Vehicle testing method, device and equipment based on intelligent agent
By integrating an agent dynamics solver into the simulation engine and utilizing an agent performance testing method in a full simulation environment, the problem of high cost and low efficiency in vehicle testing in existing technologies is solved, achieving low-cost and high-efficiency vehicle performance testing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- VOYAH AUTOMOBILE TECH CO LTD
- Filing Date
- 2026-03-10
- Publication Date
- 2026-05-15
AI Technical Summary
Existing technologies for testing vehicles in simulation environments are costly and inefficient, failing to achieve low-cost and high-efficiency performance testing.
A full-simulation intelligent agent performance testing method is adopted. By integrating an intelligent agent dynamics solver into the simulation engine, the following steps are repeatedly executed: determining scene information and intelligent agent information based on scene data, controlling the movement of the intelligent agent to obtain performance test data, and updating the simulation environment.
It enables low-cost and efficient vehicle performance testing, which is more cost-effective and efficient than hardware-in-the-loop (HIL) simulation testing.
Smart Images
Figure CN122042276A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent agent performance testing technology, and in particular to a vehicle testing method, apparatus and equipment based on intelligent agents. Background Technology
[0002] When testing vehicle performance, an intelligent agent corresponding to the vehicle can be created in a simulation environment based on a simulation system. This intelligent agent represents the vehicle in the simulation environment. Then, the vehicle's performance can be tested.
[0003] In existing technologies, when testing a vehicle in a simulation environment, control commands are obtained after acquiring simulated sensor data from the vehicle in the simulation environment. These control commands are then sent to the real vehicle, which executes the commands to complete the vehicle's performance test.
[0004] However, the above methods are based on semi-simulation and semi-realistic environments for vehicle performance testing, which leads to high testing costs and low testing efficiency. Summary of the Invention
[0005] This application provides a vehicle testing method, apparatus, and device based on intelligent agents, which can achieve the effect of testing vehicle performance at low cost and high efficiency.
[0006] In a first aspect, embodiments of this application provide a vehicle testing method based on an intelligent agent, wherein the intelligent agent represents a vehicle in a simulated environment, and the method includes:
[0007] Repeat the following steps until the overall test of the agent is completed:
[0008] Based on the scene data of the intelligent agent in the current simulation environment, scene information and intelligent agent information are determined; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment;
[0009] Based on the scene information and the agent information, when controlling the agent to move, the performance test data of the agent is obtained, and the position information of the agent after the movement is determined.
[0010] The current simulation environment is updated based on the position information of the moved agent.
[0011] In one possible implementation, based on the scene information and the agent information, when controlling the agent to move, performance test data of the agent is obtained, including:
[0012] Based on the scene information and the agent structure parameters in the agent information, a control signal for the agent is generated; wherein, the agent structure parameters characterize data on the vehicle structure of the agent;
[0013] Based on the control signal, the intelligent agent is controlled to move, and while controlling the intelligent agent to move, the performance test data of the intelligent agent is obtained.
[0014] In one possible implementation, the scene data further includes collision information; the agent information includes agent structural parameters and agent control parameters, wherein the agent structural parameters characterize data on the vehicle structure of the agent, and the agent control parameters characterize control parameters of the agent's driving process.
[0015] The collision information represents the obstacle information of the agent in the current simulation environment; determining the position information of the agent after movement includes:
[0016] Based on a preset intelligent agent dynamics solver, the collision information and the intelligent agent information are processed to determine the position information of the moved intelligent agent.
[0017] In one possible implementation, updating the current simulation environment based on the position information of the moved agent includes:
[0018] Based on the location information of the moved agent, the current simulation environment, and the agent information, the scene data in the current simulation environment is updated to update the current simulation environment.
[0019] In one possible implementation, the method further includes:
[0020] The updated scene data in the current simulation environment is transmitted to the refeeding simulation engine that processes the intelligent agent.
[0021] In one possible implementation, the method further includes:
[0022] Based on a pre-established communication channel, the scene data of the intelligent agent in the current simulation environment is acquired; wherein, the communication channel represents the communication channel between the agent and the real vehicle scene data storage library, and the real vehicle scene data storage library stores scene data for each real scene.
[0023] The acquired scene data is then denoised.
[0024] In one possible implementation, the performance test data includes braking performance test data. Based on the scene information and the agent information, when controlling the agent to move, the performance test data of the agent is obtained, including:
[0025] Based on the location information of the intelligent agent, the relative distance between the intelligent agent and the obstacle is determined;
[0026] If the relative distance is determined to be equal to a preset relative distance threshold, the braking performance of the intelligent agent is tested to obtain the braking performance test data.
[0027] Secondly, embodiments of this application provide a vehicle testing device based on an intelligent agent, wherein the intelligent agent represents a vehicle in a simulated environment, and the device includes:
[0028] Repeat the steps for each of the following modules until the overall test of the agent is completed:
[0029] The first determining module is used to determine scene information and agent information based on scene data of the agent in the current simulation environment; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment;
[0030] The control module is used to obtain the performance test data of the intelligent agent when controlling the intelligent agent to move, based on the scene information and the intelligent agent information.
[0031] The second determining module is used to determine the position information of the moved agent;
[0032] The update module is used to update the current simulation environment based on the position information of the moved agent.
[0033] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;
[0034] The memory stores computer-executed instructions;
[0035] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.
[0036] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.
[0037] The vehicle testing method, apparatus, and equipment based on intelligent agents provided in this application determine scene information and intelligent agent information based on scene data of the intelligent agent in the current simulation environment; then, based on the scene information and the intelligent agent information, when controlling the intelligent agent to move, the performance test data of the intelligent agent is obtained, and the position information of the intelligent agent after movement is determined; finally, the current simulation environment is updated based on the position information of the intelligent agent after movement. Since this solution adopts full simulation intelligent agent performance testing, it has lower cost and higher testing efficiency compared to semi-physical simulation testing. Attached Figure Description
[0038] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0039] Figure 1 A flowchart illustrating the agent-based vehicle testing method provided in this application;
[0040] Figure 2 A schematic diagram of the workflow of the intelligent agent dynamics solver provided in this application;
[0041] Figure 3 Schematic diagram of the process for updating scene data provided in this application Figure 1 ;
[0042] Figure 4 Schematic diagram of the process for updating scene data provided in this application Figure 2 ;
[0043] Figure 5 A schematic diagram of the structure of the agent-based vehicle testing device provided in this application;
[0044] Figure 6 A schematic diagram of the structure of the electronic device provided in this application.
[0045] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation
[0046] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.
[0047] In existing vehicle testing methods, the virtual scene server obtains the coordinate data of the simulated vehicle in the virtual scene and then uses this coordinate data to acquire corresponding simulation sensor data. The autonomous driving controller derives control commands for the autonomous vehicle based on the simulation sensor data and sends them to the vehicle controller. The autonomous vehicle then moves to the next position according to the control commands. Existing technology utilizes data interaction between the virtual scene simulation server and the autonomous driving controller to achieve hardware-in-the-loop (HIL) simulation testing, avoiding the need for complex vehicle dynamics models and expensive multi-DOF test benches. However, using HIL simulation testing for vehicle-in-the-loop simulation is costly. Using a real vehicle for HIL testing is inefficient.
[0048] To address the problems existing in existing technologies, the inventors, during their research on agent-based vehicle testing schemes, discovered that by integrating an agent dynamics solver into the simulation engine, vehicle performance can be tested through full simulation. Compared to hardware-in-the-loop (HIL) simulation testing, this method is lower in cost and more efficient. Specifically, the following steps are repeatedly executed using the HIL simulation engine with the integrated agent dynamics solver until the overall testing of the agent is completed: Based on the scenario data of the agent in the current simulation environment, scene information and agent information are determined; where scene data represents the data of the vehicle in the real environment corresponding to the current simulation environment; then, based on the scene information and agent information, when controlling the agent to move, the performance test data of the agent is obtained, and the position information of the moved agent is determined; based on the position information of the moved agent, the current simulation environment is updated. Since the entire testing process is carried out on a full simulation platform, it is lower in cost and more efficient than HIL simulation testing.
[0049] Based on the above inventive concept, the vehicle testing scheme based on intelligent agents in this application was designed.
[0050] The technical solution of this application and how it solves the above-mentioned technical problems will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.
[0051] Figure 1 A flowchart illustrating the agent-based vehicle testing method provided in this application is shown below. Figure 1 As shown, the method includes:
[0052] S101. Based on the scene data of the intelligent agent in the current simulation environment, determine the scene information and intelligent agent information; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment.
[0053] For example, an intelligent agent represents a vehicle in a simulated environment.
[0054] For example, the current simulation environment refers to the simulation environment in which the intelligent agent is being tested. Narrowing the interpretation of scene information, it should refer to information other than agent information and collision object information, such as information about mountains, plains, houses, factories, signal towers, forests, grasslands, rivers, lakes, streetlights, etc. Scene information can be understood as information other than agent information and collision object information. For example, scene information can also include information such as sunny weather, rain / snow, foggy weather, day / night cycle, and light intensity.
[0055] For example, scene data specifically refers to a set of data used to digitally reproduce or correspond to a specific real scene in a simulation environment. Its core purpose is to enable the re-feeding simulation engine to conduct tests based on information from the real environment. In other words, scene data is a digital twin data template of the real world in the simulation system.
[0056] Scene data includes: high-precision maps and static features, environmental parameters, events, and logical annotations.
[0057] High-precision maps and static elements: precise three-dimensional geometry of real roads, lane lines, traffic signs, building locations, and other static information obtained through LiDAR, surveying and mapping, etc.
[0058] Ambient light parameters: Record the real weather conditions at the time the scene occurs, such as visibility, rainfall, and lighting conditions, such as time, sun angle, and road conditions.
[0059] Event and logic annotation: Annotate key events that occur in real-world scenarios, such as changes in traffic lights, and define their triggering conditions and rules.
[0060] As shown above, traffic data recorded in a real city, including complex intersections and dense traffic flow, is processed into scene data and then injected back into the simulation engine. The agent's algorithm (or the agent's performance) can then be repeatedly and safely tested in highly realistic virtual scene data to evaluate the agent's performance in similar real-world environments.
[0061] S102. Based on the scene information and agent information, when controlling the agent to move, obtain the agent's performance test data and determine the agent's position information after the movement.
[0062] For example, scene information refers to the environment in which the agent is located, as discussed above. Agent information includes agent structural parameters and agent control parameters. Agent structural parameters represent data on the agent's vehicle structure, while agent control parameters represent control parameters for the agent's driving process.
[0063] For example, agent structural parameters refer to the set of inherent parameters that define the agent's physical properties, mobility, perception configuration, and control interface. These parameters determine the agent's basic form and behavioral boundaries in the simulation environment. Agent structural parameters include physical geometric parameters such as length, width, height, wheelbase, track width, and shape, which define the agent's spatial occupancy and shape for collision detection and spatial relationship calculation.
[0064] The agent's control parameters directly determine how the agent should act at any given moment. For example, longitudinal control parameters include: target speed: the speed the agent wants to achieve, e.g., 30 km / h; acceleration / deceleration commands: commands that directly control the depth of throttle or brake pedal movement, such as... (accelerate), (Braking); Target following distance: The time interval you want to maintain when following the vehicle in front, such as 2.0 seconds, which means maintaining a 2-second driving distance from the vehicle in front.
[0065] Based on scene information and agent information, when controlling the agent to move, the performance test data of the agent is obtained, and the position information of the agent after movement is determined. The entire cycle of perception, decision-making, control, physical update, and evaluation is continuously carried out in the simulation. The agent also stops smoothly and maintains a safe distance. After the test ends, the simulation engine can be fed back to summarize the performance data of the whole process, such as the minimum distance, average deceleration, and parking accuracy, and record the updated vehicle position trajectory for each frame.
[0066] S103. Update the current simulation environment based on the position information of the moved agent.
[0067] For example, the calculated new position, such as coordinates and orientation, is used to update the current simulation environment.
[0068] For example, not only can the current simulation environment be updated based on the new location, but velocity and acceleration information can also be written into the agent's state data to replace the old state. Not only the controlled agent, but other dynamic entities in the environment, such as vehicles and pedestrians, will also move according to their own control logic, and the simulation engine needs to be updated to uniformly update their positions and states.
[0069] For example, check whether the agent has interacted with other elements in the environment, such as other vehicles, pedestrians, curbs, and traffic cones, such as collisions or intrusions, and update the status flags of these interactions.
[0070] For example, based on the agent's new location, it can be determined whether a new scenario condition is triggered. For instance, if the agent enters a new road area, the simulation engine may need to load a more detailed map model. Or, if the agent reaches a preset point, the next test phase may be triggered, such as when a pedestrian suddenly appears crossing the road.
[0071] like Figure 2 As shown, the agent dynamics solver includes parameterized modeling of the agent's dynamics to obtain agent information, and modeling of colliders to obtain collider information. That is, after parameterized modeling of the agent's dynamics, the agent dynamics solver obtains the agent's geometric and physical information. It also includes modeling of colliders (other moving objects and stationary obstacles) to obtain their geometric and physical information. The modeling of the agent and colliders also includes various constraint information.
[0072] The agent dynamics solver includes a physics engine, which inputs the geometric, physical, and kinematic information of the agent, as well as the geometric, physical, and kinematic information of other moving bodies and stationary obstacles. By performing dynamics solving in the physics engine, new motion states of the agent, other moving bodies, and obstacles can be obtained and input back into the simulation engine for further calculations.
[0073] The agent algorithm sends control signals to the agent dynamics model, which then performs dynamic calculations based on the control signals to obtain the agent's kinematic state information.
[0074] like Figure 3 As shown, the recharge simulation engine includes a scene data parsing module. By parsing the scene data, the scene parsing module obtains information including agent structural parameters, agent motion state, agent constraints, and collision objects.
[0075] The scene update module provides information such as agent information, collision information, environment and event triggering status at a certain timestamp to the corresponding modules. For example, it provides the agent state update module with the geometric, physical and motion state information of the agent, provides the collision generation module with the geometric, physical and motion state information of other moving objects and obstacles, and provides the environment / state update module with the environment and event triggering status information.
[0076] The agent state update module is responsible for receiving new agent state information and passing it to the agent dynamics solver for state update; the collision body generation module provides the agent dynamics solver with the geometric, physical and motion state information of other moving bodies and obstacles to generate collision bodies; the environment / state update module provides the agent algorithm with environmental information and event trigger messages to perform control signal calculation.
[0077] like Figure 4As shown, the recharge simulation engine parses the scene data to obtain collision information, agent structural parameters, agent control parameters, and scene information. The agent algorithm generates control signals based on the agent structural parameters (or agent control parameters) and scene information. The agent dynamics solver generates new agent position information based on the collision information, control signals, and agent structural parameters. Based on the new agent position information, it generates new scene data and recharges the generated new scene data back to the recharge simulation engine.
[0078] The vehicle testing method based on intelligent agents provided in this application tests vehicle performance through full simulation. Compared with semi-physical simulation testing methods, it is lower in cost and more efficient. Specifically, after re-injecting the simulation engine and connecting the dynamics solver, the following steps are executed: determining scene information and intelligent agent information based on scene data of the intelligent agent in the current simulation environment; then, based on the scene information and intelligent agent information, obtaining the performance test data of the intelligent agent while controlling the intelligent agent to move, and determining the position information of the intelligent agent after movement; finally, based on the position information of the intelligent agent after movement, entering the current simulation environment.
[0079] In one example, based on scene information and agent information, performance test data of the agent is obtained when controlling the agent to move, including:
[0080] Based on the intelligent agent structural parameters in the scene information and intelligent agent information, control signals for the intelligent agent are generated; whereby the intelligent agent structural parameters represent the data on the vehicle structure of the intelligent agent.
[0081] Based on control signals, the agent is controlled to move, and performance test data of the agent is obtained while the agent is moving.
[0082] In one example, the performance test data includes braking performance test data. Based on scene information and agent information, when controlling the agent to move, the agent's performance test data is obtained, including:
[0083] Based on the agent's location information, determine the relative distance between the agent and obstacles;
[0084] If the relative distance is determined to be equal to the preset relative distance threshold, the braking performance of the intelligent agent is tested to obtain braking performance test data.
[0085] For example, in a static environment: a dry, straight asphalt road with good adhesion; a stationary obstacle (vehicle) 100 meters ahead; the key triggering rule: when the relative distance between the tested vehicle and the obstacle is 50 meters, the AEB system is triggered to apply maximum braking force.
[0086] Intelligent agent information: Identity, test vehicle equipped with AEB system, initial state: driving at a constant speed of 80 km / h towards the obstacle, the driver model is in a "distracted" state (no active braking); Key structural parameters: Mass, 1500 kg; Axle load distribution, 60% front, 40% rear; Peak coefficient of friction between tires and the ground: 0.8; Maximum braking force of the braking system, providing approximately... The deceleration. The sensor, a forward-facing radar, has a detection range of 150 meters and a refresh rate of 100Hz.
[0087] For example, the sensors of the intelligent agent (the vehicle under test) continuously detect obstacles ahead, and the controller calculates the relative distance and relative speed in real time. The controller algorithm (the AEB decision logic of the intelligent agent algorithm) performs pre-calculation based on scene information (the obstacle is stationary) and its own structural parameters, such as mass and the maximum braking force of the braking system: based on the current speed and mass, it calculates the theoretical stopping distance required under maximum braking; it compares this distance with the current actual distance and, following its internal strategy, sets different levels of alarms and partial braking trigger points; the key control signal generation moment: when the simulation engine determines that the relative distance between the two vehicles is equal to the preset 50-meter threshold, the AEB system is triggered, generating the strongest control signal. The vehicle dynamics model receives the maximum braking request, and the model calculates based on the intelligent agent's structural parameters: under the maximum braking request, due to the forward shift of the center of gravity, the front wheel load increases and the rear wheel load decreases. The tire model calculates the actual maximum braking force that the four tires can provide based on the vertical load and slip ratio. This force may be slightly lower than the theoretical maximum value; based on the calculated total braking force, the model calculates the actual deceleration of the vehicle according to Newton's second law, such as... And begin updating the vehicle's motion status.
[0088] For example, braking performance test data is obtained as follows: From the moment emergency braking is triggered (relative distance = 50 meters) until the vehicle comes to a complete stop, the system collects and calculates the following braking performance test data: The actual distance traveled from the trigger point (50 meters) to the vehicle coming to a complete stop is 38.5 meters; the average deceleration of the vehicle during the entire braking process is... The maximum deceleration achieved during braking is ; The total time from triggering to complete stop is 2.8 seconds; after the vehicle stops, the final distance between the front of the vehicle and the obstacle is 11.5 meters.
[0089] As shown above, in the set test scenario, the braking performance of this intelligent agent is excellent. It not only completely avoids collisions in terms of safety, but also demonstrates a high level of performance in terms of braking efficiency, response speed and stability.
[0090] In one example, the scene data also includes collision information; the agent information includes agent structural parameters and agent control parameters. The agent structural parameters represent the data on the vehicle structure of the agent, and the agent control parameters represent the control parameters of the agent's driving process.
[0091] The collision information represents the information of obstacles in the current simulation environment for the agent; it determines the position information of the agent after movement, including:
[0092] Based on a pre-defined agent dynamics solver, the collision information and agent information are processed to determine the position information of the agent after it has moved.
[0093] Please refer to the above exemplary description, which will not be repeated here. For example, the final distance between the front of the vehicle and the obstacle is 11.5 meters, which is the location information of the intelligent agent after it has moved.
[0094] In one example, the current simulation environment is updated based on the position information of the moved agent. This includes updating the scene data in the current simulation environment based on the position information of the moved agent, the current simulation environment, and the agent information.
[0095] In one example, the updated scene data in the current simulation environment is sent to the feedback simulation engine that processes the intelligent agent.
[0096] For example, continuing with the above exemplary description, if another performance of the agent is to be tested continuously, then the test of the other performance is performed on the basis of a relative distance of 11.5 meters between the agent and the obstacle.
[0097] In one example, scene data of the agent in the current simulation environment is acquired based on a pre-established communication channel; wherein, the communication channel represents the communication channel between the agent and the real vehicle scene data storage library, which stores scene data for each real scene; the acquired scene data is then denoised.
[0098] For example, denoising aims to improve the reliability and usability of data; without denoising, performance evaluations will be distorted.
[0099] This solution is a full simulation test of vehicle performance, which is lower in cost and more efficient than hardware-in-the-loop (HIL) simulation testing.
[0100] This solution provides an intelligent agent dynamics solver that includes a scheme for the intelligent agent to receive control signals and solve for the dynamic response, as well as a scheme for the intelligent agent dynamics solver to perform physical calculations after receiving scene information and intelligent agent information. This can improve the existing simulation engine's ability to solve the problem of insufficient model accuracy calculation. This solution includes three major modules: a recharge simulation engine, an intelligent agent dynamics solver, and an intelligent agent algorithm, realizing a complete intelligent agent closed-loop simulation recharge system, which can improve the realism of the recharge simulation system.
[0101] This solution provides collision information, which improves the realism of the simulation.
[0102] Figure 5 This is a schematic diagram of the structure of the agent-based vehicle testing device provided in this application, as shown below. Figure 5 As shown, the vehicle testing device 50 based on intelligent agents provided in this embodiment includes:
[0103] The first determining module 501 is used to determine scene information and intelligent agent information based on scene data of the intelligent agent in the current simulation environment; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment;
[0104] The control module 502 is used to obtain the performance test data of the intelligent agent when controlling the intelligent agent to move, based on scene information and intelligent agent information.
[0105] The second determining module 503 is used to determine the position information of the moved agent;
[0106] The update module 504 is used to update the current simulation environment based on the position information of the moved agent.
[0107] In one possible implementation, the control module 502 is further configured to:
[0108] Based on the intelligent agent structural parameters in the scene information and intelligent agent information, control signals for the intelligent agent are generated; whereby the intelligent agent structural parameters represent the data on the vehicle structure of the intelligent agent.
[0109] Based on control signals, the agent is controlled to move, and performance test data of the agent is obtained while the agent is moving.
[0110] In one possible implementation, the scene data also includes collision information; the agent information includes agent structural parameters and agent control parameters, wherein the agent structural parameters represent data on the vehicle structure of the agent, and the agent control parameters represent control parameters of the agent's driving process.
[0111] The collision information characterizes the information of obstacles in the current simulation environment for the intelligent agent; the second determination module 503 is also used for:
[0112] Based on a pre-defined agent dynamics solver, the collision information and agent information are processed to determine the position information of the agent after it has moved.
[0113] In one possible implementation, the update module 504 is further configured to:
[0114] Based on the location information of the moved agent, the current simulation environment, and the agent's information, the scene data in the current simulation environment is updated to update the current simulation environment.
[0115] In one possible implementation, the agent-based vehicle testing device 50 is further used for:
[0116] The updated scene data in the current simulation environment is transmitted to the feedback simulation engine of the processing agent.
[0117] In one possible implementation, the agent-based vehicle testing device 50 is also used for:
[0118] Based on a pre-established communication channel, the scene data of the intelligent agent in the current simulation environment is acquired; wherein, the communication channel represents the communication channel between the intelligent agent and the real vehicle scene data storage library, which stores the scene data of each real scene;
[0119] The acquired scene data is then denoised.
[0120] In one possible implementation, the performance test data includes brake performance test data, and the control module 502 is further used for:
[0121] Based on the agent's location information, determine the relative distance between the agent and obstacles;
[0122] If the relative distance is determined to be equal to the preset relative distance threshold, the braking performance of the intelligent agent is tested to obtain braking performance test data.
[0123] The vehicle testing device 50 based on intelligent agents provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described again in this embodiment.
[0124] Figure 6 A schematic diagram of the structure of the electronic device provided in this application. Figure 6As shown, the electronic device 60 provided in this embodiment includes at least one processor 601 and a memory 602. Optionally, the device 60 further includes a communication component 603. The processor 601, memory 602, and communication component 603 are connected via a bus 604.
[0125] In a specific implementation, at least one processor 601 executes computer execution instructions stored in memory 602, causing at least one processor 601 to perform the above-described method.
[0126] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.
[0127] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.
[0128] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.
[0129] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.
[0130] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.
[0131] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.
[0132] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.
[0133] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.
[0134] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.
[0135] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0136] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0137] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0138] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.
[0139] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.
Claims
1. A vehicle testing method based on intelligent agents, characterized in that, The intelligent agent represents the vehicle in the simulation environment, and the method includes: Repeat the following steps until the overall test of the agent is completed: Based on the scene data of the intelligent agent in the current simulation environment, scene information and intelligent agent information are determined; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment; Based on the scene information and the agent information, when controlling the agent to move, the performance test data of the agent is obtained, and the position information of the agent after the movement is determined. The current simulation environment is updated based on the position information of the moved agent.
2. The method according to claim 1, characterized in that, Based on the scene information and the agent information, when controlling the agent to move, performance test data of the agent is obtained, including: Based on the scene information and the agent structure parameters in the agent information, a control signal for the agent is generated; wherein, the agent structure parameters characterize data on the vehicle structure of the agent; Based on the control signal, the intelligent agent is controlled to move, and while controlling the intelligent agent to move, the performance test data of the intelligent agent is obtained.
3. The method according to claim 1, characterized in that, The scene data also includes collision information; the agent information includes agent structural parameters and agent control parameters, wherein the agent structural parameters represent data on the vehicle structure of the agent, and the agent control parameters represent control parameters of the agent's driving process. The collision information represents the information of obstacles encountered by the intelligent agent in the current simulation environment; Determining the position information of the agent after movement includes: Based on a preset intelligent agent dynamics solver, the collision information and the intelligent agent information are processed to determine the position information of the moved intelligent agent.
4. The method according to claim 1, characterized in that, Based on the position information of the moved agent, update the current simulation environment, including: Based on the location information of the moved agent, the current simulation environment, and the agent information, the scene data in the current simulation environment is updated to update the current simulation environment.
5. The method according to claim 4, characterized in that, The method further includes: The updated scene data in the current simulation environment is transmitted to the refeeding simulation engine that processes the intelligent agent.
6. The method according to any one of claims 1-5, characterized in that, The method further includes: Based on a pre-established communication channel, the scene data of the intelligent agent in the current simulation environment is acquired; wherein, the communication channel represents the communication channel between the agent and the real vehicle scene data storage library, and the real vehicle scene data storage library stores scene data for each real scene. The acquired scene data is then denoised.
7. The method according to any one of claims 1-5, characterized in that, The performance test data includes braking performance test data. Based on the scene information and the agent information, when controlling the agent to move, the performance test data of the agent is obtained, including: Based on the location information of the intelligent agent, the relative distance between the intelligent agent and the obstacle is determined; If the relative distance is determined to be equal to a preset relative distance threshold, the braking performance of the intelligent agent is tested to obtain the braking performance test data.
8. A vehicle testing device based on intelligent agents, characterized in that, The intelligent agent represents the vehicle in the simulation environment, and the device includes: Repeat the steps for each of the following modules until the overall test of the agent is completed: The first determining module is used to determine scene information and agent information based on scene data of the agent in the current simulation environment; wherein, the scene data represents the scene data of the vehicle in the real environment corresponding to the current simulation environment; The control module is used to obtain the performance test data of the intelligent agent when controlling the intelligent agent to move, based on the scene information and the intelligent agent information. The second determining module is used to determine the position information of the moved agent; The update module is used to update the current simulation environment based on the position information of the moved agent.
9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.