Methods for localizing a sound source

Abstract

Method for localizing a sound source (9.1, 9.2) wherein several microphones (2) arranged on a vehicle (1) detect sound generated by the sound source (9.1, 9.2), wherein a direction and / or position (P1, P2, P3) of the sound source (9.1, 9.2) relative to the vehicle (1) is determined based on a delay in the reception of the sound at one of the microphones (2) compared to another of the microphones (2), where the sound source (9.1, 9.2) is located as a motorized two-wheeled vehicle (9.1, 9.2), characterized in that the driving behavior of the vehicle (1) is adapted to the direction and / or position of the sound source (9.1, 9.2) by steering the vehicle (1) in a lateral direction to avoid an expected two-wheeled vehicle (9.1, 9.2) which, in the event of a traffic jam or slow-moving traffic, weaves between the vehicles (1) and moves through the traffic at a higher relative speed.

Description

[0001] The invention relates to a method for locating a sound source according to the preamble of claim 1 and to a vehicle according to the preamble of claim 3.

[0002] In the future, automated vehicles (including trucks, cars, buses, etc.) will be on the roads without drivers. Therefore, automated driving systems must take over the tasks of human drivers. This includes detecting and locating two-wheeled vehicles, especially motorcycles.

[0003] WO 2020 / 259892 A1 describes a method for generating an environment model of an autonomously controlled vehicle. In the method, sensor data is acquired by a plurality of infrastructure-side sensors in an area surrounding the vehicle. Fused sensor data is generated stationary based on the acquired sensor data. Furthermore, environment model data is generated stationary by infrastructure-side evaluation of the sensor data, whereby objects located in the area surrounding the vehicle are localized and identified. Finally, the environment model data is transmitted to the vehicle. In addition, a method for autonomously controlling a vehicle is described. An environment model generation device and a vehicle control device are also described. Furthermore, an autonomously controlled vehicle and an automated transport system are described.A junction in the road network has a core area with high priority and a peripheral area with lower priority. The junction is monitored using sensor units. In the junction area and thus also in the monitoring area, there are two vehicles, each driving towards the junction. A motorcycle is also detected moving away from the inner monitoring area. Furthermore, a person is detected in the inner monitoring area crossing a zebra crossing. Monitoring data is transmitted via radio from a communications unit to vehicles within the transmission range. To generate the environment model data, sensor data from the area surrounding the vehicle is also used. The data basis for the vehicle's environment model can be expanded, thus improving the reliability, resolution, and completeness of the environment model.

[0004] WO 2021 / 116351 A1 describes a method for positioning a sound wave signal source in a vehicle, comprising: detecting at least two sound wave signals; calculating a relative position of the sound wave signal source relative to the vehicle based on the at least two sound wave signals; obtaining a position of the vehicle; and determining position information of the sound wave signal source based on the position of the vehicle and the calculated relative position of the sound wave signal source relative to the vehicle.

[0005] US 2011 / 0175755 A1 describes a vehicle location detection device comprising: a sound source location indicating unit that calculates an attenuation amount of a sound pressure of a vehicle noise by subtracting a sound pressure detected by a sound pressure detection unit from a sound pressure indicated by vehicle sound pressure information stored in a sound pressure information storage unit, and indicates a location where another vehicle emitting a vehicle noise is expected to exist by indicating a sound source location of a sound diffracted by an obstacle, which is indicated by a diffraction location indicating unit and serves as a diffraction point, based on the calculated attenuation amount, a rule indicating an attenuation amount depending on a propagation distance of the noise, and a relationship indicated by diffraction information,stored in a diffraction information storage unit. Furthermore, a presentation unit is described that presents the specified location.

[0006] DE 10 2015 011 246 A1 describes a method for locating a signal source. An analysis device receives a vehicle position from a plurality of vehicles and at least one relative source position of the signal source determined with respect to the vehicle. Based on the respective vehicle position of each vehicle and the at least one relative source position, the analysis device determines an intersection area in which the relative source positions of at least two of the vehicles overlap. An absolute source position of the signal source is determined based on the intersection area.

[0007] US 2014 / 0025284 A1 describes systems, methods, devices, and computer-readable media for determining relative position information. A signal detection unit configured to be associated with a first vehicle includes one or more signal sensors and one or more processors and is configured to receive one or more signals from one or more signal sources that may be associated with a second vehicle. Based on the arrival times of the signal(s), a set of time values ​​may be determined and a set of distance expressions may be generated. A set of distance equations may be generated based on the set of time values ​​and the set of distance expressions, and the set of distance equations may be solved to determine one or more positions associated with the first vehicle or the one or more signal sources within a defined coordinate system.

[0008] US 2013 / 0107668 A1 describes a method for locating a sound source using multiple vehicles. The method includes transmitting acoustic inputs from multiple sensors to multiple processing modules, determining a location of each of the vehicles using at least one of the following modules: a GPS module and an inertial motion unit module located in each vehicle, processing the received acoustic inputs in each processing module, designating one of the processing modules as the master processing module and sending the processed acoustic inputs received in each processing module to the master processing module, combining the processed acoustic inputs in the master processing module, and estimating the location of the sound source based on the combined processed acoustic data.The method may further include determining whether the master processing module is functional, and in response to determining that the master processing module is inoperative, designating another one of the plurality of processing modules as the master processing module.

[0009] The invention is based on the object of providing a novel method for locating a sound source and a novel vehicle.

[0010] The object is achieved according to the invention by a method for locating a sound source having the features of claim 1 and by a vehicle having the features of claim 3.

[0011] Advantageous embodiments of the invention are the subject of the subclaims.

[0012] In a method for locating a sound source, a sound generated by the sound source is detected by several microphones mounted on a vehicle. According to the invention, the direction and / or position of the sound source relative to the vehicle is determined based on a delay in the reception of the sound at one of the microphones compared to another.

[0013] In one embodiment, the direction and / or position is checked and / or corrected taking into account a world model that is created using data acquired by an environmental sensor system of the vehicle and / or map information from a digital map and that contains information about objects from which the sound is reflected.

[0014] According to the invention, a motorized two-wheeled vehicle is localized as the sound source. A motorized two-wheeled vehicle can be distinguished from other sound sources, for example, based on characteristic engine noises.

[0015] According to the invention, the vehicle's driving behavior is adapted to the direction and / or position of the sound source. According to the invention, the vehicle is steered in a transverse direction to avoid an expected two-wheeled vehicle that weaves between vehicles in a traffic jam or slow-moving traffic and moves through the traffic at a higher relative speed. Alternatively or additionally, a notification or warning can be issued to a driver of the vehicle indicating that a two-wheeled vehicle is approaching and its location.

[0016] According to one aspect of the present invention, a vehicle is proposed, comprising a plurality of microphones and a control unit configured to carry out the method described above.

[0017] In one embodiment, the vehicle is designed as a commercial vehicle, as a passenger car or as a bus.

[0018] According to the invention, the vehicle is configured as an automated or highly automated, in particular autonomous vehicle.

[0019] In one embodiment, at least four microphones are provided on the vehicle.

[0020] In one embodiment, two microphones are arranged on a left side and on a right side of a cabin of the vehicle.

[0021] Localizing two-wheeled vehicles is challenging because they are agile and smaller than other vehicles. The present invention allows for more precise localization because sound reflections caused by occlusions can be taken into account.

[0022] Embodiments of the invention are explained in more detail below with reference to drawings.

[0023] Showing: Fig. 1 a schematic view of a vehicle, Fig. 2 a schematic view of a processing chain in which a sound source can be localized using sound signals and a world model, and Fig. 3 a schematic view of a traffic situation in which a sound source cannot be directly localized by a control unit of the vehicle due to masking and / or reflection by other objects.

[0024] Corresponding parts are provided with the same reference numerals in all figures.

[0025] In traffic jams or slow-moving traffic, it is often observed that two-wheeled vehicles, especially motorcycles, weave their way between the vehicles and move through the traffic at a higher relative speed. Because the two-wheeled vehicles drive between the vehicles, they are difficult to detect (e.g. blind spots) and are only detected relatively late (e.g. obscured objects). Visual perception is therefore hampered. However, the characteristic engine noise of two-wheeled vehicles often makes acoustic perception possible earlier. Although exact localization may not be possible, the vehicle's behavior can at least be adapted, for example by changing its position laterally to make it easier for the two-wheeled vehicle to pass and thus avoid dangerous situations. The invention therefore relates to an acoustic method for detecting and localizing two-wheeled vehicles with engines.

[0026] Fig. 1 is a schematic view of a vehicle 1, for example a commercial vehicle.

[0027] Several microphones 2 are mounted on the vehicle 1. For example, at least four microphones 2 can be used, in particular two microphones 2 on a left side and two on a right side of a cabin of the vehicle 1. The vehicle 1 has a control unit 3 in which signals from the microphones 2 are processed by software.

[0028] The software also has access to a world model 6 (shown in Fig. 2), which is created using data acquired by an environment sensor system 4 of the vehicle 1 and map information from a digital map. The world model 6 contains knowledge about the environment of the vehicle 1, for example, the positions of other vehicles and information about the road, for example, whether the vehicle 1 is in a tunnel or other structural information.

[0029] The environmental sensor system 4 can, for example, have at least one camera, at least one lidar sensor and / or at least one radar sensor.

[0030] By measuring the difference in the impact of a sound signal between at least two microphones 2, the direction of the sound can be determined. This does not always allow for the exact direction to be determined, for example, if the sound comes from a direction different from the source due to obscurations and reflections.

[0031] By using the world model 6, the sound reflections can be taken into account, i.e. a direction estimate can be corrected, since the sound direction of the signal recorded by the microphone 2 lies on a reflecting surface.

[0032] Fig. 2 is a schematic view of a processing chain 5 in which a more precise sound localization 7 can be carried out based on the sound signals and the world model 6, with which the driving behavior 8 of the vehicle 1 can be adapted.

[0033] Fig.Figure 3 is a schematic view of a traffic situation 11 in which, due to obscuration and / or reflection by other objects 10.1, 10.2, 10.3, a sound source 9.1, 9.2, for example, a motorized two-wheeled vehicle 9.1, 9.2, cannot be directly localized by the control unit 3 of the vehicle 1. In the example shown, the sound is reflected by an object 10.1 formed as a wall and / or by an object 10.2 formed as another vehicle, and is obscured by two objects 10.2, 10.3 formed as other vehicles. The world model 6, i.e., by knowing the existence of the wall, its position, and orientation, allows the determined direction of the sound to be corrected. In this case, the control unit 3 would localize the sound source 9.1 at a position P1 based on the sound received by the microphones 2, without taking the world model 6 into account. Knowing that the wall is object 10.1 from the world model 6, the control unit 3 localizes the sound source 9.1 at a position P2 based on the sound received by the microphones 2. Furthermore, the control unit 3 localizes the sound source 9.1 at a position P3 based on the sound received by the microphones 2, knowing that the other vehicle is an object 10.2 from the world model 6. List of reference symbols 1 vehicle 2 microphones 3 Control unit 4 Environmental sensors 5 Processing chain 6 World model 7 Sound localization 8 Driving behavior 9.1, 9.2 Sound source, two-wheeled vehicle 10.1 to 10.3 Object 11 Traffic situation P1 to P3 positions

Claims

[1] Method for locating a sound source (9.1, 9.2), wherein a sound generated by the sound source (9.1, 9.2) is detected by several microphones (2) arranged on a vehicle (1), wherein a direction and / or position (P1, P2, P3) of the sound source (9.1, 9.2) relative to the vehicle (1) is determined based on a delay in the reception of the sound at one of the microphones (2) compared to another of the microphones (2), wherein a motorized two-wheeled vehicle (9.1, 9.2) is localized as the sound source (9.1, 9.2), characterized by that the driving behavior of the vehicle (1) is adapted to the direction and / or position of the sound source (9.1, 9.2) by steering the vehicle (1) in a transverse direction to avoid an expected two-wheeled vehicle (9.1, 9.2) which, in the event of a traffic jam or slow-moving traffic, weaves its way between the vehicles (1) and moves through the traffic at a higher relative speed. [2] Method according to claim 1, characterized by that the direction and / or position (P1, P2, P3) is checked and / or corrected taking into account a world model (6) which is created using data acquired by means of an environmental sensor system (4) of the vehicle (1) and / or map information from a digital map and which contains information about objects (10.1, 10.2, 10.3) at which the sound is reflected. [3] Vehicle (1) comprising a plurality of microphones (2) and a control unit (3), wherein the vehicle (1) is configured as an automated or highly automated, in particular autonomous vehicle, characterized by that the control device (3) is configured to carry out the method according to one of the preceding claims. [4] Vehicle (1) according to claim 3, characterized by that the vehicle (1) is designed as a commercial vehicle, as a passenger car or as a bus. [5] Vehicle (1) according to claim 3 or 4, characterized by that at least four microphones (2) are provided on the vehicle (1). [6] Vehicle (1) according to claim 5, characterized by that two microphones (2) are arranged on a left side and on a right side of a cabin of the vehicle (1).

Images