Vehicle system for generating a noise inventory, a deployment system and an ego vehicle

Abstract

The invention relates to a vehicle system (1) for generating a noise register, for displaying ambient noise on maps or road maps, the vehicle system (1) comprising a sensor system designed to capture ambient noise as environmental data, a reference unit (3) which is connected to the sensor system for receiving environmental data and is designed to provide the received environmental data with position data, an evaluation unit (4) which is designed to subdivide the environment data into position-independent environment data and position-dependent environment data as well as unrecognized environment data on the basis of a classifier, and wherein the evaluation unit (4) is further configured to subdivide position-dependent environmental data detected by the classifier into noise categories for the generation of a position-dependent noise register. Furthermore, the invention relates to a provisioning system and an ego vehicle.

Description

[0001] The invention relates to a vehicle system for generating a noise map and for displaying ambient noise on maps or road maps. Furthermore, the invention relates to a provisioning system and an ego vehicle.

[0002] A noise map is used to depict ambient noise on maps or road maps. However, such maps primarily refer to the noise sources: road traffic, rail traffic, air traffic, and industry.

[0003] Compiling a noise inventory using stationary measurements or special measuring vehicles is very complex, not comprehensive, and only possible at discrete, usually long-ago points in time. A comprehensive and up-to-date database would be beneficial for noise abatement measures, determining land values ​​or real estate, planning stays, or finding a location for a business in quiet or busy areas. Furthermore, currently only the noise level is recorded and noise is not categorized by source, such as traffic noise, aircraft noise, people, or businesses.

[0004] Previous noise detection in the environment of vehicles refers to internal vehicle noises for detecting the vehicle condition or noises for detecting the road quality.

[0005] DE102012219631A1 discloses a device for detecting at least one road surface unevenness.

[0006] It is an object of the invention to provide a vehicle system which is improved with regard to ambient noise.

[0007] Furthermore, a further object is to provide a delivery system which is improved with regard to ambient noise and an ego vehicle.

[0008] The object is achieved by a vehicle system having the features of claim 1, a provisioning system having the features of claim 12, and an ego vehicle having the features of claim 18.

[0009] The subclaims list further advantageous measures which can be suitably combined to achieve further advantages.

[0010] The task is solved by a vehicle system for generating a noise register, for displaying ambient noise on maps or road maps, the vehicle system comprising a sensor system designed to detect ambient noise as environmental data, a reference unit which is communicatively connected to the sensor system for receiving the environmental data and is designed to provide the received environmental data with position data, an evaluation unit which is designed to subdivide the environmental data into position-independent environmental data and position-dependent environmental data as well as unrecognized environmental data on the basis of a classifier, and wherein the evaluation unit is further designed to subdivide position-dependent environmental data recognized by the classifier into noise categories for generating a position-dependent noise register.

[0011] In particular, the reference unit is designed to provide the received environmental data with a time stamp, wherein the evaluation unit is designed to subdivide the detected environmental data into position-dependent and time-dependent environmental data on the basis of the classifier, and wherein the evaluation unit is further designed to subdivide position-dependent and time-dependent environmental data detected by the classifier into noise categories for generating a position-dependent and time-dependent noise register.

[0012] Such a noise register indicates the noises depending on the position - and, if applicable, time-dependent - along the lanes / roads.

[0013] A sensor system can comprise one or more sensors. Ambient noise can be captured as acoustically executable ambient signals, which represent the ambient data or can be converted into such data, for example, through preprocessing.

[0014] The reference unit can be integrated as a software module, for example within the sensor system, or run on a separate ECU.

[0015] A classifier can, for example, be a machine learning process and / or a target-actual comparison, for example, a comparison with sounds in a reference database, or a pattern recognition algorithm. Position-independent and time-independent environmental data, for example, are not exhaustive: the siren of an emergency vehicle, human speech (where keywords such as "Hello Siri" can also be recognized), or vehicle noises emanating from the vehicle itself: drivetrain, interior, driving over speed bumps, potholes, etc. These can be stored in the classifier as stored position-independent and time-independent environmental data.

[0016] Position-dependent and time-dependent environmental data include, for example, aircraft, increased traffic noise, for example at busy intersections, increased children's voices at schools, kindergartens and playgrounds, increased pedestrian voices, for example in pedestrian zones / shopping streets, bicycle bells in conjunction with tire wear on busy cycle paths or at tunnel entrances; here, for example, higher traffic noise is characteristically present.

[0017] The vehicle system can be configured to initially store the environmental data, including position data and a timestamp, in a storage unit and only analyze it when sufficient computing capacity is available, for example, when the vehicle is stationary. If sufficient computing capacity is available, an online analysis can be performed.

[0018] After analysis, the environmental data can be discarded.

[0019] Furthermore, the vehicle can have an online map in which the position along with the noise or environmental data is simply entered.

[0020] Furthermore, such a noise register can be displayed to the driver for information.

[0021] The sensor system is designed to record the acoustic signals of the environment for a configured measurement period at fixed intervals (cyclical) or position-dependently, if position or route data are available via the vehicle data bus.

[0022] In a further development, the evaluation unit is further configured to store at least the unrecognized environmental data in a sound file. If the classifier is not yet able to classify the environmental data, it can be categorized according to structured features, for example, similar to a directory tree into higher-level and lower-level categories: vehicle exterior environment, lower-level category: traffic, siren.

[0023] In a further development, the evaluation unit is further configured to use unrecognized environmental data stored in the sound file to expand the classifier. If the classifier is configured as a neural network, it can be retrained using the unrecognized environmental data to perform this expansion, or the expansion can be performed by querying the user. For example, an unknown noise might be a siren sound during a vacation, in which case the classifier can be expanded to include this siren sound by manually querying the driver.

[0024] This allows more noise types / causes to be identified, which means that the classifier and thus also the noise register continuously grows.

[0025] Furthermore, an interface is provided for transmitting at least the unrecognized environmental data stored in the sound file to an external server. There, the classifier can be retrained using the unrecognized environmental data. Subsequently, an updated version of the classifier can be loaded from the external server into the vehicle system. With the updated classifier, for example, more sounds / environmental data can be recognized, thereby expanding the noise register. Enhanced sound detection can enable new functions in the vehicle or improve existing functions.

[0026] In a further development, a correlation unit is provided for correlating the position-dependent or position-dependent and time-dependent environmental data in the noise register with corresponding vehicle adjustment parameters. The vehicle adjustment parameters can also be stored in the noise register or correlated in another way.

[0027] For example, if there is a correlation with children's noises, such as at a kindergarten, the speed can be adjusted, the sensor systems can be made more sensitive to detecting children, and the braking system can be prepared using the corresponding correlated vehicle settings. Furthermore, a warning message can be displayed. In addition, a correlation with sign recognition can be used to achieve greater prediction accuracy.

[0028] For example, if the signal is correlated with entering a tunnel, automatic and earlier activation of the low beams can be enabled, as the human eye is deceived by brighter lighting in the tunnel entrance area. A brightness sensor, which usually controls automatic headlight control, is also deceived by changes in lighting in the entrance and exit areas of a tunnel, leading to a delayed activation of the headlight control. Such entry into a tunnel can be detected, for example, by the characteristically louder traffic noise and echo.

[0029] If the noise register shows the mapping of road irregularities, which have been detected, for example, by acoustically detectable infrastructure information such as speed bumps, curbs, potholes, etc., this information can be used to adapt the chassis, for example to adjust the damping, local evasive action, and adjustment of the driving speed.

[0030] In environments with increased traffic noise, for example, the radio volume can be adjusted.

[0031] In particular, in the case of increased noise, for example aircraft noise, a filter that filters out the characteristics of the aircraft noise can be activated or amplified in the vehicle.

[0032] In a further development, a control unit is provided for receiving the vehicle adjustment parameters based on the noise register and the current vehicle position. The control unit can, for example, adjust the actuators according to the vehicle adjustment parameters. These can be the vehicle adjustment parameters stored in the noise register or they can be taken into account. During a telephone conversation, the level of support and intervention and intensity of the assistance systems, such as ACC, Front Assist, Lane Keeping, etc., can be increased depending on the emotional nature of the conversation.

[0033] In a further embodiment, a linking unit is provided for linking the currently detected environmental data to predefined vehicle setting parameters, with a control unit being provided for receiving the vehicle setting parameters. This can be the detected position-dependent and time-dependent environmental data as well as the detected position-independent and time-independent environmental data.

[0034] For example, if sirens are detected, it might be a vehicle setting parameter that "turns off the radio." However, these vehicle setting parameters are not linked to the noise register, but only to the current detection.

[0035] In further development, the timestamp includes not only the time of day but also the time of week / day of the week. Furthermore, in addition to the time of day and day of the week, information about holidays / vacation seasons can also be included, enabling a differentiated noise inventory.

[0036] In a further embodiment, a second reference unit can be provided which is communicatively connected to the sensor system for receiving the environmental data and is designed to provide the received environmental data with georeferenced features, and wherein the evaluation unit is designed to correlate at least the position-dependent or the position-dependent and time-dependent environmental data with the georeferenced features, and wherein the evaluation unit is further designed to divide position-dependent or position-dependent and time-dependent environmental data recognized by the classifier into noise categories for generating a position-dependent and georeferenced or a position-dependent and georeferenced and time-dependent noise register.

[0037] This means that the sounds are related or correlated to georeferenced features, such as weather data, infrastructure information, health data, sociological and economic data, etc., thus enabling a differentiated noise inventory. The second reference unit can be integrated within the sensor system or, for example, on an ECU as a software module.

[0038] In a further embodiment, an interface is provided for transmitting the unrecognized environmental data and / or the recognized classified environmental data and / or the noise register to an external server. In particular, the bidirectional interface can also be configured to receive an update from a classifier so that multiple noises can be identified. For example, the environmental data can first be stored and then transmitted to the external server if sufficient computing capacity is available. In particular, the unrecognized, i.e., unclassifiable environmental data can also be transmitted to the external server to expand the classifier through retraining.

[0039] Furthermore, the object is achieved by a provision system for generating a noise register, for displaying ambient noise on maps or road maps, the provision system comprising an interface which is designed to receive ambient noise as ambient data or pre-processed ambient data from one or more detection vehicles, wherein the ambient data or pre-processed ambient data comprise at least a time stamp and position data, further comprising a computing unit which is designed to subdivide the environmental data into position-independent and time-independent environmental data and into position-dependent and time-dependent environmental data and into unrecognized environmental data on the basis of a classifier, and wherein the computing unit is further configured to divide position-dependent and time-dependent environmental data detected by the classifier into noise categories to generate a position-dependent and time-dependent noise register.

[0040] The provisioning system can be designed as a cloud, which receives corresponding environmental data or pre-processed environmental data through crowd sourcing, i.e. through many different detection vehicles.

[0041] Environmental data can be raw sensor data or preprocessed data. Environmental data can also include metadata from the vehicle, such as driving speed, status signals, etc.

[0042] The detected position-dependent and time-dependent environmental data can be used to validate previous results or expand the noise register based on the land / road map, for example, for areas that do not yet have a noise register assigned. Furthermore, this data is available to a developer, for example, to improve the classifier.

[0043] Furthermore, the computing unit can be configured to store the environmental data not recognized by the classifier in a memory unit to improve the classifier. Unrecognized environmental data can thus be used to build the classifier. If no classifier is yet available to classify the environmental data, classification can be performed according to structured features, for example, similar to a directory tree into higher-level and lower-level categories: vehicle exterior environment, or into a lower-level category: traffic, siren.

[0044] If the classifier is designed as a neural network, the unrecognized environmental data can be used to retrain or redevelop the classifier.

[0045] The results can be analyzed and corresponding environmental data can be assigned to a class in order to be recognized by the classifier in the future.

[0046] In this way, the functionality of both the classifier and the noise register grows continuously over time. More noise types / causes can be differentiated and made available to external users, as well as reacted to within the vehicle.

[0047] The noise register can be used not only for vehicles but also for other customers, for example consultants, agents and service providers, such as real estate agents, town planning offices, or the noise register can be taken into account in planning: e.g. the location of sought-after businesses in certain traffic / noise locations, the avoidance of areas where demonstrations take place.

[0048] In a further embodiment, the computing unit is designed to store unrecognized environmental data in a storage unit in order to improve the classifier.

[0049] Furthermore, the classifier is designed, for example, as a machine learning method, such as a neural network, and / or by comparison with a reference database and / or as a pattern recognition algorithm. It can be further trained using the unrecognized environmental data.

[0050] Improvement can also be achieved through automated retraining of the machine learning process or through parameterization of analytical models by optimizers. The success of the optimization can be verified using the available data and an automatic comparison of the classifier with the previous classifier. Deteriorations are discarded, thus generating a self-learning system that can be continuously improved and developed with minimal personnel input.

[0051] In a further embodiment, the computing unit is designed to correlate known position-dependent and time-dependent environmental data with vehicle parameters to be adjusted, and to store these vehicle parameters in the noise register correlated with the position-dependent and time-dependent environmental data. When such a noise register is transmitted to a vehicle, control parameters can be generated based on these vehicle parameters or they can be taken into account during generation. For example, in the case of a correlation with children's noises, such as in a kindergarten, the speed can be adjusted, the sensor systems can be made more sensitive to detecting children, and if necessary, the braking system can be prepared using the corresponding vehicle adjustment parameters. This adjustment can only take effect during the day and not at night.

[0052] Furthermore, the provisioning system can correlate the environmental data with georeferenced features to generate a position-dependent and time-dependent georeferenced noise inventory. This means that the noises are related or correlated to other georeferenced features, such as weather data, infrastructure information, health data, sociological and economic data, etc., thus enabling a differentiated noise inventory.

[0053] Using such a correlation, this data can be taken into account in route planning, for example, to avoid roads with many potholes, or in the construction industry to consider the location of businesses in certain traffic / noise areas.

[0054] Furthermore, the object is achieved by an ego vehicle comprising a navigation device for inputting a destination route and a vehicle control system, the vehicle control system comprising a coupling point for coupling to a provision system, wherein the provision system provides a noise register for displaying ambient noise as position-dependent ambient data on maps or road maps and wherein the vehicle control system is designed to receive a noise register and wherein the vehicle control system is further designed to generate the vehicle parameters depending on the current position and the noise register.

[0055] In particular, the delivery system is a delivery system as described above.

[0056] Thus, for example, the noise register can also contain vehicle parameters to be set along the target route, which can also be taken into account when generating the currently set vehicle parameters, or based on which the vehicle parameters to be set are generated by the ego vehicle. Such parameters can also include, for example, increased sensitivity for the sensor system detecting children along a playground. Furthermore, a one-time download / installation of such a noise register can be carried out, for example, for a fee. This allows special features along the route, such as kindergartens, etc., to be taken into account in advance.

[0057] The noise register can also be displayed to the driver using a display unit.

[0058] During the journey, the operation of one or more functions is changed depending on the position and the information from the noise register. For example, if it is known from the noise register that increased aircraft noise is to be expected at this position, but this is not relevant for carrying out the maneuver, a filter that filters out the characteristics of the aircraft noise can be activated or amplified. The same can be done for speech recognition. Another example is that the monitoring function uses the available data from the noise register to calculate the presence of pedestrians / children in the driving corridor, thereby increasing the sensitivity of a function for detecting pedestrians in the vicinity of the ego vehicle. This increase can extend to all sensor systems, e.g. camera, radar or lidar systems.

[0059] Furthermore, the vehicle controller can be configured to transmit the target route to the provisioning system and to receive at least the noise map along the target route, and the vehicle controller can be configured to generate the vehicle parameters depending on the current position and the noise map. This ensures that an up-to-date noise map is always available.

[0060] In addition, the computing and storage capacity is not unnecessarily burdened.

[0061] In a further development, sensors are provided for recording ambient noise as real environmental data, with the vehicle controller being configured to search for the ambient data present in the noise register in the real environmental data, depending on the position. This allows both the validation of the noise register and the validation of the sensors of the host vehicle to be achieved.

[0062] Furthermore, the vehicle controller can be configured to improve the noise map and / or the classifier based on the real environmental data. This can be done by retraining the classifier, or the detected environmental data can be added to the noise map in a position-dependent and time-dependent manner.

[0063] Furthermore, in a further embodiment, the vehicle control system can be designed to correlate the received environmental data of the noise register with georeferenced features and to generate the vehicle parameters depending on the current position and the noise register as well as the georeferenced features.

[0064] This enables better operation, especially of an autonomously operated ego vehicle.

[0065] Further features and advantages of the present invention will become apparent from the following description with reference to the accompanying figures, which schematically show: Fig. 1: a vehicle system for generating a noise register, Fig. 2: a delivery system according to the invention, Fig. 3: an inventive ego vehicle.

[0066] Fig. 1 shows a vehicle system 1 for generating a noise register, for displaying ambient noise on maps or road maps in a vehicle.

[0067] This comprises a sensor system comprising several acoustic sensors S1, S2, Sn for recording ambient noise as environmental data. Sensors for detecting acoustic signals, such as airborne and structure-borne sound, e.g., a microphone or vibration sensor, can be provided.

[0068] The recorded acoustic signals are transmitted to a reference unit 3 and read into an integrated, structured data storage system as a runtime environment. This can be configured as a ring buffer / buffer.

[0069] Alternatively, the reference unit 3 can be integrated into the sensor system as a software module.

[0070] The reference unit 3 can provide the received environmental data with a time stamp and position data.

[0071] The timestamp can include time-related characteristics such as time of day, day of the week, holiday, season, etc.

[0072] Furthermore, an evaluation unit 4 is present, which can also be formed on the ECU.

[0073] The evaluation unit 4 has a classifier that first subdivides the environmental data into position-independent and time-independent environmental data, position-dependent and time-dependent environmental data, and unrecognized environmental data. The classifier can be a machine learning method, such as a neural network. Furthermore, the classifier can be implemented as a pattern recognition algorithm.

[0074] Position-dependent and time-dependent environmental data include, for example, aircraft, increased traffic noise, for example at busy intersections, increased children's voices at schools, kindergartens and playgrounds, increased pedestrian voices, for example in pedestrian zones / shopping streets, bicycle bells in conjunction with tire wear on busy cycle paths or entrances to tunnels; here, for example, higher traffic noise is characteristically present.

[0075] Furthermore, the classifier can subdivide detected position-dependent and time-dependent environmental data into noise categories, i.e. noise classes, here for example class: children's voices, class: aircraft noise, to generate a position-dependent and time-dependent noise register.

[0076] Furthermore, the vehicle system 1 can be configured to initially store the environmental data, including position data and a time stamp, in a storage unit and only analyze it when sufficient computing capacity is available, for example, when the vehicle is stationary. If sufficient computing capacity is available, an online analysis can be performed.

[0077] Furthermore, the environmental data can be enriched with metadata from the vehicle, such as speed, status signals, etc. The environmental data can also be generated at fixed intervals or position-dependently, as well as for a configured measurement duration.

[0078] The evaluation unit 4 can be configured to cyclically store environmental data with the enriched metadata, for example, based on a configurable event (trigger), in order to enable subsequent analysis. This prevents the blocking of computing resources necessary for smooth vehicle operation, particularly in autonomous driving.

[0079] Position-dependent and time-dependent environmental data can be used to build the noise inventory. This allows an existing map, such as a road map, to be enriched with the recorded sounds at the respective locations over time.

[0080] Furthermore, the evaluation unit 4 can be designed to also store the position-independent and time-independent environmental data in a storage unit.

[0081] Position-independent and time-independent environmental data are, for example, not exhaustively listed: a siren of an emergency vehicle, human speech, where keywords such as "Hello Siri" can also be recognized, or vehicle noises emanating from the vehicle itself: drivetrain, interior, driving over speed bumps, potholes, etc. These can be stored in the classifier as stored position-independent and time-independent environmental data.

[0082] The unrecognized environmental data is also saved. If the classifier is not yet able to classify the sensor data, it can be categorized according to structured characteristics, for example, similar to a directory tree with higher-level and lower-level categories: vehicle environment outside, lower-level category: traffic, siren, etc.

[0083] This unrecognized environmental data can be used to expand the classifier. If the classifier is designed as a neural network, it can be retrained using the unrecognized environmental data to perform this expansion, or the expansion can be performed by asking the user what the noise is. For example, an unrecognized noise might be a siren on vacation, and a manual query to the driver can expand the classifier to include this siren noise.

[0084] This allows more noise types / causes to be identified, which means that the classifier and thus also the noise register continuously grows.

[0085] Furthermore, this unrecognized environmental data can be transmitted to an external server, downloaded for analysis and a corresponding extension of the classifier.

[0086] Furthermore, a correlation unit 5 is provided for correlating the position-dependent and time-dependent environmental data in the noise register with corresponding vehicle adjustment parameters. The vehicle adjustment parameters can also be stored in the noise register or correlated in another way.

[0087] If children's voices are identified as a detected noise from the noise register, these can be correlated with, for example, kindergartens / playgrounds or schools. This can be used to temporally adjust the speed at this location as a vehicle parameter, increase the sensitivity of the sensor systems for detecting children, or even issue a warning to the driver. The adjustment can be time-dependent, for example, only during the day.

[0088] If bicycle bells are identified as a noise from the noise register, this can be correlated with bicycle paths. Vehicle parameters to be set at this position, known from the noise register, include, for example, a legally prescribed distance for overtaking during maneuver planning and trajectory planning, or a minimum distance from the road edge.

[0089] If the characteristic noise that occurs when entering a tunnel is known from the noise register, the low beam headlights can be activated automatically, for example.

[0090] If the noise register indicates that increased aircraft noise is to be expected at this location, but this is not relevant for the maneuver, a filter that filters out the characteristics of the aircraft noise can be activated or amplified. The same can be done for speech recognition.

[0091] Furthermore, the detected position-dependent and time-dependent environmental data can be related or correlated to georeferenced features. Such georeferenced features include, for example, weather data, infrastructure information, health data, sociological and economic data. This allows the classifier and the vehicle parameters to be adjusted to be refined.

[0092] Such a noise register allows noises to be assigned even if sensors S1, S2, Sn fail, and to react with appropriate vehicle parameters.

[0093] Likewise, the sensitivity can be increased according to the “noises” in the noise register and thus be better recognized during a journey.

[0094] This can increase safety for both the driver and the environment in autonomous vehicles.

[0095] Furthermore, the position-independent and time-independent environmental data in the noise register can also be correlated with corresponding vehicle setting parameters. This can, for example, be increased braking readiness upon detection of a siren, switching off the radio, etc. In particular, the vehicle system 1 can be configured to detect this position-independent and time-independent environmental data in the sensor data online, i.e., in real time.

[0096] Fig. Figure 2 shows a provisioning system 6 according to the invention. This can be provided on an external server, for example, a cloud 14.

[0097] The provisioning system 6 has an interface 8. Using the interface 8, the provisioning system 6 can capture ambient noise as environmental data from multiple detection vehicles 10. This data can be recorded by the corresponding sensors S1, S2, Sn of the detection vehicle 10. This data can be raw data or preprocessed data. Processed data from the field of acoustics can be, for example, filters, frequency spectra, or signal intensity.

[0098] In particular, the transmitted environmental data includes a timestamp and position data. The timestamp can include time-related characteristics such as time of day, day of the week, holiday, season, etc. The environmental data is recorded by the vehicle at fixed intervals or based on its position and then transmitted. The environmental data can include additional georeferenced features. These can include, for example, weather data, infrastructure information, health data, or sociological or economic data. Furthermore, metadata from the recording vehicle, such as driving speed, status signals, etc., can be added to the environmental data.

[0099] The environmental data is stored, for example, in the recording vehicle 10 and transmitted via a gateway 9.

[0100] The signals georeferenced in this way are transferred through the gateway 9 to the delivery system 6 and analyzed there.

[0101] Alternatively or additionally, the provisioning system 6 may also add such georeferenced features.

[0102] For this purpose, the provisioning system 6 has a computing unit 11. The computing unit 11 is designed to use a classifier to subdivide the environmental data into position-independent and time-independent environmental data and into position-dependent and time-dependent environmental data, as well as to subdivide it into unrecognized environmental data.

[0103] Unrecognized data can be used to build the classifier. If the raw data has not yet been classified by the classifier, it can be categorized according to structured characteristics, for example, similar to a directory tree with higher-level and lower-level categories: vehicle environment exterior, lower-level category: traffic, siren, etc.

[0104] For example, if the classifier is a neural network, the unrecognized environmental data can be used to automatically retrain the neural network or parameterize analytical models using optimizers. The success of the optimization can be verified using the available data and an automatic comparison of the new classifier with the previous one. Deteriorations are discarded, allowing the generation of a self-learning system that continuously improves and evolves with minimal human intervention.

[0105] The results can be analyzed and corresponding environmental data can be assigned to a class in order to be recognized by the classifier in the future, according to the structure.

[0106] In this way, the functionality of both the classifier and the noise register grows continuously over time. More noise types / causes can be differentiated and provided georeferenced to external users, such as vehicles, and the vehicle can react to them.

[0107] Furthermore, the detected environmental data is subdivided into position-independent and time-independent environmental data and into position-dependent and time-dependent environmental data.

[0108] Position-independent and time-independent environmental data are, for example, the siren of an emergency vehicle or speech; here, in particular, the classifier can be trained to recognize keywords such as “Hey Siri,” or vehicle noises that come from the drive, interior, driving over thresholds, or potholes.

[0109] If the classifier detects such position-independent and time-independent environmental data, corresponding vehicle parameters, such as slow driving, can be stored in the noise register.

[0110] Position-dependent and time-dependent environmental data include, for example, children's voices, entering a tunnel, bicycle bells, aircraft noise, etc.

[0111] If the classifier detects position-dependent and time-dependent environmental data, this can be used to generate the noise inventory. For this purpose, the provisioning system 6 can have a map / map that is, for example, enriched with the noise / environmental data at the detected position with the timestamp.

[0112] According to the detected position-dependent and time-dependent environmental data, a correlation can be made with one or more functions / vehicle parameters, which can also be stored in the noise register.

[0113] For example, the environmental data "children's voices" can be correlated with kindergartens / schools. Based on this correlation, a temporarily reduced speed or a predefined low speed, or, for example, an increase in the sensitivity of the sensor systems with regard to the detection of children, can be stored in the noise register as vehicle parameters.

[0114] For example, the environmental data "bicycle bells" can be correlated with bicycle paths. Based on this correlation, a corresponding distance from the bicycle path or a correspondingly reduced speed can be stored in the noise register during overtaking maneuvers.

[0115] The environmental data from entering a tunnel, which can be detected, for example, by characteristic higher traffic noise and echo, can be correlated with the tunnel. The low beams can be activated based on this correlation.

[0116] A filter can be activated for the aircraft / aircraft noise environmental data, which filters out the characteristics of the aircraft noise.

[0117] Furthermore, the position-independent and time-independent environmental data in the noise register can also be correlated with corresponding vehicle setting parameters. This can, for example, be increased braking readiness upon detection of a siren, switching off the radio, etc. In particular, the vehicle system 1 can be configured to detect this position-independent and time-independent environmental data in the sensor data online, i.e., in real time.

[0118] This storage of vehicle parameters is particularly useful when using the noise register in vehicles.

[0119] Such a noise register can be used in a variety of ways. For example, if environmental data has been correlated with georeferenced data, the noise register can be taken into account in various planning and route planning processes, such as locating businesses in certain traffic / noise locations or avoiding areas where demonstrations are taking place. The noise register can also be used for insurance products and for real estate agents and urban planning offices.

[0120] The noise inventory can also be used by vehicle manufacturers to improve the acoustic design of their vehicles. The noise inventory also allows for better real-time determination of current noise emissions based on traffic volume and merging this data with the noise inventory data. For this purpose, networking with traffic control centers that record traffic volumes using traffic monitoring systems, toll gates, induction loops, etc. can be provided.

[0121] Furthermore, such a noise register with stored vehicle parameters can be used in the vehicle to increase safety, for example as pre-crash measures.

[0122] Fig. 3 shows an ego vehicle 12 according to the invention. This can be designed as an autonomously driving ego vehicle 12.

[0123] The ego vehicle 12 includes a navigation device for entering a destination route.

[0124] Furthermore, the ego vehicle 12 comprises a coupling point 13, for example a bidirectional interface for coupling to a provisioning system 6.

[0125] This has a noise register as described above for displaying ambient noise on maps or road maps.

[0126] The ego vehicle 12 is coupled to the provisioning system 6 by means of the coupling point 13.

[0127] The noise register can then be transmitted to the ego vehicle 12.

[0128] Alternatively, the target route can first be transmitted to the provisioning system 6, and only the noise map along the route can be transmitted to the ego vehicle 12. This saves storage capacity. The ego vehicle 12 can generate the vehicle parameters depending on the current position and the noise map.

[0129] If the noise register itself contains vehicle parameters linked to the environmental data, these vehicle parameters can be adopted or at least taken into account.

[0130] Thus, when the ego vehicle 12 is traveling, the vehicle parameters of one or more functions are changed depending on the position and the information from the noise register. For example, if it is known from the noise register that increased aircraft noise is to be expected at this position and time, but this is not relevant for the execution of the maneuver, the ego vehicle 12 can activate a filter that filters out the characteristics of the aircraft noise or amplify such a filter. Another example is that the sensor system S1, S2, Sn calculates the occurrence of pedestrians / children in the travel corridor based on the available data from the noise register, thereby increasing the sensitivity of a function to

[0131] The detection of pedestrians in the vicinity of the ego vehicle 12 is increased. This can also be done for sensor systems, such as camera, radar, or lidar systems. Furthermore, the radio volume can be adjusted in environments with increased traffic noise, which is also known from the noise register.

[0132] If there is a correlation with other non-acoustically detectable, georeferenced data, such as weather data or socio-economic data in the noise register, ABS, ESP or other safety systems, such as lighting systems, can be preconditioned or activated accordingly if it is known from the noise register that, for example, many children are walking along the route, for example at 1 p.m. near a primary school.

[0133] If the ego vehicle 12 has sensors S1, S2, Sn for generating environmental data, these can be transmitted to the provisioning system 6 with a time stamp and position data. Furthermore, additional metadata of the ego vehicle 12 can be added to these environmental data.

[0134] Such a transmission can also take place, for example, when the ego vehicle 12 is stationary or in the parked position, in order to avoid using up unnecessary computing capacity while driving.

[0135] This allows the ego vehicle 12 to use a noise register even in areas that are being driven through for the first time without generating it itself.

[0136] Such a noise register can, for example, be displayed to the driver, with appropriate warnings such as “Caution: Aircraft Noise”. List of reference symbols 1 vehicle system 2 users 3 Reference unit 4 Evaluation unit 5 Correlation unit 6 Provisioning system 7 external servers 8 Interface 9 Gateway 10 recording vehicles 11 Computing unit 12 Ego vehicle 13 Coupling point 14 Cloud QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] DE102012219631A1

[0005]

Claims

[1] Vehicle system (1) for generating a noise register, for displaying ambient noise on maps or road maps, the vehicle system (1) comprising a sensor system designed to detect ambient noise as environmental data, a reference unit (3) which is communicatively connected to the sensor system for receiving the environmental data and is designed to provide the received environmental data with position data, an evaluation unit (4) which is designed to subdivide the environmental data into position-independent environmental data and position-dependent environmental data as well as unrecognized environmental data on the basis of a classifier, and wherein the evaluation unit (4) is further designed to subdivide position-dependent environmental data recognized by the classifier into noise categories in order to generate a position-dependent noise register. [2] Vehicle system (1) according to claim 1, characterized by that the reference unit (3) is designed to provide the received environmental data with a time stamp, wherein the evaluation unit (4) is designed to subdivide the detected environmental data into position-dependent and time-dependent environmental data on the basis of the classifier, wherein the evaluation unit (4) is further designed to divide position-dependent and time-dependent environmental data recognized by the classifier into noise categories in order to generate a position-dependent and time-dependent noise register. [3] Vehicle system (1) according to claim 1 or 2, characterized by that the evaluation unit (4) is further designed to store at least the unrecognized environmental data in a noise file. [4] Vehicle system (1) according to claim 3, characterized bythat the evaluation unit (4) is further designed to use unrecognized environmental data stored in the noise file to expand the classifier. [5] Vehicle system (1) according to claim 4, characterized by that an interface (8) is provided for transmitting at least the unrecognized environmental data stored in the sound file to an external server (7). [6] Vehicle system (1) according to one of the preceding claims, characterized by that a correlation unit (5) is provided for correlating the position-dependent environmental data in the noise register with corresponding vehicle setting parameters. [7] Vehicle system (1) according to claim 6, characterized by that a control unit is provided for receiving the vehicle setting parameters based on the noise register and a current vehicle position. [8] Vehicle system (1) according to one of the preceding claims, characterized bythat a linking unit is provided for linking the currently detected environmental data to predetermined vehicle setting parameters, wherein a control unit is provided for receiving the vehicle setting parameters. [9] Vehicle system (1) according to one of the preceding claims, characterized by that a second reference unit is provided which is communicatively connected to the sensor system for receiving the environmental data and is designed to provide the received environmental data with georeferenced features, and wherein the evaluation unit (4) is designed to correlate at least the recognized position-dependent environmental data with the georeferenced features, and wherein the evaluation unit (4) is further designed to subdivide position-dependent environmental data recognized by the classifier into noise categories for generating a position-dependent and georeferenced noise register. [10] Vehicle system (1) according to one of the preceding claims, characterized by that an interface (8) is provided for transmitting the unrecognized environmental data and / or the recognized classified environmental data and / or the noise register to an external server (7). [11] Provision system (6) for generating a noise register, for displaying ambient noise on maps or road maps, the provision system (6) comprising an interface (8) which is designed to receive ambient noise as ambient data or pre-processed ambient data from one or more detection vehicles (10), wherein the ambient data or pre-processed ambient data comprise at least a time stamp as well as position data, further comprising a computing unit (11) which is designed, on the basis of a classifier, to subdivide the ambient data into position-independent and time-independent ambient data and into position-dependent and time-dependent ambient data and into unrecognized ambient data, and wherein the computing unit (11) is further designed toto divide position-dependent and time-dependent environmental data detected by the classifier into noise categories to generate a position-dependent and time-dependent noise register. [12] Provisioning system (6) according to claim 11, characterized by that the classifier is designed as a machine learning method and / or by comparison with a reference database and / or as a pattern recognition algorithm. [13] Provisioning system (6) according to claim 11 or 12, characterized by that the computing unit (11) is designed to store the environmental data not recognized by the classifier in a storage unit in order to improve the classifier. [14] Provisioning system (6) according to one of the preceding claims 11 to 13, characterized bythat the computing unit (11) is designed to effect a data correlation of known position-dependent and time-dependent environmental data with vehicle parameters to be set and to store these vehicle parameters in the noise register correlated with the position-dependent and time-dependent environmental data. [15] Provisioning system (6) according to one of the preceding claims 11 to 14, characterized by that the computing unit (11) is designed to provide the received environmental data with georeferenced features and wherein the computing unit (11) is designed to correlate at least the position-dependent and time-dependent environmental data with the georeferenced features. [16] Provisioning system (6) according to one of the preceding claims 11 to 15, characterized bythat the computing unit (11) is designed to transmit the noise register along the target route to the at least one vehicle when a target route of a vehicle is transmitted. [17] Ego vehicle (12) comprising a navigation device for entering a destination route and a vehicle control, the vehicle control comprising a coupling point (13) for coupling to a provision system (6), wherein the provision system (6) provides a noise register for displaying ambient noise as position-dependent ambient data on maps or road maps and wherein the vehicle control is designed to receive a noise register and wherein the vehicle control is designed to generate the vehicle parameters depending on the current position and the noise register. [18] Ego vehicle (12) according to claim 17, characterized byin that the vehicle control is designed to transmit the target route to the provisioning system (6) and to receive at least the noise register along the target route, and wherein the vehicle control is designed to generate the vehicle parameters as a function of the current position and the noise register. [19] Ego vehicle (12) according to claim 17 or 18, characterized by that sensors are provided for recording ambient noise as real ambient data and wherein the vehicle control is designed to search for the ambient data present in the noise register in the real ambient data depending on the position. [20] Ego vehicle (12) according to claim 19, characterized by that the vehicle control system is designed to improve the noise register based on the real environmental data. [21] Ego vehicle (12) according to one of the preceding claims 17 to 20, characterized bythat the vehicle control is designed to correlate the received environmental data of the noise register with georeferenced features and to generate the vehicle parameters depending on the current position and the noise register as well as the georeferenced features.

Images