Intelligent Driving Car Accident Detection System
Through the comprehensive judgment of microphone array, surround view module and surround view ultrasonic radar, the traffic accident detection problem in the case of intelligent driving cars without drivers is solved, and timely detection and notification of accidents are achieved.
Patent Information
- Application Number
- CN202110021810.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-01-08
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2041-01-08
AI Technical Summary
When an intelligent driving car has a traffic accident, it is impossible to detect and handle the accident in a timely manner because there are no drivers or passengers.
Comprehensive judgment of microphone array module, surround view module and surround view ultrasonic radar module is used to generate accident detection results through the multi-sensor fusion judgment module, and accident information reports are sent through the telematics processor.
It realizes traffic accident detection for intelligent driving cars without drivers or passengers, can promptly and accurately judge accidents, and promptly notify the owner or person in charge.
Smart Images

Figure CN114750754B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention generally relate to the field of intelligent driving technology, and more specifically, to an intelligent driving automobile accident detection system. Background Art
[0002] In recent years, as a solution to improve automobile driving safety, reduce congestion and traffic accidents, smart driving cars have gradually become a very important development direction in the automotive field.
[0003] The Society of Automotive Engineers (SAE) divides vehicle automation levels into six levels: L0-L5. Among them, L4 is highly automated driving, which means that under limited road and environmental conditions, human drivers do not need to operate the vehicle or pay attention, and all driving operations are completed by the vehicle; L5 is fully automated driving, and in all application scenarios, the vehicle completes all driving operations without the driver operating the vehicle or paying attention. In other words, when the vehicle automation level reaches L4 or L5, in actual use scenarios, there may be no one in the vehicle. For example, in the actual application scenario of the automatic valet parking (AVP: Automated Valet Parking) function, after the occupants leave the vehicle, the vehicle will automatically find a parking space and park; in the actual application scenario of the autonomous taxi (Robotaxi) function, the vehicle will automatically drive to the designated location according to the passenger's requirements; in the above two specific use scenarios, there are no passengers in the car.
[0004] Although smart driving cars have high safety, they may also have traffic accidents in complex and changing traffic environments. For smart driving cars that are not L4 or L5 autonomous driving, or traditional non-intelligent driving vehicles, when a traffic accident occurs, the driver can identify the accident immediately and handle it accordingly because there is a driver in the car. However, for L4 or L5 level smart driving vehicles, when an accident occurs, there may be no driver or passengers in the car. In this case, it is impossible to detect and handle the traffic accident in time. Summary of the invention
[0005] In order to solve the above problems in the prior art, an embodiment of the present invention provides an intelligent driving car accident detection system, which includes a microphone array module, an acoustic electronic control unit, a central control unit, a surround view module, a surround view ultrasonic radar module, a positioning module and a telematics processor, wherein the central control unit includes a multi-sensor fusion judgment module. The microphone array module is used to detect the sound source signal around the intelligent driving car. The acoustic electronic control unit is used to: determine the probability that the sound source signal is a collision sound based on the sound source signal, and obtain an acoustic collision risk assessment result according to the probability that the sound source signal is a collision sound; calculate acoustic positioning information according to the acoustic collision risk assessment result; and send the acoustic collision risk assessment result and the acoustic positioning information to the multi-sensor fusion judgment module. The surround view ultrasonic radar module is used to: detect the obstacle distance through the ultrasonic radar, determine the ultrasonic radar detection result according to the obstacle distance, and send the ultrasonic radar detection result and the ultrasonic radar identifier to the multi-sensor fusion judgment module. The surround view module is used to: collect images of the surroundings of the smart driving car through a camera, determine a visual collision risk assessment result based on the image, and send the visual collision risk assessment result and the camera identifier to the multi-sensor fusion determination module. The multi-sensor fusion determination module is used to generate an accident detection result based on one or more of the acoustic collision risk assessment result, the acoustic positioning information, the ultrasonic radar detection result, the ultrasonic radar identifier, the visual collision risk assessment result and the camera identifier. The positioning module is used to obtain the location information of the smart driving car. The telematics processor is used to transmit the accident detection result and / or the location information to a remote device.
[0006] In some embodiments, the multi-sensor fusion judgment module is further used to: when the acoustic collision risk assessment result is a high acoustic collision risk, a medium acoustic collision risk, or a low acoustic collision risk, determine the positioning sector grid of the sound source signal according to the acoustic positioning information; if the positioning sector grid is not a detection blind spot of the ultrasonic radar, obtain the ultrasonic radar detection result of the positioning sector grid of the ultrasonic radar detection result in the positioning sector grid; if the positioning sector grid is not a detection blind spot of the camera, obtain the visual collision risk assessment result of the positioning sector grid of the visual collision risk assessment result in the positioning sector grid; and obtain the accident detection result of the positioning sector grid according to whether the positioning sector grid is the detection blind spot of the ultrasonic radar, whether it is the detection blind spot of the camera, the ultrasonic radar detection result of the positioning sector grid, and the visual collision risk assessment result of the positioning sector grid.
[0007] In some embodiments, determining the positioning sector grid of the sound source signal includes: obtaining the sound source coordinate value of the sound source relative to the coordinate system of the intelligent driving car; determining an ellipse with the center of the vehicle as the origin and the detection range of the microphone array module as the boundary as the acoustic positioning area; using an angle dividing line to divide the acoustic positioning area into different positioning sectors, and the positioning sectors are numbered in a clockwise or counterclockwise direction; using a distance dividing line to divide the positioning sector into different positioning sector grids, and the positioning sector grids are numbered from far to near or from near to far from the vehicle; determining the positioning sector number and the positioning sector grid number where the sound source is located according to the sound source coordinate value.
[0008] In some embodiments, the multi-sensor fusion judgment module is further used to: obtain the ultrasonic radar detection results and the visual collision risk assessment results in each positioning sector grid when the acoustic collision risk assessment result is an acoustic non-collision; obtain the accident detection results of each positioning sector grid based on whether the each positioning sector grid is a detection blind area of the ultrasonic radar, whether it is a detection blind area of the camera, the ultrasonic radar detection results and the visual collision risk assessment results.
[0009] In some embodiments, the central control unit is used to turn on the double flash warning lights of the smart driving car through the remote information processor when the accident detection result is an accident; read the location information of the smart driving car through the positioning module; read the relevant information of the accident collected by the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, and the positioning module within a preset time period before and after the accident, and store the relevant information of the accident in the storage module of the central control unit; and generate an accident information report.
[0010] In some embodiments, the microphone array module includes a plurality of micro-electromechanical system-based microphone array modules, each of the plurality of micro-electromechanical system-based microphone array modules includes a micro-electromechanical system microphone, a microphone array module processor, and a microphone array module car audio bus transceiver, wherein the acoustic interface of the microphone array module processor is respectively connected to the output interface of the micro-electromechanical system microphone and the microphone array module car audio bus transceiver; and the plurality of micro-electromechanical system-based microphone array modules are connected sequentially via a car audio bus.
[0011] In some embodiments, the plurality of micro-electromechanical system-based microphone array modules are arranged at one or more of the following positions: on a body sheet metal between the B-pillar and the ceiling of the vehicle, on a body sheet metal between the B-pillar and the floor of the vehicle, on a body sheet metal between the A-pillar and the ceiling of the vehicle, on a body sheet metal between the A-pillar and the floor of the vehicle, on a body sheet metal between the C-pillar and the ceiling of the vehicle, on a body sheet metal between the C-pillar and the floor of the vehicle, and under the front seats.
[0012] In some embodiments, the acoustic electronic control unit includes an acoustic electronic control unit processor, a serial port-controller area network conversion module, an acoustic electronic control unit car audio bus transceiver and an acoustic electronic control unit controller area network transceiver, wherein the serial port of the acoustic electronic control unit processor is connected to the serial port-controller area network conversion module; the serial port-controller area network conversion module is connected to the acoustic electronic control unit controller area network transceiver; and the acoustic electronic control unit car audio bus transceiver is connected to the acoustic interface of the acoustic electronic control unit processor.
[0013] In some embodiments, the acoustic electronic control unit is arranged on a body sheet metal near a center armrest in the vehicle.
[0014] In some embodiments, the central control unit includes an embedded computer or a computer system on a chip, at least five central control unit controller area network transceivers and a storage module, wherein the at least five central control unit controller area network transceivers are all connected to the embedded computer or the computer system on a chip, and are suitable for being respectively connected to the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, the positioning module and the telematics processor; the storage module is connected to the storage interface of the embedded computer or the computer system on a chip.
[0015] In some embodiments, the central control unit is arranged under the center console or in the trunk of the vehicle.
[0016] In some embodiments, the surround view module includes a plurality of cameras and a surround view controller installed around the body of the smart driving car.
[0017] In some embodiments, the multiple cameras of the surround view module are arranged at one or more of the following positions: a bumper, a radiator grille, and a front windshield at the front end of a vehicle; a fender, a rearview mirror housing, a B-pillar outer trim panel, a C-pillar body sheet metal, a rear bumper, a rear windshield, and a rear cover at the rear end of a vehicle; and the surround view controller of the surround view module is arranged under the center console of the vehicle.
[0018] In some embodiments, the surround-view ultrasonic radar module includes a plurality of ultrasonic radars and a parking controller.
[0019] In some embodiments, the plurality of ultrasonic radars are arranged at a front bumper and / or a rear bumper of the vehicle, and the parking controller is disposed under a center console of the vehicle.
[0020] In some implementations, the telematics processor is further configured to receive control information sent by the remote device to control the smart driving car to turn on the double flash warning lights.
[0021] The intelligent driving car accident detection system proposed in the embodiment of the present invention can solve the problem of accident detection when there is no driver or passenger in the intelligent driving car and the vehicle has an accident. The intelligent driving car accident detection system proposed in the embodiment of the present invention uses a microphone array group, a surround view module, and a surround view ultrasonic radar module to comprehensively determine vehicle accidents, which can timely and accurately determine whether the vehicle has a fault, and can reasonably, effectively and quickly detect traffic accidents. After detecting a traffic accident, the intelligent driving car accident detection system proposed in the embodiment of the present invention can send a collision information report to the driver or the person in charge of the vehicle through a remote information processor, and can promptly notify the owner or the person in charge of the vehicle of information related to the accident. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] The above and other objects, features and advantages of the embodiments of the present invention will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:
[0023] Figure 1 A schematic diagram of the structure organization of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0024] Figure 2 A schematic block diagram showing the structural composition and connection relationship of a MEMS-based microphone array module of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0025] Figure 3 A schematic block diagram showing the structural composition and connection relationship of an acoustic electronic control unit of an intelligent driving automobile accident detection system according to an embodiment of the present invention is shown;
[0026] Figure 4 A schematic block diagram showing the structural composition and connection relationship of a central control unit of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0027] Figure 5A working diagram of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0028] Figure 6 A schematic diagram of the acoustic positioning principle of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0029] Figure 7 A schematic diagram showing an example of a fault detection report of an intelligent driving car accident detection system according to an embodiment of the present invention;
[0030] Figure 8 A schematic diagram of the structural organization of an example of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0031] Fig. 9 A schematic diagram showing sensor arrangement of an example of an intelligent driving car accident detection system according to an embodiment of the present invention is shown;
[0032] Fig.10 A schematic diagram showing the working principle of an embodiment of an intelligent driving car accident detection system according to an implementation manner of the present invention is shown.
[0033] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts. DETAILED DESCRIPTION
[0034] The principle and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way.
[0035] The embodiment of the present invention provides an intelligent driving car accident detection system. Figure 1 , which shows a schematic diagram of the structure organization of the intelligent driving car accident detection system according to an embodiment of the present invention. Figure 1 As shown, the intelligent driving car accident detection system may include a microphone array module 1, an acoustic electronic control unit (AECU) 2, a central control unit (CECU) 3, a surround view module 4, a surround view ultrasonic radar module 5, a positioning module 6 and a telematics box (T-BOX) 7. The central control unit includes a multi-sensor fusion determination module.
[0036] The microphone array module can be used to detect sound source signals around smart driving cars.
[0037] The acoustic electronic control unit can be used to: determine the probability that the sound source signal is a collision sound based on the sound source signal, and obtain an acoustic collision risk assessment result based on the probability that the sound source signal is a collision sound; calculate the acoustic positioning information based on the acoustic collision risk assessment result; and send the acoustic collision risk assessment result and the acoustic positioning information to the multi-sensor fusion judgment module.
[0038] The surround-view ultrasonic radar module can be used to: detect the obstacle distance through the ultrasonic radar, determine the ultrasonic radar detection result according to the obstacle distance, and send the ultrasonic radar detection result and the ultrasonic radar identifier to the multi-sensor fusion determination module.
[0039] The surround view module can be used to: collect images of the surroundings of the smart driving car through the camera, determine the visual collision risk assessment results based on the images, and send the visual collision risk assessment results and camera identifiers to the multi-sensor fusion judgment module.
[0040] The multi-sensor fusion determination module can be used to generate an accident detection result based on one or more of the acoustic collision risk assessment result, the acoustic positioning information, the ultrasonic radar detection result, the ultrasonic radar identifier, the visual collision risk assessment result and the camera identifier.
[0041] The positioning module can be used to obtain the location information of the smart driving car.
[0042] The telematics processor may be used to transmit accident detection results and / or location information to a remote device.
[0043] As an embodiment of the present invention, specifically, the multi-sensor fusion judgment module can be used for: when the acoustic collision risk assessment result is a high acoustic collision risk, a medium acoustic collision risk or a low acoustic collision risk, determining the positioning sector grid of the sound source signal according to the acoustic positioning information; if the positioning sector grid is not a detection blind spot of the ultrasonic radar, obtaining the ultrasonic radar detection result of the positioning sector grid in the positioning sector grid; if the positioning sector grid is not a detection blind spot of the camera, obtaining the visual collision risk assessment result of the positioning sector grid in the positioning sector grid; obtaining the accident detection result of the positioning sector grid according to whether the positioning sector grid is a detection blind spot of the ultrasonic radar, whether it is a detection blind spot of the camera, the ultrasonic radar detection result of the positioning sector grid and the visual collision risk assessment result of the positioning sector grid.
[0044] As an embodiment of the present invention, determining the positioning sector grid of the sound source signal includes: obtaining the sound source coordinate value of the sound source relative to the coordinate system of the intelligent driving car; determining an ellipse with the center of the vehicle as the origin and the detection range of the microphone array module as the boundary as the acoustic positioning area; using an angle dividing line to divide the acoustic positioning area into different positioning sectors, and the positioning sectors are numbered in a clockwise or counterclockwise direction; using a distance dividing line to divide the positioning sector into different positioning sector grids, and the positioning sector grids are numbered in a manner from far to near or from near to far from the vehicle; determining the positioning sector number and the positioning sector grid number where the sound source is located according to the sound source coordinate value.
[0045] As an embodiment of the present invention, the multi-sensor fusion judgment module is further used to: when the acoustic collision risk assessment result is acoustic non-collision, obtain the ultrasonic radar detection results and visual collision risk assessment results in each positioning sector grid; obtain the accident detection results of each positioning sector grid according to whether each positioning sector grid is a detection blind area of the ultrasonic radar, whether it is a detection blind area of the camera, the ultrasonic radar detection results and the visual collision risk assessment results.
[0046] As an embodiment of the present invention, the central control unit is used to turn on the double flash warning lights of the smart driving car through the remote information processor when the accident detection result is an accident; read the location information of the smart driving car through the positioning module; read the relevant information of the accident collected by the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, and the positioning module within a preset time period before and after the accident, and store the relevant information of the accident in the storage module of the central control unit; and generate an accident information report.
[0047] refer to Figure 2 , which shows a schematic block diagram of the structure and connection relationship of a microphone array module based on MEMS of an intelligent driving car accident detection system according to an embodiment of the present invention. The microphone array module includes a plurality of microphone array modules based on micro-electro-mechanical systems (MEMS), Figure 2The figures are respectively shown as follows: the first MEMS-based microphone array module, the second MEMS-based microphone array module, and the nth MEMS-based microphone array module. Taking the first MEMS-based microphone array module as an example, each MEMS-based microphone array module among the multiple MEMS-based microphone array modules may include multiple MEMS microphones, a microphone array module processor, and a microphone array module automotive audio bus (Autmotive Audio Bus, A2B) transceiver controller power module. Among them, the microphone array module processor may be a digital signal processor (Digital Signal Processing, DSP). The acoustic interface of the microphone array module processor is respectively connected to the output interface of the MEMS microphone and the microphone array module A2B transceiver; multiple MEMS-based microphone array modules are sequentially connected via A2B.
[0048] One embodiment of the signal output port interface of the MEMS microphone is an I2S (Integrated interchip Sound) interface. The DSP may be a DSP with an acoustic interface. Specific examples of the acoustic interface include: an I2S interface or a McBSP (Multichannel Buffered Serial Port) or a McASP (Multichannel Audio Serial Port) interface. The output interface of the MEMS microphone is connected to the acoustic interface of the DSP. The acoustic interface of the DSP is connected to the A2B transceiver. The output end of the controller power module is connected to the power input end of the DSP, the A2B transceiver and the MEMS microphone, and supplies power to them. The first MEMS-based microphone array module, the second MEMS-based microphone array module, and the nth MEMS-based microphone array module can be sequentially connected in a daisy chain connection through the A2B bus. As an example only, the model of the DSP may be, for example, TMS320C6748.
[0049] refer to Figure 3, which shows a schematic block diagram of the structural composition and connection relationship of the acoustic electronic control unit of the intelligent driving car accident detection system according to an embodiment of the present invention. The acoustic electronic control unit may include a controller power module, an acoustic electronic control unit processor, a serial port-controller area network (Controller Area Network, CAN) conversion module, an acoustic electronic control unit A2B transceiver and an acoustic electronic control unit CAN transceiver. Among them, the serial port of the acoustic electronic control unit processor is connected to the serial port-CAN conversion module; the serial port-CAN conversion module is connected to the acoustic electronic control unit CAN transceiver; the acoustic electronic control unit A2B transceiver is connected to the acoustic interface of the acoustic electronic control unit processor. The acoustic electronic control unit processor may be a DSP with an acoustic interface. Specific examples of the acoustic interface may include: I2S interface or McBSP (Multichannel Buffered Serial Port: Multichannel Buffered Serial Port) or McASP (Multichannel Audio Serial Port: Multichannel Audio Serial Port) interface. As an example, the model of the DSP may be, for example, TMS320C6748.
[0050] The internal connection relationship of the acoustic electronic control unit is as follows: the output end of the controller power module is connected to the power input end of the DSP, the serial-CAN conversion module A2B transceiver and the CAN transceiver, and supplies power to them. The serial port of the DSP is connected to the serial-CAN conversion module. The serial-CAN conversion module is connected to the CAN interface. The A2B transceiver is connected to the acoustic interface of the DSP. As an example, CAN can be CAN-FD (Controller Area Network with Flexible Data-Rate: Controller Area Network with Flexible Data Rate).
[0051] refer to Figure 4, which shows a schematic block diagram of the structural composition and connection relationship of the central control unit of the intelligent driving car accident detection system according to an embodiment of the present invention. The central control unit includes an embedded computer or a computer system on a chip, at least five central control unit CAN transceivers and a storage module. The embedded computer or computer system on a chip may be a SOC (System on Chip: System on Chip) with a CAN interface and a storage interface. CAN, for example, may be CAN-FD. The storage interface may include, for example, the following: UFS (Universal Flash Storage: Universal Flash Storage) interface, SD (Secure Digital: Secure Digital) interface and eMMC (embedded MultiMediaCard Interface: embedded Multimedia Card) interface. The model of the embedded computer or computer system on a chip may be, for example, TDA4VM. The storage module may be a mobile storage device, such as a MicroSD card. The internal connection relationship of the central control unit is: the output end of the controller power module is connected to the power input end of at least five CAN transceivers, embedded computers or computer systems on a chip, and supplies power to them. At least five CAN transceivers are respectively connected to the CAN interface of the embedded computer or the on-chip computer system, and are adapted to be respectively connected to the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, the positioning module and the telematics processor. The storage module is connected to the storage interface of the embedded computer or the on-chip computer system.
[0052] As an embodiment of the present invention, the surround view module includes a plurality of cameras and a surround view controller installed around the body of the smart driving car. The camera may be a wide-angle camera module, for example, with a horizontal viewing angle greater than 180°. The input port of the surround view controller may be an in-vehicle Ethernet interface, and the output port may be a CAN interface. The plurality of cameras may be connected to the surround view controller via the in-vehicle Ethernet interface.
[0053] As an embodiment of the present invention, the surround ultrasonic radar module includes a plurality of ultrasonic radars and a parking controller. The plurality of ultrasonic radars can be installed at the front and rear ends of the vehicle. The input port of the parking controller can be a LIN interface, and the output port can be a CAN interface. The plurality of ultrasonic radars can be connected to the parking controller via a LIN line.
[0054] The output port of the positioning module 6 may be a CAN interface, and the positioning module can output the location information of the vehicle.
[0055] As an embodiment of the present invention, the telematics processor 7 (T-BOX: Telematics BOX) can communicate with a remote device, such as a mobile phone APP (application). The output port of the telematics processor 7 can be a CAN interface. The telematics processor 7 can send the vehicle positioning information obtained from the positioning module 6 to a preset remote device. As an embodiment of the present invention, the telematics processor can also be used to receive control information sent by the remote device to control the smart driving car to turn on the double flash warning lights.
[0056] As an example only, the connection relationship between the components of the intelligent driving car accident detection system can be as follows: the A2B transceiver of the microphone array module 1 is connected to the A2B transceiver of the acoustic electronic control unit 2 through the A2B bus; the CAN transceiver of the acoustic electronic control unit 2 (AECU: Acoustic Electronic Control Unit) is connected to the CAN transceiver of the central control unit 3 (CECU: Central Electronic Control Unit) by twisted pair cables; the CAN transceiver of the central control unit 3 (CECU: Central Electronic Control Unit) is connected to the CAN interface of the telematics processor 7 (T-BOX: Telematics BOX) by twisted pair cables; the CAN interface of the surround view controller of the surround view module 4 is connected to the CAN transceiver of the central control unit 3 by twisted pair cables; the CAN interface of the parking controller of the surround view ultrasonic radar module 5 is connected to the CAN transceiver of the central control unit 3 by twisted pair cables; the CAN interface of the positioning module 6 is connected to the CAN transceiver of the central control unit 3 by twisted pair cables.
[0057] The installation location and fixing method of the components of the intelligent driving car accident detection system are as follows.
[0058] Each MEMS-based microphone array module of the microphone array group 1 can be set at one or more of the following positions on the left or right side of the vehicle: on the body sheet metal between the B-pillar and the ceiling of the vehicle, on the body sheet metal between the B-pillar and the floor of the vehicle, on the body sheet metal between the A-pillar and the ceiling of the vehicle, on the body sheet metal between the A-pillar and the floor of the vehicle, on the body sheet metal between the C-pillar and the ceiling of the vehicle, on the body sheet metal between the C-pillar and the floor of the vehicle, and under the front seat. Among them, the preferred position is the body sheet metal between the B-pillar and the ceiling of the vehicle. The fixing method can be clamping, bonding, screw connection, bolt connection or welding.
[0059] The acoustic electronic control unit 2 can be arranged on a body sheet metal part of the vehicle's center armrest attachment, and can be fixed by clamping, bonding, screwing, bolting or welding.
[0060] The central control unit 3 can be arranged under the center console of the vehicle or in the trunk. The fixing method can be clamping, bonding, screw connection, bolt connection or welding.
[0061] The multiple cameras of the surround view module 4 can be arranged around the vehicle, including but not limited to: the bumper, radiator grille, and front windshield at the front of the vehicle; the fenders, exterior mirror housings, B-pillar outer trim panels, and C-pillar body panels at the sides of the vehicle; the rear bumper, rear windshield, and rear cover at the rear of the vehicle; and the surround view controller of the surround view module 4 is fixed under the center console of the vehicle. The fixing method can be clamping, bonding, screw connection, bolt connection, or welding.
[0062] The multiple ultrasonic radars of the surround ultrasonic radar module 5 can be arranged at the front and rear ends of the vehicle, including but not limited to the front bumper and the rear bumper. The parking controller of the surround ultrasonic radar module 5 is fixed under the center console of the vehicle. The fixing method can be clamping, bonding, screw connection, bolt connection or welding.
[0063] refer to Figure 5 , which shows a working flow chart of an intelligent driving car accident detection system according to an embodiment of the present invention.
[0064] The working process of the intelligent driving car accident detection system may include: 1. accident information detection; 2. multi-sensor fusion judgment; 3. accident handling.
[0065] 1. Collision information detection
[0066] Acoustic microphone array, ultrasonic radar and camera are used to detect collision information. The specific steps are as follows:
[0067] 1.1 Acoustic Microphone Array Detection
[0068] The specific steps are as follows:
[0069] 1.1.1 Acoustic Microphone Array Sound Source Detection Each MEMS-based microphone array module of the microphone array group 1 detects the sound sources around the vehicle respectively.
[0070] 1.1.2 Collision sound recognition. The acoustic electronic control unit 2 reads the sound source signal output by each MEMS-based microphone array module of the microphone array module. Then, the acoustic collision detection intelligent algorithm is used to identify the probability Rvoice that the sound source signal is a collision sound. The value range of Rvoice is 0-100%. The larger the Rvoice value, the higher the collision risk. Rvoice=0 means that there is no collision risk at present, and Rvoice=100% means that the current collision risk is the highest.
[0071] The acoustic collision detection intelligent algorithm is obtained by training the acoustic neural network model with the traffic accident sound data set. The traffic accident sound data set is an audio file of the sound emitted when the vehicle collides. The true value of the sample in the traffic accident sound data set is the probability of the vehicle accident judged based on the sound. A specific example of the acoustic neural network model can be the BP neural network model.
[0072] 1.1.3 Acoustic collision risk assessment. The preset threshold values are RvoiceLow, RvoiceMid, and RvoiceHig, and the relationship between the threshold values is: 0≤RvoiceLow<RvoiceMid<RvoiceHig≤100%.
[0073] The acoustic electronic control unit 2 divides the possibility of collision sound into four levels: high, medium, low, and non-collision sound according to the preset threshold values RvoiceLow, RvoiceMid, and RvoiceHig:
[0074] When Rvoice≤RvoiceLow, the acoustic collision risk assessment result RvoiceOut is: non-collision;
[0075] When RvoiceLow<Rvoice≤RvoiceMid, the acoustic collision risk assessment result RvoiceOut is: low;
[0076] When RvoiceMid<Rvoice≤RvoiceHig, the acoustic collision risk assessment result RvoiceOut is: medium;
[0077] When Rvoice>RvoiceHig, the acoustic collision risk assessment result RvoiceOut is: high.
[0078] 1.1.4 Calculation of acoustic positioning information For sounds with Rvoice>RvoiceLow, the acoustic electronic control unit 2 locates the sound source according to the sound signal of the microphone array group 1. The result obtained by the sound source positioning method is the coordinate value of the sound source in the vehicle coordinate system relative to the center of the vehicle.
[0079] As an embodiment of the present invention, the acoustic positioning information may be calculated based on the above-mentioned sound source coordinate values. Figure 6 , which shows a schematic diagram of the acoustic positioning principle of the intelligent driving car accident detection system according to an embodiment of the present invention. In the process of calculating the acoustic positioning information, an ellipse is determined as the acoustic positioning area with the center of the vehicle as the origin and the detection range of the microphone array group 1 as the boundary. Then, the acoustic positioning area is divided into different positioning sectors by angle dividing lines, and the positioning sectors are numbered in a clockwise or counterclockwise direction. The positioning sector is divided into different sector grids by distance dividing lines, and the sector grids are numbered in a manner from far to near or from near to far from the vehicle. In the illustrated embodiment of this institute, the acoustic positioning area is divided into 16 positioning sectors, and the positioning sectors are numbered in a clockwise manner, and the sector to the right of the vehicle is 1 sector. Each positioning sector is divided into 4 sector grids, and the sector grids are numbered in a manner from near to far from the vehicle, and the sector grid closest to the vehicle is numbered 1. The positioning sector number and the sector grid number where the sound source is located can be determined according to the coordinate value of the sound source. In this embodiment, the sound source is in 3 sectors and 3 grids. Therefore, the acoustic positioning information PvoiceOut calculated based on the sound source coordinates is (3, 3)
[0080] 1.1.5 Sending of risk information and positioning information. The acoustic electronic control unit 2 sends the acoustic collision risk assessment result RvoiceOut and the acoustic positioning information PvoiceOut to the multi-sensor fusion determination module in the central control unit 3.
[0081] 1.2 Ultrasonic radar detection
[0082] 1.2.1 Ultrasonic radar detects the distance of obstacles. Surround-view ultrasonic radar module 5 Multiple ultrasonic radars distributed at the front and rear ends of the vehicle detect the distance Dradar of obstacles around the vehicle in real time.
[0083] 1.2.2 Ultrasonic collision risk assessment. The preset threshold values are DradarLow, DradarMid, and DradarHig. The relationship between the threshold values is: 0≤DradarLow<DradarMid<DradarHig≤DradarRange. DradarRange is the maximum distance that the ultrasonic radar can detect.
[0084] The 5 surround-view ultrasonic radar modules divide the collision risk of each ultrasonic radar into four levels: high, medium, low, and non-collision sound according to the preset distance thresholds DradarLow, DradarMid, and DradarHig:
[0085] When Dradar≤DradarLow, the ultrasonic radar sets the alarm information result RradarOut to: high;
[0086] When DradarLow<Dradar≤DradarMid, the ultrasonic radar will set the alarm information result RradarOut to: medium;
[0087] When DradarMid<Dradar≤DradarHig, the ultrasonic radar will set the alarm information result RradarOut to: low;
[0088] When Dradar>DradarHig, the ultrasonic radar will set the alarm information result RradarOut to: non-collision;
[0089] The ultrasonic radar sends an alarm message and an ultrasonic radar identifier (ID). Each ultrasonic radar in the surround ultrasonic radar module 5 defines a different ID value IDradar. The ultrasonic radar detection result RradarOut and the ultrasonic radar ID value IDradar are sent to the multi-sensor fusion judgment module.
[0090] 1.3 Camera Detection
[0091] 1.3.1 Camera Detection of Vehicle Obstacles Multiple cameras distributed around the vehicle in the surround view module 4 detect images around the vehicle in real time.
[0092] 1.3.2 Camera control unit collision risk identification, surround view module 4 uses visual intelligence algorithm to identify the probability of collision Rvisual in each camera shooting picture. Rvisual value range is 0-100%, the larger the Rvisual value, the higher the collision risk, where Rvisual = 0 means there is no collision risk at present, and Rvisual = 100% means the current collision risk is the highest.
[0093] The visual intelligence algorithm is obtained by training the visual neural network model with the traffic accident image dataset. The traffic accident image dataset is an image file taken by a camera when a vehicle collides. The true value of the sample in the traffic accident image dataset is the probability of the vehicle having an accident determined based on the image. A specific example of the visual neural network model can be a BP neural network model.
[0094] 1.3.3 Visual collision risk assessment. The threshold values RvisualLow, RvisualMid, and RvisualHig are preset, and the relationship between the threshold values is: 0≤RvisualLow<RvisualMid<RvisualHig≤100%.
[0095] The surround view module 4 divides the collision risk into four levels: high, medium, low, and non-collision according to the preset threshold values RvisualLow, RvisualMid, and RvisualHig:
[0096] When Rvisual≤RvisualLow, the visual collision risk assessment result RvisualOut is: non-collision;
[0097] When RvisualLow<Rvisual≤RvisualMid, the visual collision risk assessment result RvisualOut is: low;
[0098] When RvisualMid<Rvisual≤RvisualHig, the visual collision risk assessment result RvisualOut is: medium;
[0099] When Rvisual>RvisualHig, the visual collision risk assessment result RvisualOut is: high.
[0100] 1.3.4 The camera sends out collision warning information and camera ID. Each camera in the surround view module 4 defines a different ID value IDvisual. The visual collision risk assessment result RvisualOut and the camera ID value IDvisual are sent to the multi-sensor fusion judgment module.
[0101] 2. Multi-sensor fusion judgment
[0102] Multi-sensor fusion judgment is based on the acoustic microphone array and uses a multi-sensor fusion method to determine whether an accident has occurred. It is specifically divided into the following categories.
[0103] A) RvoiceOut is: High
[0104] When the acoustic collision risk assessment result RvoiceOut is high, the central control unit 3 searches for the detection result of the ultrasonic radar in the positioning sector grid determined by the acoustic positioning information PvoiceOut. If the positioning sector grid is not the detection blind area of the ultrasonic radar, the ultrasonic radar alarm information result RradarOut can be obtained in the positioning sector grid. If the positioning sector grid is not the detection blind area of the camera, the visual collision risk assessment result RvisualOut can be obtained in the positioning sector grid.
[0105] According to whether the positioning sector grid is a blind area, and the status of RradarOut and RvisualOut, it can be determined whether the positioning sector grid status is "accident" or "non-accident". For specific judgment rules, see the following Table 1.
[0106] Table 1. Judgment rules when the acoustic collision risk assessment result is high
[0107]
[0108] B) RvoiceOut is: Medium
[0109] When the acoustic collision risk assessment result RvoiceOut is medium, the central control unit 3 searches for the detection result of the ultrasonic radar in the positioning sector grid determined by the acoustic positioning information PvoiceOut. If the positioning sector grid is not the detection blind area of the ultrasonic radar, the ultrasonic radar alarm information result RradarOut can be obtained in the positioning sector grid. If the positioning sector grid is not the detection blind area of the camera, the visual collision risk assessment result RvisualOut can be obtained in the positioning sector grid.
[0110] According to whether the positioning sector grid is a blind area, and the status of RradarOut and RvisualOut, the positioning sector grid status can be determined as "accident" or "non-accident". The specific judgment rules are as follows:
[0111] Table 2. Judgment rules for the case where the acoustic collision risk assessment result is medium
[0112]
[0113] C) RvoiceOut is: Low
[0114] When the acoustic collision risk assessment result RvoiceOut is low, the central control unit 3 searches for the detection result of the ultrasonic radar in the positioning sector grid determined by the acoustic positioning information PvoiceOut. If the positioning sector grid is not the detection blind area of the ultrasonic radar, the ultrasonic radar alarm information result RradarOut can be obtained in the positioning sector grid. If the positioning sector grid is not the detection blind area of the camera, the visual collision risk assessment result RvisualOut can be obtained in the positioning sector grid.
[0115] According to whether the positioning sector grid is a blind area, and the status of RradarOut and RvisualOut, the positioning sector grid status can be determined as "accident" or "non-accident". The specific judgment rules are as follows:
[0116] Table 3. Judgment rules when the acoustic collision risk assessment result is low
[0117]
[0118]
[0119] D) RvoiceOut is: non-collision
[0120] When the acoustic collision risk assessment result RvoiceOut is non-collision, the central control unit 3 searches for the detection results of the ultrasonic radars in all positioning sector grids to obtain the alarm information results RradarOut and the visual collision risk assessment results RvisualOut of the ultrasonic radars in each positioning sector grid.
[0121] According to whether the positioning sector grid is a blind area, and the status of RradarOut and RvisualOut, the positioning sector grid status can be determined as "accident" or "non-accident". The specific judgment rules are as follows:
[0122] Table 4. Judgment rules for the case where the acoustic collision risk assessment result is non-collision
[0123]
[0124] 3. Accident handling
[0125] 3.1 When no accident occurs, return to 1. Collision information detection. When an accident occurs, proceed as follows:
[0126] 3.2 The central control unit 3 turns on the vehicle's double flash warning lights through the telematics processor 7.
[0127] 3.3 The central control unit 3 reads the current vehicle positioning information through the positioning module 6.
[0128] 3.4 The central control unit 3 reads the sound, photo, distance, positioning information and other data of the collision-related sensors in the acoustic electronic control unit 2, the surround view module 4, the surround view ultrasonic radar module 5 and the positioning module 6 within a preset time t seconds before and after the collision through the CAN transceiver, and stores them in the storage module of the central control unit 3. A specific example of the preset time t may be 10 seconds.
[0129] 3.5 The central control unit 3 generates a collision information report. The collision information report includes the following and some of the contents:
[0130] a) pre-set in the central control unit 3 to help the owner quickly identify the basic vehicle information of the accident vehicle, such as the license plate number, vehicle model, and vehicle color;
[0131] b) The central control unit 3 makes a clear determination through fusion judgment that the vehicle has an accident;
[0132] c) the time of the current accident given by the central control unit 3;
[0133] d) The location of the accident obtained by the central control unit 3 from the positioning module 6;
[0134] e) a reminder from the central control unit 3 that the vehicle owner needs to take action;
[0135] Optionally, the collision information report may also include a summary description of the accident:
[0136] f) A schematic diagram of the sound detection collision position given by the central control unit 3;
[0137] g) A schematic diagram of the collision position detected by the camera given by the central control unit 3;
[0138] h) A schematic diagram of the collision position detected by the ultrasonic radar given by the central control unit 3;
[0139] Optionally, the collision information report can also contain detailed accident detection data information:
[0140] i) the vehicle collision sound file in the acoustic electronic control unit 2 within a preset time t seconds before and after the collision read by the central control unit 3 through the CAN transceiver;
[0141] j) The central control unit 3 reads the vehicle collision photo file in the surround view module 4 within a preset time t seconds before and after the collision through the CAN transceiver;
[0142] k) The central control unit 3 reads the vehicle collision distance data in the surround ultrasonic radar module 5 within a preset time t seconds before and after the collision through the CAN transceiver;
[0143] refer to Figure 7 , which shows a schematic diagram of an example of a fault detection report of an intelligent driving car accident detection system according to an embodiment of the present invention. In this embodiment, the collision information report includes basic information, schematic information and detailed information. The basic information includes vehicle information, accident information and prompt information. The vehicle information includes the license plate number, vehicle model and vehicle color. The accident information includes the judgment conclusion, the time of the accident and the location of the accident. The prompt information is used to prompt the owner of the operation to be performed. The schematic information includes the collision position detected by sound, the collision position detected by camera and the collision position detected by ultrasonic radar. The schematic information illustrates the location of the accident of the owner's vehicle in the form of a schematic diagram. The detailed information includes a collision sound file, a collision photo file and a collision distance data, which are used to give the owner detailed information about the accident.
[0144] 3.6 T-BOX sends collision information report. The telematics processor 7 sends the collision information report to the preset mobile phone. The car owner can learn about the accident in time based on the collision information report sent to the mobile phone, and locate the location of the accident based on the accident location information, so that he can go to the accident location in time to deal with the accident.
[0145] refer to Figure 8, which shows a schematic diagram of the structural organization of an example of an intelligent driving car accident detection system according to an embodiment of the present invention. In this embodiment, the microphone module 1 includes two microphone modules, namely a first MEMS-based microphone array module 101 and a second MEMS-based microphone array module 102; the surround view module 4 includes a surround view controller 405 and four camera modules, and the four camera modules are: a front view camera module 401, a right camera module 402, a left camera module 403, and a rear view camera module 404; the surround view ultrasonic radar module 5 includes a parking controller 613 and 12 ultrasonic radars, and the 12 ultrasonic radars are: a right front corner ultrasonic radar 501, a right front ultrasonic radar 502, a right middle front ultrasonic radar 503, a left middle front ultrasonic radar 504, a left front ultrasonic radar 505, a left front corner ultrasonic radar 506, a right rear corner ultrasonic radar 507, a right rear ultrasonic radar 508, a right middle rear ultrasonic radar 509, a left middle rear ultrasonic radar 510, a left rear ultrasonic radar 511, and a left rear corner ultrasonic radar 512.
[0146] refer to Fig. 9 , which shows a schematic diagram of sensor arrangement of an example of an intelligent driving car accident detection system according to an embodiment of the present invention. The front camera module 401 is fixed on the radiator grille 9; the right camera module 402 and the left camera module 403 are fixed on the right rearview mirror housing 11 and the left rearview mirror housing 12 respectively; the rear camera module 404 is fixed on the rear bumper 20; the right front corner ultrasonic radar 501, the right front ultrasonic radar 502, the right middle front ultrasonic radar 503, the left middle front ultrasonic radar 504, the left front ultrasonic radar 505, and the left front corner ultrasonic radar 506 are fixed on the front bumper 8; the right rear corner ultrasonic radar 507, the right rear ultrasonic radar 508, the right middle rear ultrasonic radar 509, the left middle rear ultrasonic radar 510, the left rear ultrasonic radar 511, and the left rear corner ultrasonic radar 512 are fixed on the rear bumper 20. The fixing methods can be snap-on, welding, bonding, screw connection or bolt connection.
[0147] refer to Fig.10 , which shows a schematic diagram of the working principle of an embodiment of an intelligent driving car accident detection system according to an embodiment of the present invention. The external environment of this embodiment is: the right rear side of the vehicle hits an obstacle, a traffic accident occurs, and a clear collision sound is emitted. In the external environment, the collision obstacle is within the detection range of the right camera, the detection range of the rear ultrasonic radar, the detection range of the first microphone array module, and the detection range of the second microphone array module.
[0148] A specific example of the working process of an embodiment of the present invention is described below in combination with the above external environment.
[0149] 1. Collision information detection
[0150] 1.1 Acoustic Microphone Array Detection
[0151] The specific steps are as follows:
[0152] 1.1.1 Acoustic Microphone Array Sound Source Detection Each MEMS-based microphone array module of the microphone array group 1 detects the sound sources around the vehicle respectively.
[0153] 1.1.2 Collision sound identification The acoustic electronic control unit 2 reads the sound source signal output by the first microphone module 101 and the second microphone array group 102. Then, the acoustic collision detection intelligent algorithm is used to identify the probability Rvoice that the sound source signal is a collision sound.
[0154] 1.1.3 Acoustic collision risk assessment. The acoustic electronic control unit 2 performs an assessment based on the preset threshold values RvoiceLow, RvoiceMid, and RvoiceHig, and identifies that Rvoice>RvoiceHig. The acoustic collision risk assessment result RvoiceOut is: high;
[0155] 1.1.4 Calculation of acoustic positioning information The acoustic electronic control unit 2 locates the sound source according to the sound signals of the first microphone module 101 and the second microphone array group 102. The acoustic positioning information PvoiceOut finally outputted is (2, 2).
[0156] 1.1.5 Sending of risk information and positioning information. The acoustic electronic control unit 2 sends the acoustic collision risk assessment result RvoiceOut and the acoustic positioning information PvoiceOut to the multi-sensor fusion determination module in the central control unit 3.
[0157] 1.2 Ultrasonic radar detection
[0158] 1.2.1 Ultrasonic radar detects obstacle distance. The 12 ultrasonic radars distributed at the front and rear ends of the vehicle detect the distance of obstacles around the vehicle in real time.
[0159] 1.2.2 Ultrasonic collision risk assessment. The surround ultrasonic radar module 5 performs an assessment based on the preset threshold values DradarLow, DradarMid, and DradarHig. The right rear ultrasonic radar 507 (ultrasonic radar ID value IDradar=507) identifies DradarLow<Dradar≤DradarMid, and the ultrasonic radar sets the alarm information result RradarOut as: medium; the other ultrasonic radars identify Dradar>DradarHig, and the ultrasonic radar sets the alarm information result RradarOut as: non-collision.
[0160] The ultrasonic radar sends an alarm message and the ultrasonic radar ID. All ultrasonic radar detection results RradarOut and the ultrasonic radar ID value IDradar are sent to the multi-sensor fusion judgment module.
[0161] 1.3 Camera Detection
[0162] 1.3.1 Camera Detection of Vehicle Obstacles The four cameras distributed around the vehicle in the surround view module 4 detect the images around the vehicle in real time.
[0163] 1.3.2 Camera control unit collision risk identification, the surround view module 4 uses a visual intelligence algorithm to identify the probability Rvisual of a collision in the images captured by each camera.
[0164] 1.3.3 Visual collision risk assessment. Based on the preset threshold values RvisualLow, RvisualMid, and RvisualHig, the right camera module 502 (camera ID value IDvisual=502) identifies that when Rvisual>RvisualHig, the visual collision risk assessment result RvisualOut is: high; the other camera modules identify that Rvisual≤RvisualLow, and the visual collision risk assessment result RvisualOut is: non-collision.
[0165] 1.3.4 The camera sends out collision warning information and camera ID. The visual collision risk assessment result RvisualOut and camera ID value IDvisual are sent to the multi-sensor fusion judgment module.
[0166] 2. Multi-sensor fusion judgment
[0167] Based on the acoustic microphone array, a multi-sensor fusion method is used to determine whether an accident has occurred.
[0168] The acoustic collision risk assessment result RvoiceOut is high, the central control unit 3 searches the acoustic positioning information PvoiceOut to determine that the detection result of the ultrasonic radar 607 in the positioning sector grid (2 sectors and 2 grids) is RradarOut=medium, and the visual collision risk assessment result RvisualOut=high.
[0169] According to the status of RradarOut and RvisualOut, according to the rules in Table 1 above, it can be determined that the status of the positioning sector grid (2 sectors and 2 grids) is "accident".
[0170] 3. Accident handling
[0171] 3.1 If the conclusion is that an accident has occurred, proceed as follows:
[0172] 3.2 The central control unit 3 activates the vehicle's double flash warning lights through the telematics processor 7.
[0173] 3.3 The central control unit 3 reads the current vehicle positioning information through the positioning module 6.
[0174] 3.4 The central control unit 3 reads the sound, photos, distance, positioning information and other data of the collision-related sensors in the acoustic electronic control unit 2, the surround view module 4, the surround view ultrasonic radar module 5 and the positioning module 6 within 10 seconds before and after the collision through the CAN transceiver, and stores them in the storage module of the central control unit 3.
[0175] 3.5 The central control unit 3 generates a collision information report. For an example of a collision information report, see Figure 6 .
[0176] 3.6 T-BOX sends a collision information report.
[0177] The intelligent driving car accident detection system proposed in the embodiment of the present invention can solve the problem of accident detection when there is no driver or passenger in the intelligent driving car and the vehicle has an accident. The intelligent driving car accident detection system proposed in the embodiment of the present invention uses a microphone array group, a surround view module, and a surround view ultrasonic radar module to comprehensively determine vehicle accidents, which can timely and accurately determine whether the vehicle has a fault, and can reasonably, effectively and quickly detect traffic accidents. After detecting a traffic accident, the intelligent driving car accident detection system proposed in the embodiment of the present invention can send a collision information report to the driver or the person in charge of the vehicle through a remote information processor, and can promptly notify the owner or the person in charge of the vehicle of information related to the accident.
[0178] The intelligent automobile accident detection system proposed in the embodiments of the present invention has the prospect of being equipped in future mass-produced intelligent driving cars.
[0179] For the purpose of illustration, the foregoing description of the embodiments of the present invention has been given, which is not exhaustive nor intended to limit the present invention to the disclosed exact form. It will be appreciated by those skilled in the art that various changes can be made without departing from the scope of the present invention, and that the elements therein can be replaced with equivalents. In addition, without departing from the basic scope of the present invention, many modifications can be made so that specific situations or materials are adapted to the teachings of the present invention. Therefore, the present invention is not intended to be limited to the specific embodiments disclosed as the best mode for realizing the present invention, and the present invention will include all embodiments falling within the scope of the appended claims.
Claims
1. An intelligent driving car accident detection system, characterized in that: The intelligent driving car accident detection system includes a microphone array module, an acoustic electronic control unit, a central control unit, a surround view module, a surround view ultrasonic radar module, a positioning module and a telematics processor, wherein the central control unit includes a multi-sensor fusion determination module, The microphone array module is used to detect sound source signals around the smart driving car; The acoustic electronic control unit is used to: determine the probability that the sound source signal is a collision sound based on the sound source signal, and obtain an acoustic collision risk assessment result according to the probability that the sound source signal is a collision sound; calculate acoustic positioning information according to the acoustic collision risk assessment result; and send the acoustic collision risk assessment result and the acoustic positioning information to the multi-sensor fusion determination module; The surround ultrasonic radar module is used to: detect the obstacle distance through the ultrasonic radar, determine the ultrasonic radar detection result according to the obstacle distance, and send the ultrasonic radar detection result and the ultrasonic radar identifier to the multi-sensor fusion determination module; The surround view module is used to: collect images of the surroundings of the smart driving car through a camera, determine a visual collision risk assessment result based on the image, and send the visual collision risk assessment result and a camera identifier to the multi-sensor fusion determination module; The multi-sensor fusion determination module is used to generate an accident detection result according to one or more of the acoustic collision risk assessment result, the acoustic positioning information, the ultrasonic radar detection result, the ultrasonic radar identifier, the visual collision risk assessment result and the camera identifier; The positioning module is used to obtain the location information of the smart driving car; The telematics processor is used to transmit the accident detection result and / or the location information to a remote device, Wherein, the multi-sensor fusion determination module is further used for: When the acoustic collision risk assessment result is a high acoustic collision risk, a medium acoustic collision risk, or a low acoustic collision risk, determining a positioning sector grid of the sound source signal according to the acoustic positioning information; If the positioning sector grid is not a detection blind area of the ultrasonic radar, obtaining the ultrasonic radar detection result of the positioning sector grid in which the ultrasonic radar detection result is located in the positioning sector grid; If the positioning sector grid is not a detection blind area of the camera, obtaining the positioning sector grid visual collision risk assessment result of the visual collision risk assessment result in the positioning sector grid; The accident detection result of the positioning sector grid is obtained according to whether the positioning sector grid is a detection blind area of the ultrasonic radar, whether it is a detection blind area of the camera, the ultrasonic radar detection result of the positioning sector grid, and the visual collision risk assessment result of the positioning sector grid. Wherein, determining the positioning sector grid of the sound source signal includes: Obtaining a sound source coordinate value of a sound source relative to a coordinate system of the intelligent driving vehicle; An ellipse is determined with the center of the vehicle as the origin and the detection range of the microphone array module as the boundary, as the acoustic positioning area; Dividing the acoustic positioning area into different positioning sectors by using angle dividing lines, and the positioning sectors are numbered in a clockwise or counterclockwise direction; The positioning sector is divided into different positioning sector grids by using distance dividing lines, and the positioning sector grids are numbered in a manner from far to near or from near to far according to the distance from the vehicle; The positioning sector number and the positioning sector grid number where the sound source is located are determined according to the sound source coordinate value.
2. The intelligent driving car accident detection system according to claim 1, characterized in that: The multi-sensor fusion determination module is further used for: When the acoustic collision risk assessment result is acoustic non-collision, obtaining the ultrasonic radar detection result and the visual collision risk assessment result in each positioning sector grid; The accident detection results of each positioning sector grid are obtained according to whether each positioning sector grid is a detection blind area of the ultrasonic radar, whether it is a detection blind area of the camera, the ultrasonic radar detection result and the visual collision risk assessment result.
3. The intelligent driving car accident detection system according to claim 1, characterized in that: The central control unit is used for, when the accident detection result is an accident, Turning on the double flash warning lights of the smart driving car through the telematics processor; Reading the location information of the smart driving car through the positioning module; Reading relevant information of the accident collected by the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, and the positioning module within a preset time period before and after the accident, and storing the relevant information of the accident in the storage module of the central control unit; Generate incident information report.
4. The intelligent driving car accident detection system according to claim 1, characterized in that: The microphone array module includes a plurality of microphone array modules based on micro-electromechanical systems, each of the plurality of microphone array modules based on micro-electromechanical systems includes a micro-electromechanical system microphone, a microphone array module processor, and a microphone array module car audio bus transceiver, wherein: The acoustic interface of the microphone array module processor is respectively connected to the output interface of the micro-electromechanical system microphone and the microphone array module car audio bus transceiver; The plurality of micro-electromechanical system-based microphone array modules are sequentially connected via a car audio bus.
5. The intelligent driving car accident detection system according to claim 4, characterized in that: The multiple micro-electromechanical system-based microphone array modules are arranged at one or more of the following positions: on a body sheet metal between the B-pillar and the ceiling of the vehicle, on a body sheet metal between the B-pillar and the floor of the vehicle, on a body sheet metal between the A-pillar and the ceiling of the vehicle, on a body sheet metal between the A-pillar and the floor of the vehicle, on a body sheet metal between the C-pillar and the ceiling of the vehicle, on a body sheet metal between the C-pillar and the floor of the vehicle, and under the front seats.
6. The intelligent driving car accident detection system according to claim 1, characterized in that: The acoustic electronic control unit includes an acoustic electronic control unit processor, a serial port-controller area network conversion module, an acoustic electronic control unit car audio bus transceiver and an acoustic electronic control unit controller area network transceiver, wherein: The serial port of the acoustic electronic control unit processor is connected to the serial port-controller area network conversion module; The serial port-controller area network conversion module is connected to the acoustic electronic control unit controller area network transceiver; The acoustic electronic control unit car audio bus transceiver is connected to the acoustic interface of the acoustic electronic control unit processor.
7. The intelligent driving car accident detection system according to claim 1, characterized in that: The acoustic electronic control unit is arranged on a body sheet metal member near a center armrest in the vehicle.
8. The intelligent driving car accident detection system according to claim 1, characterized in that: The central control unit includes an embedded computer or a computer system on a chip, at least five central control unit controller area network transceivers and a storage module, wherein: The at least five central control unit controller area network transceivers are all connected to an embedded computer or a computer system on a chip, and are adapted to be connected to the acoustic electronic control unit, the surround view module, the surround view ultrasonic radar module, the positioning module and the telematics processor, respectively; The storage module is connected to a storage interface of the embedded computer or computer system on chip.
9. The intelligent driving car accident detection system according to claim 1, characterized in that: The central control unit is arranged under the center console or in the trunk of the vehicle.
10. The intelligent driving car accident detection system according to claim 1, characterized in that: The surround view module includes a plurality of cameras and a surround view controller installed around the body of the smart driving car.
11. The intelligent driving car accident detection system according to claim 10, characterized in that: The multiple cameras of the surround view module are arranged at one or more of the following positions: a bumper, a radiator grille, and a front windshield at the front end of the vehicle; a fender, an exterior mirror housing, a B-pillar outer trim panel, a C-pillar body sheet metal, a rear bumper, a rear windshield, and a rear cover at the side of the vehicle; and, The surround view controller of the surround view module is arranged under the center console of the vehicle.
12. The intelligent driving car accident detection system according to claim 1, characterized in that: The surround-view ultrasonic radar module includes a plurality of ultrasonic radars and a parking controller.
13. The intelligent driving car accident detection system according to claim 12, characterized in that: The plurality of ultrasonic radars are arranged at the front bumper and / or the rear bumper of the vehicle, and The parking controller is arranged under the center console of the vehicle.
14. The intelligent driving car accident detection system according to claim 1, characterized in that: The telematics processor is also used to receive control information sent by the remote device to control the smart driving car to turn on the double flash warning lights.
