Simulation test method and device, equipment, storage medium and product

Abstract

The embodiment of the invention provides a simulation test method, device and equipment, a storage medium and a product, and the method comprises the steps: obtaining a natural driving environment where a tested vehicle is located, and the natural driving environment comprises real traffic flow data; determining the criticality of a background vehicle in the natural driving environment, wherein the criticality represents the influence degree of the background vehicle on the tested vehicle; in the plurality of background vehicles, determining the background vehicle with the criticality meeting a preset condition as a key background vehicle; adjusting the action distribution of the key background vehicle, and generating a dangerous traffic flow; and according to the dangerous traffic flow, performing a simulation test on the tested vehicle. Through the above steps, the test efficiency of the vehicle in the automatic driving process can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle networking, and in particular relates to a simulation test method, device, equipment, storage medium and product. Background Art

[0002] Autonomous driving technology is considered an effective means to reduce traffic congestion and accidents. According to a study by the Rand Corporation of the United States: From a statistical point of view, autonomous vehicles need to be tested for at least 11 billion miles in real or virtual environments to prove that the autonomous driving system is more reliable than human drivers; however, mileage-based testing methods have the disadvantages of long testing cycles, low efficiency, and high costs. For the commercialization of autonomous vehicles, it is obviously not enough to rely solely on mileage-based testing methods. Simulation testing methods based on traffic flow scenarios are indispensable.

[0003] Existing simulation test methods include traffic flow simulation test methods based on neural networks. It mainly refers to inputting real traffic flow data into a specific neural network model to obtain a simulated traffic flow including vehicle interactions, and to achieve targeted generation of traffic flow under complex road conditions. This type of test method is based on real traffic flow data, so that the background vehicles in the simulation scene can adjust their own driving strategies according to the real-time control strategies of the simulation scene test vehicles during the test, thereby simulating real traffic conditions more timely and accurately. However, this type of method has high requirements on the algorithm, the calculation process in the construction of the simulation scene is relatively complex, and the scenes in the real road are infinitely rich, extremely complex, and unpredictable. Therefore, when using this method to conduct a large number of scene tests, since there are fewer dangerous scenes in the real traffic flow, a large number of tests are required to ensure that the test scene covers dangerous scenes, resulting in low test efficiency. Summary of the invention

[0004] The embodiments of the present application provide a simulation test method, apparatus, device, storage medium and product to solve the problem of low test efficiency of existing autonomous driving tests.

[0005] In a first aspect, an embodiment of the present application provides a simulation test method, the method comprising:

[0006] Acquire a natural driving environment where the tested vehicle is located, wherein the natural driving environment includes real traffic flow data;

[0007] Determine the criticality of background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of the background vehicles on the tested vehicle;

[0008] Among the plurality of background vehicles, determining a background vehicle whose criticality meets a preset condition as a critical background vehicle;

[0009] Adjusting the motion distribution of the key background vehicles to generate a dangerous traffic flow;

[0010] According to the dangerous traffic flow, a simulation test is performed on the vehicle under test.

[0011] In a second aspect, an embodiment of the present application provides a simulation test device, the device comprising:

[0012] A first acquisition module is used to acquire a natural driving environment where the tested vehicle is located, wherein the natural driving environment includes real traffic flow data;

[0013] A first determination module is used to determine the criticality of the background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of the background vehicles on the tested vehicle;

[0014] A second determining module is used to determine, among the plurality of background vehicles, a background vehicle whose criticality meets a preset condition as a critical background vehicle;

[0015] An adjustment module, used for adjusting the movement distribution of the key background vehicles to generate a dangerous traffic flow;

[0016] The test module is used to perform a simulation test on the vehicle under test according to the dangerous traffic flow.

[0017] In a third aspect, an embodiment of the present application provides a terminal device, the device comprising: a processor and a memory storing computer program instructions;

[0018] When the processor executes the computer program instructions, the simulation test method of the first aspect is implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the simulation test method of the first aspect is implemented.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the simulation test method of the first aspect.

[0021] The simulation test method, device, equipment, storage medium and product of the embodiment of the present invention, in this embodiment, by selecting key background vehicles in the natural driving environment of the tested vehicle, and then adjusting the action distribution of the key background vehicles, so that the driving behavior of the key background vehicles can be dangerous to the tested vehicle, generating a dangerous driving flow. Finally, the tested vehicle is simulated and tested using the dangerous driving flow, so that the behavior of the key background vehicles in the natural driving traffic flow can be changed, thereby increasing the dangerous driving behavior that has a risk of collision with the test vehicle, thereby effectively improving the efficiency of testing and verification of the autonomous driving decision control algorithm, and truly and efficiently testing the performance of the autonomous driving vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0023] Figure 1 It is a flowchart of the simulation test method provided in the embodiment of the present application;

[0024] Figure 2 is a schematic diagram of a test scenario provided in an embodiment of the present application;

[0025] Figure 3 is a structural schematic diagram of a simulation test device provided in an embodiment of the present application;

[0026] Figure 4 It is a schematic diagram of the structure of the terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0028] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0029] In order to solve the problems of the prior art, the embodiments of the present application provide a simulation test method, device, equipment, storage medium and product. The following first describes an example of the simulation test method provided by the embodiments of the present application.

[0030] Figure 1 FIG. 1 is a flow chart of a simulation test method provided by an embodiment of the present application. Figure 1 As shown, the method includes:

[0031] Step 101, obtaining a natural driving environment where the vehicle under test is located, wherein the natural driving environment includes real traffic flow data;

[0032] In this embodiment, the natural driving environment refers to the real traffic flow in which the tested vehicle is located. In the process of generating the natural driving environment, a data-driven method is used in this embodiment to simulate the natural behavior pattern of the background vehicle. The basic idea is to model the real driving traffic flow with a Markov decision process, calculate the natural distribution of vehicle driving behavior from the real driving data, and then sample the longitudinal and lateral driving behaviors of the vehicle through the natural driving model completed by modeling.

[0033] Specifically, in one embodiment, obtaining the natural driving environment of the tested vehicle includes:

[0034] Determine background vehicles in the area where the tested vehicle is located, and a first state of the background vehicles at each time point within a preset time step;

[0035] Constructing a natural joint distribution of the states, the natural joint distribution including the second state corresponding to the tested vehicle at each time point, and the third state and action corresponding to the background vehicle at each time point;

[0036] The natural joint distribution corresponding to each of the time points is sampled to obtain the natural driving environment, where the natural driving environment includes the movement distribution of the background vehicles in the real driving traffic flow.

[0037] In this embodiment, in the real traffic flow data, in addition to the tested vehicle, there are other background vehicles in the area where the tested vehicle is located. The real traffic flow data also includes the first state of the background vehicles at each time point, and the first state may include the state of position, speed, etc. The background vehicles and the first state can be represented by a first model. Specifically,

[0038] The Markov decision process (MDP) and probabilistic graphical model (PGM) can be used to model the natural driving environment using the spatiotemporal independence relationship between variables, as shown in Equation 1.

[0039]

[0040] Where x i,j represents the state (such as position and speed) of the i-th background vehicle (background vehicle) at the j-th time point, N represents the number of background vehicles in the region of interest (ROI), T represents the total number of preset simulation time steps, and X represents the state space, that is, the state set of the positions and speeds of N vehicles at a total of T moments.

[0041] By constructing a natural joint distribution (denoted as x~P(x)), the action of each background vehicle can be calculated by sampling.

[0042] Assuming that P(x) has the Markov property, its joint distribution factor is simplified as shown in Equation 2.

[0043]

[0044] Among them, the time point k=0,...,T, the state and action space are shown in Equation 3 and Equation 4 respectively.

[0045] s(k)=[s0(k),s1(k)…,s N (k)](3)

[0046] u(k)=[u1(k),…u N (k)](4)

[0047] Among them, s0 represents the second state (position, speed, etc.) of the vehicle under test, s i (i=1, ..., N) represents the third state of the i-th background vehicle, u irepresents the action (longitudinal acceleration, etc.) of the i-th background vehicle. Then the natural driving environment is generated by sampling as u(k)~P(u(k)|s(k)), where sampling can be performed by the following method:

[0048] Assuming that all background vehicles choose their actions independently, P(u(k)|s(k)) can be calculated by Equation 5.

[0049]

[0050] Among them, P(u(k)|s(k)) is the action distribution in the natural driving environment, u(k) is the action of the background vehicle at the kth time point, s(k) is the state of the background vehicle at the kth time point, and u i (k) is the action of the i-th background vehicle at the k-th time point, and N is the number of background vehicles.

[0051] Assuming that the action selection of the background vehicle depends only on the state of itself and the vehicle immediately preceding it, then P(u i (k)|s(k)) can be approximated as It can be calculated through the state-action distribution in the actual collected traffic flow data.

[0052] In this embodiment, by accurately capturing the state changes of background vehicles and constructing a natural joint distribution containing multiple vehicle states and actions, it is possible to highly simulate real traffic scenes and accurately capture the dynamic changes in the traffic environment, thus ensuring the accuracy and reliability of subsequent test results.

[0053] Step 102, determining the criticality of background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of background vehicles on the tested vehicle;

[0054] In this embodiment, after the natural driving environment is generated, since traffic accidents usually occur between two vehicles, in order to further generate vehicles with dangerous driving behaviors around the vehicle under test, this embodiment will select a vehicle that is most likely to collide with the vehicle at each time point as the key background vehicle, and increase its safety risk with the vehicle under test by modifying its driving behavior.

[0055] In order to determine the key background vehicles, it is first necessary to calculate the key degree of each background vehicle in the natural driving environment. Specifically, in one embodiment, determining the key degree of the background vehicles in the natural driving environment includes:

[0056] Determine, by using the action distribution, a plurality of first actions of the background vehicle in the natural driving traffic flow, first occurrence probabilities respectively corresponding to the plurality of first actions, and first risks respectively corresponding to the plurality of first actions;

[0057] For each second action in the plurality of first actions, determining an action criticality of the second action by using a second occurrence probability of the second action and a second risk of the second action;

[0058] The sum of the action criticality of each second action is determined as the criticality of the background vehicle.

[0059] In this embodiment, the method for calculating the criticality of the vehicle's action is shown in Formula 6.

[0060]

[0061] Among them, A i represents the accident between the i-th background vehicle and the detected vehicle, P(u i |s) is the probability of the action occurring in the current state of the vehicle obtained by sampling the natural driving traffic flow (i.e., the first probability), P(A i |s,u i ) is the first risk, indicating that given a state-action pair (s,u i ) probability of an accident.

[0062] In order to calculate the value of the first risk, the behavior of the vehicle under test can be approximated by the following and lane changing models in the microscopic traffic flow model (parameter optimization has been completed through machine learning and other methods). Let S i represents the accident between the i-th background vehicle and the detected vehicle. At this time, the first risk calculation method is shown in Formula 7.

[0063]

[0064] Where P(u0|s) represents the probability that the vehicle under test chooses action u0 in state s. Represents event S i The vehicle state at the time of occurrence. P(u0|s) can be approximately calculated using the proposed model. Calibration can also be performed in simulation using the proposed model.

[0065] After obtaining the first risk corresponding to each first action, for each second action in the plurality of first actions, the action criticality of the second action may be determined by the above formula 6.

[0066] Furthermore, the criticality of each background vehicle is calculated as shown in Formula 8, which is the sum of the criticalities of all second actions that the background vehicle may take in state s.

[0067]

[0068] In this embodiment, by evaluating the criticality of each background vehicle, critical background vehicles can be identified, and thus dangerous traffic flows can be generated by changing the driving behavior of critical background vehicles, thereby achieving a true and efficient test of the performance of the autonomous driving vehicle.

[0069] Step 103, among the plurality of background vehicles, determining the background vehicle whose criticality meets a preset condition as a critical background vehicle;

[0070] In this embodiment, the preset condition may be that the criticality is the largest and greater than a set threshold. And since most accidents involve only two vehicles, in this embodiment, at most one critical background vehicle is considered at the same time.

[0071] Step 104, adjusting the motion distribution of the key background vehicles to generate a dangerous traffic flow;

[0072] In one embodiment, adjusting the motion distribution of the key background vehicles to generate a dangerous traffic flow includes:

[0073] Generate an action probability distribution of the dangerous driving traffic flow according to the criticality of the key background vehicle, the action criticality, the action distribution, and the action distribution of the background vehicles other than the key background vehicle;

[0074] The action probability distribution is adjusted to generate a dangerous traffic flow of the detected vehicle.

[0075] In this embodiment, the action distribution defines the distribution of key background vehicles in different states, including position, speed, acceleration, etc. The action probability distribution reflects the probability of key background vehicles taking different actions in a given state. By adjusting these probabilities, the probability of certain dangerous actions (such as sudden lane changes and emergency braking) can be increased, thereby creating dangerous traffic flows.

[0076] Specifically, the action probability distribution is determined by the following expression:

[0077]

[0078] Among them, q(u|s) is the action probability distribution, ε is the weight of the natural distribution, P(u c |s) is the action distribution of key background vehicles, C c (s) is the criticality of the key background vehicle, V(u c |s) is the action criticality of the key background vehicle, P(u i |s) is the action distribution of the background vehicles except the key background vehicles.

[0079] Step 105: Perform a simulation test on the vehicle under test according to the dangerous traffic flow.

[0080] In this embodiment, by selecting key background vehicles relative to the vehicle under test in the natural driving environment, and then adjusting the action distribution of the key background vehicles, the driving behavior of the key background vehicles can cause danger to the vehicle under test, generating a dangerous driving flow. Finally, the vehicle under test is simulated using the dangerous driving flow, so that the behavior of the key background vehicles in the natural driving traffic flow can be changed to increase the dangerous driving behavior that poses a risk of collision to the test vehicle, thereby effectively improving the efficiency of testing and verification of the autonomous driving decision control algorithm, and truly and efficiently testing the performance of the autonomous driving vehicle.

[0081] In an embodiment of the present application, after performing a simulation test on the tested vehicle according to the dangerous traffic flow, the method further includes:

[0082] In the event that a dangerous accident occurs to the tested vehicle, the test scene and test data corresponding to the dangerous accident are saved.

[0083] In this embodiment, see Figure 2 , Figure 2 For test scenario information, the above-mentioned test scenarios may include scenarios such as traffic environment elements and vehicle information. For example, traffic environment elements may include weather and lighting, static road information and dynamic road information; vehicle information may include test targets and background vehicle behaviors, etc.

[0084] The test data may be various driving data of the vehicle corresponding to the test scene, such as speed, heading angle, position, etc. The vehicle includes the tested vehicle and background vehicles.

[0085] In this embodiment, by saving the test scenarios and test data when a dangerous accident occurs to the tested vehicle during the simulation test process, a dangerous traffic flow scenario library can be constructed, which is convenient for subsequent direct testing using the scenario library to improve test efficiency.

[0086] In an embodiment of the present application, after performing a simulation test on the tested vehicle according to the dangerous traffic flow, the method further includes:

[0087] Obtaining collision events, headway of a single vehicle, and vehicle driving information of the tested vehicle during the simulation test;

[0088] The decision-making planning algorithm of the tested vehicle is evaluated using the collision event, the single-vehicle headway and the vehicle driving information.

[0089] In this embodiment, after completing the simulation test of the vehicle under test, it is also necessary to evaluate the control algorithm used in the simulation test in different dimensions, including safety, efficiency, energy saving and comfort. The specific evaluation method is as follows:

[0090] Since collaborative decision-making / control algorithms contain evaluation indicators of different evaluation dimensions, a comprehensive subjective and objective evaluation method should be adopted. In all indicator systems, algorithm basic performance indicators such as latency, algorithm service scope, and service vehicle upper limit are pass indicators. Only when the algorithm basic performance meets the service scenario requirements and the algorithm basic performance test is passed, the subsequent algorithm evaluation process can be carried out.

[0091] Safety is the most important indicator in collaborative decision-making / control algorithms. The algorithm evaluation method can be based on the table. Among them, a1,…a n is the score assigned to different types of violations (the value is positive), and different values ​​can be taken according to the severity of the violation type. t is the proportion of the total travel time of a single vehicle headway less than the safety threshold, and b is the weight coefficient of the headway score (the value is positive). The final score of the algorithm is the sum of the first-level indicator score and the second-level indicator score, and the lowest score is 0.

[0092] Table 1 Safety index evaluation method

[0093]

[0094] The evaluation models for the two types of algorithms, high efficiency and energy saving, are relatively similar. In each test process, under the same scenario and scenario parameters, the algorithm to be tested and the benchmark algorithm are tested simultaneously and the values ​​of various indicators are calculated. The improvement of each indicator compared with the benchmark indicator is used as the basis for evaluation. The evaluation method can be referred to in the table, and the secondary indicator items can be adjusted according to the actual situation.

[0095] Table 2 Evaluation methods for high efficiency, energy saving and comfort indexes

[0096]

[0097]

[0098] Among them, the secondary indicators are all improvement percentages, and the weight coefficients in the secondary indicator scores must meet

[0099] Formula 1.

[0100]

[0101] The final score of each first-level indicator is the sum of the scores of each second-level indicator, with the maximum value being the highest score and the minimum value being 0. When the value of each second-level indicator is greater than 100%, it is taken as 100%. When the value of a second-level indicator item is negative, there is no lower limit for the score of each sub-item.

[0102] The evaluation method of comfort can be referred to in the table. The final comfort evaluation result is the sum of the primary index and each secondary index, with the highest score being C and the lowest being 0.

[0103] Table 3 Comfort evaluation index

[0104]

[0105] The assignment weight coefficient in the above evaluation method can be determined by combining the entropy method and the order relationship analysis method, and the fuzzy comprehensive evaluation method is used to conduct a comprehensive quantitative evaluation of the collaborative decision-making / control algorithm.

[0106] In another embodiment of the present application, a simulation test platform is also proposed. The vehicle-road-cloud integrated system simulation platform is mainly composed of scene library construction, traffic flow generation, high-precision map, simulation engine, standardized service interface, test report generation and data bus and other functions. Among them, the scene library construction aims to provide continuous high-coverage scene testing for tests such as autonomous driving / networked cloud-controlled driving through the construction of dangerous traffic flows. The simulation engine can use simulation tools such as SUMO and panosim and complete the high-precision map import function. The platform provides the same standardized algorithm service interface as the cloud control basic platform, and can support various algorithm tests that meet the interface protocol. The platform supports full virtual simulation, and can also obtain vehicle real-time status information, roadside perception information and traffic light information by docking with the cloud control basic platform data bus, and then simulate the combination of virtual and real. After the test, a test evaluation report of the algorithm can be automatically generated.

[0107] like Figure 3 As shown, the embodiment of the present application further provides a simulation test device 300, and the simulation test device 300 includes:

[0108] A first acquisition module 301 is used to acquire a natural driving environment where the tested vehicle is located, wherein the natural driving environment includes real traffic flow data;

[0109] A first determination module 302 is used to determine the criticality of the background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of the background vehicles on the tested vehicle;

[0110] A second determining module 303 is used to determine, among the plurality of background vehicles, a background vehicle whose criticality meets a preset condition as a critical background vehicle;

[0111] An adjustment module 304, used to adjust the motion distribution of the key background vehicles to generate a dangerous traffic flow;

[0112] The testing module 305 is used to perform a simulation test on the vehicle under test according to the dangerous traffic flow.

[0113] Optionally, the first acquisition module 301 includes:

[0114] A first determination submodule, used to determine background vehicles in the area where the detected vehicle is located, and a first state of the background vehicles at each time point within a preset time step;

[0115] A construction submodule, used to construct a natural joint distribution of the states, wherein the natural joint distribution includes the second state corresponding to the test vehicle at each time point, and the third state and action corresponding to the background vehicle at each time point;

[0116] The sampling submodule is used to sample the natural joint distribution corresponding to each time point to obtain the natural driving environment, wherein the natural driving environment includes the movement distribution of the background vehicles in the real driving traffic flow.

[0117] Optionally, the sampling submodule is specifically used for:

[0118] Using the first formula to sample the natural joint distribution corresponding to each time point to obtain the natural driving environment;

[0119] The first formula is:

[0120]

[0121] Among them, P(u(k)|s(k)) is the action distribution in the natural driving environment, u(k) is the action of the background vehicle at the kth time point, s(k) is the state of the background vehicle at the kth time point, and u i (k) is the action of the i-th background vehicle at the k-th time point, and N is the number of background vehicles.

[0122] Optionally, the first determining module 302 includes:

[0123] A second determination submodule is used to determine, by using the action distribution, a plurality of first actions of the background vehicle in the natural driving traffic flow, first occurrence probabilities respectively corresponding to the plurality of first actions, and first risks respectively corresponding to the plurality of first actions;

[0124] A third determination submodule, configured to determine, for each second action in the plurality of first actions, an action criticality of the second action by using a second occurrence probability of the second action and a second risk of the second action;

[0125] The fourth determining submodule is used to determine the sum of the action criticality of each second action as the criticality of the background vehicle.

[0126] Optionally, the adjustment module 304 includes:

[0127] A generating submodule, for generating an action probability distribution of a dangerous driving traffic flow according to the criticality of the critical background vehicle, the action criticality, the action distribution, and the action distribution of the background vehicles other than the critical background vehicle;

[0128] The adjustment submodule is used to adjust the action probability distribution to generate a dangerous traffic flow of the vehicle under test.

[0129] Optionally, the simulation test device 300 is specifically used for:

[0130] In the event that a dangerous accident occurs to the tested vehicle, the test scene and test data corresponding to the dangerous accident are saved.

[0131] Optionally, the simulation test device 300 further includes:

[0132] The second acquisition module is used to obtain the collision event, single vehicle headway and vehicle driving information of the tested vehicle during the simulation test;

[0133] An evaluation module is used to evaluate the decision-making planning algorithm of the tested vehicle by using the collision event, the single-vehicle headway and the vehicle driving information.

[0134] Figure 4 A schematic diagram of the hardware structure of a terminal device provided in an embodiment of the present application is shown.

[0135] The terminal device may include a processor 401 and a memory 402 storing computer program instructions.

[0136] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0137] Memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, memory 402 may include a removable or non-removable (or fixed) medium. In appropriate cases, memory 402 may be inside or outside of an integrated gateway disaster recovery device. In a specific embodiment, memory 402 is a non-volatile solid-state memory.

[0138] In certain embodiments, the memory 402 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage media device, an optical storage media device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0139] The processor 401 implements any one of the simulation test methods in the above embodiments by reading and executing computer program instructions stored in the memory 402 .

[0140] In one example, the terminal device may further include a communication interface 403 and a bus 410. Figure 4 As shown, the processor 401, the memory 402, and the communication interface 403 are connected via a bus 410 and communicate with each other.

[0141] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0142] Bus 410 includes hardware, software or both, and the components of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, first video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0143] In addition, in combination with the simulation test method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the simulation test methods in the above embodiment is implemented.

[0144] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0145] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0146] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0147] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each square block in the flowchart and / or block diagram and the combination of each square block in the flowchart and / or block diagram can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable simulation test device to produce a machine so that these instructions executed via the processor of a computer or other programmable simulation test device enable the realization of the function / action specified in one or more square blocks of the flowchart and / or block diagram. This processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It can also be understood that each square block in the block diagram and / or flowchart and the combination of the square blocks in the block diagram and / or flowchart can also be realized by the dedicated hardware that performs the specified function or action, or can be realized by the combination of dedicated hardware and computer instructions.

[0148] The above are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A simulation test method, characterized in that: The method comprises: Acquire a natural driving environment where the tested vehicle is located, wherein the natural driving environment includes real traffic flow data; Determine the criticality of background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of the background vehicles on the tested vehicle; Among the plurality of background vehicles, determining a background vehicle whose criticality meets a preset condition as a critical background vehicle; Adjusting the motion distribution of the key background vehicles to generate a dangerous traffic flow; According to the dangerous traffic flow, a simulation test is performed on the vehicle under test.

2. The simulation test method according to claim 1, characterized in that: The obtaining of the natural driving environment of the tested vehicle includes: Determine background vehicles in the area where the tested vehicle is located, and a first state of the background vehicles at each time point within a preset time step; Constructing a natural joint distribution of the states, the natural joint distribution including the second state corresponding to the tested vehicle at each time point, and the third state and action corresponding to the background vehicle at each time point; The natural joint distribution corresponding to each of the time points is sampled to obtain the natural driving environment, where the natural driving environment includes the movement distribution of the background vehicles in the real driving traffic flow.

3. The simulation test method according to claim 2, characterized in that: The sampling of the natural joint distribution corresponding to each time point to obtain the natural driving environment includes: Using the first formula to sample the natural joint distribution corresponding to each of the time points to obtain the natural driving environment; The first formula is: Among them, P(u(k)|s(k)) is the action distribution in the natural driving environment, u(k) is the action of the background vehicle at the kth time point, s(k) is the state of the background vehicle at the kth time point, and u i (k) is the action of the i-th background vehicle at the k-th time point, and N is the number of preset background vehicles.

4. The simulation test method according to claim 1, characterized in that: Determining the criticality of background vehicles in the natural driving environment includes: Determine, by using the action distribution, a plurality of first actions of the background vehicle in the natural driving traffic flow, first occurrence probabilities respectively corresponding to the plurality of first actions, and first risks respectively corresponding to the plurality of first actions; For each second action in the plurality of first actions, determining an action criticality of the second action by using a second occurrence probability of the second action and a second risk of the second action; The sum of the action criticality of each second action is determined as the criticality of the background vehicle.

5. The simulation test method according to claim 1, characterized in that: The step of adjusting the movement distribution of the key background vehicles to generate a dangerous traffic flow includes: Generate an action probability distribution of the dangerous driving traffic flow according to the criticality of the key background vehicle, the action criticality, the action distribution, and the action distribution of the background vehicles other than the key background vehicle; The action probability distribution is adjusted to generate a dangerous traffic flow of the detected vehicle.

6. The simulation test method according to claim 5, characterized in that: The action probability distribution is determined by the following expression: Among them, q(u|s) is the action probability distribution, ε is the weight of the natural distribution, P(u c |s) is the action distribution of key background vehicles, C c (s) is the criticality of the key background vehicle, V(u c |s) is the action criticality of the key background vehicle, P(u i |s) is the action distribution of the background vehicles except the key background vehicles.

7. The simulation test method according to claim 1, characterized in that: After performing a simulation test on the tested vehicle according to the dangerous traffic flow, the method further comprises: In the event that a dangerous accident occurs to the tested vehicle, the test scene and test data corresponding to the dangerous accident are saved.

8. The simulation test method according to claim 1, characterized in that: After performing a simulation test on the tested vehicle according to the dangerous traffic flow, the method further comprises: Obtaining collision events, headway of a single vehicle, and vehicle driving information of the tested vehicle during the simulation test; The decision-making planning algorithm of the tested vehicle is evaluated using the collision event, the single-vehicle headway time and the vehicle driving information.

9. A simulation test device, characterized in that: The device comprises: A first acquisition module is used to acquire a natural driving environment where the tested vehicle is located, wherein the natural driving environment includes real traffic flow data; A first determination module is used to determine the criticality of the background vehicles in the natural driving environment, wherein the criticality is expressed as the degree of influence of the background vehicles on the tested vehicle; A second determining module is used to determine, among the plurality of background vehicles, a background vehicle whose criticality meets a preset condition as a critical background vehicle; An adjustment module, used for adjusting the movement distribution of the key background vehicles to generate a dangerous traffic flow; The test module is used to perform a simulation test on the vehicle under test according to the dangerous traffic flow.

10. A terminal device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the simulation test method according to any one of claims 1 to 8 is implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the simulation test method according to any one of claims 1 to 8 is implemented.

12. A computer program product, characterized in that The computer product comprises a computer program, and when the computer program is executed by a processor, the simulation test method according to any one of claims 1 to 8 is implemented.

Images