Vehicle-road cooperation virtual-real fusion simulation test method, device, equipment and medium

Abstract

The invention discloses a vehicle-road cooperation virtual-real fusion simulation test method, device, equipment and medium, and the method comprises the steps: setting a test event in a virtual simulation environment, generating virtual sensor detection data, sending the data to a cloud control platform, carrying out the fusion processing, generating a standard vehicle-road cooperation message, and issuing the message to a real road side unit, and the real road side unit forwards the message to the real vehicle-mounted unit, the real vehicle-mounted unit analyzes the message, generates a vehicle control instruction and returns the vehicle control instruction to the virtual simulation environment, and the state of the virtual vehicle model is updated according to the vehicle control instruction in the virtual simulation environment and response verification is completed. According to the method, a virtual-real combined vehicle-road cooperation test link is constructed under a unified framework, and a complete closed loop from virtual sensor sensing, cloud fusion processing, roadside forwarding, vehicle-mounted analysis to virtual vehicle response is realized, so that the test coverage rate and controllability of a vehicle-road cooperation function in a complex traffic scene are effectively improved; and the test efficiency and the verification precision are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traffic control and simulation testing, and particularly relates to a virtual-real fusion simulation testing method, device, equipment and medium for vehicle-road cooperation. BACKGROUND

[0002] With the development of intelligent networked vehicle technology, the vehicle-road cooperation system gradually becomes an important direction to improve road traffic safety and operation efficiency. By deploying various sensing devices and communication units on the vehicle side and the road side, vehicles can obtain real-time road environment information, traffic event information and cooperative control instructions, thereby improving the processing capability for complex traffic scenarios. However, in the development of related technologies, there are still obvious deficiencies in the virtual-real fusion vehicle-road cooperation testing.

[0003] In the prior art, the testing of vehicle-road cooperation functions usually relies on real vehicles, real roads and real traffic environments. Due to the problems such as limited site scale, complex testing conditions and difficulty in reproducing extreme scenarios, it is difficult to comprehensively cover various vehicle-road cooperation event scenarios. For sudden traffic events, multi-vehicle conflict behaviors and road facility abnormalities, there is a lack of safe, controllable and reproducible testing means. In addition, real road testing relies on a large number of hardware deployments, which is high in cost and long in cycle, and is difficult to support continuous iteration and large-scale verification requirements.

[0004] On the other hand, existing simulation testing methods mostly focus on vehicle-side single simulation or road-side single simulation, and often lack cooperative verification mechanisms among vehicles, road-side units and cloud control platforms, and cannot truly reflect the information flow and control link among devices, networks and roles. The existing simulation environment cannot simultaneously simulate virtual sensor output, road-side device processing logic and vehicle-side execution behavior, so that the vehicle-road cooperation full-link testing is incomplete, and the entire closed-loop process from sensing information to control execution cannot be verified.

[0005] In addition, traditional simulation systems mostly rely on one-way data flow, lack of interactive links between real vehicle units, real road-side units and cloud control platforms, and cannot accurately verify the transmission reliability and delay influence of control instructions between different devices. At the same time, the response of vehicle control instructions in the virtual environment lacks systematic evaluation methods, and it is difficult to verify whether the vehicle behavior under the control link conforms to the preset traffic event. SUMMARY

[0006] The main purpose of the present application is to provide a virtual-real fusion simulation testing method, device, equipment and storage medium for vehicle-road cooperation, aiming at solving the technical problem that the prior art cannot realize vehicle-road cooperation full-link virtual-real fusion testing covering perception, fusion, communication and control execution, so as to verify the effectiveness and reliability of the complete control link.

[0007] To achieve the above object, the application provides a kind of vehicle-road cooperation virtual-real fusion simulation test method, comprising: Deploy real roadside unit, real vehicle-mounted unit and cloud control platform, and establish the data transmission link between the real roadside unit, the real vehicle-mounted unit, the cloud control platform and virtual simulation environment; Setting test event in virtual simulation environment; In response to the test event, virtual sensor detection data is generated in the virtual simulation environment through the virtual sensor model, and the virtual sensor detection data is sent to the cloud control platform through the data transmission link between virtual simulation environment and cloud control platform; The virtual sensor detection data received by the cloud control platform is fused to generate a standard vehicle-road cooperation message, and the standard vehicle-road cooperation message is sent to the real roadside unit; The standard vehicle-road cooperation message is forwarded to the real vehicle-mounted unit by the real roadside unit; The real vehicle-mounted unit parses the standard vehicle-road cooperation message to generate a vehicle control instruction, and sends the vehicle control instruction to the virtual simulation environment through the data transmission link between the real vehicle-mounted unit and the virtual simulation environment; The state of the virtual vehicle model is updated according to the vehicle control instruction in the virtual simulation environment, and the response of the virtual vehicle model to the test event is verified.

[0008] Further, to achieve the above object, the application provides a kind of vehicle-road cooperation virtual-real fusion simulation test device, comprising: Communication link configuration module, for deploying real roadside unit, real vehicle-mounted unit and cloud control platform, and establishing the data transmission link between the real roadside unit, the real vehicle-mounted unit, the cloud control platform and virtual simulation environment; Test event configuration module, for setting test event in virtual simulation environment; Virtual sensor data generation module, for generating virtual sensor detection data in the virtual simulation environment through the virtual sensor model in response to the test event, and sending the virtual sensor detection data to the cloud control platform through the data transmission link between virtual simulation environment and cloud control platform; Cloud fusion and message generation module, for fusing the virtual sensor detection data received by the cloud control platform to generate a standard vehicle-road cooperation message, and sending the standard vehicle-road cooperation message to the real roadside unit; Message forwarding module, for forwarding the standard vehicle-road cooperation message to the real vehicle-mounted unit by the real roadside unit; The vehicle-mounted instruction generation module is configured to parse the standard vehicle-road cooperation message by the real vehicle-mounted unit, generate a vehicle control instruction, and send the vehicle control instruction to the virtual simulation environment through a data transmission link between the real vehicle-mounted unit and the virtual simulation environment. The virtual vehicle response verification module is configured to update a state of the virtual vehicle model in the virtual simulation environment according to the vehicle control instruction, and verify a response of the virtual vehicle model to the test event.

[0009] Further, to achieve the above-mentioned purpose, the present application also provides a computer device, which comprises a memory, a processor, and a vehicle-road cooperation virtual-real fusion simulation test program stored in the memory and executable on the processor, and the vehicle-road cooperation virtual-real fusion simulation test program implements the steps of the vehicle-road cooperation virtual-real fusion simulation test method when executed by the processor.

[0010] Further, to achieve the above-mentioned purpose, the present application also provides a computer readable storage medium, which stores a vehicle-road cooperation virtual-real fusion simulation test program, and the vehicle-road cooperation virtual-real fusion simulation test program implements the steps of the vehicle-road cooperation virtual-real fusion simulation test method when executed by a processor.

[0011] Beneficial effects: The present application discloses a vehicle-road cooperation virtual-real fusion simulation test method, device, equipment and medium, which comprises the following steps: setting a test event in a virtual simulation environment, generating virtual sensor detection data and sending the virtual sensor detection data to a cloud control platform, generating a standard vehicle-road cooperation message through fusion processing and delivering the standard vehicle-road cooperation message to a real roadside unit, forwarding the standard vehicle-road cooperation message to a real vehicle-mounted unit by the real roadside unit, analyzing the message by the real vehicle-mounted unit to generate a vehicle control instruction and returning the vehicle control instruction to the virtual simulation environment, updating a virtual vehicle model state in the virtual simulation environment according to the vehicle control instruction and completing response verification. The present application constructs a virtual-real combined vehicle-road cooperation test link under a unified framework, realizes a complete closed loop from virtual sensor perception, cloud fusion processing, roadside forwarding, vehicle-mounted analysis to virtual vehicle response, thereby effectively improving test coverage, controllability and reproducibility of vehicle-road cooperation function in a complex traffic scene, reducing dependence on real road resources, and improving test efficiency and verification accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0012] The present application will be further described below in combination with the drawings and embodiments, and the drawings show: Figure 1 An application environment schematic diagram of the vehicle-road cooperation virtual-real fusion simulation test method in an embodiment of the present application; Figure 2 A flowchart of an embodiment of the vehicle-road cooperation virtual-real fusion simulation test method of the present application; Figure 3 A functional module schematic diagram of a preferred embodiment of the vehicle-road cooperation virtual-real fusion simulation test device of the present application; Figure 4 A structural schematic diagram of a computer device in an embodiment of the present application; Figure 5 Another structural schematic diagram of a computer device in an embodiment of the present application; Figure 6 Another flowchart of a vehicle-road cooperation virtual-real fusion simulation test method in an embodiment of the present application. DETAILED DESCRIPTION

[0013] It should be understood that the specific embodiments described herein are merely intended to explain the present application and are not intended to limit the present application.

[0014] The vehicle-road cooperation virtual-real fusion simulation test method provided by the embodiments of the present application can be applied in an application environment such as Figure 1 , wherein a client communicates with a server through a network. The server can set a test event in a virtual simulation environment through the client, generate virtual sensor detection data and send it to a cloud control platform, generate a standard vehicle-road cooperation message through fusion processing and distribute it to a real roadside unit, which forwards it to a real vehicle-mounted unit, the real vehicle-mounted unit analyzes the message to generate a vehicle control instruction and returns it to the virtual simulation environment, and the virtual simulation environment updates the virtual vehicle model state according to the vehicle control instruction and completes response verification. The present application constructs a virtual-real combined vehicle-road cooperation test link under a unified framework, realizes a complete closed loop from virtual sensor perception, cloud fusion processing, roadside forwarding, vehicle-mounted analysis to virtual vehicle response, thereby effectively improving the test coverage, controllability and reproducibility of vehicle-road cooperation functions in complex traffic scenarios, reducing the dependence on real road resources, and improving the test efficiency and verification accuracy. The client can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices. The server can be implemented by an independent server or a server cluster composed of multiple servers. The present application will be described in detail through specific embodiments.

[0015] Please refer to Figure 2 , Figure 2 A flowchart of an embodiment of the vehicle-road cooperation virtual-real fusion simulation test method provided by the present application. It should be noted that although a logical sequence is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that described herein.

[0016] As shown in Figure 2 , the vehicle-road cooperation virtual-real fusion simulation test method proposed by the present application includes the following steps: S10, deploy a real roadside unit, a real vehicle-mounted unit and a cloud control platform, and establish a data transmission link between the real roadside unit, the real vehicle-mounted unit, the cloud control platform and a virtual simulation environment; In the embodiment, deploying the real roadside unit, the real vehicle-mounted unit and the cloud control platform involves connecting different types of physical devices to the same communication system. The real roadside unit is usually installed in a roadside pole or a fixed structure, and has a communication module, a processing module and a protocol adaptation module, which are used to undertake the functions of message relay and environmental information pushing in the network. The real vehicle-mounted unit is installed in the vehicle and interacts with the outside world through a positioning module, a communication module and a message analysis module, and undertakes the tasks of receiving and returning information at the vehicle end. The cloud control platform is deployed in a central server or an edge computing node, and processes multi-source data from the roadside and the vehicle-mounted unit through a computing resource, a network access module and a data management module.

[0017] Establishing the data transmission link between the three and the virtual simulation environment requires building a communication mapping relationship across devices and networks. The data transmission link is realized through a combination of physical networks, wireless networks or virtual networks, and unifies the real devices and the virtual environment in the same data interaction system. During the link establishment process, the communication protocol, address mapping and data format conversion need to be configured so that the data generated by the real device can be recognized in the virtual environment, and the data output by the virtual environment can also be received and processed by the real device. Through the establishment of the link, the real device can be used as an input or output node of the virtual environment, so that data can flow continuously between the real hardware and the virtual environment.

[0018] By deploying the real roadside unit, the real vehicle-mounted unit and the cloud control platform and building the data transmission link for the virtual simulation environment, the real hardware can participate in the virtual test, realize synchronous interaction across networks and devices, and make the virtual environment reflect the processing behavior of the real hardware, thereby improving the accuracy and reproducibility of the verification and reducing the cost of building a real test scene.

[0019] Further, before establishing the data transmission link, the method further comprises the steps of: S001, building a virtual simulation environment, the virtual simulation environment comprising a virtual vehicle model and a virtual sensor model; In this embodiment, the construction of the virtual simulation environment involves a joint creation process of three levels of spatial modeling, logical modeling, and interactive modeling. The virtual simulation environment is used to carry vehicle behavior calculation, sensor data generation, and event triggering logic, and therefore needs to have spatiotemporal continuity, data-driven, and interactivity. The construction of the virtual environment usually starts from a three-dimensional simulation engine or a digital twin framework, and generates a runnable digital space through basic objects such as terrain grid, road structure, lane topology, traffic facilities, and environment light source. For example, the geometric structure of the road network can be generated by importing GIS data, or a diversified road form can be generated by parameterized modeling. Dynamic elements of the environment include wind field, light, rain, snow, etc. These parameters affect the detection characteristics of the sensor, so when constructing the virtual environment, environmental attribute parameters are added to enable the environment to dynamically adjust the physical behavior.

[0020] The virtual vehicle model is composed of a vehicle dynamics model, a vehicle control interface, and a vehicle appearance structure. The vehicle dynamics model is used to describe the motion law of the vehicle, including longitudinal speed change calculation, lateral steering angle change, tire ground mechanics relationship, and vehicle body attitude change. When implementing this model, a lightweight dynamic model based on kinematics equation can be used, or a multi-rigid-body model can be used to achieve more detailed vehicle dynamic response. The vehicle control interface is used to receive parameters such as acceleration, steering angle, and brake amount in the vehicle control instruction, and apply these parameters to the dynamics model to drive the vehicle motion. The vehicle appearance structure is mainly used to adapt to the sampling process of the sensor, so that the virtual sensor can identify the vehicle contour, shape boundary, and material characteristics.

[0021] The virtual sensor model is used to simulate the detection behavior of the real sensing device in the virtual environment. The virtual sensor model includes perception field configuration, detection distance range, resolution control, sampling frequency setting, and noise model injection elements. For example, the laser radar model generates a three-dimensional point cloud by emitting point cloud rays in space and calculating the intersection with the environment object; the millimeter wave radar model simulates distance, speed, and angle information based on the reflection characteristics of the target; the camera model obtains an image frame through a rendering engine, and adds optical characteristics such as distortion, exposure, and blur degree. The virtual sensor model needs to be linked with the virtual vehicle model when constructing, for example, the camera needs to be aligned with the vehicle body coordinate system, and the laser radar needs to be simulated with the installation inclination of the vehicle, to ensure that the sensor output structure is consistent with the real hardware.

[0022] The construction of the virtual simulation environment not only includes the generation of digital objects, but also includes the establishment of a running mechanism. The running mechanism includes a time stepping system, a physical simulation system, and an event triggering mechanism. The time stepping system controls the update frequency of the environment simulation, so that the vehicle dynamics calculation, sensor sampling, and event judgment are run under the same time reference. The physical simulation system is used to support behaviors such as collision detection, object motion, and light propagation, so that the environment has the ability to have real physical responses. The event triggering mechanism judges whether the vehicle position, speed, or scene change meets the triggering condition according to the preset condition, and enables the sensor model to generate corresponding data.

[0023] Through the joint construction of the above elements, the virtual simulation environment can form an extensible, interactive, and vehicle-road cooperative data flow carrying running space, so that a closed-loop linkage relationship is formed between the vehicle, the sensor, and the environment, and the data generation and control execution requirements in the subsequent virtual-real fusion test process are met.

[0024] By constructing a virtual simulation environment containing a vehicle motion model and a sensor detection model, the vehicle state change and perception output can be obtained in the digital space at the same time, so that the subsequent test can complete scene generation and data acquisition without relying on real roads and real hardware, thereby reducing the test risk, improving the construction efficiency, and realizing the simulation verification ability of multiple scenes and repeatability.

[0025] S20, setting a test event in the virtual simulation environment; In this embodiment, setting a test event in the virtual simulation environment can be understood as defining a condition set for triggering traffic interaction in the constructed simulation scene. The test event can be configured based on a traffic behavior model, a road facility attribute, or a vehicle interaction rule, for triggering a target condition in a specific road scene. This processing can load an event template through a scene editing module, configure the triggering condition, target object, and duration through a parameter setting module, and bind the occurrence position and execution mode of the event on the simulation time axis, so that the test event has repeatability and controllability.

[0026] The preset event type can be selected through a scene editing interface, and the triggering speed threshold, position point, and involved object of the event can be set in a parameter panel, so that the event is automatically generated when the simulation time reaches the condition. The event instance can be added to the simulation environment as an independent object, loaded by an event scheduler before simulation running, and triggered according to the definition during simulation execution. The user can preview the running track of the event after setting it, to ensure that it meets the test requirements.

[0027] The embodiment can accurately reproduce the traffic state to be verified in the scene by setting a test event in a virtual simulation environment. Through controllable trigger conditions and repeatable event definitions, a specific traffic scene can be simulated without real road testing, so that the subsequent testing process is more stable, the cost is lower, and more boundary conditions can be covered.

[0028] S30, in response to the test event, generating virtual sensor detection data in the virtual simulation environment through the virtual sensor model, and sending the virtual sensor detection data to the cloud control platform through a data transmission link between the virtual simulation environment and the cloud control platform; In the embodiment, after detecting the test event in the virtual simulation environment, the simulation system starts the virtual sensor model according to the event trigger signal. The virtual sensor model generates detection data consistent with the real sensor form by sampling the simulation scene object, spatial position and motion state. The detection data can include target position, speed, perception region point cloud, image frame or radar echo data according to the model setting, and the format is consistent with the real hardware output, so that the subsequent processing module can be directly used. The generated data is encoded and packaged through the established data transmission link and sent to the cloud control platform, so that the cloud control platform can obtain input data consistent with the real acquisition process in the simulation environment.

[0029] In the simulation system, a plurality of types of virtual sensor models are preconfigured, including camera models, radar models and lidar models. After the test event is triggered, the simulation engine queries the current event coverage area and synchronously calls the corresponding sensor model to generate data. Taking the camera model as an example, image frames can be generated based on the set field of view angle, resolution and frame rate; taking the radar model as an example, a virtual echo matrix can be generated according to the set transmission frequency and scanning angle. The generated data is packaged into a unified data structure through the transmission interface of the simulation system, and is sent to the cloud control platform through the data transmission link.

[0030] The embodiment can generate data consistent with the real acquisition format by calling the virtual sensor model after the test event is triggered, and send the data to the cloud control platform through the data transmission link, so that the cloud control platform can obtain stable, controllable and repeatable data input in the simulation test stage. By replacing the real hardware with the virtual sensor model, various working conditions can be simulated without consuming real device resources, improving the testing flexibility, reducing the test cost, and improving the coverage ability of complex scenes.

[0031] S40, fusing the received virtual sensor detection data through the cloud control platform to generate a standard vehicle-road cooperation message, and sending the standard vehicle-road cooperation message to the real roadside unit; In this embodiment, after the cloud control platform receives the virtual sensor detection data transmitted by the virtual simulation environment, it first parses the data according to the communication format, loads the target detection information, environment state information and space-time index information contained therein into the data processing module in a structured form. The parsing link usually relies on the field dictionary and parsing logic, which is used to decompose the data into time, position, direction, target category, motion characteristics and other basic elements. After completing the parsing, the cloud control platform synchronizes the data from different virtual sensor models in time sequence through the fusion processing logic, realizes the spatial correlation based on the coordinate calibration parameters, and generates a unified environment expression according to the data confidence and feature priority. The environment expression obtained by fusion is used to construct a standard car-road cooperation message, and the message content includes the motion state of the object element and the change trend of the environment element. The cloud control platform completes field filling, content organization and verification generation according to the cooperative communication format, and sends the message through the communication link established with the real roadside unit, so that the real roadside unit obtains structured cooperative information from the virtual scene.

[0032] In actual deployment, the cloud control platform can integrate a time reference calibration module to synchronize data from various virtual sensor models according to a unified reference time axis, avoiding inconsistent scene interpretation due to differences in update frequency between sensor models. For spatial processing, the pixel coordinates of virtual cameras, polar coordinates of virtual radars and three-dimensional coordinates of virtual laser point clouds can be unified to the same reference coordinate system through pre-set coordinate mapping parameters in the virtual scene. The fusion logic can use a weighted feature superposition method to jointly process data from virtual laser models with high-precision distance perception capability and virtual vision models with strong classification and recognition capability. The fused environment content is organized into a standardized message structure through the car-road cooperation communication rule, and is encoded, verified and reported. The message is sent to the real roadside unit through the communication interface of the cloud control platform, so that the real roadside unit can receive car-road cooperation information in the same format as the actual measured environment under virtual test conditions.

[0033] This embodiment generates a standard car-road cooperation message by fusing and processing virtual sensor detection data through the cloud control platform, forming a data-level consistent cooperative communication path between the virtual environment and the real road infrastructure. Fusion processing can improve the completeness and consistency of virtual perception data, thereby simulating the information interaction process of the real traffic scene without the participation of real vehicles.

[0034] S50, forwarding the standard car-road cooperation message to the real vehicle-mounted unit through the real roadside unit; In this embodiment, after the real roadside unit receives the standard car-road cooperation message, it needs to parse, verify and forward prepare the message through the internal communication processing logic. After the message enters the roadside communication module, first of all, according to the message structure rule, the field check and content check are performed to confirm that the message meets the forwarding conditions in terms of field format, data integrity and content sequence. In the parsing process, the roadside unit will disassemble the message into vehicle state data items, road event data items and auxiliary index data items, and load these data items into the local cache. After completing the parsing, a short-range communication link is established with the real vehicle-mounted unit through the pre-configured communication interface. The short-range communication link is usually based on the wireless direct connection protocol, which is conducive to stable data transmission within a limited range. The roadside unit re-packages the message content according to the transmission rules of the link and executes the message sending according to the queue scheduling logic, so that the standard car-road cooperation message can be received by the vehicle-mounted unit in time. During the forwarding process, the roadside unit can adjust the sending order according to the message priority or time stamp order to ensure the timeliness and continuity of the forwarding process.

[0035] In actual operating environment, the real roadside unit can deploy multiple communication modules to adapt to the access capability of the vehicle terminal. For example, it can complete forwarding based on wireless direct connection communication mode, or send based on the polling communication mechanism between roadside and vehicle. For different communication modes, the roadside unit can configure a buffer queue to handle high-concurrency messages. When the generation frequency of standard car-road cooperation messages is high, the roadside unit can reduce the single transmission load by segmenting and packaging, thereby improving the overall transmission stability. In some road environments, if there is an obstruction or weak coverage area between the vehicle position and the roadside unit, the roadside unit can enable power adjustment strategy to improve link quality or switch through the standby communication module. For field verification, the message content can be verified for completeness using a check sequence or hash digest, and then the valid message is packaged into a structure format that can be parsed by the vehicle, and forwarding is performed.

[0036] This embodiment transmits the standard car-road cooperation message through the real roadside unit, so that the environmental information generated by the virtual scenario can enter the vehicle side in a real communication mode, thereby simulating the process of the vehicle receiving cooperative information in a real road environment. This forwarding mode completes the vehicle's parsing logic verification and behavior response verification without relying on real traffic events, reduces the dependence on external scenarios in the testing process, improves the real consistency of the verification link, and helps to build a reproducible, scalable and low-risk linkage test environment.

[0037] S60, parsing the standard car-road cooperation message through the real vehicle-mounted unit, generating vehicle control instructions, and sending the vehicle control instructions to the virtual simulation environment through a data transmission link between the real vehicle-mounted unit and the virtual simulation environment; In this embodiment, the real vehicle-mounted unit will complete field parsing and semantic disassembly through the internal message processing program after receiving the standard car-road cooperation message. The message usually contains vehicle target position, road event state, path guidance indication and cooperative behavior trigger information. After receiving the message, the vehicle-mounted unit inputs it into the message parsing module and parses it field by field according to the message format rule. The parsing process includes structure format unpacking, field content verification, instruction intention judgment and time sequence label identification operations to ensure that the target intention contained in the message can be accurately extracted. After completing the parsing, the vehicle-mounted unit generates vehicle control instructions according to the behavior prompt content in the message. The vehicle control instructions include acceleration and deceleration parameters, direction adjustment parameters, brake trigger parameters or path offset correction parameters, and the control instructions are combined through the internal control logic module to make the control instructions executable in time dimension and action dimension. The generated control instructions need to be returned to the virtual simulation environment through the data transmission link, so the vehicle-mounted unit will complete the packaging, encoding and sending preparation of the control instructions again, package them into a control instruction format that can be recognized by the virtual simulation environment, and complete the sending through the existing data transmission link.

[0038] This embodiment completes the parsing of the standard car-road cooperation message and the generation of the control instruction through the vehicle-mounted unit, so that the decision logic of the vehicle side can participate in the complete response link of the virtual scene, and the virtual scene has the real vehicle-mounted processing characteristics. After the vehicle control instruction is returned to the virtual simulation environment through the data transmission link, the virtual vehicle model can be driven to perform actions in the virtual environment, thereby forming a closed-loop control link.

[0039] S70, updating the state of the virtual vehicle model in the virtual simulation environment according to the vehicle control instruction, and verifying the response of the virtual vehicle model to the test event.

[0040] In this embodiment, the virtual simulation environment will input the vehicle control instruction into the internal motion solving module of the virtual vehicle model after receiving the vehicle control instruction, which is used to drive the dynamic state change of the vehicle model. The vehicle control instruction usually contains speed adjustment parameters, direction adjustment parameters, brake execution parameters or path correction parameters, which correspond to the speed state, position state and attitude state of the virtual vehicle model. The virtual simulation environment will modify the dynamic equation input of the vehicle model according to the parameter value of the control instruction, so that the three-dimensional position, speed vector and attitude angle of the vehicle model will change continuously with time. The updated state is recorded by the state recording module of the simulation environment, which is used to build the dynamic trajectory of the virtual vehicle model.

[0041] After the state update of the virtual vehicle model is completed, the virtual simulation environment compares the latest state with the trigger condition, preset event rule or scene parameter contained in the test event, so as to determine whether the response of the vehicle meets the expectation. For example, the test event can contain obstacle position, environmental signal trigger, road geometry change or traffic rule change, and the virtual simulation environment can determine whether the response of the vehicle meets the requirements of the event according to the motion trajectory, speed change trend or direction adjustment behavior of the vehicle model after the state update. Through the corresponding relationship between the state data and the event data, the verification of the response behavior of the vehicle model can be completed.

[0042] In this embodiment, the virtual simulation environment updates the state of the virtual vehicle model according to the vehicle control instruction, so that the test link can completely verify the behavior response of the vehicle to the test event, and does not rely on the real vehicle to perform actual actions. The change of the vehicle state in the virtual environment can be accurately recorded and quickly verified, and forms a closed loop response link with the test event, thereby improving the restoration degree of the virtual test scene to the real behavior and laying a stable data foundation for subsequent simulation analysis.

[0043] In one embodiment, the above step S20 comprises: S201, setting a vehicle-road cooperation event, configuring an obstacle area, a construction area and a special vehicle in the virtual simulation environment; S202, setting a vehicle-vehicle cooperation event, configuring a multi-vehicle interaction scene and a conflict event in the virtual simulation environment; S203, setting an event trigger condition according to the configuration of the vehicle-road cooperation event and the vehicle-vehicle cooperation event; S204, configuring an event influence parameter based on the event trigger condition; S205, integrating the vehicle-road cooperation event, the vehicle-vehicle cooperation event, the event trigger condition and the event influence parameter to generate test event configuration information; S206, loading the test event configuration information into an event management system of the virtual simulation environment; S207, starting a test event monitoring process.

[0044] In this embodiment, when setting a test event in the virtual simulation environment, a group of event entities that can be scheduled and monitored need to be defined based on a unified three-dimensional road scene model and a traffic element model. The test event is a configurable scene unit here, which describes the trigger condition, spatial range, participating object and duration through parameterization, and is used to drive the vehicle behavior, roadside state and traffic flow state to change as expected. In order to realize the verification process of vehicle-road integration, the event configuration covers both vehicle-road cooperation events and vehicle-vehicle cooperation events.

[0045] The vehicle-road coordination event is implemented by calibrating the obstacle area, construction area and special vehicle configuration in the simulation road network. The obstacle area can be represented in the form of polygon or grid in the road coordinate system, recording the obstacle type, occupied lane, length, width, height and effective time, which is used to simulate falling objects, malfunctioning vehicles, temporary closures and other situations. The construction area adds construction phase, operation boundary, diversion area, speed limit interval and temporary traffic sign information based on the obstacle area, which is used to describe the decrease of road capacity and the reduction of passable range in different time periods. Special vehicles are used to represent vehicles with different passing priority or behavior patterns, such as wrecker vehicles, engineering vehicles, emergency vehicles, etc., whose priority, typical trajectory and different right-of-way rules from ordinary vehicles are specified by separate vehicle attribute fields. After loading these configurations, the virtual simulation environment can dynamically generate or update the obstacle, construction equipment and special vehicle entities in the scene within the specified time window.

[0046] The vehicle-vehicle coordination event is constructed by configuring multi-vehicle interaction scenarios and conflict events to establish the spatial and temporal coupling relationship between vehicles. The multi-vehicle interaction scenario is achieved by assigning initial lane position, initial speed, acceleration and deceleration strategy, lane changing strategy and following parameters to multiple virtual vehicles, so that the vehicles form a controllable dense traffic structure in the intersection area or merging section. The conflict event is achieved by defining the potential collision point, conflict time window and expected trajectory of two or more vehicles in the time-space domain, so that the vehicles produce high-risk scenarios such as rush, strong merging and sudden deceleration near intersections, ramps or curves. By jointly modeling multi-vehicle interaction scenarios and conflict events, the virtual simulation environment can generate a variety of complex vehicle-vehicle coordination behaviors under the same road structure, to investigate the response capability of the road end and the vehicle end under different risk levels.

[0047] The event triggering condition is used to control the activation time and activation mode of the test event on the simulation time axis, which is defined depending on the configuration results of the vehicle-road coordination event and the vehicle-vehicle coordination event. The triggering condition can be based on time, such as activating the event when the simulation time reaches a certain time; it can also be based on space, such as triggering the event when the vehicle enters the boundary of a certain obstacle area or construction area; it can also be based on state, such as starting the event when the vehicle flow density reaches a preset threshold, the vehicle spacing decreases to a critical value or the roadside device detects a specific sign change. The virtual simulation environment monitors the vehicle running state, road state and roadside device detection results in real time, expresses these conditions as logical expressions or rule nodes, and marks the corresponding event as activated when the conditions are met.

[0048] The event influence parameters are used to describe the quantitative effects of events on the traffic environment and vehicle behavior, including speed limit, lane closure ratio, available lane set, visibility decay coefficient, perception noise intensity, warning range, control suggestion type, etc. The influence parameters of cooperative vehicle infrastructure events can limit the maximum speed of vehicles passing through the construction area, adjust the lane availability status, change the visible area and measurement error distribution of virtual sensors; the influence parameters of cooperative vehicle-vehicle events can change the safety time interval, minimum distance and acceleration upper limit of the car-following model, thereby forming a more tense interaction relationship in the conflict area. These parameters are associated with specific events in a structured configuration form, so that each event activation can produce repeatable environmental disturbances and behavior constraints in the simulation environment.

[0049] The test event configuration information is a structured description of cooperative vehicle infrastructure events, cooperative vehicle-vehicle events, event trigger conditions and event influence parameters. This configuration information is usually organized in a unified configuration format, such as based on a hierarchical object model or a key-value pair structure, which encapsulates event identification, event type, spatial range, time constraint, participating entity, trigger logic and influence parameter. The event management system of the virtual simulation environment will establish a life cycle management entry for each event after loading the test event configuration information, responsible for state switching, condition listening and influence application during simulation running. The loading process includes parsing configuration content, checking parameter validity, establishing index structure and registering events to the monitoring queue, so that the event management system can quickly access related configurations in the simulation main loop.

[0050] The test event monitoring process runs in the event management system for a long time, and scans the state variables in the simulation environment through periodic or event-driven methods. The monitoring process checks vehicle position, speed, road occupancy and roadside device detection results, and matches them with the trigger conditions of each test event. When the conditions are met, the event state is updated to the active state, and the simulation environment is adjusted according to the event influence parameters, such as modifying road signs, adjusting vehicle available paths or updating virtual sensor output constraints. When the event enters the completion or invalid state, the monitoring process records the trigger time, duration and actual influence range, providing a data basis for subsequent result analysis. Through this monitoring mechanism, the virtual simulation environment can uniformly manage and execute different types of test events under unified configuration.

[0051] The embodiment can build various reproducible complex traffic scenes in the same environment by structured configuration of car-road cooperation events and car-car cooperation events in a virtual simulation environment, clear trigger conditions, adjustable event influence parameters, and loadable test event configuration information, so that the test process no longer depends on real roads and real traffic flow, and rich scene coverage of car-road cooperation functions can be completed, and an event lifecycle control mechanism that can be uniformly managed is formed through an event management system and a monitoring process, thereby providing a stable, repeatable, and extensible event scene basis for subsequent perception verification, control verification, and closed-loop testing.

[0052] In one embodiment, the above step S30 comprises: S301, when detecting that the trigger of the test event is detected, activating a virtual sensor model in the virtual simulation environment; S302, simulating a laser radar perception function by the virtual sensor model to generate three-dimensional point cloud data; S303, simulating a millimeter wave radar perception function by the virtual sensor model to generate target trajectory data; S304, simulating a camera perception function by the virtual sensor model to generate image acquisition data; S305, integrating the three-dimensional point cloud data, the target trajectory data, and the image acquisition data to generate virtual sensor detection data; S306, encoding the virtual sensor detection data into a standard communication protocol format; S307, sending the encoded virtual sensor detection data to the cloud control platform through a data transmission link between the virtual simulation environment and the cloud control platform.

[0053] In the embodiment, in response to a test event trigger, an event management module in the virtual simulation environment first generates a trigger signal according to the previously configured test event monitoring conditions. The trigger signal can be composed of time conditions, space conditions, or traffic state conditions, such as detecting that a virtual vehicle trajectory enters a specified conflict area, the simulation time reaches a preset time, or a virtual road side unit detects an event marker. The event management module writes the trigger signal into a unified state bus or message queue after determining that the trigger condition is met, and the sensor management module polls or subscribes to the trigger signal. When the sensor management module detects that the trigger state changes from not triggered to triggered, the activation operation of the virtual sensor model is performed, including allocating computing resources for the laser radar model, the millimeter wave radar model, and the camera model, resetting the internal state, starting the sampling timer and aligning the sampling period with the simulation time step, so that the virtual simulation environment enters the event-driven perception data generation phase.

[0054] When generating three-dimensional point cloud data, the virtual sensor model performs ray casting operations on three-dimensional road scene and traffic participant geometry models through a laser radar simulation submodule. The simulation submodule generates a series of scanning time points on the simulation time axis according to the set number of beams, vertical field of view, horizontal scanning angle velocity, and measurement frequency; at each scanning time point, intersection operations are performed for each laser beam to calculate the intersection position and distance between the laser beam and the scene geometry. Combined with the reflection material property and the incident angle information, a point cloud data set containing spatial coordinates, distance, reflection intensity, and time stamp is generated, and ranging noise, missing detection probability, and occlusion effects can be superimposed as needed to make the point cloud data close to the actual laser radar output in terms of spatial distribution and density.

[0055] The millimeter wave radar simulation submodule is used to generate target trajectory data by selecting a moving target entity in the simulation scene and performing target-level modeling to construct the output content. The submodule first calculates the radial distance, radial velocity, and azimuth angle relative to the millimeter wave radar installation position based on the real trajectory of a vehicle, pedestrian, or other moving body on the simulation time axis, and then organizes the observation results at consecutive time points into a target detection sequence. To reflect the target tracking capability of the millimeter wave radar, data association and trajectory tracking algorithms can be executed in the simulation to assemble discrete target detection results into target trajectory data with target numbers, trajectory point sequences, velocity vectors, and existence probabilities. At the same time, distance resolution, velocity resolution, and angle resolution parameters can be introduced when generating the data to apply probabilistic loss to distant targets or occluded targets, thereby reflecting the detection characteristics of the millimeter wave radar in complex road scenes.

[0056] The camera simulation submodule is responsible for generating image acquisition data. Through the rendering engine in the virtual simulation environment, a virtual camera coordinate system is constructed at the vehicle installation position, and parameters such as focal length, resolution, field of view, and frame rate are set to render the three-dimensional scene into a two-dimensional image sequence. The rendering process takes into account road markings, traffic signs, surrounding vehicles, pedestrians, and weather lighting conditions, and calculates the color value and brightness information of each pixel through shaders, lighting models, and material parameters. To be closer to the real camera acquisition results, motion blur, uneven exposure, image noise, and lens distortion can be superimposed after rendering the output, and a time stamp and camera pose are added to each frame of image for subsequent multi-source data alignment and three-dimensional reconstruction.

[0057] The integration process of virtual sensor detection data is completed by a fusion packaging module, which reads three-dimensional point cloud data, target trajectory data and image acquisition data in each simulation sampling period, and aligns based on a unified time base and a unified coordinate system. Time alignment can be achieved by differencing or interpolating the timestamps of various data with the simulation master clock to ensure that different sensor outputs correspond to the same or acceptable time error range of the scene state; spatial alignment converts data in the radar coordinate system and camera coordinate system to the vehicle coordinate system or world coordinate system through the pre-configured extrinsic matrix. After alignment, the point cloud set, target list and image frame index are integrated into a structured data object, which includes sensor type markers, data time range, coordinate system definition and data block index, so that the cloud control platform can directly parse various perception results and their relationships when receiving data.

[0058] To achieve seamless connection with the vehicle-road cooperative communication stack, the virtual sensor detection data is converted into a standard communication protocol format by a protocol encoding module after generation. The encoding module selects the target protocol family according to the configuration, such as the V2X message system for vehicle-road cooperation, maps the perception results to the fields in the standard message structure, including the target position field, speed field, road danger prompt field, and environmental perception summary field. During encoding, the module fills in the message header, message body and check field according to the protocol specification, encodes the multi-source data into binary sequences or structured text form according to fixed or extensible data segments, while retaining necessary extension markers to support additional perception information. After encoding, the standard vehicle-road cooperative uplink message carrying the virtual sensor detection data is obtained, providing an input interface for the cloud control platform's subsequent fusion processing, situation analysis and instruction generation.

[0059] The data transmission link is responsible for data forwarding between the virtual simulation environment and the cloud control platform. The transmission module obtains the encoded virtual sensor detection data from the protocol encoding module, encapsulates it into a transmission layer or network layer data unit, and sends it to the cloud control platform access end through the pre-set communication interface. The communication interface can be mapped to a local inter-process communication channel, a simulation internal virtual network channel, or a transmission connection on the actual network stack for different deployment forms. The transmission module needs to attach a link-level identifier, such as the source node identifier, destination node identifier and link sequence number, to each message before transmission to support message recombination and timing recovery by the cloud control platform. At the same time, through the state monitoring mechanism, the sending time, retransmission times and queue length are recorded to ensure the continuity and integrity of link transmission when the data volume increases or the simulation time is compressed, so that the virtual sensor detection data is consistent in time with the development process of the test event.

[0060] Through the above steps, the embodiment can construct perception data flow close to the real vehicle-road cooperative scenario without relying on real sensor hardware, realize controllable input driving of cloud side processing capability, on the one hand, can systematically cover the perception data combination form under multiple complex events, verify the adaptability and robustness of the cloud control side algorithm to multi-source data, on the other hand, can flexibly replay, amplify or compress the space-time characteristics of perception data in different scenes by adjusting the virtual sensor parameters and link configuration, provide stable, repeatable and expandable input environment for subsequent cooperative control verification.

[0061] In one embodiment, the above step S40 comprises: S401, receiving, by the cloud control platform, virtual sensor detection data transmitted through a data transmission link between the virtual simulation environment and the cloud control platform; S402, performing data quality checking and format standardization processing on the received virtual sensor detection data; S403, performing multi-source data fusion processing on the virtual sensor detection data after data quality checking and format standardization processing, to generate an environment perception fusion result; S404, performing traffic situation analysis based on the environment perception fusion result, to generate a traffic situation analysis result including traffic events and risk levels; S405, generating standard car-road cooperative message content according to the traffic situation analysis result; S406, encapsulating the standard car-road cooperative message content into a message format conforming to a communication protocol; S407, sending the encapsulated standard car-road cooperative message through a data transmission link between the cloud control platform and the real roadside unit.

[0062] In the embodiment, the cloud control platform first receives virtual sensor detection data through a data transmission link between the virtual simulation environment. The receiving process can be completed by an access gateway module. The access gateway maintains a communication connection with the virtual simulation environment side at the transmission layer or the session layer, unpacks the messages arriving on the link into data frames or data packets, and according to the source identifier, timestamp, message type and other fields in the data header, transfers the virtual sensor detection data to the perception data access module. The perception data access module records the receiving sequence, marks the out-of-order, repeated or lost data packets for use in the subsequent quality checking stage, and generates a platform internal receiving time mark for each batch of arriving data, to build a unified time reference inside the cloud control platform.

[0063] In the data quality check and format standardization process, the quality check module performs a series of rule checks on the virtual sensor probe data. Quality checks can include field integrity checks to confirm the presence of mandatory fields and the length of fields meets expectations; numerical validity checks to verify that coordinate values, speed values, acceleration values, and confidence levels fall within a predetermined range; and time continuity checks to identify severely delayed or time-wound data batches by comparing the difference between data timestamps and platform reception times. Data that fails a certain type of check can be flagged as abnormal and placed in a quarantine queue or discarded queue, and will not be involved in subsequent fusion. The format standardization module, under the premise that the quality check is passed, converts virtual sensor probe data from different sources into a unified data description structure, such as using the same coordinate system, uniform unit system, uniform floating point precision, and uniform field naming method, and aligns the configuration of special fields for different types of sensors, creating a consistent input interface for multi-source data fusion.

[0064] Multi-source data fusion processing is based on virtual sensor probe data that has undergone quality checks and format standardization. The fusion engine first aligns data from different perception sources such as lidar, millimeter wave radar, and vision sensors according to the time axis, and through window matching or interpolation, point cloud data, target trajectory data, and image acquisition data within the same time window are grouped into the same fusion period. Subsequently, the fusion engine maps the outputs of each perception source to the vehicle coordinate system or the global road coordinate system according to the pre-configured coordinate transformation relationship, ensuring that spatial positions can be directly compared. On this basis, target-level fusion can be performed to associate and match point cloud clustering results, radar target tracking results, and visual recognition objects to generate a fusion target entry for the same physical target; or grid-level or voxel-level fusion can be performed to superimpose spatial occupancy information from different sensors onto a unified grid map. The fusion engine sets weights or confidence levels for each perception source during processing, adjusts the influence of different perception sources in the fusion result based on device characteristics and scene conditions, and finally generates an environment perception fusion result containing road geometry, static obstacles, dynamic targets, lane structure, and occupancy status information.

[0065] The traffic situation analysis is performed on the basis of the fusion result of the environment perception. The situation analysis module first analyzes the topological relationship between the road elements and the traffic participants from the fusion result, and identifies the possible conflict areas and approaching events by judging the relative distance, relative speed and predicted trajectory of the virtual vehicle and the preceding vehicle, oncoming vehicle, intersection approaching vehicle, pedestrian, etc. Then, the situation analysis module matches the current fusion environment with the predefined cooperative vehicle infrastructure event and cooperative vehicle-vehicle event categories in the event configuration, and generates traffic events such as congestion ahead, approaching construction area, accident ahead, temporary speed limit, special vehicle priority, etc. when the threshold condition is met. On the basis of the traffic event identification, the situation analysis module calculates the risk level, which can be comprehensively evaluated by multiple indicators, such as the risk score calculated on the basis of the collision time margin, deceleration requirement, lateral deviation degree and environmental complexity, and then the score is divided into different level ranges to obtain the traffic situation analysis result containing the traffic event type and the corresponding risk level, which provides the basis for the downstream message generation.

[0066] The standard cooperative vehicle infrastructure message content is constructed on the basis of the traffic situation analysis result. The message content generation module extracts the event type, event spatial range, affected lane set, recommended speed, recommended lane change behavior or detour suggestion, etc. from the traffic situation analysis result, and combines them with the current virtual vehicle position, cloud control platform time and road basic information to construct a set of structured fields, which will be mapped into the control prompt information broadcast by the real roadside unit and parsed by the real vehicle unit. Different content templates can be used for different risk levels in the message content generation process, such as more detailed speed control suggestions and priority information for high risk, and general reminders and lightweight suggestions for low risk, to avoid message overload or information redundancy.

[0067] In the communication protocol encapsulation stage, the message encapsulation module fills the standard cooperative vehicle infrastructure message content into the message body defined by the selected cooperative vehicle infrastructure communication protocol specification, and constructs the corresponding message header. The message header can include message identification, version information, sender identification, target area identification, timestamp and check field, and the message body part organizes the event position, event type, risk level, action range, suggestion information, etc. according to the field order and coding method specified by the protocol. In the encapsulation process, the numerical fields are quantitatively encoded, and the string or enumeration fields are encoded and mapped according to the protocol specification, and the check sum or signature field is appended at the end of the message to enable the real roadside unit to perform integrity verification when receiving. Finally, the standard cooperative vehicle infrastructure message conforming to the communication protocol format is formed, which can be directly transmitted and parsed on the existing cooperative vehicle infrastructure communication stack.

[0068] The message sending stage is executed by a downlink scheduling module in the cloud control platform. The scheduling module selects a suitable sending channel according to the data transmission link configuration between the real road side unit, and pushes the encapsulated standard vehicle-road cooperation message to the corresponding road side unit target address. Before sending, one or more target road side units can be selected according to the role range or road identifier in the message content, and the message is added to the corresponding downlink queue; the sending time and retransmission state are recorded during the sending process, and the same event updated multiple times in a short time can be merged or updated to reduce the link load. At the same time, the downlink scheduling module can set priorities for different messages according to the risk level, to ensure that high-risk event-related messages can be sent to the real road side unit first when multiple messages are sent at the same time, further supporting real-time requirements.

[0069] Through the above steps, the multi-source perception data generated by the virtual environment can be converted into structured standard cooperative messages after cloud-side fusion and situation analysis, so that the content received by the real road side unit and the real vehicle-mounted unit in the test scene is consistent with the message format when the real road is deployed, thereby verifying the full-link behavior from the cloud control platform to the road side equipment without modifying the existing communication protocol and vehicle-mounted terminal. The quality check, standardization and fusion analysis processing in the cloud control platform can pre-eliminate the influence of input data noise, format difference and time misalignment on the subsequent control link, so that the influence of different strategies on the cooperative message generation result can be systematically investigated by adjusting the fusion strategy, risk level division method and message content template in the same simulation scene, providing reproducible, controllable and close-to-real-network deployment form data processing and message output capability for vehicle-road cooperation testing.

[0070] In one embodiment, the above step S50 comprises: S501, receiving a standard vehicle-road cooperation message from the cloud control platform by the real road side unit; S502, performing message integrity check and protocol compatibility check on the received standard vehicle-road cooperation message; S503, adapting the standard vehicle-road cooperation message that passes the message integrity check and protocol compatibility check to a wireless communication protocol format; S504, broadcasting the adapted standard vehicle-road cooperation message to the real vehicle-mounted unit through a wireless communication channel; S505, monitoring the message forwarding state by the real road side unit and recording the forwarding log information, and executing an error handling mechanism when the forwarding is abnormal.

[0071] In this embodiment, the real roadside unit assumes the edge forwarding role between the cloud control platform and the real vehicle-mounted unit in the entire communication link, and needs to have both wired or optical fiber backhaul connection capability with the cloud and transmission and scheduling capability supporting vehicle-road cooperation wireless communication. The standard vehicle-road cooperation message sent by the cloud control platform first enters the uplink access module inside the real roadside unit through the backhaul link. The module performs link layer and network layer unpacking on the arrived data frame, restores the standard vehicle-road cooperation message in the payload, and transfers the message to the roadside application processing unit according to the source identifier, message type, timestamp and other fields in the message header. In this process, local receiving time, receiving interface identifier and other metadata can be supplemented for each message inside the real roadside unit to provide basic information for subsequent log recording and state monitoring.

[0072] Message integrity check is performed by a check module in the real roadside unit, which is used to confirm that the standard vehicle-road cooperation message transmitted from the cloud control platform to the roadside has not been truncated, tampered or bit error in the transmission process. The check module can verify according to the checksum field at the end of the message, the frame-level cyclic redundancy check field or the digital signature field of the upper layer, and compare whether the calculation result is consistent with the check value carried by the message. At the same time, it can also check whether the message length is consistent with the header mark, whether the sequence number is continuous, etc. If any of the checks does not meet the requirements, the message is marked as invalid and enters the discard or isolation path, and does not enter the wireless side broadcast process. The protocol compatibility check is compared between the content of the standard vehicle-road cooperation message and the vehicle-road cooperation protocol configuration supported by the real roadside unit, such as checking whether the message version number belongs to the supported version, whether the message type is in the configured business type set, whether the field arrangement, encoding method and unit system conform to the current enabled communication protocol specification. When detecting version mismatch, field missing or encoding inconsistency, compatibility exception flag can be triggered, and whether to attempt protocol adaptation, perform degradation processing or directly reject forwarding can be decided according to the configuration.

[0073] After the message integrity check and protocol compatibility check are passed, the protocol adaptation module in the real roadside unit converts the standard vehicle-road cooperation message into the wireless communication protocol format. This process includes mapping the message content into the corresponding frame structure field according to the selected vehicle-road cooperation wireless protocol (such as a certain vehicle-road cooperation wireless frame structure specification), such as filling event location, event type, risk level, action range, suggestion information, etc. into the pre-defined fields or information elements in the protocol, and quantizing, compressing or encoding the numerical values according to the requirements of the wireless side. The protocol adaptation module also needs to select transmission power, modulation and coding method, repetition number and priority identifier and other parameters according to the wireless physical layer and link layer configuration, and submit these parameters and the encoded message to the wireless transmission and scheduling module.

[0074] The wireless communication channel carries a wireless broadcast process between the real roadside unit and the real vehicle unit. The real roadside unit broadcasts the adapted standard car-road cooperation message on a pre-configured wireless frequency band through a transmitter and an antenna subsystem. The wireless transmission scheduling module manages the queue and priority of the message according to the risk level, business type and current channel occupation of the message. High-risk event-related messages can be placed in a high-priority queue and preferentially occupy time slots for broadcast when channel resources are limited. Broadcast transmission can use periodic broadcast or event-driven mode, and immediately trigger transmission when the event state is updated. In order to improve the receiving probability of the real vehicle unit, the real roadside unit can repeatedly send the same standard car-road cooperation message several times within a short time window, or staggered transmission at different time offsets to reduce the influence of instantaneous interference.

[0075] The message forwarding state monitoring is run by a monitoring module inside the real roadside unit. The monitoring module establishes a state entry for each message passing through the wireless transmission scheduling module, records the time when the message enters the sending queue, the actual transmission time, the transmission times and the channel congestion indication and other information. If the wireless protocol or system deployment conditions allow, confirmation information, retransmission request or statistical feedback from the real vehicle unit side can also be collected, so as to indirectly infer whether the message is successfully received by most real vehicle units in the target area. The forwarding log information is continuously written into the storage unit during the monitoring process, and the log can include message identification, event type, risk level, transmission power, frequency point configuration, transmission result statistics, abnormal marker and other contents, which is convenient for subsequent analysis and troubleshooting of the car-road cooperation link performance. When the monitoring module detects forwarding exceptions, such as continuous multiple transmission failures, long-term high-occupancy state of the channel or feedback from the real vehicle unit indicating that the message loss rate is too high, the error handling mechanism is triggered. Error handling can include increasing the transmission priority of this type of message, adjusting the transmission power or modulation and coding parameters, switching to a backup frequency band, feeding back abnormal state information to the cloud control platform, or terminating the forwarding of a single message that cannot be repaired and recording the detailed reasons in the log, so as to avoid the abnormal state continuously affecting the broadcast of subsequent messages.

[0076] Through the above steps, the uplink message is subjected to integrity and compatibility screening before entering the wireless broadcast link, avoiding abnormal or non-compliant messages entering the vehicle side and reducing interference with the decision logic of the real vehicle unit.

[0077] In one embodiment, the above step S60 includes: S601, receiving a standard car-road cooperation message from the real roadside unit through the real vehicle unit; S602, performing message parsing processing on the received standard car-road cooperation message to extract traffic event information and road condition data; S603, performing risk assessment analysis based on the traffic event information, the road condition state data, and a current running state parameter of the vehicle; S604, generating a vehicle control strategy through a decision algorithm based on a result of the risk assessment analysis; S605, generating a vehicle control instruction according to the vehicle control strategy; S606, encapsulating the vehicle control instruction into a data format recognizable by a simulation system; S607, sending the encapsulated vehicle control instruction to the virtual simulation environment through a data transmission link between the real vehicle-mounted unit and the virtual simulation environment.

[0078] In the embodiment, when the real vehicle-mounted unit receives a standard car-road cooperation message from a real roadside unit, first, physical layer and link layer access are completed through a vehicle-mounted side wireless communication interface, and a wireless frame containing the standard car-road cooperation message is extracted from a broadcast channel by using a pre-configured wireless frequency band, time slot resource, and synchronization mechanism. The receiving link demodulates, channel decodes, and error code detects the signal, and delivers the data frame passing the bottom layer check to an upper layer processing unit for unpacking. After removing the physical layer and link layer header information, the business load carrying the standard car-road cooperation message is restored, and the message is put into a vehicle-mounted message buffer queue according to the message identifier, timestamp, and serial number fields, to provide input for subsequent analysis and decision processing.

[0079] The message analysis processing link is executed in an analysis module inside the real vehicle-mounted unit. The module analyzes the message header and message body content field by field according to the structure definition of the standard car-road cooperation message, and splits and arranges the information logically belonging to the traffic event and the data reflecting the road running condition. The traffic event information can correspond to the event type, event location, event duration, and affected lane number fields, and is used to describe situations such as obstacle appearance, construction road occupation, traffic control, or abnormal behavior of the front vehicle. The road condition state data can cover the current road section speed limit value, real-time average vehicle speed, lane occupancy state, weather condition marker, road surface adhesion condition marker, and the like, and is used to depict the background traffic environment. The analysis module maps these data into an internal data structure, uniformly converts them into a format in a vehicle local coordinate system or a geographic coordinate system, and aligns the contents of different messages according to the timestamp field, so as to obtain a structured traffic event information and road condition state data set.

[0080] The vehicle current running state parameters are obtained by the vehicle unit through the vehicle bus, positioning module and vehicle body sensor, including the vehicle current position, driving speed, acceleration, heading angle, yaw rate and the longitudinal and lateral distance from the front target and other indicators. The real vehicle unit inputs the current running state parameters, traffic event information and road condition data into the risk assessment analysis unit to build a risk assessment input vector corresponding to the current scene. The risk assessment analysis unit can calculate the time margin of the vehicle to reach the event affected area while maintaining the existing maneuvering behavior, estimate the collision probability under different braking intensity and steering angle conditions, and also apply weights to environmental factors such as road wetness and limited visibility to give risk level, recommended deceleration range and lateral avoidance requirements and other evaluation results. The evaluation process can be repeated in a fixed period to track the changes of the risk level in the vehicle motion process in real time.

[0081] When the decision algorithm generates vehicle control strategies based on the risk assessment analysis results, it will input the risk level, time margin, road geometry and surrounding traffic participant distribution as input, and output decision schemes for vehicle longitudinal and lateral control. The control strategy can include target speed, target acceleration, target lane, lane keeping and deceleration, lane changing and bypassing, parking in a safe area, etc. It can also be accompanied by strategy effective duration, gradual transition requirements and ride comfort constraints. For scenarios with multiple feasible strategies, the decision algorithm can consider safety, traffic efficiency and vehicle dynamic limits through a cost function to compare and select the appropriate strategy output under the premise of meeting safety constraints. The generated vehicle control strategy is stored in a structured form, clearly distinguishing between longitudinal control, lateral control and strategy metadata, to facilitate subsequent conversion to specific vehicle control instructions.

[0082] The vehicle control instructions are constructed by the control instruction generation module inside the real vehicle unit according to the vehicle control strategy. The control instructions can give target acceleration, target braking force, target steering angle, target lateral offset and other parameters at a finer granularity, or organize them into a discrete control sequence according to time slices, marking the effective time and duration of each control segment. In the virtual-real fusion test scene, the control instructions do not directly act on the real vehicle actuators, but are used as driving inputs to the virtual vehicle model, so the control instruction generation module organizes the control parameters into a structure consistent with the simulation system's expected input interface, such as using a unified coordinate system, standardized unit system, and adding timestamps and instruction sequence numbers for synchronization, to ensure that the virtual simulation environment can apply these instructions at the correct time.

[0083] The data format recognizable by the simulation system is generated by an encoding and packaging module. According to the interface agreement disclosed by the virtual simulation environment, the module maps the vehicle control instruction into a message format under a specific field layout and encoding rule. In the packaging process, the control instruction can be placed into a unified message header structure, wherein the message header includes fields such as instruction type, instruction version, vehicle identification, timestamp, and the message body includes a continuous or discrete control parameter array. The encoding and packaging module can support multiple data representation methods, such as binary structured representation, tag-based hierarchical representation, or message queue-oriented serialization, and select the specific format to be used by configuration, so as to adapt to different virtual simulation platforms.

[0084] The data transmission link undertakes the responsibility of transmitting the vehicle control instruction from the real vehicle unit to the virtual simulation environment in this process. After the real vehicle unit completes the packaging, the control instruction is handed over to the communication output module, which selects the transmission path according to the link type and can realize indirect forwarding through the backhaul link shared with the cloud control platform, or directly connect the virtual simulation environment access gateway through a dedicated network channel. The communication output module can allocate a unique identifier for each control instruction message, and implement retransmission counting and sending timeout management. In the link with the confirmation mechanism, the receiving feedback from the virtual simulation environment or the intermediate node can also be recorded, and the sending result is written into the local log for analyzing the reliability and delay characteristics of the instruction in the transmission process, so as to distinguish the response difference caused by the control logic and the uncertainty caused by the communication link in the subsequent test evaluation.

[0085] Through the above steps, the embodiment realizes the processing closed loop of completing the perception result interpretation, risk assessment and control decision by the real vehicle unit under the virtual-real fusion architecture, so that the control input received by the virtual vehicle model is consistent with the actual decision logic of the real vehicle unit, thereby preserving the complexity and diversity of the real vehicle control strategy in the test process, and safely reproducing the control behavior in the virtual environment, which provides a solid foundation for the subsequent fine analysis of the reliability, response rationality and strategy of the vehicle-road cooperation control link.

[0086] In one embodiment, the above step S70 includes: S701, receiving a vehicle control instruction from the real vehicle unit through the virtual simulation environment; S702, performing instruction analysis processing on the received vehicle control instruction to extract control parameters; S703, updating the motion state of the virtual vehicle model according to the control parameters to obtain an updated virtual vehicle model state, wherein the motion state includes position coordinates, motion speed and / or motion direction; S704, simulate a dynamic behavior response of the vehicle in the test event scenario based on the updated virtual vehicle model state; S705, collect behavior response data of the virtual vehicle model, the behavior response data including trajectory data and / or interaction data; S706, perform performance index analysis on the collected behavior response data to obtain an analysis result; S707, generate a test verification report based on the analysis result to record a complete response process of the virtual vehicle model to the test event.

[0087] In the embodiment, when the virtual simulation environment updates the state of the virtual vehicle model according to the vehicle control instruction and verifies the response to the test event, first, the vehicle control instruction from the real vehicle-mounted unit is received in the simulation communication access module. The receiving process establishes a session through a pre-configured network interface and a data transmission link, unpacks the control instruction packet in the uplink data stream from the transmission layer and the session layer into a unified data structure inside the simulation environment, and performs verification and sorting according to the vehicle identifier, the instruction sequence number and the time stamp, so as to ensure that the vehicle control instruction entering the simulation operation module has a clear time sequence relationship and source information.

[0088] The instruction analysis process is performed in an instruction analysis unit inside the virtual simulation environment. According to the message format definition of the vehicle control instruction, the instruction analysis unit performs field-level analysis on the control instruction, extracts the target speed, target acceleration, braking force parameters related to vehicle longitudinal control, the target steering angle, target lateral offset related to lateral control, and the effective time, duration time related to execution time, and uniformly maps them into a set of control parameters. The instruction analysis unit can perform unit conversion and coordinate system conversion to convert the control amount from the vehicle-mounted unit side into a standard unit system used by the virtual vehicle model, and apply default parameters or interpolation completion strategy to the default fields to form a complete control parameter set, providing input for subsequent motion state update.

[0089] The motion state of the virtual vehicle model is represented in the simulation calculation core as a set of updateable physical quantities. Position coordinates are used to describe the planar or spatial position of the vehicle in the simulation road coordinate system, which can be represented in world coordinates, road centerline coordinates plus lateral offset, etc. Motion speed is used to describe the instantaneous speed of the vehicle in the current direction. Motion direction is used to describe the heading angle or orientation vector of the vehicle. When updating the state, the simulation calculation core calculates the new position coordinates, speed, and direction based on the virtual vehicle model state at the previous simulation time and the control parameters at the current time through discrete time integration or numerical solution, and corrects excessive control changes under the condition of considering vehicle dynamics constraints and tire side slip restrictions to obtain the updated virtual vehicle model state consistent with the vehicle dynamics characteristics. The update process can support fixed time steps or adaptive time steps to balance numerical stability and computational efficiency.

[0090] The test event scenario is internally represented in the virtual simulation environment as a collection of road geometry, traffic facilities, other participating vehicles, and static or dynamic obstacles. The simulation of dynamic behavior response is completed by the scene execution engine. At each simulation time step, the scene execution engine aligns the vehicle's position in the test event scenario to the road topology based on the updated virtual vehicle model state, determines the spatial relationship with obstacles, construction areas, special vehicles, and conflict vehicles, and drives the virtual vehicle to evolve according to the longitudinal and lateral behavior specified by the control parameters. During the motion process, the scene execution engine continuously detects whether the virtual vehicle has entered the event impact area, whether it has collided with other traffic participants, and whether it has violated the preset speed limit or lane constraints, thereby forming a complete dynamic process of the test event trigger and response.

[0091] The behavior response data is generated in the simulation data collection module, which extracts trajectory data and interaction data from the virtual vehicle model and the scene execution engine according to a preset sampling period. Trajectory data can include position coordinate sequences, speed change curves, acceleration curves, and direction change curves on a continuous time axis, which are used to describe the driving trajectory and dynamics changes of the virtual vehicle in the test event scenario. Interaction data can include distance records with front, rear, and lateral vehicles, the closest distance to obstacles, and relative position changes with traffic signals or road signs, which are used to reflect the interaction between the virtual vehicle and the surrounding environment. The simulation data collection module can attach simulation time stamps and event flag bits when collecting data to indicate which data segments belong to the test event impact interval, providing a filtering basis for subsequent analysis.

[0092] The performance index analysis is performed in a data analysis unit, which takes the behavior response data as input and calculates a set of performance indexes according to pre-set evaluation rules. The performance indexes can include minimum collision distance, whether collision occurs, deceleration level before entering the event influence area, average speed of the vehicle passing through the event area, lane deviation during the event, response delay time, control execution smoothness, etc. The data analysis unit can perform interpolation and filtering processing on the trajectory data to obtain smooth speed and displacement curves, perform interval statistics on the interaction data to extract key event points, and match different index combinations according to different test event types, so as to obtain analysis results covering safety, smoothness and compliance.

[0093] The test verification report is constructed in a report generation unit, which is based on the analysis results, and the test event identification, the initial state of the virtual vehicle model, the key parameters of the vehicle control instruction, the feature segments of the behavior response data and the performance index analysis results are structured and organized. The report generation unit can divide the test process into the pre-event preparation stage, the event triggering to vehicle response stage, the vehicle passing through the event area or completing the avoidance stage, and attach the corresponding trajectory segments and index values to each stage to record the complete response process of the virtual vehicle model to the test event. The generated test verification report can be output in the form of structured data file or human-readable text, so as to be used for automatic comparison of the performance of different control strategies, or manual review and archival management by the test personnel.

[0094] Through the above steps, the embodiment can quantitatively verify the actual effect of the vehicle control instruction in typical and extreme event scenarios without relying on real vehicles and real roads, which not only ensures that the motion evolution of the virtual vehicle model is consistent with the real vehicle control logic, but also presents the response quality of the control strategy to different test events in the form of structured indexes and traceable reports, thereby providing an efficient, reproducible and fine-grained analysis verification means for the function verification and safety evaluation of the vehicle-road cooperation system in the virtual-real fusion environment.

[0095] In an embodiment, a vehicle-road cooperation virtual-real fusion simulation test device is provided, which corresponds to the vehicle-road cooperation virtual-real fusion simulation test method in the above embodiment. Referring to Figure 3 , Figure 3 A functional module schematic diagram of a preferred embodiment of the vehicle-road cooperation virtual-real fusion simulation test device of the present application. The communication link configuration module 10, the test event configuration module 20, the virtual sensor data generation module 30, the cloud fusion and message generation module 40, the message forwarding module 50, the vehicle-mounted instruction generation module 60 and the virtual vehicle response verification module 70. The detailed description of each functional module is as follows: The communication link configuration module 10 is configured to deploy a real road side unit, a real vehicle-mounted unit and a cloud control platform, and establish a data transmission link between the real road side unit, the real vehicle-mounted unit, the cloud control platform and a virtual simulation environment. The test event configuration module 20 is configured to set a test event in the virtual simulation environment. The virtual sensor data generation module 30 is configured to generate virtual sensor detection data in the virtual simulation environment through the virtual sensor model in response to the test event, and send the virtual sensor detection data to the cloud control platform through a data transmission link between the virtual simulation environment and the cloud control platform. The cloud fusion and message generation module 40 is configured to fuse the received virtual sensor detection data through the cloud control platform, generate a standard car-road cooperation message, and send the standard car-road cooperation message to the real road side unit. The message forwarding module 50 is configured to forward the standard car-road cooperation message to the real vehicle-mounted unit through the real road side unit. The vehicle-mounted instruction generation module 60 is configured to analyze the standard car-road cooperation message through the real vehicle-mounted unit, generate a vehicle control instruction, and send the vehicle control instruction to the virtual simulation environment through a data transmission link between the real vehicle-mounted unit and the virtual simulation environment. The virtual vehicle response verification module 70 is configured to update the state of the virtual vehicle model according to the vehicle control instruction in the virtual simulation environment, and verify the response of the virtual vehicle model to the test event.

[0096] In one embodiment, a computer device is provided, which can be a server, and an internal structure diagram thereof can be as shown in Figure 4 The computer device includes a processor, a memory, a network interface and a database connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium, an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the computer device is configured to communicate with an external client through a network connection. The computer program is executed by the processor to implement the functions or steps of a car-road cooperation virtual-real fusion simulation test method server side.

[0097] In one embodiment, a computer device is provided, which can be a client, and an internal structure diagram thereof can be as shown in Figure 5As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected by a system bus. Among them, the processor of the computer device is used to provide computing and control capability. The memory of the computer device includes a non-volatile storage medium, an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with the external server through the network connection. The computer program is executed by the processor to implement the functions or steps of the client side of the vehicle-road cooperative virtual-real fusion simulation test method.

[0098] In an embodiment, a vehicle-road cooperative virtual-real fusion simulation test method is provided, and the overall workflow of the method can refer to Figure 6 After the system is started, it enters the simulation environment construction stage. On the one hand, based on three-dimensional road scene modeling, real road network information is imported, lane structure, intersection topology and ramp connection relationship are reconstructed in the computer, and a virtual road model containing main road, auxiliary road and multiple types of intersection is formed, and intersection geometry and lane attributes are generated. On the other hand, traffic signs and signal control units are arranged in the virtual scene, and through the configuration of traffic light timing scheme, lane release scheme and pedestrian phase strategy, combined with the parameter setting of weather and lighting conditions, the environment of rain, fog, night and backlight is rendered, and a three-dimensional road simulation environment that can be used for vehicle-road cooperative perception and control verification is formed.

[0099] At the same time of road scene initialization, the system constructs vehicle-road side device model and vehicle side model. The road side part completes the intersection RSU unit modeling through the road test device simulation process, and hangs multiple types of perception device models such as laser radar, millimeter wave radar and camera around the RSU, configures the installation position, field of view range, detection distance, refresh frequency and error noise parameters of each type of sensor, and establishes the binding relationship between the traffic light control logic and the signal light instance in the road scene, so that the road side device can generate original perception data and signal control data consistent with the real deployment form in the simulation environment. The vehicle part describes the longitudinal acceleration and deceleration and lateral turning characteristics by establishing a vehicle dynamics model, hangs an OBU system model and a multi-source vehicle-mounted sensor model on the vehicle body coordinate system, configures the installation parameters and output interfaces of inertial navigation, camera, vehicle-mounted millimeter wave radar and other devices, and loads behavior control algorithms in the vehicle model, so that the virtual vehicle has adjustable driving behaviors such as acceleration, lane change, following and obstacle avoidance in the simulation environment, providing a basis for subsequent vehicle-vehicle cooperative and vehicle-road cooperative event injection.

[0100] After the virtual device model is initialized, the system enters the loop device access stage. In the RSU device access process, the real RSU is accessed to the network where the simulation computer is located through Ethernet or industrial switch, the communication address and data interface of the RSU are configured in the simulation platform, the virtual sensor data output channel is mapped with the uplink receiving interface of the real RSU, and the real sensor data input channel is reserved for accessing the data stream collected by the external radar or camera. Through the two-way data switching mechanism, the virtual data and the real data can be switched, so that the roadside processing logic can be repeatedly run under different data source conditions and compared and verified, and the processing result is sent to the outside through the V2X protocol stack. In the OBU device access process, the real OBU is networked with the simulation computer through the vehicle bus simulation interface or the Ethernet interface, the communication channel of the OBU for receiving messages from the RSU and the cloud control platform is configured, and the mapping relationship between the OBU information collection and the simulation vehicle control interface is established, so that the vehicle infrastructure integration message received by the OBU can be converted into acceleration, deceleration, lane change or warning prompt control of the virtual vehicle, and the running state of the virtual vehicle is fed back to the simulation platform in the form of vehicle end state information. In the cloud control platform access process, the cloud control software components are deployed on the independent server, the communication link between the RSU and the OBU is configured, the virtual sensor detection data output by the simulation platform is accessed, processed and distributed, so as to realize the data closed loop between the cloud control platform and the virtual simulation environment, the real RSU and the real OBU.

[0101] After the virtual and real devices on the road side and the vehicle side are accessed, the system enters the warning event simulation and cooperative test execution stage. Through the warning event simulation module, the vehicle infrastructure integration event and the vehicle-vehicle cooperative event can be configured based on the constructed road scene and device model, the road side perception data construction, the cloud event information processing, the OBU receiving processing result and the result issuing to the simulation vehicle are connected in series to form a complete link, realizing the cross-device and cross-network linkage from the virtual sensor, the real RSU, the cloud control platform to the real OBU and the virtual vehicle. In the vehicle infrastructure integration event test, the response effect of the vehicle under the control strategy of early warning, active deceleration, lane change and the like can be verified by driving the virtual or real RSU to aggregate multi-sensor data, generate event information and broadcast to the target vehicle through the cloud control platform or directly through V2X, around the scenes such as red light intrusion, speed limit in construction area and obstacle occupation; in the vehicle-vehicle cooperative event test, the multi-vehicle interaction relationship and conflict working condition can be configured, and the position, speed and intention information of multiple virtual vehicles or real vehicles are exchanged through the communication between OBUs, to verify the cooperative control effect of vehicles in queue following, emergency braking and collision avoidance.

[0102] In the process of continuously performing early warning event simulation and collaborative testing, the simulation platform records the RSU end data, cloud control platform transmission data and OBU end control results, and uniformly collects vehicle running trajectories, acceleration changes, minimum vehicle distance and collision determination in different test rounds. Indexes such as, at the end of the test task, output test records covering multiple event scenarios, provide a basis for subsequent analysis of car-road cooperation algorithms, communication link delays, device deployment schemes, etc. When the preset test case set is executed or the upper control instruction ends the current test, the system will archive the current workflow and end the operation.

[0103] Through the above process, the virtual-real fusion simulation test system of car-road cooperation simultaneously introduces three-dimensional road scenes, roadside device models, vehicle dynamics models, and real RSUs, real OBUs, and cloud control platforms in a single workflow. It realizes in-loop commissioning of virtual and real devices, builds car-road cooperation events and car-car cooperation events in the same simulation environment and triggers early warning event simulation. It makes virtual sensor detection data, cloud processing results and vehicle control responses form a repeatable configuration closed-loop test link. Thus, without relying on large-scale real road closed test sites, it verifies and evaluates the consistency of car-road cooperation systems in multiple scenarios, multiple devices and multiple links, and improves the safety, controllability and coverage of the test process.

Claims

1. A simulation test method for vehicle-road cooperative virtual-real fusion, characterized in that, Includes the following steps: Deploy real roadside units, real vehicle-mounted units, and cloud control platforms, and establish data transmission links between the real roadside units, the real vehicle-mounted units, the cloud control platform, and the virtual simulation environment; Set up test events in a virtual simulation environment; In response to the test event, virtual sensor detection data is generated in the virtual simulation environment using the virtual sensor model, and the virtual sensor detection data is sent to the cloud control platform through the data transmission link between the virtual simulation environment and the cloud control platform. The cloud control platform performs fusion processing on the received virtual sensor detection data to generate a standard vehicle-road cooperative message, and sends the standard vehicle-road cooperative message to the real roadside unit. The standard vehicle-road cooperative message is forwarded to the actual vehicle unit through the actual roadside unit; The real vehicle-mounted unit parses the standard vehicle-road cooperative message to generate vehicle control commands, and sends the vehicle control commands to the virtual simulation environment through the data transmission link between the real vehicle-mounted unit and the virtual simulation environment. In the virtual simulation environment, the state of the virtual vehicle model is updated according to the vehicle control command, and the response of the virtual vehicle model to the test event is verified.

2. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, Configure test events in a virtual simulation environment, including: Configure vehicle-road cooperative events and set up obstacle areas, construction areas, and special vehicles in the virtual simulation environment; Configure vehicle-to-vehicle collaborative events, and configure multi-vehicle interaction scenarios and conflict events in a virtual simulation environment; Based on the configuration of the vehicle-road cooperative events and vehicle-to-vehicle cooperative events, set the event triggering conditions; Configure event impact parameters based on the event triggering conditions; Integrate the vehicle-road cooperative events, vehicle-to-vehicle cooperative events, event triggering conditions, and event impact parameters to generate test event configuration information; The test event configuration information is loaded into the event management system of the virtual simulation environment; Start the test event monitoring process.

3. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, In response to the test event, virtual sensor detection data is generated in the virtual simulation environment using the virtual sensor model, and the virtual sensor detection data is sent to the cloud control platform via the data transmission link between the virtual simulation environment and the cloud control platform, including: When the test event is detected, the virtual sensor model in the virtual simulation environment is activated; The virtual sensor model is used to simulate the sensing function of lidar and generate three-dimensional point cloud data. The virtual sensor model simulates the sensing function of millimeter-wave radar to generate target trajectory data. The virtual sensor model simulates the camera's sensing function to generate image acquisition data. Integrate the 3D point cloud data, the target trajectory data, and the image acquisition data to generate virtual sensor detection data; The data detected by the virtual sensor is encoded into a standard communication protocol format; The encoded virtual sensor detection data is sent to the cloud control platform through the data transmission link between the virtual simulation environment and the cloud control platform.

4. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, The cloud control platform fuses the received virtual sensor detection data to generate a standard vehicle-road cooperative message, and sends the standard vehicle-road cooperative message to the real roadside unit, including: The cloud control platform receives virtual sensor detection data transmitted through the data transmission link between the virtual simulation environment and the cloud control platform. The received virtual sensor detection data is subjected to data quality verification and format standardization processing; Multi-source data fusion processing is performed on virtual sensor detection data after data quality verification and format standardization to generate environmental perception fusion results; Traffic situation analysis is performed based on the environmental perception fusion results, generating traffic situation analysis results including traffic events and risk levels; Standard vehicle-road cooperative message content is generated based on the traffic situation analysis results; The standard vehicle-road cooperative message content is encapsulated into a message format that conforms to the communication protocol; The encapsulated standard vehicle-road cooperative message is sent through the data transmission link between the cloud control platform and the actual roadside unit.

5. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, The standard vehicle-road cooperative message is forwarded to the actual vehicle-mounted unit through the actual roadside unit, including: The actual roadside unit receives standard vehicle-road cooperative messages from the cloud control platform. Perform message integrity verification and protocol compatibility checks on the received standard vehicle-road cooperative messages; Adapt standard vehicle-road cooperative messages that have passed message integrity verification and protocol compatibility checks to wireless communication protocol formats; The adapted standard vehicle-road cooperative message is broadcast to the actual vehicle unit via a wireless communication channel. The real roadside unit monitors the message forwarding status and records forwarding log information, and executes an error handling mechanism when forwarding is abnormal.

6. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, The real vehicle-mounted unit parses the standard vehicle-road cooperative message to generate vehicle control commands, and sends the vehicle control commands to the virtual simulation environment through the data transmission link between the real vehicle-mounted unit and the virtual simulation environment, including: The actual vehicle-mounted unit receives standard vehicle-road cooperative messages from the actual roadside unit. The received standard vehicle-road cooperative messages are parsed to extract traffic event information and road condition data; Risk assessment and analysis are performed based on the traffic incident information, road condition data, and vehicle current operating status parameters. Based on the risk assessment and analysis results, a vehicle control strategy is generated through a decision-making algorithm. Generate vehicle control commands based on the vehicle control strategy; The vehicle control commands are encapsulated into a data format that the simulation system can recognize; The packaged vehicle control commands are sent to the virtual simulation environment through the data transmission link between the real vehicle unit and the virtual simulation environment.

7. The vehicle-road cooperative virtual-real fusion simulation test method as described in claim 1, characterized in that, In the virtual simulation environment, updating the state of the virtual vehicle model according to the vehicle control command and verifying the virtual vehicle model's response to the test event includes: The virtual simulation environment receives vehicle control commands from the real vehicle unit. The received vehicle control commands are parsed to extract control parameters; The motion state of the virtual vehicle model is updated according to the control parameters to obtain the updated virtual vehicle model state, wherein the motion state includes position coordinates, motion speed and / or motion direction. Based on the updated virtual vehicle model state, simulate the dynamic behavior response of the vehicle in the test event scenario; Collect behavioral response data of the virtual vehicle model, including trajectory data and / or interaction data; The collected behavioral response data is subjected to performance index analysis to obtain analysis results; A test verification report is generated based on the analysis results to record the complete response process of the virtual vehicle model to the test event.

8. A vehicle-road cooperative virtual-real fusion simulation test device, characterized in that, The vehicle-road cooperative virtual-real fusion simulation test device includes: The communication link configuration module is used to deploy real roadside units, real vehicle-mounted units, and cloud control platforms, and to establish data transmission links between the real roadside units, the real vehicle-mounted units, the cloud control platform, and the virtual simulation environment. The test event configuration module is used to set up test events in a virtual simulation environment; The virtual sensor data generation module is used to respond to the test event by generating virtual sensor detection data in the virtual simulation environment through the virtual sensor model, and sending the virtual sensor detection data to the cloud control platform through the data transmission link between the virtual simulation environment and the cloud control platform. The cloud fusion and message generation module is used to fuse the received virtual sensor detection data through the cloud control platform, generate standard vehicle-road cooperative messages, and send the standard vehicle-road cooperative messages to the real roadside unit. The message forwarding module is used to forward the standard vehicle-road cooperative message to the real vehicle unit through the real roadside unit; The vehicle command generation module is used to parse the standard vehicle-road cooperative message through the real vehicle unit, generate vehicle control commands, and send the vehicle control commands to the virtual simulation environment through the data transmission link between the real vehicle unit and the virtual simulation environment. The virtual vehicle response verification module is used to update the state of the virtual vehicle model in the virtual simulation environment according to the vehicle control command, and to verify the response of the virtual vehicle model to the test event.

9. A computer device, characterized in that, The computer device includes a memory, a processor, and a vehicle-road cooperative virtual-real fusion simulation test program stored in the memory and capable of running on the processor. When the vehicle-road cooperative virtual-real fusion simulation test program is executed by the processor, it implements the steps of the vehicle-road cooperative virtual-real fusion simulation test method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a vehicle-road cooperative virtual-real fusion simulation test program, which, when executed by a processor, implements the steps of the vehicle-road cooperative virtual-real fusion simulation test method as described in any one of claims 1-7.

Images