to the external data storage 200, to provide either data from the environmental sensor module 10, i.e., from the optical sensor 11, or data from the external data storage 200.

[0093] The external data storage 200 may be a centralized or decentralized data system and includes several different databases 210, 220, 230, each of which is a database relating to environmental conditions in the vehicle's environment, in particular environmental conditions that affect visibility conditions in the vehicle's environment.

[0094] The databases may include a database 210 of meteorological data, a database 220 of solar data such as solar position and / or solar angle, and a database 230 of geographic 2D and / or 3D earth surface models.

[0095] In this embodiment, the data interface 21a receives data having detection results from the optical sensor 11 and data from one or more databases 210, 220, 230 in the external data storage 200, and the data is provided to the computing unit 21b of the visibility determination device 21, which is configured to determine characteristic visibility parameters according to the data and transmit the determined parameters to the model selection device 23.

[0096] At least two driver assistance models are stored in the model storage device 22 corresponding to different visibility conditions in the vehicle's environment, such as a first model corresponding to "good" visibility conditions and a second model corresponding to "poor" visibility conditions, and in particular the models correspond to different illumination levels in the vehicle's environment.

[0097] The model selection device 23 is coupled to the model storage device 22 and the visibility determination device 21 and is configured to select a main driving assistance model from at least two driving assistance models stored in the model storage device 22 according to the characteristic visibility parameters determined and transmitted by the visibility determination device 21.

[0098] Thereby, an optimal driving assistance model is selected based on the current visibility conditions in the vehicle's environment. The selected / selected primary assistance model is then provided to the control module 30, which is coupled to at least the assistance model module 20 and the environmental sensor module 10.

[0099] The control module is configured to receive the primary assistance model from the assistance model module 20 and environmental data from the environmental sensor module 10 and generate control signals based on that data, whereby the environmental data is fed into the primary assistance model, i.e., into the respective algorithm of that model, and then use the output of that model (algorithm) to generate at least one control signal.

[0100] The generated control signal may then be transmitted to an actuator of the driver assistance system 100 itself, which actuator may be controlled by the control signal output by the control module 30, which may be, for example, an acoustic or optical notification device 40 configured to provide a notification to the vehicle operator.

[0101] Furthermore, the generated control signal may be transmitted to an actuator 300 of the vehicle, which actuator 300 may in particular be part of the vehicle's drive system or steering system and may be controlled by the control signal output by the control module 30, so that issuing a control signal to the actuator 300 may actively influence the vehicle's driving behavior.

[0102] The exemplary driver assistance system 100 offers the advantage of providing various driver assistance models and selecting one of these driver assistance models as the primary assistance model that is most appropriate under current visibility conditions to process detected environmental data and issue control signals based thereon.

[0103] Therefore, instead of providing one advanced assistance model that takes into account various scenarios, the proposed system presents various "smaller" assistance models, from which the most suitable one is selected prior to processing the sensed environmental data. In this way, the delay of the entire method can be significantly reduced, enabling highly accurate real-time control.

[0104] The fact that a computing unit is shown alone for the visibility determination device 21 does not imply that other devices or modules, such as the model selection device 23 or the control module 30, do not have a commonly shared or separate computing unit.

[0105] FIG. 2 shows a schematic layout of a second embodiment of a driver assistance system 100 according to the invention.

[0106] The basic arrangement is essentially similar to that of the first embodiment of FIG. 1, so that the repetition of identical parts / components will be omitted.

[0107] The second embodiment differs from the first embodiment by the inclusion of a global positioning device 50 as part of the driver assistance system 100, which is configured to determine the current global position of the vehicle, in particular by communicating with a satellite-based positioning system.

[0108] The global positioning device 50 may thereby be coupled to the visibility determination device 21 and to the control module 30 .

[0109] The visibility determination device 21 is configured to access data in an external data source 200 based on the current location based on the current location transmitted by the global positioning device to receive accurate location specific data, resulting in improved determination of characteristic visibility parameters.

[0110] Additionally, the current position is also available to the control module 30 to generate control signals that are further based on the current position.

[0111] FIG. 3 shows a flow chart of an embodiment of a method for controlling a vehicle according to the invention.

[0112] In step S1, at least two driver assistance models are provided that correspond to different visibility conditions in the vehicle's environment.

[0113] In step S2, characteristic visibility parameters are determined that describe the current visibility conditions in the vehicle's environment. For this purpose, step S2 comprises a sub-step S2.1 of optically detecting the environmental conditions of the vehicle's environment and a sub-step S2.2 of determining the characteristic visibility parameters depending on the detection results of said sub-step S2.1.

[0114] In step S3, the characteristic visibility parameter of step S2 is compared to a critical threshold.

[0115] If the characteristic visibility parameter exceeds the threshold, the method continues to step S9, otherwise the method continues to step S4.

[0116] In step S4, a primary driving assistance model is selected from the at least two driving assistance models provided in step S1 depending on the determined characteristic visibility parameters of step S2.

[0117] For that purpose, the characteristic visibility parameter is compared in substep S4.1 with a first predetermined range of values ​​and with a second predetermined range of values.

[0118] If the parameter is located within the first range of values, the method proceeds to sub-step S4.2a, in which a first model of the at least two driving assistance models of step S1 is selected as the primary assistance model.

[0119] If the parameter is located in the second range of values, the method proceeds to sub-step S4.2b, in which a second model of the at least two driving assistance models of step S1 is selected as the primary assistance model.

[0120] For example, a first model may correspond to a case of "good" visibility conditions and a second model may correspond to a case of "bad" visibility conditions, and in particular the models may correspond to different illumination levels in the vehicle's environment.

[0121] In step S5, environmental data of the vehicle's environment is provided, including optically detecting environmental conditions of the vehicle's environment.

[0122] In step S6, control signals based on the provided environmental data of step S5 in conjunction with the selected primary assistance model of step S4 are generated and output to actuators, in particular to actuators of the vehicle, such as actuators of the drive system or of the steering system.

[0123] In step S7, the method is paused for a predetermined duration before the method proceeds to step S8.

[0124] In step S8, a criterion is evaluated whether the currently used primary support model is to be re-evaluated. If not, the method proceeds to step S5, corresponding to the loop case S5-S6-S7.

[0125] If a re-evaluation is to be performed, the method exits the loop S5-S6-S7 and proceeds again to step S2.

[0126] The criteria used in step S8 can be time-based or iteration-based, i.e. by comparing the number of iterations or time the method has already spent in the loop S5-S6-S7 with respective predetermined limit values.

[0127] If the method proceeds to step S9, the driver assistance system is switched to standby mode in step S9, followed by a pause of a predetermined duration in S10 before the method proceeds again to step S2.

[0128] The method shown in FIG. 3 offers the advantage of providing the opportunity to select from a variety of assistance models (step S4) the main assistance model that is most appropriate for processing the detected environmental data and issuing control signals based thereon (steps S5 and S6).

[0129] Therefore, instead of providing one advanced assistance model that takes into account various visibility scenarios, the proposed method presents various "smaller" assistance models, from which the most suitable one is selected prior to processing the detected environmental data, thus significantly reducing the delay of the overall method.

[0130] It is emphasized that the present invention is in no way limited to the embodiments and their features described above: the invention also includes variants of these embodiments, in particular those resulting from variants and / or combinations of individual or several features of the described embodiments, within the scope of protection of the independent claims.

[0131] 10 Environmental sensor module 11 Optical sensor 12 Further sensor 20 Assistance model module 21 Visibility determination device 21a Data interface 21b Computing unit 22 Model storage device 23 Model selection device 30 Control module 31 Path planning module 40 Notification device 50 Global positioning device 100 Driver assistance system 200 External data storage 210 Database with weather data 220 Database with sun data 230 Database with geographical surface data 300 Vehicle actuators

Claims

1. A driving assistance system (100) for a vehicle, comprising: a control module (30) configured to generate a control signal; an environmental sensor module (10) configured to provide environmental data of the vehicle's environment based on detection results of at least one optical sensor (11); and an assistance model module (20) for providing a driving assistance model to the control module (30), wherein the assistance model module (20) comprises: a model storage device (22) in which at least two driving assistance models corresponding to different visibility conditions in the vehicle's environment are stored; a visibility determination device (21) configured to determine a characteristic visibility parameter describing a current visibility condition in the vehicle's environment; and a model selection device (23) configured to select a main driving assistance model from the at least two driving assistance models stored in the model storage device (22) depending on the characteristic visibility parameter determined by the visibility determination device (21). The control module (30) is coupled to at least the assistance model module (20) and the environmental sensor module (10) and is configured to generate a control signal based on the data provided by the environmental sensor module (10) in combination with the selected primary assistance model provided by the assistance model module (20).

2. The driving assistance system (100) of claim 1, wherein the visibility determination device (21) is coupled to the environmental sensor module (10) and configured to determine the characteristic visibility parameter depending on the detection result of the at least one optical sensor (11).

3. The driving assistance system (100) of claim 1, wherein the at least one optical sensor (11) is a camera.

4. The driving assistance system (100) of claim 1, wherein the visibility determination device (21) comprises a data interface connectable to an external data source containing data relating to environmental conditions of the vehicle's environment, and is configured to process the data received from the external data source (200) via the data interface and to determine the characteristic visibility parameter accordingly.

5. A driving assistance system (100) as described in claim 4, wherein the visibility determination device (21) is configured to determine the characteristic visibility parameter depending on the detection result of the at least one optical sensor (11) and depending on the data received from the external data source (200), in particular depending on a weighted average value of the received data.

6. The driving assistance system (100) of claim 4, wherein the data received from the external data source (200) includes real-time data regarding sun position and / or sun angle.

7. The driving assistance system (100) of claim 4, wherein the data received from the external data source (200) comprises geographic 2D and / or 3D ground data.

8. The driving assistance system (100) of claim 4, wherein the data received from the external data source (200) includes real-time weather data.

9. The driving assistance system (100) of claim 1, further comprising a global positioning device (50) configured to determine a current global position of the vehicle, in particular by communicating with a satellite-based positioning system.

10. A driving assistance system (100) as described in claim 9, wherein the visibility determination device (21) is configured to access data in an external data source (200) and receive location-specific data based on a global position determined by the global positioning device (50).

11. The driving assistance system (100) of claim 1, wherein the visibility conditions described by the characteristic visibility parameters are brightness level, field of view, brightness contrast, or illumination level, or a combination of two or more of the foregoing, in the vehicle's environment.

12. The driving assistance system (100) of claim 1, wherein the model selection device (23) is configured to select a first model of the at least two models as a primary assistance model if the determined characteristic visibility parameter is within a first predetermined range, and to select a second model of the at least two models as a primary assistance model if the determined characteristic visibility parameter is within a second predetermined range.

13. The driving assistance system (100) of claim 1, wherein the model selection device (23) is configured to select the primary assistance model by a neural network based method and / or by a machine learning method.

14. The driving assistance system (100) of claim 1, wherein the assistance model module (20) is further configured to switch the driving assistance system (100) to a standby mode when the determined characteristic visibility parameter exceeds a critical threshold, and during the standby mode, the environmental sensor module (10) is specifically switched off.

15. The driving assistance system (100) according to claim 1, wherein the control module (30) comprises a path planning device configured to determine path data for a path to be followed by the vehicle based on the data provided by the environmental sensor module (10) in combination with the selected primary assistance model provided by the assistance model module (20), and configured to generate the control signals based on the determined path data, in particular for controllable actuators of a drive system and / or a steering system of the vehicle.

16. The driving assistance system (100) of claim 1, wherein the at least two driving assistance models stored in the model storage device (22) are vehicle localization models, each of which includes an algorithm for determining the position of the vehicle relative to its environment, in particular, the algorithm being a visual odometry algorithm.

17. A driving assistance system (100) as described in claim 1, wherein the at least two driving assistance models stored in the model storage device (22) are perception models, each of which includes an algorithm for determining the position of an object in the vehicle's environment, in particular, the algorithm being an obstacle detection algorithm.

18. The driving assistance system (100) of claim 1, wherein the environmental sensor module (10) comprises at least one further sensor (12), in particular an ultrasonic sensor or a LIDAR sensor, configured to detect an environmental condition of the vehicle's environment.

19. The driving assistance system (100) according to claim 1, further comprising at least one actuator (40) controllable by the control signal generated by the control module (30), in particular the at least one actuator (40) being an optical and / or acoustic notification device configured to notify a driver of the vehicle, and / or the control module (30) being capable of being coupled to at least one actuator (300) of the vehicle controllable by the control signal generated by the control module (30), in particular the at least one actuator (300) of the vehicle being part of a drive system of the vehicle or of a steering system of the vehicle.

20. A vehicle equipped with a driving assistance system (100) according to any one of claims 1 to 19.

21. A method for controlling a vehicle by a driving assistance system (100) of the vehicle, comprising the steps of: providing at least two driving assistance models corresponding to different visibility conditions in an environment of the vehicle; providing environmental data of the environment of the vehicle including a sub-step of optically detecting environmental conditions of the environment of the vehicle; determining a characteristic visibility parameter describing a current visibility condition in the environment of the vehicle; selecting a primary driving assistance model from the provided at least two driving assistance models in response to the determined characteristic visibility parameter; and generating a control signal based on the provided environmental data in conjunction with the selected primary assistance model.

22. The method according to claim 21, further comprising the step of outputting the control signal to an actuator (300) of the vehicle that is controllable by the control signal, in particular the actuator (300) being part of a drive system and / or a steering system of the vehicle.

23. A computer program product for use in a driving assistance system (100) according to any one of claims 1 to 19, comprising instructions which, when executed, cause the driving assistance system (100) to carry out the method according to claim 21.

Images