Security evaluation method and device, storage medium and computer device

Abstract

In the safety evaluation method and device, the storage medium and the computer device provided in the present application, the real trajectory and the plurality of predicted trajectories of the obstacle in the real scene are first obtained, and then the real trajectory and each predicted trajectory are mapped into the same coordinate system to calculate the trajectory distance between each predicted trajectory and the real trajectory at each time point, so that the disturbance time point in the simulation scene corresponding to the real scene can be determined according to each trajectory distance and the time point corresponding to each trajectory distance. The disturbance time point is used to indicate that the obstacle starts to move according to the future trajectory output by the preset obstacle interaction model in the simulation scene, so as to perform real interaction with the vehicle on the future trajectory. According to the disturbance time point, the vehicle is simulated and tested multiple times to obtain a simulation result, and then a simulation safety index corresponding to the simulation result can be determined to evaluate the safety of the vehicle in the simulation scene. In this way, the safety of the simulation scene can be accurately evaluated.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a safety assessment method, apparatus, storage medium, and computer equipment. Background Technology

[0002] With the continuous development of technology, autonomous vehicles are gradually becoming a reality. Autonomous vehicles operate in a wide variety of scenarios, and simulation testing can be used to evaluate the performance of autonomous vehicle algorithms in these diverse environments. Simulation testing allows for the evaluation of autonomous driving algorithms based on behavior within a given scenario, thus assessing their performance in areas such as safety. To better simulate real-world scenarios and test the performance of autonomous driving algorithms, simulation platforms can be constructed using real data, simulated data, and expert experience.

[0003] In simulation testing, obstacle interaction models are a crucial component. These models, trained with large datasets based on the behavior of obstacles in the original scene, generate a reasonable future trajectory distribution for the obstacles. One of these trajectories is then selected for sampling, enabling interaction with autonomous vehicles.

[0004] Safety assessment is a key indicator for evaluating the behavior of autonomous vehicles and an important metric in simulation testing. However, current simulation testing primarily relies on single-simulation results for safety assessment, making it difficult to comprehensively evaluate the safety of autonomous vehicles across all scenarios. Summary of the Invention

[0005] The purpose of this application is to address at least one of the aforementioned technical deficiencies, particularly the difficulty in comprehensively assessing the safety of autonomous vehicles across all scenarios in the prior art.

[0006] Firstly, this application provides a security assessment method, the method comprising: Obtain the actual trajectory of obstacles in a real-world scene, as well as multiple predicted trajectories; The actual trajectory and each of the predicted trajectories are mapped to the same coordinate system, and the trajectory distance between each predicted trajectory and the actual trajectory is calculated at each time point; Based on each trajectory distance and the time point corresponding to each trajectory distance, a disturbance time point in the simulation scene corresponding to the real scene is determined. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model, so as to interact with the vehicle in a real way on the future trajectory. Based on the disturbance time points, the vehicle was subjected to multiple simulation tests, and the simulation results were obtained. A simulation safety index corresponding to the simulation results is determined to assess the safety of the vehicle in the simulation scenario.

[0007] In one embodiment, the step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance includes: Among the various trajectory distances, determine whether there exists at least one trajectory distance that is less than a preset distance threshold; If at least one of the trajectory distances is less than the preset distance threshold, then the time point corresponding to the trajectory distance less than the preset distance threshold is taken as a candidate time point; Among the candidate time points, the earliest candidate time point is selected as the disturbance time point.

[0008] In one embodiment, the step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance further includes: If there is no trajectory distance less than the preset distance threshold, then the time point corresponding to the trajectory distance with the smallest value is taken as the disturbance time point.

[0009] In one embodiment, the process by which the preset obstacle interaction model outputs the future trajectory includes: The future trajectory is obtained by inputting the scene information corresponding to the disturbance time point in the simulation scene, along with a preset random seed, into the preset obstacle interaction model.

[0010] In one embodiment, if there is only one simulation scenario, and the simulation result includes the first number of simulation tests in which the vehicle collides with the obstacle, and the total number of first simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the first simulation tests to the total number of the first simulation tests.

[0011] In one embodiment, if the simulation scenario is singular, and the simulation result includes the number of second simulation tests in which the distance between the vehicle and the obstacle is less than a preset distance threshold and the vehicle and the obstacle do not collide, and the total number of second simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the second simulation tests to the total number of the second simulation tests.

[0012] In one embodiment, if there are multiple simulation scenarios, and multiple simulation tests are performed on each simulation scenario, and the simulation results include the number of third simulation tests in each simulation scenario where the vehicle collides with the obstacle, the total number of third simulation tests corresponding to each simulation scenario, and the total number of simulation scenarios, then the step of determining the simulation safety index corresponding to the simulation results includes: For each simulation scenario, the ratio of the number of third simulation tests in that simulation scenario to the total number of third simulation tests is calculated to obtain the scene collision rate of that simulation scenario; Based on the scene collision rate of each simulation scene, a target simulation scene with a scene collision rate less than a preset collision rate threshold is determined from each simulation scene. The total number of target simulation scenarios is determined, and the ratio of the total number of target simulation scenarios to the total number of simulation scenarios is calculated to obtain the simulation safety index.

[0013] Secondly, this application provides a security assessment device, the device comprising: The trajectory acquisition module is used to acquire the actual trajectory of obstacles in a real scene as well as multiple predicted trajectories; The trajectory distance calculation module is used to map the real trajectory and each of the predicted trajectories to the same coordinate system, and to calculate the trajectory distance between each predicted trajectory and the real trajectory at each time point; The disturbance time point determination module is used to determine the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model, so as to interact with the vehicle in a real way on the future trajectory. The simulation testing module is used to perform multiple simulation tests on the vehicle based on the disturbance time points and obtain the simulation results; The simulation safety index determination module is used to determine the simulation safety index corresponding to the simulation results in order to evaluate the safety of the vehicle in the simulation scenario.

[0014] Thirdly, this application provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the security assessment method as described in any of the above embodiments.

[0015] Fourthly, this application provides a computer device, including: one or more processors, and a memory; The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the security assessment method described in any of the above embodiments.

[0016] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: The safety assessment method, apparatus, storage medium, and computer equipment provided in this application first acquire the actual trajectory of the obstacle in a real-world scenario and multiple predicted trajectories. Then, the actual trajectory and each predicted trajectories are mapped to the same coordinate system to calculate the trajectory distance between each predicted trajectory and the actual trajectory at each time point. Based on these trajectory distances and the corresponding time points, a disturbance time point in the simulated scenario corresponding to the real-world scenario can be determined. This disturbance time point indicates that the obstacle in the simulated scenario begins to move according to the future trajectory output by a preset obstacle interaction model, enabling realistic interaction with the vehicle on the future trajectory. Based on this disturbance time point, multiple simulation tests are performed on the vehicle to obtain simulation results. The simulation safety index corresponding to the simulation results can then be determined to assess the vehicle's safety in the simulated scenario. By determining the disturbance time point—that is, the time when the obstacle begins to move according to the future trajectory output by the obstacle interaction model—the behavior of obstacles in real-world scenarios can be simulated more accurately. This improves the realism and credibility of simulation tests, and further evaluates the vehicle's ability to cope and its safety in simulated scenarios. Multiple simulation tests can be conducted in the simulated scenario based on the disturbance time point, obtaining the simulation results for each test, thus providing a more comprehensive understanding of the vehicle's performance under different simulation tests. Through comprehensive analysis of the simulation results, the safety assessment results can be quantified into a simulation safety index, allowing for a more accurate evaluation of the vehicle's safety in simulated scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0018] Figure 1 A schematic flowchart illustrating the security assessment method provided in this application embodiment; Figure 2 A flowchart illustrating the determination of the disturbance time point provided in this application embodiment; Figure 3 A flowchart illustrating the process of determining the simulation safety index provided in this application embodiment; Figure 4This is a schematic diagram of the structure of the security assessment device provided in the embodiments of this application; Figure 5 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] In one embodiment, this application provides a security assessment method. The following embodiments illustrate this method applied to a computer device. It is understood that the computing device refers to a device with data processing capabilities, and may be, but is not limited to, a personal laptop, a desktop laptop, a single server, or a server cluster. Figure 1 As shown, the method of this application may include the following steps: S101: Obtain the actual trajectory of the obstacle in the real scene and multiple predicted trajectories.

[0021] In this context, an obstacle refers to an object, pedestrian, or other vehicle that is close to the vehicle. In one embodiment, the obstacle may be the object, pedestrian, or other vehicle closest to the vehicle. In another embodiment, the obstacle may be an object, pedestrian, or other vehicle that is less than a preset threshold distance from the vehicle. The preset threshold can be selected based on actual conditions, for example, it could be 0.3m or 0.4m. A real-world scenario refers to an objective, real-world scene. A real trajectory refers to the actual movement path of the obstacle in a real-world scenario. A predicted trajectory refers to a trajectory that infers or estimates the possible movement paths of the obstacle in a real-world scenario.

[0022] The method for obtaining the real trajectory of an obstacle in a real-world scene can be chosen according to the actual situation, and this application does not impose specific restrictions on it. For example, sensors can be installed on the obstacle to collect its position and motion information. When the obstacle moves in the real-world scene, the sensor data sent by the obstacle's sensors is acquired, and the real trajectory of the obstacle is generated based on the sensor data. Alternatively, computer vision technology can be used to detect and track the obstacle, and the position and motion trajectory of the obstacle can be extracted by analyzing images or video streams to obtain the real trajectory of the obstacle in the real-world scene.

[0023] The method for obtaining multiple predicted trajectories of an obstacle can be chosen according to the actual situation, and this application does not impose specific restrictions on it. For example, it can use a trajectory prediction model built based on machine learning or deep learning to obtain multiple predicted trajectories of the obstacle. Alternatively, it can utilize physical models and kinematic principles, combined with the initial state of the obstacle and environmental conditions, to infer multiple predicted trajectories of the obstacle.

[0024] S102: Map the actual trajectory and each of the predicted trajectories to the same coordinate system, and calculate the trajectory distance between each predicted trajectory and the actual trajectory at each time point.

[0025] The trajectory distance refers to the distance between the predicted trajectory and the actual trajectory at a given time point. The calculation method for the trajectory distance can be selected according to the actual situation, such as Euclidean distance or Manhattan distance. The time point can be a discrete time point, such as 0.1s or 0.2s.

[0026] It is understandable that both the actual trajectory and each predicted trajectory contain the positional information of the obstacle at each time point. The coordinate systems used for the actual trajectory and each predicted trajectory can be the same or different. By mapping the actual trajectory and each predicted trajectory to the same coordinate system, the trajectory distance between each predicted trajectory and the actual trajectory at each time point can be calculated. The coordinate system can be chosen according to the actual situation; for example, it can be the world coordinate system, or it can be the image coordinate system.

[0027] In one example, the trajectory distance can be calculated using the following expression:

[0028] In the formula, Indicates the distance of the trajectory. Indicates the start time point. Indicates the end time point, ( ) indicates an obstacle t The two-dimensional coordinates of the actual trajectory at any given moment, , ) indicates an obstacle t The two-dimensional coordinates of each moment in the predicted trajectory. n for and The absolute value of the difference. Using the average of the trajectory distances corresponding to multiple consecutive time points as the trajectory distance at the starting time point can reduce the impact of random noise or outliers at a single time point on the trajectory distance calculation, resulting in a smoother and more stable trajectory distance result.

[0029] S103: Based on each of the trajectory distances and the time point corresponding to each trajectory distance, determine the disturbance time point in the simulation scene corresponding to the real scene. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model, so as to interact with the vehicle in a real way on the future trajectory.

[0030] The simulation scenario refers to a specific scenario that simulates a real-world environment, including the definition of various elements and parameters, to simulate and observe the behavior and evolution of a vehicle in a given situation. The preset obstacle interaction model is a model used to describe the interactions between objects and obstacles in the simulation scenario. The future trajectory refers to the actual path that an obstacle will take after the disturbance point in time.

[0031] In this step, the disturbance time point can be determined based on the distance of each trajectory and the time point corresponding to each trajectory distance. Furthermore, the disturbance time point can reflect the true and specific intention of the obstacle, so as to ensure that the true intention of the obstacle in the real scene can be maintained during the simulation process. Thus, when the disturbance time point is reached during the simulation process, the obstacle can move according to the future trajectory output by the preset obstacle interaction model to interact with the vehicle in a real way on the future trajectory.

[0032] In one example, before the perturbation point, the obstacle can move along the real trajectory in the simulation scene, and after the perturbation point, it deviates from the real trajectory and moves along the future trajectory.

[0033] S104: Based on the disturbance time point, perform multiple simulation tests on the vehicle and obtain the simulation results.

[0034] In this step, after determining the disturbance time point, the vehicle is subjected to multiple simulation tests to obtain the simulation results of the vehicle in each simulation test.

[0035] It is understandable that in each simulation test, the obstacle will move along its future trajectory after the disturbance point, and the vehicle will react accordingly to the change in the obstacle's movement path. Through multiple tests, the vehicle's behavior in various scenarios can be simulated, and the simulation results can provide a more comprehensive understanding of the vehicle's performance under different simulation tests.

[0036] S105: Determine the simulation safety index corresponding to the simulation results to evaluate the safety of the vehicle in the simulation scenario.

[0037] Among them, the simulation safety index refers to the indicator used to evaluate the safety of vehicles in simulation scenarios.

[0038] In one embodiment, the simulation safety index can be calculated based on the actual data contained in the simulation results. For example, the potential risk of collision can be calculated by analyzing the proximity and distance between the vehicle and the obstacle, and the simulation safety index can be obtained. Alternatively, the vehicle's trajectory, acceleration, and steering behavior can be analyzed after the obstacle changes its path to evaluate its stability and handling, and the simulation safety index can be obtained. Furthermore, the vehicle's response to emergency situations and braking performance can be evaluated after the obstacle changes its path to obtain the simulation safety index.

[0039] In one example, the obtained simulation safety index can be mapped to a predefined safety level, such as low, medium, and high safety, and the vehicle's safety in the simulation scenario can be evaluated based on the safety level. In another example, the vehicle's safety in the simulation scenario can also be evaluated based on the numerical value of the simulation safety index; for example, a higher simulation safety index indicates a safer vehicle in the simulation scenario, while a lower index indicates a more dangerous vehicle.

[0040] In the above embodiments, the actual trajectory of the obstacle in the real scene and multiple predicted trajectories are first obtained. The actual trajectory and each predicted trajectories are then mapped to the same coordinate system to calculate the trajectory distance between each predicted trajectory and the actual trajectory at each time point. Based on these trajectory distances and the corresponding time points, a disturbance time point in the simulation scene corresponding to the real scene can be determined. This disturbance time point indicates that the obstacle in the simulation scene will begin to move according to the future trajectory output by a preset obstacle interaction model, enabling realistic interaction with the vehicle on the future trajectory. Based on this disturbance time point, multiple simulation tests are performed on the vehicle to obtain simulation results. The simulation safety index corresponding to the simulation results can then be determined to assess the vehicle's safety in the simulation scene. By determining the disturbance time point—that is, the time when the obstacle begins to move according to the future trajectory output by the obstacle interaction model—the behavior of obstacles in real-world scenarios can be simulated more accurately. This improves the realism and credibility of simulation tests, and further evaluates the vehicle's ability to cope and its safety in simulated scenarios. Multiple simulation tests can be conducted in the simulated scenario based on the disturbance time point, obtaining the simulation results for each test, thus providing a more comprehensive understanding of the vehicle's performance under different simulation tests. Through comprehensive analysis of the simulation results, the safety assessment results can be quantified into a simulation safety index, allowing for a more accurate evaluation of the vehicle's safety in simulated scenarios.

[0041] like Figure 2 As shown, in one embodiment, the step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance includes: S201: Among the various trajectory distances, determine whether there exists at least one trajectory distance that is less than a preset distance threshold; S202: If there is at least one trajectory distance that is less than the preset distance threshold, then the time point corresponding to the trajectory distance that is less than the preset distance threshold is taken as a candidate time point; S203: Among the candidate time points, the earliest candidate time point is selected as the disturbance time point.

[0042] Specifically, when the distance between the predicted trajectory and the actual trajectory at a certain point in time is very small, the predicted trajectory and the actual trajectory are quite similar. It can be assumed that the time point reflects a certain true intention, and therefore the time point can be used as the perturbation time point.

[0043] First, a preset distance threshold is set. This preset distance threshold can be determined according to the actual situation, and this application does not impose specific restrictions on it. Based on the relationship between each trajectory distance and the preset distance threshold, trajectory distances smaller than the preset distance threshold are extracted from each trajectory distance, and the time points corresponding to the extracted trajectory distances are used as candidate time points. Finally, among the candidate time points, the earliest candidate time point is selected as the perturbation time point in chronological order.

[0044] In this embodiment, by setting the disturbance time point as early as possible, the sudden changes or sudden appearance of obstacles can be simulated more realistically. This allows for a more accurate evaluation of the autonomous vehicle's response to unpredictable events in simulation testing, further improving the realism and credibility of the test. By using a preset distance threshold to filter trajectory distances, the number of candidate time points that need to be evaluated can be reduced, thereby saving computational resources and time and improving simulation efficiency.

[0045] In one embodiment, the step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance further includes: If there is no trajectory distance less than the preset distance threshold, then the time point corresponding to the trajectory distance with the smallest value is taken as the disturbance time point.

[0046] Specifically, in real-world scenarios, autonomous vehicles may encounter various uncertainties and unexpected situations, requiring corresponding reactions and actions. Therefore, even if the trajectory distance is not less than a preset distance threshold, disturbances need to be introduced into the simulation. By selecting the time point corresponding to the trajectory distance with the smallest value as the disturbance time point, a certain degree of disturbance can be introduced into the simulation, increasing its realism.

[0047] In this embodiment, the time point corresponding to the trajectory distance with the smallest value is selected as the disturbance time point. By introducing a disturbance at this time point, that is, changing the movement path of the obstacle so that it moves according to the future trajectory, the performance and safety of the vehicle can be tested.

[0048] In one embodiment, the process by which the preset obstacle interaction model outputs the future trajectory includes: The future trajectory is obtained by inputting the scene information corresponding to the disturbance time point in the simulation scene, along with a preset random seed, into the preset obstacle interaction model.

[0049] Scene information refers to the data used in the preset obstacle interaction model to describe obstacles, vehicles, and the simulation environment. Scene information may include, but is not limited to, the position and speed of the vehicle, the position and speed of the current obstacle, the position and speed of surrounding obstacles, and map information. A random seed is the initial value or parameter used when generating a pseudo-random number sequence. It can be considered the starting point of the random number generator; different random number sequences can be obtained by setting different random seeds.

[0050] Specifically, in simulation testing, when the simulation time reaches the disturbance time point, all scene information at that time point and a preset random seed are input into the obstacle interaction model to obtain the future trajectory of the obstacle.

[0051] In this embodiment, inputting scene information into the obstacle interaction model helps it understand the real situation of the current simulation scene. Inputting a random seed into the obstacle interaction model allows for different future trajectories in each simulation test, adding complexity to the simulation test. This ensures that the simulation scene maintains both realism and complexity, better testing the vehicle's response in the simulation scene and further comprehensively evaluating the vehicle's safety throughout the entire scenario.

[0052] In one embodiment, if there is only one simulation scenario, and the simulation result includes the first number of simulation tests in which the vehicle collides with the obstacle, and the total number of first simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the first simulation tests to the total number of the first simulation tests.

[0053] Specifically, if multiple simulation tests are conducted in a single simulation scenario, and the simulation results include the number of first simulation tests in which the vehicle collides with an obstacle, as well as the total number of first simulation tests corresponding to the simulation scenario, then the ratio of the number of first simulation tests to the total number of first simulation tests is calculated. Based on the calculated ratio, a simulation safety index can be derived, which represents the collision rate in the simulation scenario. It can be understood that the higher the simulation safety index, the higher the collision rate between the vehicle and obstacles in that simulation scenario, and therefore the lower the vehicle's safety in that simulation scenario.

[0054] In this embodiment, the simulation safety index obtained by calculating the ratio of the first number of simulation tests to the total number of first simulation tests is a simple and intuitive indicator that can provide a preliminary assessment of simulation safety.

[0055] In one embodiment, if the simulation scenario is singular, and the simulation result includes the number of second simulation tests in which the distance between the vehicle and the obstacle is less than a preset distance threshold and the vehicle and the obstacle do not collide, and the total number of second simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the second simulation tests to the total number of the second simulation tests.

[0056] Specifically, if multiple simulation tests are conducted in a single simulation scenario, and the simulation results include the number of second simulation tests where the distance between the vehicle and the obstacle is less than a preset distance threshold, and the total number of second simulation tests corresponding to the simulation scenario, then the ratio of the number of second simulation tests to the total number of second simulation tests is calculated. Based on the calculated ratio, a simulation safety index can be derived. This simulation safety index represents the probability of the vehicle getting too close to an obstacle in the simulation scenario. It can be understood that the higher the simulation safety index, the higher the probability of the vehicle getting too close to an obstacle in the simulation scenario, and therefore the lower the vehicle's safety in that simulation scenario. The preset distance threshold can be determined according to actual conditions, and this application does not impose specific restrictions on it.

[0057] In this embodiment, the simulation safety index obtained by calculating the ratio of the second simulation test count to the total number of second simulation tests is a simple and intuitive indicator that can provide a preliminary assessment of simulation safety.

[0058] like Figure 3As shown, in one embodiment, if there are multiple simulation scenarios, and multiple simulation tests are performed on each simulation scenario, and the simulation results include the number of third simulation tests in each simulation scenario where the vehicle collides with the obstacle, the total number of third simulation tests corresponding to each simulation scenario, and the total number of simulation scenarios for the multiple simulation scenarios, then the step of determining the simulation safety index corresponding to the simulation results includes: S301: For each simulation scenario, calculate the ratio of the number of third simulation tests in that simulation scenario to the total number of third simulation tests to obtain the scene collision rate of that simulation scenario; S302: Based on the scene collision rate of each of the simulation scenarios, determine the target simulation scenario whose scene collision rate is less than a preset collision rate threshold from each of the simulation scenarios; S303: Determine the total number of target simulation scenarios and calculate the ratio of the total number of target simulation scenarios to the total number of simulation scenarios to obtain the simulation safety index.

[0059] Specifically, if there are multiple simulation scenarios, and multiple simulation tests are conducted in each scenario, and the simulation results include the number of third simulation tests in each scenario where the vehicle collides with an obstacle, the total number of third simulation tests for each scenario, and the total number of simulation scenarios, then the ratio of the number of third simulation tests for each scenario to the total number of third simulation tests is calculated to obtain the scenario collision rate for each simulation scenario. Based on the scenario collision probability of each simulation scenario, target simulation scenarios with a scenario collision probability less than a preset collision rate threshold are identified, thereby allowing the total number of target simulation scenarios to be counted. The ratio of the total number of target simulation scenarios to the total number of simulation scenarios is then calculated to obtain the simulation safety index. The preset collision rate threshold can be determined based on actual conditions, and this application does not impose specific restrictions on it. It can be understood that a higher simulation safety index indicates a higher collision rate for the vehicle in different simulation scenarios, and therefore a lower safety level for the vehicle in different simulation scenarios.

[0060] In this embodiment, by calculating the scene collision rate and setting a preset collision rate threshold, the safety of each simulation scene can be objectively evaluated. Applying the simulation safety index to multiple simulation scenes allows for a more comprehensive assessment of the vehicle's safety status under different conditions in various scenarios.

[0061] The security assessment apparatus provided in the embodiments of this application is described below. The security assessment apparatus described below can be referred to in correspondence with the security assessment method described above. For example Figure 4 As shown, this application provides a security assessment device, which may include the following structure: The trajectory acquisition module 401 is used to acquire the real trajectory of the obstacle in the real scene and multiple predicted trajectories; The trajectory distance calculation module 402 is used to map the real trajectory and each of the predicted trajectories to the same coordinate system, and to calculate the trajectory distance between each predicted trajectory and the real trajectory at each time point; The disturbance time point determination module 403 is used to determine the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model so as to interact with the vehicle in a real way on the future trajectory. The simulation test module 404 is used to perform multiple simulation tests on the vehicle based on the disturbance time points and obtain simulation results; The simulation safety index determination module 405 is used to determine the simulation safety index corresponding to the simulation result in order to evaluate the safety of the vehicle in the simulation scenario.

[0062] In one embodiment, the disturbance time point determination module 403 includes: The trajectory distance determination unit is used to determine, among the various trajectory distances, whether there is at least one trajectory distance that is less than a preset distance threshold; The candidate time point determination unit is used to select the time point corresponding to the trajectory distance that is less than the preset distance threshold as a candidate time point if there is at least one trajectory distance that is less than the preset distance threshold. The first disturbance time point determination unit is used to select the earliest candidate time point as the disturbance time point among the candidate time points.

[0063] In one embodiment, the disturbance timing determination module 403 further includes: The second disturbance time point determination unit is used to determine the time point corresponding to the trajectory distance with the smallest value as the disturbance time point if there is no at least one trajectory distance less than the preset distance threshold.

[0064] In one embodiment, the disturbance time point determination module 403 includes: The future trajectory acquisition unit is used to input the scene information corresponding to the disturbance time point in the simulation scene and the preset random seed into the preset obstacle interaction model to obtain the future trajectory.

[0065] In one embodiment, if there is only one simulation scenario, and the simulation results include the first number of simulation tests in which the vehicle collides with the obstacle, and the total number of first simulation tests corresponding to the simulation scenario, then the simulation safety index determination module 405 includes: The first simulation safety index calculation unit is used to calculate the ratio of the number of the first simulation tests to the total number of the first simulation tests, and obtain the simulation safety index.

[0066] In one embodiment, if there is only one simulation scenario, and the simulation results include the number of second simulation tests in which the distance between the vehicle and the obstacle is less than a preset distance threshold and the vehicle and the obstacle do not collide, and the total number of second simulation tests corresponding to the simulation scenario, then the simulation safety index determination module 405 includes: The second simulation safety index determination unit is used to calculate the ratio of the number of the second simulation tests to the total number of the second simulation tests to obtain the simulation safety index.

[0067] In one embodiment, if there are multiple simulation scenarios, and multiple simulation tests are performed on each simulation scenario, and the simulation results include the number of third simulation tests in each simulation scenario where the vehicle collides with the obstacle, the total number of third simulation tests corresponding to each simulation scenario, and the total number of simulation scenarios for the multiple simulation scenarios, then the simulation safety index determination module 405 includes: The scene collision rate calculation unit is used to calculate the ratio of the number of third simulation tests to the total number of third simulation tests for each simulation scene, so as to obtain the scene collision rate of the simulation scene. The target simulation scene determination unit is used to determine, based on the scene collision rate of each simulation scene, a target simulation scene whose scene collision rate is less than a preset collision rate threshold from each simulation scene. The third simulation safety index determination unit is used to determine the total number of target simulation scenarios and calculate the ratio of the total number of target simulation scenarios to the total number of simulation scenarios to obtain the simulation safety index.

[0068] In one embodiment, this application also provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the security assessment method as described in any of the above embodiments.

[0069] In one embodiment, this application also provides a computer device storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the security assessment method as described in any of the above embodiments.

[0070] Indicatively, such as Figure 5 As shown, Figure 5 This is a schematic diagram of the internal structure of a computer device 500 provided in an embodiment of this application. The computer device 500 can be provided as a server. (Refer to...) Figure 5 The computer device 500 includes a processing component 502, which further includes one or more processors, and memory resources represented by memory 501 for storing instructions, such as application programs, that can be executed by the processing component 502. The application programs stored in memory 501 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 502 is configured to execute instructions to perform the security assessment methods of any of the above embodiments.

[0071] The computer device 300 may also include a power supply component 503 configured to perform power management of the computer device 500, a wired or wireless network interface 504 configured to connect the computer device 500 to a network, and an input / output (I / O) interface 505. The computer device 500 may operate on an operating system stored in memory 501, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0072] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0073] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this document, "a," "an," "the," "the," and "its" may also include plural forms unless the context clearly indicates otherwise. "Multiple" refers to at least two, such as 2, 3, 5, or 8, etc. "And / or" includes any and all combinations of the related listed items.

[0074] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0075] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A security assessment method, characterized in that, The method includes: Obtain the actual trajectory of obstacles in a real-world scene, as well as multiple predicted trajectories; The actual trajectory and each of the predicted trajectories are mapped to the same coordinate system, and the trajectory distance between each predicted trajectory and the actual trajectory is calculated at each time point; Based on each trajectory distance and the time point corresponding to each trajectory distance, a disturbance time point in the simulation scene corresponding to the real scene is determined. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model, so as to interact with the vehicle in a real way on the future trajectory. Based on the disturbance time points, the vehicle was subjected to multiple simulation tests, and the simulation results were obtained. A simulation safety index corresponding to the simulation results is determined to assess the safety of the vehicle in the simulation scenario.

2. The security assessment method according to claim 1, characterized in that, The step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each trajectory distance and the time point corresponding to each trajectory distance includes: Among the various trajectory distances, determine whether there exists at least one trajectory distance that is less than a preset distance threshold; If at least one of the trajectory distances is less than the preset distance threshold, then the time point corresponding to the trajectory distance less than the preset distance threshold is taken as a candidate time point; Among the candidate time points, the earliest candidate time point is selected as the disturbance time point.

3. The security assessment method according to claim 2, characterized in that, The step of determining the disturbance time point in the simulation scene corresponding to the real scene based on each trajectory distance and the time point corresponding to each trajectory distance further includes: If there is no trajectory distance less than the preset distance threshold, then the time point corresponding to the trajectory distance with the smallest value is taken as the disturbance time point.

4. The security assessment method according to claim 1, characterized in that, The process by which the preset obstacle interaction model outputs the future trajectory includes: The future trajectory is obtained by inputting the scene information corresponding to the disturbance time point in the simulation scene, along with a preset random seed, into the preset obstacle interaction model.

5. The security assessment method according to claim 1, characterized in that, If the simulation scenario is singular, and the simulation result includes the first number of simulation tests in which the vehicle collides with the obstacle, and the total number of first simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the first simulation tests to the total number of the first simulation tests.

6. The security assessment method according to claim 1, characterized in that, If the simulation scenario is singular, and the simulation result includes the number of second simulation tests in which the distance between the vehicle and the obstacle is less than a preset distance threshold and the vehicle and the obstacle do not collide, and the total number of second simulation tests corresponding to the simulation scenario, then the step of determining the simulation safety index corresponding to the simulation result includes: The simulation safety index is obtained by calculating the ratio of the number of the second simulation tests to the total number of the second simulation tests.

7. The security assessment method according to claim 1, characterized in that, If there are multiple simulation scenarios, and multiple simulation tests are performed on each simulation scenario, and the simulation results include the number of third simulation tests in each simulation scenario where the vehicle collides with the obstacle, the total number of third simulation tests corresponding to each simulation scenario, and the total number of simulation scenarios, then the step of determining the simulation safety index corresponding to the simulation results includes: For each simulation scenario, the ratio of the number of third simulation tests in that simulation scenario to the total number of third simulation tests is calculated to obtain the scene collision rate of that simulation scenario; Based on the scene collision rate of each simulation scene, a target simulation scene with a scene collision rate less than a preset collision rate threshold is determined from each simulation scene. The total number of target simulation scenarios is determined, and the ratio of the total number of target simulation scenarios to the total number of simulation scenarios is calculated to obtain the simulation safety index.

8. A safety assessment device, characterized in that, The device includes: The trajectory acquisition module is used to acquire the actual trajectory of obstacles in a real scene as well as multiple predicted trajectories; The trajectory distance calculation module is used to map the real trajectory and each of the predicted trajectories to the same coordinate system, and to calculate the trajectory distance between each predicted trajectory and the real trajectory at each time point; The disturbance time point determination module is used to determine the disturbance time point in the simulation scene corresponding to the real scene based on each of the trajectory distances and the time point corresponding to each trajectory distance. The disturbance time point is used to indicate that the obstacle in the simulation scene starts to move according to the future trajectory output by the preset obstacle interaction model, so as to interact with the vehicle in a real way on the future trajectory. The simulation testing module is used to perform multiple simulation tests on the vehicle based on the disturbance time points and obtain the simulation results; The simulation safety index determination module is used to determine the simulation safety index corresponding to the simulation results in order to evaluate the safety of the vehicle in the simulation scenario.

9. A storage medium, characterized in that: The storage medium stores computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the security assessment method as described in any one of claims 1 to 7.

10. A computer device, characterized in that, include: One or more processors, and memory; The memory stores computer-readable instructions that, when executed by the one or more processors, perform the steps of the security assessment method as described in any one of claims 1 to 7.

Images