Vehicle scratch early warning system and damage positioning method based on dynamic risk assessment

Abstract

The application relates to the field of automobile active safety technology and provides a vehicle scratching early warning system and damage positioning method based on dynamic risk assessment, which comprises the following steps: detecting the contact distance between an object and a vehicle body in real time through capacitive sensing vehicle clothing, and starting a three-level risk identification model according to the contact distance; obtaining the relative speed of the object based on a millimeter wave radar and obtaining the close time through an ultrasonic sensor; calculating the threat level according to a real-time risk scoring algorithm; triggering multi-sensor fusion analysis when the threat level is greater than a threshold value, and generating a visual report containing the three-dimensional coordinates of the damage position. The application adopts a ring vehicle deployment scheme of capacitive sensing vehicle clothing and a 4D millimeter wave radar, realizes hardware innovation, adopts a dynamic risk scoring model and a damage 3D mapping engine, initiates a damage 3D mapping algorithm based on physical parameters, and dynamically evaluates the risk, and adopts trigger type directional monitoring, which not only saves storage space, but also improves privacy.

Description

Technical Field

[0001] This invention relates to the field of active safety technology for automobiles, and in particular to a vehicle collision warning system and damage location method based on dynamic risk assessment. Background Technology

[0002] In existing technologies, the Sentinel mode is often used to automatically monitor the surrounding environment when the vehicle is parked. However, the Sentinel mode (such as Tesla) relies on continuous recording and moving object detection.

[0003] In the prior art, the pure vision solution used in patent US20200198581A is difficult to accurately quantify contact parameters, the simple motion detection in patent CN113442834A cannot distinguish threat levels, vehicle environment monitoring technology cannot achieve centimeter-level damage location under low power consumption, and lacks a risk dynamic assessment engine based on physical parameters. Summary of the Invention

[0004] The purpose of this invention is to provide a vehicle collision warning system and damage location method based on dynamic risk assessment, which uses dynamic risk assessment to achieve damage location.

[0005] This invention is implemented as follows: a vehicle collision warning system and damage location method based on dynamic risk assessment, comprising:

[0006] The capacitive sensing car cover detects the contact distance between objects and the car body in real time, and activates a three-level risk identification model based on the contact distance.

[0007] The relative velocity of an object is obtained using millimeter-wave radar, and the time of approach is obtained using an ultrasonic sensor.

[0008] Threat level is calculated based on a real-time risk scoring algorithm;

[0009] When the threat level exceeds a threshold, multi-sensor fusion analysis is triggered, and a visualization report containing the three-dimensional coordinates of the damaged area is generated.

[0010] Preferably, the step of activating the three-level risk identification model based on contact distance specifically includes:

[0011] When the contact distance is greater than 0.5m, it is considered low risk and data recording is not activated;

[0012] When the contact distance is 0.3-0.5m, it is determined to be a medium risk, and the ultrasonic sensor is activated for scanning;

[0013] When the contact distance is less than 0.3m, it is judged as high risk and triggers multi-sensor fusion analysis.

[0014] Preferably, the calculation formula of the real-time risk scoring algorithm is:

[0015] Risk = dv × t × k object

[0016] Where Risk is the risk score, v is the relative velocity of the object, t is the approach time of the object, d is the minimum contact distance between the object and the vehicle body, and k object The risk factor of the object.

[0017] Preferably, the multi-sensor fusion analysis specifically includes:

[0018] Electrode positioning for capacitive car wraps: The center point of the contact area (x, y, z) is determined by the ID coordinates of the trigger electrode. c ,y c );

[0019] Millimeter-wave radar data fusion: obtaining the object's azimuth angle θ and the distance d from the object to the vehicle. r Calculate the spatial offset:

[0020] Δx=d r ×cosθ

[0021] Δy=d r ×sinθ

[0022] Where Δx is the X-axis projection compensation amount of the millimeter radar wave, and Δy is the Y-axis projection compensation amount of the millimeter radar wave.

[0023] Ultrasonic time-domain calibration: Correcting distance data d based on the continuous proximity time t. u :

[0024]

[0025] Where, d u ′ represents the moving average filter value for ultrasonic ranging, n is the size of the sliding window, and d u,i Let v be the distance between the object and the vehicle measured at the i-th sampling time. i Let Δt be the instantaneous velocity of the object at the i-th sampling time. i The time interval between adjacent sampling points;

[0026] The formula for calculating the coordinates of the damaged area in the visualization report of the three-dimensional coordinates is as follows:

[0027] Three-dimensional coordinate mapping: The weighted least squares method is used to generate the coordinates (x, y, z) of the damaged area;

[0028]

[0029] b = [x c +Δx,y c +Δy,du ′] T

[0030] Where A is the sensor position matrix, A T Let A be the transpose of A, W be the weight matrix, and b be the observation vector.

[0031] Preferably, the visualization report includes: a three-dimensional reconstruction of the contact object's motion trajectory, a predicted damage probability value, and characteristic information of the suspected object.

[0032] A vehicle collision warning system based on dynamic risk assessment, characterized in that it includes:

[0033] The perception layer uses multiple sensors to detect the contact distance between objects and the vehicle body in real time, and establishes a three-level risk contact identification model.

[0034] The edge computing layer uses a real-time risk scoring algorithm to assess and calculate the risk of scratches;

[0035] The interaction layer is used to push reports containing 3D location maps and information about suspected objects.

[0036] Preferably, the sensing layer includes a capacitive sensing car cover, a millimeter-wave radar, and an ultrasonic sensor;

[0037] A capacitive sensing car cover is arranged around the car. The capacitive sensing car cover is composed of a distributed electrode array and is used to detect the contact distance between objects and the car body in real time.

[0038] A 4D millimeter-wave radar array is used to obtain the relative velocity of objects;

[0039] An ultrasonic sensor is used to determine the time it takes for an object to come close.

[0040] Preferably, the edge computing layer employs a real-time risk scoring algorithm to assess and calculate the risk of scraping, specifically including:

[0041] The Renesas R-Car V3H chip runs a dynamic risk assessment engine;

[0042] Real-time execution risk formula: Risk = dv × t × k object ;

[0043] Dynamic configuration of object risk coefficients; risk coefficient k for metal objects. metal =1.5, the risk coefficient k of the organism biology =0.8.

[0044] Preferably, the interaction layer pushes a report containing a 3D location map and information about the suspect, specifically including:

[0045] One-click generation of insurance claim materials module;

[0046] Intelligent cost estimation and prompts for repairs;

[0047] Images of suspects captured by a directional activation camera.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] 1. This invention adopts a vehicle surround deployment scheme combining capacitive sensing car cover and 4D millimeter-wave radar, achieving hardware innovation;

[0050] 2. This invention employs a dynamic risk scoring model and a damage 3D mapping engine, pioneering a damage 3D mapping algorithm based on physical parameters to dynamically assess risk;

[0051] 3. This invention employs trigger-based targeted monitoring, identifying only potential threatening behaviors and saving only risk event characteristic data. By replacing the original video storage architecture with characteristic data, it not only saves storage space but also improves privacy. Attached Figure Description

[0052] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention;

[0054] Figure 2 This is a flowchart of the three-level risk identification model in Embodiment 1 of the present invention;

[0055] Figure 3 This is a flowchart of the technical chain of Embodiment 2 of the present invention;

[0056] Figure 4 This is a flowchart of the process of Embodiment 2 of the present invention. Detailed Implementation

[0057] To better understand the technical content of this invention, the technical solutions of this invention are further described and explained below with reference to specific embodiments, but are not limited thereto. The technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0058] Example 1

[0059] refer to Figure 1 and Figure 2 The vehicle collision damage localization method based on dynamic risk assessment includes:

[0060] S1: The capacitive sensing car cover detects the contact distance between objects and the car body in real time, and activates a three-level risk identification model based on the contact distance;

[0061] The activation of the three-level risk identification model based on contact distance specifically includes:

[0062] When the contact distance is greater than 0.5m, it is considered low risk and data recording is not activated;

[0063] When the contact distance is 0.3-0.5m, it is determined to be a medium risk, and the ultrasonic sensor is activated for scanning;

[0064] When the contact distance is less than 0.3m, it is judged as high risk and triggers multi-sensor fusion analysis.

[0065] S2: The relative velocity of the object is obtained based on millimeter-wave radar, and the proximity time is obtained by ultrasonic sensor;

[0066] S3: Calculates the threat level based on a real-time risk scoring algorithm;

[0067] The calculation formula for the real-time risk scoring algorithm is:

[0068] Risk = dv × t × k object

[0069] Where Risk is the risk score, v is the relative velocity of the object, t is the approach time of the object, d is the minimum contact distance between the object and the vehicle body, and k object Risk factor of an object

[0070] S4: When the threat level exceeds the threshold, trigger multi-sensor fusion analysis and generate a visualization report containing the three-dimensional coordinates of the damaged area.

[0071] Multi-sensor fusion analysis specifically includes:

[0072] Electrode positioning for capacitive car wraps: The center point of the contact area (x, y, z) is determined by the ID coordinates of the trigger electrode. c ,y c );

[0073] Millimeter-wave radar data fusion: obtaining the object's azimuth angle θ and the distance d from the object to the vehicle. r Calculate the spatial offset:

[0074] Δx=d r ×cosθ

[0075] Δy=d r ×sinθ

[0076] Where Δx is the X-axis projection compensation amount of the millimeter radar wave, and Δy is the Y-axis projection compensation amount of the millimeter radar wave.

[0077] Ultrasonic time-domain calibration: Correcting distance data d based on the continuous proximity time t. u :

[0078]

[0079] Where, d u ′ represents the moving average filter value for ultrasonic ranging, n is the size of the sliding window, and d u,i Let v be the distance between the object and the vehicle measured at the i-th sampling time. i Let Δt be the instantaneous velocity of the object at the i-th sampling time. i This represents the time interval between adjacent sampling points.

[0080] The visualization report includes: a 3D reconstruction of the contact object's trajectory, a predicted damage probability, and characteristic information of the suspected object.

[0081] The formula for calculating the coordinates of the damaged area in the visualization report of the three-dimensional coordinates is as follows:

[0082] Three-dimensional coordinate mapping: The weighted least squares method is used to generate the coordinates (x, y, z) of the damaged area;

[0083]

[0084] b = [x c +Δx,y c +Δy,d u ′] T

[0085] Where A is the sensor position matrix, A T Let A be the transpose of A, W be the weight matrix, and b be the observation vector.

[0086] A vehicle collision warning system based on dynamic risk assessment, characterized in that it includes:

[0087] The perception layer uses multiple sensors to detect the contact distance between objects and the vehicle body in real time, and establishes a three-level risk contact identification model.

[0088] The sensing layer includes a capacitive sensing car cover, millimeter-wave radar, and ultrasonic sensors;

[0089] A capacitive sensing car cover is arranged around the car. The capacitive sensing car cover is composed of a distributed electrode array and is used to detect the contact distance between objects and the car body in real time.

[0090] A 4D millimeter-wave radar array is used to obtain the relative velocity of objects;

[0091] An ultrasonic sensor is used to determine the time it takes for an object to come close.

[0092] Multimodal threat behavior analysis using multiple sensors:

[0093] sensor Detection parameters Threat determination criteria millimeter-wave radar Relative speed > 5km / h Rapid approach means high risk ultrasonic sensor Continuous close proximity time > 2 seconds Suspicious loitering behavior Capacitive car cover <![CDATA[Contact area > 10 cm 2 > Eliminate interference from small objects such as flying insects

[0094] The edge computing layer uses a real-time risk scoring algorithm to assess and calculate the risk of scratches;

[0095] The edge computing layer employs a real-time risk scoring algorithm to assess and calculate the risk of minor collisions, specifically including:

[0096] The Renesas R-Car V3H chip runs a dynamic risk assessment engine;

[0097] Real-time execution risk formula: Risk = dv × t × k object ;

[0098] Dynamic configuration of object risk coefficients; risk coefficient k for metal objects. metal =1.5, the risk coefficient k of the organism biology =0.8.

[0099] The interaction layer is used to push reports containing 3D location maps and information about suspected objects;

[0100] The interaction layer pushes a report containing a 3D location map and information about the suspect, specifically including:

[0101] One-click generation of insurance claim materials module;

[0102] Intelligent cost estimation and prompts for repairs;

[0103] Images of suspects captured by a directional activation camera.

[0104] The essential difference between this invention and the Sentinel mode:

[0105] Technical dimension Tesla Sentry Mode This plan Innovation Breakthrough Working Logic Continuous recording + saving triggered by moving objects Dynamic risk assessment triggers targeted monitoring Power consumption reduced by 92% Identify target All moving objects Identify only potential threatening behaviors False alarm rate decreased by 87%. Output Video clip Damage site illustration + risk event report Information density increased by 300% Privacy protection Recording the entire process can easily infringe on privacy. Only save risk event characteristic data Compliant with strict GDPR standards

[0106] Example 2

[0107] refer to Figure 3 and Figure 4 Roadside parking scenario:

[0108] 1. A delivery tricycle passes close to the right side of the vehicle at 8 km / h (distance < 0.3m);

[0109] 2. System Trigger:

[0110] Targeted activation of the right-view camera to capture the license plate of the three-wheeled vehicle;

[0111] The sensor performs multimodal threat behavior analysis, including close-range detection of the capacitive car cover, contact time of 1.8 seconds → dynamic risk assessment engine → millimeter-wave radar behavior analysis, where the millimeter-wave radar determines if the speed is greater than the risk threshold, for example: the sensor detects an object rapidly approaching from 0.2m away;

[0112] The processor then calculates the risk value; for example, a risk value of 86 indicates a high risk.

[0113] If a high risk is detected, the high-definition camera is activated to take directional pictures of the approaching object. For example, the right-view camera is activated to capture the license plate of the tricycle and a vector image of the damage location is generated by 3D mapping of the damaged area.

[0114] 3. Push notifications to the user's phone when they return:

[0115] Send visual, targeted reports to the user terminal, such as "Please check the right rear door".

[0116]

[0117] The above-described embodiments are only some embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for locating vehicle scratch damage based on dynamic risk assessment, characterized in that, The capacitive sensing car cover composed of a distributed electrode array arranged in a ring shape detects the contact distance between an object and the car body in real time, and starts a three-level risk identification model according to the contact distance; The relative speed of the object is obtained based on a millimeter wave radar, and the approach time is obtained based on an ultrasonic sensor; The threat level is calculated according to a real-time risk scoring algorithm; When the threat level is greater than a threshold value, multi-sensor fusion analysis is triggered, and a visual report containing three-dimensional coordinates of the damage site is generated; The multi-sensor fusion analysis specifically includes: The damage site coordinate calculation formula in the visual report of the three-dimensional coordinates is: Capacitive car cover electrode positioning: the center point of the contact area is determined by triggering the ID coordinates of the electrode ; Millimeter wave radar data fusion: obtaining object azimuth and object distance to vehicle , calculating spatial offset ； ； wherein, is a compensation amount for the X-axis projection of the millimeter radar wave, is a compensation amount for the Y-axis projection of the millimeter radar wave; Ultrasonic time domain calibration: according to continuous proximity time Corrected distance data : ； wherein, is a moving average filter value for the ultrasonic ranging, is a size of the moving window, is a distance between the object and the vehicle measured at the sample time, is an instantaneous speed of the object at the sample time, is a time interval of adjacent sample points; The three-level risk identification model started according to the contact distance specifically includes: Three-dimensional coordinate mapping: weighted least squares method is used to generate coordinates of the damage site ; ； ； wherein, is a sensor position matrix, is the transpose of is a weight matrix, is an observation vector.

2. The dynamic risk assessment based vehicle ding damage localization method of claim 1, wherein, When the contact distance is greater than 0.5 m, it is determined as low risk, and data recording is not activated; When the contact distance is 0.3-0.5 m, it is determined as medium risk, and the ultrasonic sensor is activated for scanning; When the contact distance is less than 0.3 m, it is determined as high risk, and multi-sensor fusion analysis is triggered. The calculation formula of the real-time risk scoring algorithm is:

3. The dynamic risk assessment based vehicle ding damage localization method of claim 1, wherein, The visual report includes: three-dimensional reconstruction diagram of the motion trajectory of the contacted object, damage probability prediction value and suspicious object feature information. ； wherein, is a risk score value, is a relative speed of the object, is a closing time of the object, is a minimum contact distance of the object from the vehicle body, is a risk coefficient of the object.

4. The dynamic risk assessment based vehicle ding damage localization method of claim 1, wherein, It includes:

5. A vehicle scratch warning system based on dynamic risk assessment, implementing any of the methods of claims 1-4, characterized in that, The perception layer is composed of a capacitive sensing car cover composed of a distributed electrode array arranged in a ring shape, a 4D millimeter wave radar array and an ultrasonic sensor, which is used to detect the contact distance between an object and the car body in real time, and establish a three-level risk identification model; The interaction layer is used to push a report containing a three-dimensional positioning map and suspicious object information, and has the functions of one-key generation of insurance report materials, intelligent estimation of repair cost, and directional activation of a camera to capture the image of a suspicious object. The edge computing layer adopts a dynamic risk assessment engine running on a R-Car V3H chip of Ressence to evaluate and calculate the scratch risk through the real-time risk scoring algorithm, wherein the risk coefficient of the object is dynamically configured, the risk coefficient of the metal object , and the risk coefficient of the living body . ​

Images