Automatic driving safety detection method and device and computer program product

Abstract

The invention relates to an automatic driving safety detection method and device and a computer program product, and the method comprises the steps: obtaining the speed vector of a vehicle at the current moment and the speed vector of each target when a static or dynamic target exists in a preset region around the vehicle; acquiring a virtual collision speed of the vehicle relative to each target according to the speed vector of the vehicle and the speed vector of each target; acquiring virtual collision kinetic energy of the vehicle relative to each target according to the mass of the vehicle and the virtual collision speed; acquiring a virtual collision risk of the vehicle relative to each target according to the virtual collision kinetic energy; and comparing the at least one virtual collision risk corresponding to the at least one target, and obtaining a detection result according to the virtual collision risk with the maximum risk. According to the invention, collision risk assessment of dynamic and static targets is effectively integrated, and comprehensive and quantitative safety performance assessment is provided for an automatic driving system.

Description

Technical Field

[0001] This application relates to the field of vehicle emergency rescue technology, specifically to an autonomous driving safety detection method, device, and computer program product. Background Technology

[0002] As the automotive industry transforms towards electrification and intelligentization, autonomous driving has made rapid progress in the past few years. Autonomous driving uses sensors installed on vehicles to perceive their surroundings and plan reasonable driving routes based on this information. It can, to a certain extent, or even completely, replace the driver in performing driving operations. Ensuring road safety is the biggest challenge for autonomous driving. Quantitatively assessing driving safety is an important means of testing and verifying autonomous driving, helping to continuously improve its safety performance and enhance the user's driving experience.

[0003] In actual driving, vehicles face both dynamic target collision risks and static target collision risks such as roadside curbs and median strips. However, most current autonomous driving safety detection methods are based on single dynamic target calculations, ignoring static target calculations and lacking a unified approach to calculating both dynamic and static targets. Summary of the Invention

[0004] The purpose of this application is to propose an autonomous driving safety detection method and device, as well as a computer program product, to achieve a unified safety assessment of the vehicle relative to dynamic and static targets.

[0005] To achieve the above objectives, according to the first aspect of this application, an autonomous driving safety detection method is provided, comprising:

[0006] When there is at least one target in a preset area around the vehicle, the vehicle's current velocity vector and the velocity vector of each target are obtained; the target can be a static target or a dynamic target.

[0007] Based on the vehicle's velocity vector and the velocity vector of each target, obtain the virtual collision velocity of the vehicle relative to each target;

[0008] Obtain the vehicle's mass, and based on the vehicle's mass and the vehicle's virtual collision velocity relative to each target, obtain the vehicle's virtual collision kinetic energy relative to each target.

[0009] The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision kinetic energy of the vehicle relative to each target.

[0010] The detection result is obtained by comparing at least one virtual collision risk corresponding to the at least one target and the at least one virtual collision risk with the highest risk.

[0011] According to a second aspect of this application, an autonomous driving safety detection device is provided, the device including a module for performing the method described in the first aspect of this application.

[0012] According to a third aspect of this application, an autonomous driving safety detection device is provided, comprising:

[0013] A communication interface used for communicating with other electronic devices;

[0014] Memory is used to store computer program instructions;

[0015] A processor for executing the computer program instructions to support the implementation of the method according to the first aspect of this application based on the autonomous driving safety detection device.

[0016] According to a fourth aspect of this application, a computer program product is provided, including computer program instructions that instruct a computer device to perform an operation corresponding to the method described in the first aspect of this application.

[0017] The autonomous driving safety detection method, apparatus, and computer program product proposed in this application have the following beneficial effects:

[0018] By comprehensively acquiring and calculating the target trajectories of the vehicle and all targets (including dynamic and static targets) in a preset area around the vehicle, a unified method is used to obtain the virtual collision speed of the vehicle relative to each target, and the virtual collision kinetic energy of the vehicle relative to each target is obtained based on the virtual collision speed of the vehicle relative to each target. Finally, the virtual collision risk of the vehicle relative to each target is quantified based on the virtual collision kinetic energy of the vehicle relative to each target, and the detection result is obtained based on the virtual collision risk with the highest risk, thus realizing a unified safety assessment of the vehicle relative to dynamic and static targets. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings required in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an autonomous driving safety detection method according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram showing the velocity vector angle between the vehicle and surrounding targets in an embodiment of this application.

[0022] Figure 3 This is a schematic diagram of the preset area in the embodiments of this application. Detailed Implementation

[0023] The detailed description of the accompanying drawings is intended to illustrate the present embodiments of this application and is not intended to represent only the forms in which this application can be implemented. It should be understood that the same or equivalent functions can be accomplished by different embodiments intended to be included within the spirit and scope of this application.

[0024] See Figure 1 One embodiment of this application provides an autonomous driving safety detection method, including the following steps:

[0025] Step S10: When there is at least one target in a preset area around the vehicle, obtain the vehicle's velocity vector and the velocity vector of each target at the current moment; the target is a static target or a dynamic target.

[0026] Specifically, assuming the current time is time y, the vehicle's velocity vector at time y is denoted as... The speed of the vehicle at time y is denoted as The heading angle is denoted as Then we have:

[0027]

[0028] Assuming there are n targets around the vehicle, the velocity vector of the nth target at time t is expressed as: The velocity of the nth target at time t is denoted as . The heading angle is denoted as Then we have:

[0029]

[0030] Based on the above method, the vehicle's velocity vector and the velocity vector of each target at the current moment can be obtained.

[0031] Step S20: Based on the vehicle's velocity vector and the velocity vector of each target, obtain the virtual collision speed of the vehicle relative to each target.

[0032] Specifically, in this embodiment, the angle between the vehicle's velocity vector and the velocity vector of each target is obtained based on the vehicle's velocity vector and the velocity vector of each target. Furthermore, based on the vehicle's velocity vector, the velocity vector of each target, and the angle between the vehicle's velocity vector and the velocity vector of each target, the virtual collision speed of the vehicle relative to each target is obtained. Figure 2 As shown, the vehicle velocity vector at time t... and the velocity of the nth target The included angle between them is denoted as According to the formula for the dot product of vectors:

[0033]

[0034] Therefore:

[0035]

[0036] Based on the above, the angle between the velocity vector of the vehicle and each target at time t can be calculated.

[0037] Furthermore, the virtual collision velocity of the vehicle relative to the nth target at time t is denoted as... Then we have:

[0038]

[0039] Based on the above, the virtual collision velocity of the vehicle relative to each target at time t can be calculated.

[0040] Step S30: Obtain the vehicle's mass, and based on the vehicle's mass and the vehicle's virtual collision velocity relative to each target, obtain the vehicle's virtual collision kinetic energy relative to each target.

[0041] Specifically, in this embodiment, the kinetic energy formula is simplified to E = 1 / 2MV. 2 Let the virtual collision kinetic energy of the vehicle relative to the nth target at time t be denoted as... Then we have:

[0042]

[0043] Where M0 is the vehicle mass.

[0044] Based on the above, the virtual collision kinetic energy of the vehicle relative to each target at time t can be calculated.

[0045] Step S40: Obtain the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target.

[0046] Specifically, the virtual collision kinetic energy of the vehicle relative to each target represents the energy that may be released when the vehicle collides with each target under the assumed collision conditions. The greater the energy, the more severe the consequences of the collision. Based on this, the method in this embodiment quantifies the virtual collision risk based on the conversion of virtual collision kinetic energy. The greater the virtual collision kinetic energy, the higher the corresponding virtual collision risk; the smaller the virtual collision kinetic energy, the lower the corresponding virtual collision risk.

[0047] Step S50: Compare at least one virtual collision risk corresponding to the at least one target, and obtain the detection result based on the virtual collision risk with the highest risk.

[0048] Specifically, the one with the highest calculated virtual collision risk is identified as follows:

[0049]

[0050] Among them, R t Let t be the highest virtual collision risk among multiple targets around the vehicle. Let R be the virtual collision kinetic energy of the vehicle relative to the nth target at time t, where n = 1, 2, 3…. t Obtain the test results.

[0051] This embodiment of the method comprehensively acquires and calculates the target trajectories (including dynamic and static) of the vehicle and all targets (including dynamic and static targets) in a preset area around the vehicle. It uses a unified method to obtain the virtual collision speed of the vehicle relative to each target, and obtains the virtual collision kinetic energy of the vehicle relative to each target based on the virtual collision speed of the vehicle relative to each target. Finally, it quantifies the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target, and obtains the detection result based on the virtual collision risk with the highest risk, thus realizing a unified safety assessment of the vehicle relative to dynamic and static targets.

[0052] In some specific embodiments, step S40 includes:

[0053] Step S411: Obtain the target type for each target, and determine the damage sensitivity factor for each target based on the target type.

[0054] Step S412: Obtain the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target and the damage sensitivity factor of each target.

[0055] Specifically, current autonomous driving safety testing lacks an assessment of the severity of collision damage. For example, a low-speed minor collision between two vehicles results in minimal damage, but a high-speed head-on collision between a regular car and a heavy truck can cause severe vehicle damage. Using collision probability to represent the intensity of collision damage fails to accurately reflect the degree of safety risk. Therefore, the method in this embodiment combines target type and virtual collision kinetic energy to assess virtual collision risk.

[0056] For example, target types can include dynamic targets such as small vehicles, trucks, pedestrians, and bicycles, as well as static targets such as traffic warning objects and curbs. The damage sensitivity factors corresponding to each type of target are shown in Table 1 below.

[0057] Table 1 Values ​​of Damage Sensitivity Factors

[0058]

[0059] Taking the nth target as a pedestrian as an example, the virtual collision risk of the vehicle relative to the nth target at time t is denoted as... Let k be the loss sensitivity factor for the nth target.n k n =1.5, then It can be represented as:

[0060]

[0061] This embodiment of the method no longer uses collision probability as the sole evaluation criterion, but instead comprehensively assesses collision risk by considering both virtual collision kinetic energy and target type. This allows for a more precise reflection of the degree of safety risk. For example, even a low-probability collision event involving a truck will be assessed as having a very high safety risk.

[0062] In some specific embodiments, step S40 includes:

[0063] Step S421: Obtain the time decay factor of the vehicle relative to each target at each moment within a preset future time period.

[0064] Specifically, the method in this embodiment calculates the possible collisions within a certain time period. Typically, the collision risk decays rapidly over time, meaning that the autonomous driving system or the driver has enough reaction time to take evasive action. The calculation of the time decay factor can be based on a certain function, such as an exponential function or an inverse proportional function.

[0065] For example, the time decay factor of any target at each time step is Q. T Let (m) represent the number of elements, where m = 1, 2, 3…N. Then we have:

[0066] Q T (m)=e -am△t

[0067] Where a is the calibration parameter, m is the number of forward prediction time steps, Δt is the step size, representing the time interval of each time step, and e is the base of the natural logarithm (approximately 2.71828).

[0068] Step S422: Based on the virtual collision kinetic energy of the vehicle relative to each target and the time decay factor of the vehicle relative to each target at each moment, obtain the virtual collision risk of the vehicle relative to each target at each moment.

[0069] Let the virtual collision risk of the vehicle relative to the nth target at time "t+m△t" be denoted as but

[0070] It can be represented as:

[0071]

[0072] In another example, the loss sensitivity factor k from the aforementioned embodiments can also be combined. n,but It can be represented as:

[0073]

[0074] By introducing a time decay factor, the collision risk between the vehicle and different targets at a specific point in time can be assessed more accurately, thereby improving the accuracy and practicality of autonomous driving safety detection.

[0075] Step S423: Obtain the virtual collision risk of the vehicle relative to each target based on the virtual collision risk of the vehicle relative to each target at each moment.

[0076] Specifically, taking the nth target as an example, the largest of the multiple virtual collision risks of the vehicle relative to the nth target at multiple moments within the preset future time period is selected as the virtual collision risk of the vehicle relative to each target. The virtual collision risk of the vehicle relative to the nth target can be obtained by the following formula:

[0077]

[0078] in, Let m represent the virtual collision risk of the vehicle relative to the nth target at each moment, where m = 1, 2, 3...N.

[0079] In some specific embodiments, step S40 includes:

[0080] Step S431: Obtain the distance attenuation factor of the vehicle relative to each target at each moment within a preset future time period.

[0081] Specifically, similar to the time decay factor, the distance decay factor is used to quantify the degree to which the potential collision risk decreases as the distance increases. Generally speaking, the farther away from the target, the smaller the probability of a collision, and therefore the risk will decrease accordingly. The distance decay factor can be calculated based on a function, such as an exponential function or an inverse proportional function.

[0082] Step S432: Based on the virtual collision kinetic energy of the vehicle relative to each target and the distance attenuation factor of the vehicle relative to each target at each moment, obtain the virtual collision risk of the vehicle relative to each target at each moment.

[0083] Specifically, Q is defined D (m) is the distance attenuation factor of the vehicle relative to a certain target at time “t+m△t”, where m = 1, 2, 3…N. The virtual collision risk of the vehicle relative to the nth target at time “t+m△t” is denoted as… but It can be represented as:

[0084]

[0085] In some examples, the loss sensitivity factor k from the aforementioned embodiments can also be combined. n and / or time decay factor, then It can be represented in any of the following forms:

[0086]

[0087] Step S543: Obtain the virtual collision risk of the vehicle relative to each target based on the virtual collision risk of the vehicle relative to each target at each moment.

[0088] Specifically, taking the nth target as an example, the largest of the multiple virtual collision risks of the vehicle relative to the nth target at multiple moments within the preset future time period is selected as the virtual collision risk of the vehicle relative to each target. The virtual collision risk of the vehicle relative to the nth target can be obtained by the following formula:

[0089]

[0090] in, Let m represent the virtual collision risk of the vehicle relative to the nth target at each moment, where m = 1, 2, 3...N.

[0091] In some specific embodiments, step S431 includes:

[0092] Step S5411: Obtain the outer contour of the vehicle and its trajectory within a preset future time period.

[0093] Specifically, the vehicle trajectory includes the centroid position of the vehicle within a preset future time period, and the position of each point on the outer contour of the vehicle can be determined based on its centroid position.

[0094] Step S4312: Obtain the outer contour of each target and its target trajectory within a preset future time period.

[0095] Specifically, the target trajectory includes the centroid position of the target within a preset future time period, and the position of each point on the target's outer contour can be determined based on its centroid position.

[0096] Step S4313: Based on the outer contour of the vehicle and its trajectory within a preset future time period, and the outer contour of each target and its trajectory within a preset future time period, obtain the minimum distance between the vehicle and each target at each moment.

[0097] Specifically, based on the centroid coordinates and external dimensions, the coordinate set of all outer contour points of the vehicle and the nth target at time t on the two-dimensional plane is obtained, and denoted as follows: and Then, using the formula for the distance between two points, we iterate through all the outer contour points of the vehicle and the nth target at time t to calculate the distance between them. Then we have:

[0098]

[0099] Assume the minimum distance between the vehicle and the target's outer contour at time t is D. t Then we have:

[0100]

[0101] Based on the above, a series of effective risk distances D can be obtained. t+m△t , m=1,2,3...N, T=N·Δt.

[0102] Step S4314: Obtain the distance attenuation factor of the vehicle relative to each target at each time based on the minimum distance of the vehicle relative to each target at each time.

[0103] Specifically, the distance attenuation factor of the vehicle relative to a certain target at time "t+m△t" can be expressed as:

[0104]

[0105] Where b is the calibration parameter, m is the number of forward prediction time steps, Δt is the step size, representing the time interval of each time step, and e is the base of the natural logarithm (approximately 2.71828).

[0106] In some specific embodiments, step S40 includes:

[0107] Step S441: Obtain the orientation correction factor of the vehicle relative to each target at each moment.

[0108] Specifically, the method in this embodiment introduces a direction correction factor, which can more accurately assess the impact of the vehicle's motion characteristics in a specific direction on collision risk, thereby improving the accuracy of autonomous driving safety detection.

[0109] Step S442: Based on the virtual collision kinetic energy of the vehicle relative to each target and the orientation correction factor of the vehicle relative to each target at each moment, obtain the virtual collision risk of the vehicle relative to each target at each moment.

[0110] Specifically, the direction correction factor at each time step is denoted as I. t (m), m=1,2,3…N, then the virtual collision risk of the vehicle relative to the nth target at each moment. It can be represented in any of the following forms:

[0111]

[0112] Step S443: Obtain the virtual collision risk of the vehicle relative to each target based on the virtual collision risk of the vehicle relative to each target at each moment.

[0113] Specifically, taking the nth target as an example, the largest of the multiple virtual collision risks of the vehicle relative to the nth target at multiple moments within the preset future time period is selected as the virtual collision risk of the vehicle relative to each target. The virtual collision risk of the vehicle relative to the nth target can be obtained by the following formula:

[0114]

[0115] in, Let m represent the virtual collision risk of the vehicle relative to the nth target at each moment, where m = 1, 2, 3...N.

[0116] In some specific embodiments, step S441 includes:

[0117] Step S4411: When the minimum distance between the vehicle and any target at any time is less than or equal to the minimum distance between the vehicle and the target at the previous time, the direction correction factor of the vehicle relative to each target at that time is 1.

[0118] Step S4412: When the minimum distance between the vehicle and any target at any time is greater than the minimum distance between the vehicle and the target at the previous time, the direction correction factor of the vehicle relative to each target at that time is 0.

[0119] Specifically, in this embodiment, I t (m) can be represented as:

[0120]

[0121] Based on the above formula, the orientation correction factor of the vehicle relative to each target at each moment can be obtained.

[0122] In some specific embodiments, the center of the preset area is the center of the vehicle, the horizontal range is a preset range value, and the vertical range is determined according to the vehicle speed at the current moment.

[0123] Specifically, to simplify model computation and improve operational efficiency, security calculations are performed only on targets within a preset area, such as... Figure 3 As shown, with the vehicle's geometric center O as the reference point, the horizontal range of the target selection area is [-5.625m, 5.625m], and the vertical range is [-L, L]. We take L = max{5, v0 * THW}, where v0 is the vehicle speed (m / s) and THW is the calibrable distance parameter of the model.

[0124] It's important to note that the longitudinal range is related to vehicle speed because the faster the vehicle travels, the longer the reaction time and braking distance required, thus necessitating a longer detection range to ensure safety. For example, if an autonomous vehicle is currently traveling at a high speed, the longitudinal range of the preset area will be expanded to detect obstacles further ahead earlier, providing sufficient time and space for avoidance or deceleration. Conversely, if the vehicle is traveling at a slower speed, such as in congested urban areas, the longitudinal range can be relatively smaller because the vehicle's speed and reaction distance are both shorter.

[0125] Another embodiment of this application provides an autonomous driving safety detection device, including a module for performing the autonomous driving safety detection method described in the above embodiments. This device can be a hardware device, a software device, or a combination of hardware and software.

[0126] Another embodiment of this application also provides another autonomous driving safety detection device, including:

[0127] A communication interface used for communicating with other electronic devices;

[0128] Memory is used to store computer program instructions;

[0129] A processor is configured to execute the computer program instructions to support the autonomous driving safety detection device in implementing the methods described in the above embodiments.

[0130] In this embodiment, the memory mainly includes a program storage area and a data storage area. The program storage area can store the operating device, applications required for at least one function, etc., and the data storage area can store related data, etc. Furthermore, the memory can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard disk, a smart media card (SMC), a secure digital card (SD), and a flash card, or other volatile solid-state storage devices.

[0131] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor can be any conventional processor. The processor is the control center of the autonomous driving safety detection device and connects to various parts of the autonomous driving safety detection device using various interfaces and lines.

[0132] This application also provides a computer program product, including computer program instructions, which instruct a computer device to perform operations corresponding to the methods described in the above embodiments.

[0133] Specifically, the computer program product includes a series of computer program instructions that can instruct a computer device to execute the autonomous driving safety detection method described in this application. These instructions are code written in a computer program that defines how to perform specific operations. In this embodiment, these instructions are used to execute the autonomous driving safety detection method of the above embodiments.

[0134] These program instructions are designed to be loaded onto a computer device and to instruct the device to perform specific operations, which refer to the various steps in the autonomous driving safety detection method described in the above embodiments.

[0135] In this way, the computer program product provides a complete software solution that can run on various computer devices to implement the autonomous driving safety detection method described in the above embodiments.

[0136] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technological improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for detecting the safety of autonomous driving, characterized in that, include: When there is at least one target in a preset area around the vehicle, obtain the vehicle's velocity vector and the velocity vector of each target at the current moment; The target can be a static target or a dynamic target; Based on the vehicle's velocity vector and the velocity vector of each target, obtain the virtual collision velocity of the vehicle relative to each target; Obtain the vehicle's mass, and based on the vehicle's mass and the vehicle's virtual collision velocity relative to each target, obtain the vehicle's virtual collision kinetic energy relative to each target. The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision kinetic energy of the vehicle relative to each target. The detection result is obtained by comparing at least one virtual collision risk corresponding to the at least one target and the at least one virtual collision risk with the highest risk.

2. The method according to claim 1, characterized in that, The process of obtaining the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target includes: Obtain the target type for each target, and determine the damage sensitivity factor for each target based on the target type; The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision kinetic energy of the vehicle relative to each target and the damage sensitivity factor of each target.

3. The method according to claim 1, characterized in that, The process of obtaining the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target includes: Obtain the time decay factor of the vehicle relative to each target at each moment within a preset future time period; Based on the virtual collision kinetic energy of the vehicle relative to each target and the time decay factor of the vehicle relative to each target at each moment, the virtual collision risk of the vehicle relative to each target at each moment is obtained. The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision risk of the vehicle relative to each target at each moment.

4. The method according to claim 1, characterized in that, The process of obtaining the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target includes: Obtain the distance decay factor of the vehicle relative to each target at each moment within a preset future time period; Based on the virtual collision kinetic energy of the vehicle relative to each target and the distance attenuation factor of the vehicle relative to each target at each moment, the virtual collision risk of the vehicle relative to each target at each moment is obtained. The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision risk of the vehicle relative to each target at each moment.

5. The method according to claim 4, characterized in that, The step of obtaining the distance attenuation factor of the vehicle relative to each target at each moment within a preset future time period includes: Obtain the vehicle's outer contour and its trajectory within a preset future time period; Obtain the outer contour of each target and its trajectory within a preset future time period; Based on the vehicle's outer contour and its trajectory within a preset future time period, as well as the outer contour of each target and its trajectory within a preset future time period, the minimum distance between the vehicle and each target at each moment is obtained. The distance attenuation factor of the vehicle relative to each target at each moment is obtained based on the minimum distance of the vehicle relative to each target at each moment.

6. The method according to claim 1, characterized in that, The process of obtaining the virtual collision risk of the vehicle relative to each target based on the virtual collision kinetic energy of the vehicle relative to each target includes: Obtain the orientation correction factor of the vehicle relative to each target at each moment; Based on the virtual collision kinetic energy of the vehicle relative to each target and the orientation correction factor of the vehicle relative to each target at each moment, the virtual collision risk of the vehicle relative to each target at each moment is obtained. The virtual collision risk of the vehicle relative to each target is obtained based on the virtual collision risk of the vehicle relative to each target at each moment.

7. The method according to claim 6, characterized in that, The process of obtaining the orientation correction factor at each moment includes: If the minimum distance between the vehicle and any target at any given moment is less than or equal to the minimum distance between the vehicle and that target at the previous moment, then the direction correction factor of the vehicle relative to each target at that moment is 1. If the minimum distance between the vehicle and any target at any given moment is greater than the minimum distance between the vehicle and that target at the previous moment, then the direction correction factor of the vehicle relative to each target at that moment is 0.

8. The method according to claim 1, characterized in that, The center of the preset area is the vehicle center, the horizontal range is a preset range value, and the vertical range is determined according to the vehicle speed at the current moment.

9. An autonomous driving safety detection device, characterized in that, The apparatus includes a module for performing the method according to any one of claims 1 to 8.

10. An autonomous driving safety detection device, characterized in that, include: A communication interface used for communicating with other electronic devices; Memory is used to store computer program instructions; A processor for executing the computer program instructions to support the apparatus in implementing the method according to any one of claims 1 to 8.

11. A computer program product, characterized in that, It includes computer program instructions that instruct a computer device to perform an operation corresponding to the method as described in any one of claims 1 to 8.

Images