Vehicle collision event identification method, device, equipment and medium

Abstract

The present invention discloses a vehicle collision event identification method, device, equipment and medium. The method includes: when an unmanned vehicle has a collision event, a cloud control center obtains the collision basic data corresponding to the collision event according to the preset collision probability category and monitors the connection information uploaded by each device component in real time; each collision basic data is used to calculate the collision probability corresponding to the corresponding collision probability category; when any connection information is lost, the corresponding device component information is determined as the fault component information; corresponding maintenance tasks are created according to the collision degree intervals where each collision probability is located; and a maintenance priority sequence corresponding to the collision event is generated according to the matching results of each maintenance task and the fault component information. Thus, the respective collision probabilities are calculated through multi-dimensional collision basic data, and corresponding maintenance tasks are generated according to each collision probability, which effectively improves the timeliness of vehicle collision response and improves road safety.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving scene recognition technology, and in particular to a vehicle collision event recognition method, device, equipment and medium. Background Art

[0002] With the development of intelligent driving technology and intelligent transportation systems, technologies such as vehicle collision detection and corresponding alarm processing have also been developed. Vehicle collision detection and alarm systems are usually used to detect whether a collision event has occurred and for subsequent emergency alarms. The application of this system in autonomous driving technology can significantly reduce the waiting time for rescue personnel and provide data support for subsequent vehicle maintenance and preservation of accident evidence.

[0003] The vehicle collision detection scheme in the prior art usually includes a ground sensing coil or an ultrasonic sensor, etc., which determines whether a vehicle collision occurs by whether the traffic flow is affected.

[0004] However, since there may be no people in the driverless vehicle when it is in autonomous driving operation, when a collision occurs, it is difficult for the driverless vehicle to perceive the occurrence of the accident through the sensors used for driving. The long response time further reduces road safety. Summary of the invention

[0005] The present invention provides a vehicle collision event identification method, device, equipment and medium, which solves the technical problem that when an unmanned vehicle is in an automatic driving operation state, there may be no people in the vehicle. When a vehicle collision occurs, the unmanned vehicle is difficult to perceive the occurrence of an accident by relying on the sensors used for driving, and the response time is too long, which further reduces the safety of the road.

[0006] A vehicle collision event recognition method provided in a first aspect of the present invention is applied to a cloud control center, and the method comprises:

[0007] When a collision event occurs to the unmanned vehicle, basic collision data corresponding to the collision event is obtained according to a preset collision probability category and connection information uploaded by each of the device components is monitored in real time;

[0008] The collision basic data are respectively used to calculate the collision probability corresponding to the corresponding collision probability category;

[0009] When any of the connection information is lost, the corresponding device component information is determined as faulty component information;

[0010] Creating corresponding maintenance tasks respectively according to the collision degree intervals where the collision probabilities are located;

[0011] A maintenance priority sequence corresponding to the collision event is generated according to the matching results of each maintenance task and the faulty component information.

[0012] Optionally, the collision probability categories include emergency braking categories, obstacle deformation categories, collision sound categories, protection triggering categories, and disconnection collision categories; and the steps of respectively using the collision basic data to calculate the collision probability corresponding to the corresponding collision probability category include:

[0013] Extracting speed related information from the collision basic data to calculate the emergency braking probability corresponding to the emergency braking category;

[0014] Extracting an obstacle point cloud from the collision basic data, and calculating an obstacle deformation probability corresponding to the obstacle deformation category;

[0015] From the moment of occurrence of the collision event, extracting the rubber strip voltage value and the AEB trigger signal in a preset pre-event interval from the collision basic data, and determining the protection trigger probability corresponding to the protection trigger category;

[0016] Extracting a noise Hertz value from the collision basic data, calculating a ratio between the noise Hertz value and a preset noise threshold, and obtaining a collision sound probability corresponding to the collision sound category;

[0017] When a communication disconnection event occurs, a disconnection collision probability corresponding to the disconnection collision category is calculated according to the speed related information and the historical collision probability distribution information.

[0018] Optionally, the step of extracting speed-related information from the collision basic data to calculate the emergency braking probability corresponding to the emergency braking category includes:

[0019] Extracting speed related information from the collision basic data; wherein the speed related information includes vehicle deceleration and braking percentage;

[0020] Calculating a deceleration ratio between the vehicle deceleration and a preset deceleration constant value;

[0021] Calculating a percentage difference between a preset percentage extreme value and the braking percentage;

[0022] The multiplication value between the deceleration ratio, the percentage difference and a preset deceleration probability factor is calculated to obtain the emergency braking probability corresponding to the emergency braking category.

[0023] Optionally, the step of extracting an obstacle point cloud from the collision basic data and calculating an obstacle deformation probability corresponding to the obstacle deformation category includes:

[0024] Extracting an obstacle point cloud from the collision basic data; the obstacle point cloud includes first point cloud data corresponding to the occurrence time of the collision event and multiple frames of second point cloud data before the occurrence time;

[0025] Determine a distance change percentage based on the first point cloud data and multiple frames of the second point cloud data;

[0026] Counting the number of second sample points whose distance change percentage is greater than a preset change threshold;

[0027] If the number of the second sample points is greater than a preset deformation quantity threshold, determining the obstacle deformation probability corresponding to the obstacle deformation category as a first preset value;

[0028] If the number of the second sample points is less than or equal to a preset deformation quantity threshold, the obstacle deformation probability corresponding to the obstacle deformation category is determined as a second preset value.

[0029] Optionally, the step of determining the distance change percentage according to the first point cloud data and multiple frames of the second point cloud data includes:

[0030] Sampling a plurality of frames of the second point cloud data according to a preset size to obtain a plurality of first sample points;

[0031] respectively calculating a first distance between each of the first sample points and the center of gravity of the second point cloud data;

[0032] Selecting second sample points corresponding to each of the first sample points from the first point cloud data;

[0033] respectively calculating a second distance between each of the second sample points and the centroid of the point cloud data;

[0034] Calculate the distance difference between each of the second distances and the first distance;

[0035] The absolute value of the ratio between the distance difference and the first distance is calculated to obtain the distance change percentage.

[0036] Optionally, the protection triggering probability includes a rubber strip deformation probability and a device triggering probability; and the step of extracting a rubber strip voltage value and an AEB triggering signal in a preset pre-event interval from the collision basic data from the moment of occurrence of the collision event to determine the protection triggering probability corresponding to the protection triggering category includes:

[0037] From the moment of occurrence of the collision event, extracting the rubber strip voltage value and the AEB trigger signal in a preset pre-event interval from the collision basic data;

[0038] If it is monitored that the voltage value of the rubber strip changes from a positive value to a zero value within the pre-event interval, the deformation probability of the rubber strip is determined to be a third preset value;

[0039] If it is monitored that the AEB trigger signal occurs at the occurrence time, the device trigger probability is determined to be the third preset value.

[0040] Optionally, when a communication disconnection event occurs, the step of calculating the disconnection collision probability corresponding to the disconnection collision category according to the speed-related information and the historical collision probability distribution information includes:

[0041] When a communication disconnection event occurs, obtaining the last frame distance information between the unmanned vehicle and the obstacle;

[0042] According to the lane where the driverless vehicle is located, the speed-related information and the oncoming speed information of the oncoming vehicle within a preset time period are traced back;

[0043] Calculating a collision value corresponding to the unmanned vehicle according to the speed related information and the oncoming speed information;

[0044] The historical probability distribution information is retrieved according to the collision value, and the interval probability of the collision value is determined as the disconnection collision probability corresponding to the disconnection collision category.

[0045] Optionally, the speed-related information further includes a maximum deceleration of the vehicle and an instantaneous speed of the vehicle, and the oncoming speed information includes an oncoming maximum deceleration and an oncoming instantaneous speed; the step of calculating the collision value corresponding to the unmanned vehicle according to the speed-related information and the oncoming speed information includes:

[0046] Calculating the vehicle deceleration displacement according to the maximum deceleration of the vehicle and the instantaneous speed of the vehicle;

[0047] Calculating the opposite deceleration displacement according to the opposite maximum deceleration and the opposite instantaneous speed;

[0048] Calculating a displacement sum value of the vehicle deceleration displacement and the opposing deceleration displacement;

[0049] The difference between the displacement sum value and the distance information is calculated to obtain a collision value corresponding to the unmanned vehicle.

[0050] Optionally, the step of generating a maintenance priority sequence corresponding to the collision event according to the matching results of each maintenance task and the faulty component information includes:

[0051] Invoke the remote cockpit to connect to the unmanned vehicle according to each maintenance task, and obtain the real-time fault image corresponding to each maintenance task;

[0052] It is determined whether the real-time fault image matches the fault component information, and the maintenance tasks are sorted in descending order according to the number of matches, so as to generate a maintenance priority sequence corresponding to the collision event.

[0053] Optionally, the method further comprises:

[0054] When receiving a plurality of maintenance tasks respectively created by a plurality of the unmanned vehicles, selecting a plurality of intermediate maintenance tasks with the highest priority from the plurality of maintenance tasks;

[0055] Counting the number of intermediate maintenance tasks associated with each of the unmanned vehicles;

[0056] The intermediate maintenance tasks are inserted at the first position of the maintenance priority sequence according to the task quantity from high to low.

[0057] A second aspect of the present invention provides a vehicle collision event recognition device, which is applied to a cloud control center, and the device includes:

[0058] An information acquisition module, used for acquiring collision basic data corresponding to the collision event according to a preset collision probability category and monitoring the connection information uploaded by each of the device components in real time when a collision event occurs to the unmanned vehicle;

[0059] A collision probability calculation module, used to respectively calculate the collision probability corresponding to the corresponding collision probability category using each of the collision basic data;

[0060] A fault component information locating module, used for determining the corresponding device component information as the fault component information when any of the connection information is lost;

[0061] A maintenance task creation module, used to create corresponding maintenance tasks according to the collision degree intervals where the collision probabilities are located;

[0062] The maintenance priority sequence generating module is used to generate a maintenance priority sequence corresponding to the collision event according to the matching result of each maintenance task and the faulty component information.

[0063] A third aspect of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the vehicle collision scene recognition method as described in any one of the first aspect of the present invention.

[0064] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the vehicle collision scene recognition method as described in any one of the first aspects of the present invention.

[0065] It can be seen from the above technical solutions that the present invention has the following advantages:

[0066] When a collision occurs in an unmanned vehicle, the cloud control center obtains the collision basic data corresponding to the collision event according to the preset collision probability category and monitors the connection information uploaded by each device component in real time; each collision basic data is used to calculate the collision probability corresponding to the corresponding collision probability category; when any connection information is lost, the corresponding device component information is determined as the faulty component information; corresponding maintenance tasks are created according to the collision degree interval of each collision probability; and the maintenance priority sequence corresponding to the collision event is generated according to the matching results of each maintenance task and the faulty component information. Thus, the respective collision probabilities are calculated through multi-dimensional collision basic data, and the maintenance tasks are further sorted in combination with the faulty component information monitored in real time, and the maintenance priority sequence corresponding to the collision event is generated, which effectively improves the timeliness of vehicle collision response and improves road safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0068] Figure 1 A flowchart of a vehicle collision event identification method provided in Embodiment 1 of the present invention;

[0069] Figure 2 A flowchart of a vehicle collision event identification method provided in Embodiment 2 of the present invention;

[0070] Figure 3 This is a device block diagram of a vehicle collision event recognition device provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0071] The embodiments of the present invention provide a vehicle collision event identification method, device, equipment and medium, which are used to solve the technical problem that when an unmanned vehicle is in an automatic driving operation state, there may be no people in the vehicle. When a vehicle collision occurs, the unmanned vehicle is difficult to perceive the occurrence of an accident by relying on the sensors used for driving, and the response time is too long, resulting in further reduced road safety.

[0072] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] See also Figure 1 , Figure 1 A flowchart of the steps of a vehicle collision event identification method provided in Embodiment 1 of the present invention.

[0074] A vehicle collision event recognition method provided by the present invention is applied to a cloud control center, comprising:

[0075] Step 101, when a collision event occurs to the unmanned vehicle, basic collision data corresponding to the collision event is obtained according to a preset collision probability category and connection information uploaded by each device component is monitored in real time;

[0076] A collision event is an incident in which the autonomous vehicle detects that a collision has occurred but is unable to determine the severity of the collision.

[0077] The collision probability categories refer to multiple categories that measure the severity of the collision. Each category uses probability to express its severity. The categories include but are not limited to emergency braking category, obstacle deformation category, collision sound category, protection trigger category and disconnection collision category.

[0078] Basic collision data refers to specific types of data obtained through various sensors installed on unmanned vehicles, including but not limited to deceleration, braking percentage, rubber strip voltage value, collision sound, AEB trigger signal, airbag pop-up signal and disconnection signal, etc.

[0079] Connection information refers to the topics signals transmitted by various devices on the unmanned vehicle to the host of the autonomous driving system. The devices include but are not limited to cameras, lidars, chassis or doors, etc. The host periodically obtains component information and locates the corresponding components to accurately locate the fault location caused by the collision.

[0080] In an embodiment of the present invention, when a collision event occurs in an unmanned vehicle, basic collision data corresponding to the collision event may be acquired according to a preset collision probability category to obtain a data basis for subsequent collision probability calculation.

[0081] At the same time, the cloud control center monitors the connection information uploaded by each device component in real time through the host on the unmanned vehicle to determine in real time whether any device component has a disconnection failure.

[0082] Step 102, using each collision basic data to calculate the collision probability corresponding to the corresponding collision probability category;

[0083] After the collision basic data is obtained, corresponding data are extracted from the collision basic data according to each collision probability category, and the corresponding collision probabilities are calculated respectively.

[0084] Step 103, when any connection information is lost, the corresponding device component information is determined as faulty component information;

[0085] At the same time, when any connection information is lost, it indicates that the host has detected that a device component of the unmanned vehicle has a disconnection failure. At this time, the device component corresponding to the connection information can be determined as a faulty component, and the corresponding device component information can be determined as faulty component information.

[0086] It should be noted that device component information refers to the descriptive information corresponding to each device component on the unmanned vehicle, including but not limited to its properties and historical usage.

[0087] Step 104, creating corresponding maintenance tasks according to the collision degree intervals of the collision probabilities;

[0088] In the embodiment of the present invention, after each collision probability is obtained, each collision probability can be classified into a collision degree interval to which it belongs, and a corresponding maintenance task can be created according to each collision degree interval.

[0089] It should be noted that the collision degree intervals can be divided into high (greater than 80%), medium (greater than 50%) and low (greater than 30%) according to the processing priority. The collision probability in each collision degree interval generates corresponding maintenance tasks, waiting for subsequent priority sorting maintenance.

[0090] Step 105: Generate a maintenance priority sequence corresponding to the collision event according to the matching results of each maintenance task and the faulty component information.

[0091] After each maintenance task is created, in order to verify its authenticity and maintenance importance, the maintenance task and faulty component information can be matched separately to determine the authenticity of the maintenance task and at the same time, each maintenance task can be statistically sorted according to the collision degree range in which it is located to generate a maintenance priority sequence corresponding to the collision event.

[0092] It should be noted that the maintenance priority sequence refers to sequential information sorted from high to low according to the collision severity range. If multiple unmanned vehicles collide at the same time, they can be further prioritized according to the number of maintenance tasks in the high collision severity range to ensure that the maintenance tasks handled by the maintenance personnel are the highest priority tasks.

[0093] In an embodiment of the present invention, when a collision occurs in an unmanned vehicle, the cloud control center obtains the collision basic data corresponding to the collision event according to the preset collision probability category and monitors the connection information uploaded by each device component in real time; each collision basic data is used to calculate the collision probability corresponding to the corresponding collision probability category; when any connection information is lost, the corresponding device component information is determined as the faulty component information; corresponding maintenance tasks are created according to the collision degree intervals of each collision probability; and the maintenance priority sequence corresponding to the collision event is generated according to the matching results of each maintenance task and the faulty component information. Thus, the respective collision probabilities are calculated through multi-dimensional collision basic data, and the maintenance tasks are further sorted in combination with the faulty component information monitored in real time, and the maintenance priority sequence corresponding to the collision event is generated, which effectively improves the timeliness of vehicle collision response and improves road safety.

[0094] See also Figure 2 , Figure 2 A flowchart of the steps of a vehicle collision event identification method provided in Embodiment 2 of the present invention.

[0095] A vehicle collision event recognition method provided by the present invention is applied to a cloud control center, comprising:

[0096] Step 201, when a collision event occurs to the unmanned vehicle, basic collision data corresponding to the collision event is obtained according to a preset collision probability category and connection information uploaded by each device component is monitored in real time;

[0097] In the embodiment of the present invention, the specific implementation process of step 201 is similar to step 101 and will not be repeated here.

[0098] Optionally, the collision probability categories include an emergency braking category, an obstacle deformation category, a collision sound category, a protection triggering category, and a disconnection collision category.

[0099] Step 202, extracting speed-related information from the collision basic data to calculate the emergency braking probability corresponding to the emergency braking category;

[0100] Optionally, step 202 may include the following sub-steps:

[0101] Extracting speed-related information from the collision basic data; wherein the speed-related information includes vehicle deceleration and braking percentage;

[0102] Calculating a deceleration ratio between a vehicle deceleration and a preset deceleration constant value;

[0103] Calculate the percentage difference between the preset percentage extreme value and the braking percentage;

[0104] The multiplication value between the deceleration ratio, the percentage difference and the preset deceleration probability factor is calculated to obtain the emergency braking probability corresponding to the emergency braking category.

[0105] In an embodiment of the present invention, speed-related information may be first extracted from the collision basic data, wherein the speed-related information includes the vehicle deceleration and braking percentage of the unmanned vehicle when braking. By calculating the deceleration ratio between the vehicle deceleration and the preset deceleration constant value, further calculating the percentage difference between the preset percentage mechanism and the braking percentage, and calculating the multiplication value between the deceleration ratio, the percentage difference and the preset deceleration probability factor, the emergency braking probability corresponding to the emergency braking category is obtained.

[0106] It should be noted that, for the same braking ratio, when the deceleration is greater, the probability of a collision is greater and the braking is more severe, which means that the vehicle is more likely to have an emergency response. For the same deceleration, the lower the braking percentage, the greater the probability of braking, because the deceleration caused by not braking can prove that the vehicle is not decelerating due to braking, and is more likely to be caused by hitting an obstacle.

[0107] The deceleration probability factor can be obtained by statistical calculation based on historical collision events. For example, in 100 driving processes, 70 collisions occurred when (x / -15)*(1-y%)>90, then the deceleration probability factor is 0.9*(70 / 100), where x is the vehicle deceleration, -15 is the deceleration constant, 1 is the percentage extreme value, and y% is the braking percentage.

[0108] Step 203, extracting the obstacle point cloud from the collision basic data, and calculating the obstacle deformation probability corresponding to the obstacle deformation category;

[0109] Optionally, step 203 may include the following sub-steps S11-S15:

[0110] S11, extracting an obstacle point cloud from the collision basic data; the obstacle point cloud includes first point cloud data corresponding to the time when the collision event occurs and multiple frames of second point cloud data before the time when the collision event occurs;

[0111] The obstacle point cloud refers to a set of point data of the surface of the obstacle in front of the vehicle detected by the laser radar on the unmanned vehicle. It includes the first point cloud data corresponding to the moment of the collision event, and multiple frames of second point cloud data within a preset number of frames before the moment of the occurrence.

[0112] In an embodiment of the present invention, when an unmanned vehicle collides, it is usually accompanied by obstacle deformation. At this time, in order to evaluate the obstacle deformation, an obstacle point cloud can be extracted from the collision basic data to obtain the data basis of the obstacle before the collision event and at the time of the collision event.

[0113] S12, determining a distance change percentage according to the first point cloud data and multiple frames of second point cloud data;

[0114] Further, S12 may include the following sub-steps:

[0115] Sampling the second point cloud data of multiple frames according to a preset size to obtain multiple first sample points;

[0116] Calculate the first distance between each first sample point and the center of gravity of the second point cloud data respectively;

[0117] Selecting second sample points corresponding to each first sample point from the first point cloud data;

[0118] Calculate the second distance between each second sample point and the centroid of the point cloud data respectively;

[0119] Calculate the distance difference between each second distance and the first distance;

[0120] The absolute value of the ratio between the distance difference and the first distance is calculated to obtain the distance change percentage.

[0121] In the embodiment of the present invention, according to the preset size, for example, 10 cm 2 Sample points are extracted from each frame of the second point cloud data to obtain a plurality of first sample points corresponding to each frame of the second point cloud data. The first distances between each first sample point and the center of gravity of the second point cloud data are calculated to obtain the first distances corresponding to each first sample point to represent the approximate shape of the obstacle before the vehicle collision event occurs.

[0122] At the same time, after obtaining the first point cloud data, select the second sample point after the first sample point experiences a collision event, calculate the second distance between the second sample point and the center of the point cloud data, calculate the distance difference between each second distance and the first distance, and determine whether a point cloud offset occurs.

[0123] The ratio between the distance difference and the first distance is further calculated and the absolute value is taken to obtain the distance change percentage to measure the degree of deformation of the point cloud shape of the obstacle.

[0124] S13, counting the number of second sample points whose distance change percentage is greater than a preset change threshold;

[0125] S14. If the number of the second sample points is greater than a preset deformation quantity threshold, the obstacle deformation probability corresponding to the obstacle deformation category is determined as a first preset value;

[0126] S15: If the second number of sample points is less than or equal to a preset deformation quantity threshold, the obstacle deformation probability corresponding to the obstacle deformation category is determined as a second preset value.

[0127] In the embodiment of the present invention, the number of second sample points whose distance change percentage is greater than the preset change threshold is counted as the second sample point number. If the second sample point number is greater than the preset deformation amount threshold, it indicates that the shape of the obstacle has undergone a significant change. At this time, the obstacle deformation probability corresponding to the obstacle deformation category can be determined as the first preset value. If the second sample point number is less than or equal to the deformation amount threshold, it indicates that the obstacle has not undergone a significant change. At this time, the obstacle deformation probability corresponding to the obstacle deformation category can be determined as the second preset value.

[0128] It should be noted that the first preset value can be 1, and the second preset value can be set to 0. The point cloud data can be acquired when the camera radar recognizes that the distance between the vehicle and the obstacle is 0. At the same time, the point cloud can effectively distinguish between people and obstacles by recognizing the shape, reducing the false alarm caused by people getting too close to the car. The preset change threshold can be set to 30%, and the deformation quantity threshold can be set to 30% of the number of distance change percentages. The embodiment of the present invention does not limit the specific numerical setting.

[0129] Step 204, from the moment of occurrence of the collision event, extract the rubber strip voltage value and the AEB trigger signal in the preset pre-event interval from the collision basic data, and determine the protection trigger probability corresponding to the protection trigger category;

[0130] In an example of the present invention, the protection triggering probability includes the strip deformation probability and the device triggering probability, and step 204 may include the following sub-steps:

[0131] From the moment of the collision event, the rubber strip voltage value and AEB trigger signal in the preset pre-event interval are extracted from the collision basic data;

[0132] If it is monitored that the voltage value of the rubber strip changes from a positive value to a zero value in the pre-event interval, the probability of deformation of the rubber strip is determined to be a third preset value;

[0133] If it is monitored that the AEB trigger signal appears at the occurrence time, the device trigger probability is determined to be a third preset value.

[0134] The preset pre-event interval refers to a period of time before the occurrence time, which can be set to 3s, etc. The specific setting value is not limited here. The Autonomous Emergency Braking system (AEB) is an active automotive safety technology, mainly composed of three modules, including the control module (ECU), the ranging module, and the braking module. The core of the ranging module includes microwave radar, face recognition technology and video system, etc. It can provide safe, accurate and real-time images and road condition information of the road ahead.

[0135] In this embodiment, when a collision occurs, the unmanned vehicle itself will also deform due to the mutual force, and AEB will be triggered. Generally, unmanned vehicles have collision rubber strips on the outside, and the voltage value of the rubber strips will change when they are deformed.

[0136] Therefore, from the moment the collision event occurs, the rubber strip voltage value and the AEB trigger signal in the preset pre-event interval are extracted from the collision basic data to determine whether the AEB is triggered. If the rubber strip voltage value is monitored to change from a positive value to a zero value in the pre-event interval, it indicates that the collision rubber strip of the vehicle has deformed, resulting in a voltage change. At this time, the rubber strip deformation probability can be determined as the third preset value, which can be 1.

[0137] At the same time, if the AEB trigger signal is detected to appear at the time of occurrence, it indicates that the unmanned vehicle has triggered the automatic emergency braking operation. At this time, the device trigger probability can be determined as the third preset value.

[0138] In addition, in addition to the AEB trigger signal, it is also possible to determine whether the airbag of the driverless vehicle is triggered to pop out at the time of occurrence. If so, the device trigger probability can also be determined as a third preset value.

[0139] Step 205, extracting the noise Hertz value from the collision basic data, calculating the ratio between the noise Hertz value and the preset noise threshold, and obtaining the collision sound probability corresponding to the collision sound category;

[0140] In a specific implementation, a collision event is often accompanied by a loud sound. Therefore, the noise Hertz value can be extracted from the collision basic data, and its ratio to the preset noise threshold can be calculated to obtain the collision sound probability corresponding to the collision sound category.

[0141] Among them, the noise threshold can be set to 110hz.

[0142] Step 206, when a communication disconnection event occurs, the disconnection collision probability corresponding to the disconnection collision category is calculated according to the speed related information and the historical collision probability distribution information;

[0143] Optionally, step 206 may include the following sub-steps S21-S24:

[0144] S21. When a communication disconnection event occurs, obtain the last frame distance information between the unmanned vehicle and the obstacle;

[0145] S22, retracing the speed-related information within a preset time period and the oncoming speed information of the oncoming vehicle according to the lane where the driverless vehicle is located;

[0146] In an embodiment of the present invention, when a collision event is more serious, it is easy to cause the vehicle to stall instantly and be unable to make its own decisions. The stable long connection between the vehicle and the cloud control center is disconnected. At this time, a collision event is also likely to occur. Therefore, when a communication disconnection event occurs, the last frame distance information between the unmanned vehicle and the obstacle can be obtained, and according to the lane where the unmanned vehicle is located, the speed-related information within a preset time period and the oncoming speed information of the oncoming vehicles in the lane can be traced back.

[0147] It should be noted that the type of the specific content of the opposite speed information is consistent with the speed-related information.

[0148] S23, calculating a collision value corresponding to the unmanned vehicle according to speed related information and oncoming speed information;

[0149] Furthermore, the speed-related information also includes the maximum deceleration of the vehicle and the instantaneous speed of the vehicle, and the opposite speed information includes the opposite maximum deceleration and the opposite instantaneous speed. S23 may include the following sub-steps:

[0150] Calculate the vehicle deceleration displacement according to the vehicle's maximum deceleration and the vehicle's instantaneous speed;

[0151] According to the opposite maximum deceleration and the opposite instantaneous speed, the opposite deceleration displacement is calculated;

[0152] Calculate the displacement and value of the vehicle deceleration displacement and the opposite deceleration displacement;

[0153] The difference between the displacement sum value and the distance information is calculated to obtain the collision value corresponding to the unmanned vehicle.

[0154] In the embodiment of the present invention, the vehicle deceleration displacement s may be calculated according to the maximum deceleration of the vehicle and the instantaneous speed of the vehicle. The specific calculation formula may be as follows:

[0155]

[0156] a is the maximum deceleration of the vehicle -10m / s2, and v is the instantaneous speed of the vehicle.

[0157] For the opposite deceleration displacement, it can be obtained by substituting the opposite maximum deceleration and the opposite instantaneous speed into the corresponding positions of the above formula.

[0158] After calculating the vehicle deceleration displacement and the opposite deceleration displacement, the displacement sum value between the two can be further calculated, and then the difference between the displacement sum value and the distance information can be calculated to obtain the collision value corresponding to the unmanned vehicle.

[0159] S24. Retrieve historical probability distribution information according to the collision value, and determine the interval probability of the collision value as the disconnection collision probability corresponding to the disconnection collision category.

[0160] The historical probability distribution information refers to the probability distribution of collision events of unmanned vehicles that have been counted in historical road tests.

[0161] In the specific implementation, since the unmanned vehicle must have undergone multiple simulation runs and road tests before operation, the historical probability distribution information can be obtained, and the historical probability distribution information is retrieved according to the collision value to determine the interval probability of the collision value in the historical probability distribution information, and determine it as the disconnected collision probability corresponding to the disconnected collision category.

[0162] Step 207, when any connection information is lost, the corresponding device component information is determined as faulty component information;

[0163] Step 208, creating corresponding maintenance tasks according to the collision degree intervals of the collision probabilities;

[0164] In the embodiment of the present invention, the specific implementation process of steps 207-208 is similar to that of steps 103-104, and will not be repeated here.

[0165] Step 209: Generate a maintenance priority sequence corresponding to the collision event according to the matching results of each maintenance task and the faulty component information.

[0166] In another example of the present invention, step 209 may include the following sub-steps:

[0167] According to each maintenance task, the remote cockpit is called to connect to the unmanned vehicle to obtain the real-time fault image corresponding to each maintenance task;

[0168] Determine whether the real-time fault image matches the fault component information, sort the maintenance tasks from high to low according to the number of matches, and generate a maintenance priority sequence corresponding to the collision event.

[0169] In this embodiment, after multiple maintenance tasks are established, the remote cockpit can be called to connect to the unmanned vehicle according to the requirements of each maintenance task to obtain the real-time fault image corresponding to each maintenance task, and further compare the real-time fault image and the fault component information. Specifically, the image comparison can be performed between the fault component image in the fault component information and the real-time fault image. If the comparison similarity is higher and the similar positions are more, it means that the two are more matched. The number of matches is counted, and the maintenance tasks are sorted from high to low to generate a maintenance priority sequence corresponding to the collision event.

[0170] In another embodiment of the present invention, the method further comprises the following steps:

[0171] The method also includes:

[0172] When receiving multiple maintenance tasks respectively created by multiple unmanned vehicles, selecting multiple intermediate maintenance tasks with the highest priority from the multiple maintenance tasks;

[0173] Count the number of tasks associated with each unmanned vehicle intermediate maintenance task;

[0174] Insert the intermediate maintenance tasks at the top of the maintenance priority sequence according to the task quantity from high to low.

[0175] It is worth mentioning that when facing a collision incident involving multiple unmanned vehicles, they can be matched according to the priority of their associated maintenance tasks, that is, the priority and quantity of each maintenance task are counted, and the unmanned vehicles with the highest priority and the largest number are placed at the top of the maintenance priority sequence and the sequence is updated according to the preset time interval.

[0176] In an embodiment of the present invention, when a collision occurs in an unmanned vehicle, the cloud control center obtains the collision basic data corresponding to the collision event according to the preset collision probability category and monitors the connection information uploaded by each device component in real time; each collision basic data is used to calculate the collision probability corresponding to the corresponding collision probability category; when any connection information is lost, the corresponding device component information is determined as the faulty component information; corresponding maintenance tasks are created according to the collision degree intervals of each collision probability; and the maintenance priority sequence corresponding to the collision event is generated according to the matching results of each maintenance task and the faulty component information. Thus, the respective collision probabilities are calculated through multi-dimensional collision basic data, and the maintenance tasks are further sorted in combination with the faulty component information monitored in real time, and the maintenance priority sequence corresponding to the collision event is generated, which effectively improves the timeliness of vehicle collision response and improves road safety.

[0177] See also Figure 3 , Figure 3 A structural block diagram of a vehicle collision event recognition device according to a third embodiment of the present invention is shown.

[0178] An embodiment of the present invention provides a vehicle collision event recognition device, which is applied to a cloud control center. The device includes:

[0179] The information acquisition module 301 is used to acquire the collision basic data corresponding to the collision event according to the preset collision probability category and monitor the connection information uploaded by each device component in real time when a collision event occurs to the unmanned vehicle;

[0180] A collision probability calculation module 302, used to calculate the collision probability corresponding to the corresponding collision probability category using each collision basic data;

[0181] The fault component information locating module 303 is used to determine the corresponding device component information as the fault component information when any connection information is lost;

[0182] A maintenance task creation module 304 is used to create corresponding maintenance tasks according to the collision degree intervals of each collision probability;

[0183] The maintenance priority sequence generating module 305 is used to generate a maintenance priority sequence corresponding to the collision event according to the matching results of each maintenance task and the faulty component information.

[0184] Optionally, the collision probability category includes an emergency braking category, an obstacle deformation category, a collision sound category, a protection trigger category, and a disconnection collision category; the collision probability calculation module 302 includes:

[0185] The emergency braking probability calculation submodule is used to extract speed-related information from the collision basic data and calculate the emergency braking probability corresponding to the emergency braking category;

[0186] The obstacle deformation probability calculation submodule is used to extract the obstacle point cloud from the collision basic data and calculate the obstacle deformation probability corresponding to the obstacle deformation category;

[0187] The protection trigger probability calculation submodule is used to extract the rubber strip voltage value and AEB trigger signal in the preset pre-event interval from the collision basic data from the moment of occurrence of the collision event, and determine the protection trigger probability corresponding to the protection trigger category;

[0188] A collision sound calculation submodule is used to extract the noise Hertz value from the collision basic data, calculate the ratio between the noise Hertz value and the preset noise threshold, and obtain the collision sound probability corresponding to the collision sound category;

[0189] The disconnection collision probability calculation submodule is used to calculate the disconnection collision probability corresponding to the disconnection collision category according to speed-related information and historical collision probability distribution information when a communication disconnection event occurs.

[0190] Optionally, the emergency braking probability calculation submodule is specifically used for:

[0191] Extracting speed-related information from the collision basic data; wherein the speed-related information includes vehicle deceleration and braking percentage;

[0192] Calculating a deceleration ratio between a vehicle deceleration and a preset deceleration constant value;

[0193] Calculate the percentage difference between the preset percentage extreme value and the braking percentage;

[0194] The multiplication value between the deceleration ratio, the percentage difference and the preset deceleration probability factor is calculated to obtain the emergency braking probability corresponding to the emergency braking category.

[0195] Optionally, the obstacle deformation probability calculation submodule includes:

[0196] A point cloud extraction unit is used to extract an obstacle point cloud from the collision basic data; the obstacle point cloud includes first point cloud data corresponding to the time when the collision event occurs and multiple frames of second point cloud data before the time when the collision event occurs;

[0197] a distance change percentage determination unit, configured to determine the distance change percentage based on the first point cloud data and the plurality of frames of second point cloud data;

[0198] A sample point number counting unit, used to count the number of second sample points whose distance change percentage is greater than a preset change threshold;

[0199] a first preset value assigning unit, configured to determine the obstacle deformation probability corresponding to the obstacle deformation category as a first preset value if the number of the second sample points is greater than a preset deformation quantity threshold;

[0200] The second preset value assigning unit is used to determine the obstacle deformation probability corresponding to the obstacle deformation category as a second preset value if the second sample point number is less than or equal to a preset deformation quantity threshold.

[0201] Optionally, the distance change percentage determination unit is specifically used for:

[0202] Sampling the second point cloud data of multiple frames according to a preset size to obtain multiple first sample points;

[0203] Calculate the first distance between each first sample point and the center of gravity of the second point cloud data respectively;

[0204] Selecting second sample points corresponding to each first sample point from the first point cloud data;

[0205] Calculate the second distance between each second sample point and the centroid of the point cloud data respectively;

[0206] Calculate the distance difference between each second distance and the first distance;

[0207] The absolute value of the ratio between the distance difference and the first distance is calculated to obtain the distance change percentage.

[0208] Optionally, the protection triggering probability includes the strip deformation probability and the device triggering probability; the protection triggering probability calculation submodule is specifically used for:

[0209] From the moment of the collision event, the rubber strip voltage value and AEB trigger signal in the preset pre-event interval are extracted from the collision basic data;

[0210] If it is monitored that the voltage value of the rubber strip changes from a positive value to a zero value in the pre-event interval, the probability of deformation of the rubber strip is determined to be a third preset value;

[0211] If it is monitored that the AEB trigger signal appears at the occurrence time, the device trigger probability is determined to be a third preset value.

[0212] Optionally, the disconnection collision probability calculation submodule includes:

[0213] A distance information acquisition unit, used to acquire the last frame of distance information between the unmanned vehicle and the obstacle when a communication disconnection event occurs;

[0214] A speed-related information backtracking unit, used to backtrack speed-related information within a preset time period and oncoming speed information of oncoming vehicles according to the lane where the driverless vehicle is located;

[0215] A collision value calculation unit, used to calculate a collision value corresponding to the unmanned vehicle according to speed-related information and oncoming speed information;

[0216] The disconnection collision probability determination unit is used to retrieve the historical probability distribution information according to the collision value, and determine the interval probability of the collision value as the disconnection collision probability corresponding to the disconnection collision category.

[0217] Optionally, the speed-related information further includes the maximum deceleration of the vehicle and the instantaneous speed of the vehicle, and the oncoming speed information includes the oncoming maximum deceleration and the oncoming instantaneous speed; the collision value calculation unit is specifically used for:

[0218] Calculate the vehicle deceleration displacement according to the vehicle's maximum deceleration and the vehicle's instantaneous speed;

[0219] According to the opposite maximum deceleration and the opposite instantaneous speed, the opposite deceleration displacement is calculated;

[0220] Calculate the displacement and value of the vehicle deceleration displacement and the opposite deceleration displacement;

[0221] The difference between the displacement sum value and the distance information is calculated to obtain the collision value corresponding to the unmanned vehicle.

[0222] Optionally, the maintenance priority sequence generating module 305 is specifically used for:

[0223] According to each maintenance task, the remote cockpit is called to connect to the unmanned vehicle to obtain the real-time fault image corresponding to each maintenance task;

[0224] Determine whether the real-time fault image matches the fault component information, sort the maintenance tasks from high to low according to the number of matches, and generate a maintenance priority sequence corresponding to the collision event.

[0225] Optionally, the module also includes:

[0226] A task screening module, for selecting a plurality of intermediate maintenance tasks with the highest priority from the plurality of maintenance tasks when receiving a plurality of maintenance tasks respectively created by a plurality of unmanned vehicles;

[0227] The task quantity statistics module is used to count the task quantity of each unmanned vehicle associated intermediate maintenance task;

[0228] The sequence insertion module is used to insert the intermediate maintenance tasks into the first position of the maintenance priority sequence according to the task quantity from high to low.

[0229] An embodiment of the present invention provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the vehicle collision scene recognition method as described in any embodiment of the present invention.

[0230] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed, the vehicle collision scene recognition method as described in any embodiment of the present invention is implemented.

[0231] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0232] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0233] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0234] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0235] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0236] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A vehicle collision event recognition method, characterized in that: Applied to a cloud control center, the method includes: When a collision occurs in an unmanned vehicle, the collision basic data corresponding to the collision event is obtained according to the preset collision probability categories and the connection information uploaded by each device component is monitored in real time; the collision probability categories include emergency braking category, obstacle deformation category, collision sound category, protection trigger category and disconnection collision category; The collision basic data are respectively used to calculate the collision probability corresponding to the corresponding collision probability category; When any of the connection information is lost, the corresponding device component information is determined as faulty component information; Creating corresponding maintenance tasks respectively according to the collision degree intervals where the collision probabilities are located; Generating a maintenance priority sequence corresponding to the collision event according to a matching result between each maintenance task and the faulty component information; The step of generating a maintenance priority sequence corresponding to the collision event according to the matching results of each maintenance task and the faulty component information comprises: Invoke the remote cockpit to connect to the unmanned vehicle according to each maintenance task, and obtain the real-time fault image corresponding to each maintenance task; It is determined whether the real-time fault image matches the fault component information, and the maintenance tasks are sorted in descending order according to the number of matches, so as to generate a maintenance priority sequence corresponding to the collision event.

2. The method according to claim 1, characterized in that The step of respectively using each of the collision basic data to calculate the collision probability corresponding to the corresponding collision probability category comprises: Extracting speed related information from the collision basic data to calculate the emergency braking probability corresponding to the emergency braking category; Extracting an obstacle point cloud from the collision basic data, and calculating an obstacle deformation probability corresponding to the obstacle deformation category; From the moment of occurrence of the collision event, extracting the rubber strip voltage value and the AEB trigger signal in a preset pre-event interval from the collision basic data, and determining the protection trigger probability corresponding to the protection trigger category; Extracting a noise Hertz value from the collision basic data, calculating a ratio between the noise Hertz value and a preset noise threshold, and obtaining a collision sound probability corresponding to the collision sound category; When a communication disconnection event occurs, a disconnection collision probability corresponding to the disconnection collision category is calculated according to the speed related information and the historical collision probability distribution information.

3. The method according to claim 2, characterized in that The step of extracting speed-related information from the collision basic data and calculating the emergency braking probability corresponding to the emergency braking category comprises: Extracting speed related information from the collision basic data; wherein the speed related information includes vehicle deceleration and braking percentage; Calculating a deceleration ratio between the vehicle deceleration and a preset deceleration constant value; Calculating a percentage difference between a preset percentage extreme value and the braking percentage; The multiplication value between the deceleration ratio, the percentage difference and a preset deceleration probability factor is calculated to obtain the emergency braking probability corresponding to the emergency braking category.

4. The method according to claim 2, characterized in that: The step of extracting the obstacle point cloud from the collision basic data and calculating the obstacle deformation probability corresponding to the obstacle deformation category comprises: Extracting an obstacle point cloud from the collision basic data; the obstacle point cloud includes first point cloud data corresponding to the occurrence time of the collision event and multiple frames of second point cloud data before the occurrence time; Determine a distance change percentage based on the first point cloud data and multiple frames of the second point cloud data; Counting the number of second sample points whose distance change percentage is greater than a preset change threshold; If the number of the second sample points is greater than a preset deformation quantity threshold, determining the obstacle deformation probability corresponding to the obstacle deformation category as a first preset value; If the number of the second sample points is less than or equal to a preset deformation quantity threshold, the obstacle deformation probability corresponding to the obstacle deformation category is determined as a second preset value.

5. The method according to claim 4, characterized in that The step of determining the distance change percentage according to the first point cloud data and multiple frames of the second point cloud data comprises: Sampling a plurality of frames of the second point cloud data according to a preset size to obtain a plurality of first sample points; respectively calculating a first distance between each of the first sample points and the center of gravity of the second point cloud data; Selecting second sample points corresponding to each of the first sample points from the first point cloud data; respectively calculating a second distance between each of the second sample points and the centroid of the point cloud data; Calculate the distance difference between each of the second distances and the first distance; The absolute value of the ratio between the distance difference and the first distance is calculated to obtain the distance change percentage.

6. The method according to claim 2, characterized in that The protection triggering probability includes the rubber strip deformation probability and the device triggering probability; the step of extracting the rubber strip voltage value and the AEB triggering signal in the preset pre-event interval from the collision basic data from the moment of occurrence of the collision event to determine the protection triggering probability corresponding to the protection triggering category includes: From the moment of occurrence of the collision event, extracting the rubber strip voltage value and the AEB trigger signal in a preset pre-event interval from the collision basic data; If it is monitored that the voltage value of the rubber strip changes from a positive value to a zero value within the pre-event interval, the deformation probability of the rubber strip is determined to be a third preset value; If it is monitored that the AEB trigger signal occurs at the occurrence time, the device trigger probability is determined to be the third preset value.

7. The method according to claim 3, characterized in that The step of calculating the disconnection collision probability corresponding to the disconnection collision category according to the speed related information and the historical collision probability distribution information when a communication disconnection event occurs comprises: When a communication disconnection event occurs, obtaining the last frame distance information between the unmanned vehicle and the obstacle; According to the lane where the driverless vehicle is located, the speed-related information and the oncoming speed information of the oncoming vehicle within a preset time period are traced back; Calculating a collision value corresponding to the unmanned vehicle according to the speed related information and the oncoming speed information; The historical collision probability distribution information is retrieved according to the collision value, and the interval probability of the collision value is determined as the disconnection collision probability corresponding to the disconnection collision category.

8. The method according to claim 7, characterized in that The speed related information also includes the maximum deceleration of the vehicle and the instantaneous speed of the vehicle, and the opposite speed information includes the maximum deceleration of the opposite direction and the instantaneous speed of the opposite direction; The step of calculating the collision value corresponding to the unmanned vehicle according to the speed-related information and the oncoming speed information comprises: Calculating the vehicle deceleration displacement according to the maximum deceleration of the vehicle and the instantaneous speed of the vehicle; Calculating the opposite deceleration displacement according to the opposite maximum deceleration and the opposite instantaneous speed; Calculating a displacement sum value of the vehicle deceleration displacement and the opposing deceleration displacement; The difference between the displacement sum value and the distance information is calculated to obtain a collision value corresponding to the unmanned vehicle.

9. The method according to claim 1, characterized in that: The method further comprises: When receiving a plurality of maintenance tasks respectively created by a plurality of the unmanned vehicles, selecting a plurality of intermediate maintenance tasks with the highest priority from the plurality of maintenance tasks; Counting the number of intermediate maintenance tasks associated with each of the unmanned vehicles; The intermediate maintenance tasks are inserted at the first position of the maintenance priority sequence according to the task quantity from high to low.

10. A vehicle collision event recognition device, characterized in that: Applied to a cloud control center, the device comprises: An information acquisition module is used to acquire the collision basic data corresponding to the collision event according to the preset collision probability categories and monitor the connection information uploaded by each device component in real time when a collision event occurs to the unmanned vehicle; the collision probability categories include emergency braking category, obstacle deformation category, collision sound category, protection trigger category and disconnection collision category; A collision probability calculation module, used to respectively calculate the collision probability corresponding to the corresponding collision probability category using each of the collision basic data; A fault component information locating module, used for determining the corresponding device component information as the fault component information when any of the connection information is lost; A maintenance task creation module, used to create corresponding maintenance tasks according to the collision degree intervals where the collision probabilities are located; A maintenance priority sequence generating module, used for generating a maintenance priority sequence corresponding to the collision event according to the matching result between each maintenance task and the fault component information; The maintenance priority sequence generation module is specifically used for: Invoke the remote cockpit to connect to the unmanned vehicle according to each maintenance task, and obtain the real-time fault image corresponding to each maintenance task; It is determined whether the real-time fault image matches the fault component information, and the maintenance tasks are sorted in descending order according to the number of matches, so as to generate a maintenance priority sequence corresponding to the collision event.

11. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the vehicle collision event recognition method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the vehicle collision event recognition method according to any one of claims 1 to 9 is implemented.

Images