Method for ascertaining the direction from which an emergency vehicle is approaching a vehicle

Abstract

The invention relates to a method for ascertaining an emergency vehicle direction from which an emergency vehicle is approaching a vehicle (1). A special acoustic signal emitted by the emergency vehicle is detected by means of a microphone assembly (2) which has a plurality of microphones (2.1 to 2.n) arranged on the vehicle (1) in a mutually spaced manner. The microphone assembly (2) is calibrated for the respective current position of the vehicle (1) in a location-specific manner in that an acoustic test signal (TS) is emitted into the surroundings of the vehicle (1) at the respective current position, reflections (R) of the test signal (TS) produced in the surroundings are detected by the microphone assembly (2), and a sound map (K) is generated on the basis of the reflections (R), the amplitude thereof, and the direction thereof relative to the vehicle (1), said sound map indicating to what extent the emitted test signal (TS) is reflected back to the vehicle (1) from different directions. Received acoustic signals are weighted when ascertaining the emergency vehicle direction on the basis of the sound map (K) generated for the current position by means of the microphones (2.1 to 2.n) of the microphone assembly (2), said microphones being oriented in different directions.

Description

[0001] Method for determining a direction from which an emergency vehicle is approaching a vehicle

[0002] The invention relates to a method for determining a direction from which an emergency vehicle approaches a vehicle, according to the preamble of claim 1.

[0003] For automated, particularly highly automated or autonomous vehicles, and for the operation of driver assistance systems, it is important to detect emergency vehicles in the vehicle's vicinity. This is achieved, among other things, by acoustically detecting an emergency sound signal using external microphones mounted on the vehicle. This allows appropriate driving actions to be initiated upon detection of an emergency vehicle, for example, an automated driving task can be terminated by handing over to a driver of the vehicle, or an emergency lane can be formed appropriately in a complex situation.

[0004] The unpublished German patent application with the file number

[0005] 102023 000 565.7 describes a method for determining the reception direction of a special acoustic signal emitted by an emergency vehicle for a vehicle with a microphone array. Based on signals detected by an environmental sensor system and information from a digital map, sound-reflecting surfaces in the vehicle's surroundings are identified and taken into account when determining the reception direction.

[0006] The unpublished DE 10 2022 002 500 A1 describes a method for directional calibration of microphones arranged on a vehicle using a reference noise, wherein the reference noise is generated by a sound source arranged at a predetermined position on the vehicle and recorded by the microphones. DE 102020 213 185 A1 discloses a method for assisting a driver of a vehicle in perceiving external noises. This method involves capturing audio signals using several external microphones provided on the vehicle, filtering out frequency ranges of emergency vehicle sirens from the audio signals, and determining the direction of movement of an audio source outputting the filtered audio signals.

[0007] From DE 102018 207 758 A1 a detection device for detecting an acoustic signal in the environment of a vehicle is known, wherein the detection device has a plurality of external microphones arranged above a cabin of the vehicle.

[0008] DE 102018217 891 A1 describes a method for operating a

[0009] Driver assistance systems are known in which acoustic signals from an emergency vehicle are recorded using several microphones and, depending on this, information is given to a driver or a driving intervention is carried out.

[0010] The invention is based on the object of providing a novel method for determining a direction from which an emergency vehicle is approaching a vehicle.

[0011] The object is achieved according to the invention by a method which has the features specified in claim 1.

[0012] Advantageous embodiments of the invention are the subject of the subclaims.

[0013] In a method for determining the direction from which an emergency vehicle is approaching a vehicle, a special acoustic signal emitted by the emergency vehicle is detected using a microphone array comprising a plurality of microphones arranged at a distance from one another on the vehicle. In an evaluation of the acoustic signals received by the microphone array, a reception direction of the special signal is determined, taking into account sound-reflecting surfaces present in the vehicle's surroundings, and the direction of the emergency vehicle is determined based on the reception direction.

[0014] According to the invention, the microphone arrangement is calibrated location-specifically for a respective current position of the vehicle by - emitting an acoustic test signal into the environment of the vehicle at a respective current position,

[0015] - reflections of the test signal occurring in the environment are recorded by the microphone arrangement,

[0016] - depending on the reflections, their amplitude and relative direction to the vehicle, a sound map is generated which indicates the degree to which the emitted test signal is reflected back to the vehicle from different directions, and

[0017] - depending on the sound map generated for the current position, acoustic signals received by the microphones of the microphone array aligned in different directions are weighted when determining the direction of the emergency vehicle.

[0018] A sound map, which can also be referred to as a noise map, is understood in this case to be a map representation in which an acoustic reflection behavior in the environment of the vehicle is shown.

[0019] Using this method, sound-reflecting objects present in the vehicle's surroundings, particularly in an automated, for example, highly automated or autonomously operated vehicle, can be detected simply and reliably based on reflections. Taking these sound-reflecting objects into account during the acoustic detection of the vehicle's surroundings enables particularly precise acoustic detection of a special signal from an emergency vehicle and, consequently, particularly precise detection of the emergency vehicle's direction, while significantly reducing the number of false-positive detections due to reflections. This results in particularly reliable detection of emergency vehicles, especially in urban traffic scenarios with a high number of acoustically reflective objects.This, in turn, results in improved situation recognition and improved system response in the vehicle, for example, transferring a driving task from an automated driving function to a driver to ensure the vehicle's driving is appropriate to the situation, such as forming an emergency lane. This allows automated driving functions to be designed with exceptional safety in mind, and enables the particularly safe implementation of regulatory requirements for automated driving.

[0020] According to one possible embodiment of the method, the sound map is used to map reflection properties of the vehicle's surroundings within a 360° area around the vehicle. This enables comprehensive mapping of reflection properties in the vehicle's surroundings, allowing precise, high-resolution detection of the emergency vehicle's direction in all directions around the vehicle.

[0021] According to another possible embodiment of the method, a high-frequency noise, in particular a clicking noise, is emitted as the acoustic test signal. Such a noise is particularly suitable as a test signal.

[0022] According to another possible embodiment of the method, the reflections are recorded using at least four microphones arranged at four corners or four sides of the vehicle. This allows for a simple setup of the microphone array while simultaneously providing a large coverage area.

[0023] According to a further possible embodiment of the method, the volume of a signal emitted by the respective microphone is adjusted by means of the weighting, so that detected noises from areas in which reflective objects were detected are taken into account less when determining the direction of the emergency vehicle.

[0024] According to another possible embodiment of the method, the propagation times between the emission of the test signal and the reception of the reflections are determined, and the distances between the sound-reflecting surfaces and the respective microphone are determined based on these propagation times. This allows the distances to the sound-reflecting surfaces to be determined without additional hardware effort.

[0025] According to another possible embodiment of the method, the generated sound map is validated using environmental data acquired by at least one camera, at least one lidar, and / or at least one radar. This can further improve the reliability of determining the direction of the emergency vehicle.

[0026] Embodiments of the invention are explained in more detail below with reference to drawings.

[0027] In the drawings: Fig. 1 shows schematically a sequence of a method for determining an emergency vehicle direction from which an emergency vehicle approaches a vehicle,

[0028] Fig. 2 shows a schematic plan view of a traffic situation during an emission of an acoustic test signal by a vehicle,

[0029] Fig. 3 schematically shows a plan view of the traffic situation according to Figure 2 and a sound map and

[0030] Fig. 4 schematically shows a plan view of the traffic situation and the sound map according to Figure 3 during a determination of correction values ​​for weighting acoustic signals received by means of microphones of a microphone arrangement.

[0031] Corresponding parts are provided with the same reference numerals in all figures.

[0032] Figure 1 shows a sequence of a possible embodiment of a method for determining an emergency vehicle direction from which an emergency vehicle approaches a vehicle 1 shown in more detail in Figures 2 to 4.

[0033] Vehicle 1 is specifically designed for automated, for example, highly automated or autonomous ferry operations. During such automated ferry operations, it is important to reliably detect emergency vehicles in the vicinity of vehicle 1 in order to be able to respond appropriately.

[0034] For such detection, it is provided that a direction from which the emergency vehicle is approaching the vehicle 1 is determined. This is done by detecting a special acoustic signal emitted by the emergency vehicle using a microphone arrangement 2, also shown in more detail in Figures 2 to 4. For this purpose, the microphone arrangement 2 comprises a plurality of microphones 2.1 to 2.n. In an evaluation of acoustic signals received by the microphone arrangement 2, a reception direction of the special signal is then determined, taking into account sound-reflecting surfaces A1 to Ax present in the vicinity of the vehicle 1, and the direction of the emergency vehicle is determined based on the reception direction.

[0035] In order to determine the sound-reflecting surfaces A1 to Ax and their acoustic reflection behavior, it is provided that in a method step S1, an acoustic test signal TS is emitted into the environment of the vehicle 1 at a current position by means of at least one external loudspeaker 3 of the vehicle 1, which is also shown in more detail in Figures 2 to 3.

[0036] Reflections R of the test signal TS occurring in the environment at the sound-reflecting surfaces A1 to Ax are recorded in a further method step S2 by means of the microphone arrangement 2.

[0037] In a further method step S3, a sound map K is generated depending on the reflections R as well as their amplitude and relative direction to the vehicle 1, which indicates the degree to which the transmitted test signal TS is reflected back to the vehicle 1 from different directions.

[0038] Subsequently, in a further method step S4, the acoustic signals received by the microphones 2.1 to 2.n of the microphone array 2, which are oriented in different directions, are weighted to determine the direction of the emergency vehicle, depending on the sound map K generated for the current position. Thus, the microphone array 2 is calibrated location-specifically for the respective current position of the vehicle 1.

[0039] Figure 2 shows a plan view of a traffic situation during an emission of the acoustic test signal TS through the external loudspeaker 3 of the vehicle 1 and visualizes a possible embodiment of the method step S1 described in Figure 1.

[0040] In the traffic situation, vehicle 1 is on a roadway and is moving towards an intersection. To the left of vehicle 1 there are several objects 01 to 04 in the form of trees. Diagonally to the left behind vehicle 1 there is an object 05 in the form of a vehicle, and to the right of vehicle 1 there are several objects 06 to 0m in the form of buildings. In particular, during ferry operation of vehicle 1, the acoustic test signal TS, for example a high-frequency noise, in particular a clicking noise, is emitted into the surroundings of vehicle 1 at a current position by means of the external loudspeaker 3 of vehicle 1.

[0041] This test signal TS is applied to sound-reflecting surfaces A1 to Ax of the

[0042] Objects 01 to 0m are reflected, with objects 01 to 04, formed as trees, exhibiting medium attenuation and reflectivity, and objects 06 to 0m, formed as buildings, exhibiting high reflectivity. In the forward direction of vehicle 1, there is an open field with low reflectivity. Object 05, located to the left behind vehicle 1 and also formed as a vehicle, does exhibit high reflectivity, but due to its small size, it only represents a medium reflection R.

[0043] The reflections R of the test signal TS occurring at the sound-reflecting surfaces A1 to Ax are then recorded by the microphone arrangement 2, for example by means of at least four microphones 2.1 to 2.n arranged at four corners or four sides of the vehicle 1.

[0044] Based on these reflection properties of objects O1 to Om, a sound map K shown in Figure 3 is generated depending on the reflections R as well as their amplitude and relative direction to vehicle 1. This map indicates the degree to which the transmitted test signal TS is reflected back to vehicle 1 from different directions. Reflections R of the test signal TS originating from objects O1 to Om with high reflectivity have a higher amplitude than reflections R originating from objects O1 to Om with low reflectivity.

[0045] In a possible embodiment, additionally, the propagation times occurring between the emission of the test signal TS and the reception of the reflections R can be determined, and the propagation times can be used to determine distances between the sound-reflecting surfaces A1 to Ax of the objects 01 to Om and the respective microphone 2.1 to 2.n.

[0046] As can be seen from Figure 3, the sound map K is used to map reflection properties of the surroundings of the vehicle 1 in a range of 360° around the vehicle 1, whereby the reflections R are shown in a direction-resolved manner and are shown with higher amplitude at objects 01 to Om with high reflectivity and with lower amplitude at objects 01 to Om with low reflectivity.

[0047] This means that it is detected whether reflections R occur in a direction. In one possible embodiment, the directions relative to vehicle 1, for example, "front," "rear," "left," and "right," can be divided into different reflection levels. The reflection levels are, for example, "no reflection," "low reflection," "medium reflection," and "high reflection."

[0048] In another possible embodiment, the generated sound map K is validated using environmental data acquired by at least one camera, at least one lidar, and / or at least one radar. If a camera, radar, and / or lidar with 360° detection is available, a 360° sound map can also be created using these sensors. This can be determined in advance in experiments to determine the influence on the acoustic reflections R and then also create a 360° sound map with the reflection levels.

[0049] The weighting of the acoustic signals received by the microphones 2.1 to 2.n of the microphone arrangement 2, which are oriented in different directions, in the subsequent method step S4 as a function of the sound map K generated for the current position when determining the direction of the emergency vehicle is shown in Figure 4. In particular, the weighting is used to adjust the volume of a signal output by the respective microphone 2.1 to 2.n.

[0050] This means that, based on the sound map K with the reflection levels, the signals captured by microphones 2.1 to 2.n when detecting the direction of the emergency vehicle are calculated using correction factors F1 to F8. These correction factors F1 to F8, represented as arrows, are adapted to the various reflection levels in tests, for example. This can significantly increase the performance of direction detection, as direction detection is improved by taking objects O1 to Om and their reflectivity into account, particularly in urban traffic scenarios.

[0051] If the microphones 2.1 to 2.n were not calibrated using the correction factors F1 to F8, a special signal from an emergency vehicle, for example, would be perceived more strongly from the right and too weakly from the front or rear due to the high reflectivity of the objects 06 to Om, which are designed as buildings.

[0052] In a free environment, which is therefore optimal for acoustic detection using microphone array 2, all microphones 2.1 to 2.n would exhibit no influence in any direction, so that the detected reflections R would be evenly distributed in amplitude in all directions. This is represented by a circle C extending 360° around vehicle 1.

[0053] The determination of the correction factors F1 to F8 and the calibration of the

[0054] Microphones 2.1 to 2.n are based on deviations of the sound map K from this circle C. An average of the amplitudes of the reflections R recorded by all microphones 2.1 to 2.n is used to determine a radius of the circle C and for normalization.

[0055] It follows that the areas to the right of vehicle 1 are reduced or attenuated due to the high amplitudes of the reflections R resulting from the high reflectivity with correction factors F5 to F7 having a value of less than "1", for example "0.5".

[0056] The areas in front of and behind the vehicle 1 are increased or amplified with correction factors F4, F8, which have a value of more than "1", for example "2", due to the low amplitudes of the reflections R resulting from the low reflectivity.

[0057] The areas to the left of vehicle 1 are increased or amplified to a lesser extent due to the mean amplitudes of the reflections R resulting from the mean reflectivity with correction factors F1 to F3, which have a value of more than "1", for example "1.25".

[0058] This creates 360° correction factors F1 to F8, which are used to calibrate microphones 2.1 to 2.n. The environment detection of microphones 2.1 to 2.n is then set to "active."

[0059] If a special acoustic signal from an emergency vehicle is detected, microphones 2.1 to 2.n are less influenced by the environment. The following example illustrates an application of the described method. For example, an emergency vehicle, such as a fire engine, approaches vehicle 1 from the right. The area surrounding vehicle 1 is clear in all directions, but there is a concrete barrier to the left.

[0060] Without the concrete barrier present, the direction detection would clearly assign the received special signal to the right and thus correctly. With the concrete barrier present, the direction detection would detect incorrect and false-positive reflections R, which would lead to an incorrect directional assignment of the emergency vehicle to the left due to the reflections R at the concrete barrier. By creating the sound map K as described above, it is known that there are no or only slight reflections R in front of, behind and to the right of vehicle 1 and that the reflections R are high to the left of vehicle 1. Using the sound map K, the microphones 2.1 to 2.n can now be weighted with the correction factors F1 to F8, so that the volume of the microphones 2.1 to 2.n, which point to the left of vehicle 1, is reduced by means of at least one correction factor F1 to F8.This would give greater consideration to acoustic signals from the right and the algorithm would detect the direction of the emergency vehicle more reliably to the right, i.e. correctly.

Claims

Patent claims 1. A method for determining an emergency vehicle direction from which an emergency vehicle approaches a vehicle (1), wherein - a special acoustic signal emitted by the emergency vehicle is detected by means of a microphone arrangement (2) comprising a plurality of microphones (2.1 to 2.n) arranged at a distance from one another on the vehicle (1), and - in an evaluation of acoustic signals received by means of the microphone arrangement (2), taking into account sound-reflecting surfaces (A1 to Ax) present in the surroundings of the vehicle (1), a reception direction of the special signal is determined and the direction of the emergency vehicle is determined on the basis of the reception direction, characterized in that the microphone arrangement (2) is calibrated in a location-specific manner for a respective current position of the vehicle (1) by - an acoustic test signal (TS) is emitted into the surroundings of the vehicle (1) at a current position, - reflections (R) of the test signal (TS) occurring in the environment are detected by means of the microphone arrangement (2), - depending on the reflections (R) as well as their amplitude and relative direction to the vehicle (1), a sound map (K) is generated which indicates the degree to which the transmitted test signal (TS) is reflected back to the vehicle (1) from different directions, and - depending on the sound map (K) generated for the current position, acoustic signals received by means of the microphones (2.1 to 2.n) of the microphone arrangement (2) aligned in different directions are weighted when determining the direction of the emergency vehicle.

2. Method according to claim 1, characterized in that reflection properties of the surroundings of the vehicle (1) in a range of 360 ° around the vehicle (1) are mapped by means of the sound map (K).

3. Method according to claim 1 or 2, characterized in that a high-frequency noise, in particular a clicking noise, is emitted as the acoustic test signal (TS).

4. Method according to one of the preceding claims, characterized in that the reflections (R) are detected by means of at least four microphones (2.1 to 2.n) arranged at four corners or four sides of the vehicle (1).

5. Method according to one of the preceding claims, characterized in that by means of the weighting a volume of a signal output by means of the respective microphone (2.1 to 2.n) is adjusted.

6. Method according to one of the preceding claims, characterized in that transit times occurring between the emission of the test signal (TS) and the reception of the reflections (R) are determined and, based on the transit times, distances between the sound-reflecting areas (A1 to Ax) and the respective microphone (2.1 to 2.n).

7. Method according to one of the preceding claims, characterized in that the generated sound map (K) is checked for plausibility by means of data from the environment which are recorded by means of at least one camera, at least one lidar and / or at least one radar.

Images