An automatic driving cloud control test system and method

Abstract

The application relates to an automatic driving cloud control test system and method, which comprises a cloud control platform, a test special network, simulated traffic participants and a tested vehicle installed with a VBOX, wherein the cloud control platform is used for issuing test tasks, scheduling and planning control of the simulated traffic participants, recording and real-time supervision of test data; the test special network is used for uploading vehicle state information and issuing test instructions; the simulated traffic participants are composed of cloud control simulated non-motor vehicles and cloud control simulated dummy vehicles; and the VBOX is used for real-time acquisition of the tested vehicle state information. Compared with the prior art, the application can simulate the cluster of the simulated traffic participants and the continuous interaction of the test vehicle, complete the automatic driving vehicle test which is high in integrity, close to the real traffic environment, covers the high-risk, low-probability and difficult-to-reproduce scene, and has the advantages of low cost, flexible motion control of the simulated traffic participants, real-time cloud control, simple test process and easy operation.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving testing technology, and in particular to an autonomous driving cloud control testing system and method. Background Technology

[0002] Closed-site testing is a crucial step in the testing and verification of autonomous driving. However, existing closed-site testing has many problems in terms of testing equipment and methods. Specifically, the testing equipment often uses towed mobile balloons, vehicle-controlled mobile dummies, and dummy vehicles. Such testing equipment has problems such as difficulty in expanding control interfaces, inflexible motion control, and high testing costs, resulting in poor interactivity of the autonomous driving test system and difficulty in reflecting the intelligence level of autonomous vehicles.

[0003] In addition, existing testing methods are mainly aimed at testing specific autonomous driving functions. The test scenarios have problems such as fragmented working conditions, simple scenario conditions, and fragmented functional testing, which make it difficult to reflect the professional intelligent capabilities of autonomous vehicles in complex traffic environments. Furthermore, the testing process requires a test safety officer to reset the scenario and assist in the test, which poses safety risks, consumes manpower and resources, and has low testing efficiency.

[0004] With the rapid development of autonomous driving technology, closed-site testing exhibits low-level and low-intelligence characteristics, failing to systematically evaluate high-level and highly intelligent autonomous vehicles. Therefore, there is an urgent need for a fully automated autonomous vehicle testing system and method that can handle variable parameters and multi-variable scenarios. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology by providing an autonomous driving cloud control test system and method. This system can simulate the continuous interaction between traffic participants and test vehicles, and complete autonomous vehicle tests that are highly complete, close to the real traffic environment, and cover high-risk, low-probability, and difficult-to-reproduce scenarios. It also has the advantages of low cost, flexible simulation of traffic participant motion control, real-time cloud control, simple test process, and easy operation.

[0006] The objective of this invention can be achieved through the following technical solution: an autonomous driving cloud control test system, comprising a cloud control platform, a dedicated test network, simulated traffic participants, and a test vehicle equipped with a VBOX (Vehicle Box). The cloud control platform is used to issue test tasks, schedule and plan the control of simulated traffic participants, and record and monitor test data in real time.

[0007] The dedicated test network is used to upload vehicle status information and issue test commands.

[0008] The simulated traffic participants consist of cloud-controlled simulated non-motorized vehicles and cloud-controlled simulated dummy vehicles;

[0009] The VBOX is used to acquire the status information of the vehicle under test in real time.

[0010] Furthermore, the cloud control platform includes a test task configuration module, a cloud control algorithm module, a test data storage module, and a test video monitoring module. The test task configuration module is used to set the test scenario, traffic flow, and intensity of conflict.

[0011] The cloud control algorithm module schedules simulated traffic participants and plans global paths according to the test task; and generates challenging test conditions for the test vehicle by optimizing local trajectories based on the real-time location of the test vehicle, while avoiding the risk of collisions between simulated traffic participants.

[0012] The test data storage module is used to receive and store the real-time status information of the vehicle under test and the trajectory data of simulated traffic participants during the test.

[0013] The test video monitoring module is used to collect video data of the entire test process of the vehicle under test and record the entire test process.

[0014] Furthermore, the tested vehicle status information includes position, speed, heading, and acceleration.

[0015] Furthermore, the test network includes a dedicated test base station and a 5G communication network. The dedicated test base station is deployed within the test site to achieve full coverage of the test site.

[0016] The 5G communication network mentioned is specifically a 5G network provided by operators for autonomous driving testing.

[0017] Furthermore, the cloud control platform is deployed on the test server in the test site management room and is connected to the test video monitoring equipment and dedicated test base station in the test site via a fiber optic network.

[0018] Furthermore, the cloud-controlled simulated non-motorized vehicle includes a high-load-bearing omnidirectional motion chassis and a simulated target object. The omnidirectional motion chassis consists of a first 5G communication module, a first chassis information acquisition module, a motor control module, and a first power supply module. The first chassis information acquisition module is used to collect the omnidirectional motion chassis pose information and omnidirectional motion chassis status information, and upload them to the cloud control platform through the first 5G communication module. The motor control module receives real-time control commands through the first 5G communication module and generates steering and motion commands accordingly to achieve omnidirectional motion control.

[0019] The simulated targets include mannequins and bicycles.

[0020] Furthermore, the cloud-controlled simulated vehicle includes a fully steerable chassis and a removable car cover. The fully steerable chassis is equipped with a second 5G communication module, a second chassis information acquisition module, a motion control module, and a second power supply module. The second chassis information acquisition module is used to collect the full steerable chassis's position and posture information and status information, and uploads it to the cloud control platform through the second 5G communication module. The motion control module receives real-time control commands through the second 5G communication module to realize chassis steering and drive / braking control.

[0021] Furthermore, the removable car cover is made of polyethylene foam material, which is beneficial for impact cushioning.

[0022] Furthermore, the VBOX includes a positioning module, a third 5G communication module, a test data acquisition module, a parameter configuration module, and a third power supply module. The positioning module is used to collect the positioning information of the vehicle under test and upload it to the cloud control platform through the third 5G communication module, and the cloud control platform generates test conditions.

[0023] The test data acquisition module is used to transmit the real-time acquired test data to the test data storage module for storage and recording via the third 5G communication module;

[0024] The parameter configuration module is used to configure the size parameter information of the vehicle under test in order to achieve vehicle calibration.

[0025] An autonomous driving cloud control testing method includes the following steps:

[0026] S1. Install VBOX on the vehicle under test and configure and calibrate the parameters;

[0027] S2. Based on the test requirements and test level proposed by the vehicle under test, the cloud control platform generates the test scenario task and driving route, delivers the generated test route to the vehicle under test, and completes the test preparation work.

[0028] S3. Power on the VBOX of the simulated traffic participants and the vehicle under test and check whether each module is working properly; if it is normal, proceed to step S4; otherwise, power off the device and return to re-execute step S3.

[0029] S4. Based on the test scenario and driving route, the cloud control platform determines the type and number of simulated traffic participants required for the test, dispatches the simulated traffic participants to their respective test starting points, and presets routes.

[0030] S5. After the simulated traffic participants arrive at their respective designated locations, the cloud control platform issues preset control commands, and the simulated traffic participants begin to move, generating a test traffic environment.

[0031] S6, the cloud control platform enables test data storage, test video monitoring and visualization functions, and issues a start test command based on the on-site status of the test site;

[0032] S7. The vehicle under test is tested according to the preset driving route. During the test, the cloud control platform collects and records test data and video data in real time.

[0033] And based on the driving status of the tested vehicle, the cloud-controlled simulation of traffic participants' trajectories is continuously optimized;

[0034] S8. After the test ends or is terminated abnormally, the tested vehicle drives out of the test area, and the cloud control platform dispatches the simulated traffic participants back to the storage location.

[0035] S9. Based on test data and video playback, design multi-objective evaluation indicators, combine expert scores, and use the analytic hierarchy process to complete the comprehensive performance evaluation of autonomous vehicles.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] This invention utilizes 5G communication and cloud computing technologies to establish a cloud-controlled testing system for autonomous vehicles. The system includes a cloud control platform, a dedicated testing network, simulated traffic participants, and a test vehicle equipped with a VBOX. The cloud control platform is the system's core, responsible for issuing test tasks, scheduling and planning control of simulated traffic participants throughout the testing process, recording test data, and real-time monitoring. The dedicated testing network serves as the medium for system data interaction, responsible for uploading vehicle status information and issuing test commands. The cloud-controlled simulated traffic participants are the execution link for generating dynamic traffic scenarios, consisting of cloud-controlled simulated non-motorized vehicles (pedestrians, bicycles, etc.) and cloud-controlled simulated dummy vehicles. The test vehicle's VBOX is responsible for acquiring the test vehicle's status information in real time. This enables continuous interaction between the simulated traffic participant cluster and the test vehicle, completing highly complete tests that closely resemble real traffic environments, covering high-risk, low-probability, and difficult-to-reproduce scenarios. It also offers advantages such as low cost, flexible motion control of simulated traffic participants, real-time cloud control, a simple testing process, and ease of operation.

[0038] This invention targets closed testing grounds for autonomous driving. It uses test scenarios and parametric conditions as system inputs and, based on testing requirements, enables cloud-controlled scheduling of simulated traffic participants to meet the performance requirements of test vehicles in real traffic environments. It performs trajectory planning for vehicle clusters and remotely controls simulated traffic participants in the cloud, achieving a collaborative, scenario-rich, fully unmanned, and automated testing solution, providing a new solution for the field of autonomous vehicle testing. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the system framework of the present invention;

[0040] Figure 2 This is a schematic diagram of the method flow of the present invention;

[0041] Figure 3 This is a vehicle route map for an autonomous vehicle roundabout test scenario in the embodiment;

[0042] Figure 4 This is a schematic diagram illustrating a scenario where an autonomous vehicle is driving around an island and another vehicle is merging into it, as shown in the example.

[0043] The markings in the diagram are as follows: 1. Cloud control platform; 2. Dedicated test network; 3. Simulated traffic participants; 4. Test vehicle. Detailed Implementation

[0044] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0045] Example

[0046] like Figure 1 As shown, an autonomous driving cloud control testing system includes a cloud control platform 1, a dedicated testing network 2, simulated traffic participants 3, and a test vehicle 4 equipped with a VBOX. The cloud control platform 1 is the center of the testing system, responsible for issuing test tasks, scheduling and planning control of simulated traffic participants throughout the testing process, recording test data, and real-time monitoring. The dedicated testing network 2 is the medium for data interaction within the testing system, responsible for uploading vehicle status information and issuing test commands. The cloud-controlled simulated traffic participants 3 are the execution link for generating dynamic traffic scenarios, consisting of cloud-controlled simulated non-motorized vehicles (pedestrians, bicycles, etc.) and cloud-controlled simulated dummy vehicles. The VBOX on the test vehicle 4 is responsible for acquiring the real-time status information of the test vehicle 4.

[0047] Specifically, the cloud control platform 1 consists of a test task configuration module, a cloud control algorithm module, a test data storage module, and a test video monitoring module. The test task configuration module is used to set the test scenario, traffic flow, and conflict intensity. The cloud control algorithm module mainly schedules simulated traffic participants and plans global paths based on the test task; in addition, based on the real-time location of the test vehicle, it generates challenging test conditions for the tested vehicle by optimizing local trajectories, while avoiding the risk of collisions between simulated traffic participants. The test data storage module is used to collect real-time status information of the tested vehicle during the test, including position, speed, heading, acceleration, etc., and simultaneously collect trajectory data of the cloud-controlled simulated traffic participants. This data information is saved to the test database of the storage module for record storage. The test video monitoring module collects video data of the entire test process of the tested vehicle and records the entire test process. The cloud control platform is deployed on the test server in the test site management room and is connected to the test video monitoring equipment and the dedicated 5G test base station in the test site via a fiber optic network.

[0048] Test Network 2 consists of dedicated test base stations and a 5G communication network. The dedicated test base stations are deployed within the test site, providing full coverage. The 5G communication network is a 5G network provided by an operator specifically for autonomous driving testing.

[0049] The cloud-controlled simulated traffic participant 3 consists of cloud-controlled simulated non-motorized vehicles (pedestrians, bicycles, etc.) and cloud-controlled simulated dummy vehicles. The cloud-controlled simulated non-motorized vehicles consist of a high-load-bearing omnidirectional motion chassis and simulated target objects. The omnidirectional motion chassis comprises a first 5G communication module, a first chassis information acquisition module, a motor control module, and a first power supply module; the simulated target objects include simulated dummies and simulated bicycles. During testing, the real-time omnidirectional motion chassis pose and status information collected by the first chassis information acquisition module is uploaded to the cloud control platform via the first 5G communication module. Real-time control commands are obtained through the first 5G communication module, and steering and motion commands are generated by the motor control module via Controller Area Network (CAN) communication, enabling omnidirectional motion control. The cloud-controlled simulated dummy vehicle consists of a fully steerable chassis and a detachable car cover. The fully steerable chassis is equipped with a second 5G communication module, a second chassis information acquisition module, a motion control module, and a second power supply module. The detachable car cover is made of polyethylene foam material, which is beneficial for collision cushioning. During testing, the real-time full-drive chassis pose information and full-drive chassis status information collected by the second chassis information acquisition module are uploaded to the cloud control platform through the second 5G communication module. Real-time control commands are obtained through the second 5G communication module, and the vehicle control unit (VCU), i.e. the motion control module, realizes chassis steering and drive / braking control through CAN communication.

[0050] The VBOX on the vehicle under test (4) consists of a high-precision positioning module, a third-party 5G communication module, a test data acquisition module, a parameter configuration module, and a third-party power supply module, suitable for various vehicle models. During testing, the VBOX is installed at the top center of the vehicle under test (4), and the measured vehicle length, width, wheelbase, and other parameters are calibrated using the parameter configuration module. During testing, the vehicle positioning information is uploaded to the cloud control platform in real time via the third-party 5G communication module, facilitating the generation of test conditions by the cloud control algorithm module. Simultaneously, the test data acquisition module uploads the test data to the test data storage module in real time, completing the data recording of the test process.

[0051] like Figure 2 As shown, this invention also proposes an autonomous driving cloud control testing method, the specific process of which includes the following steps:

[0052] (1) Install VBOX on the vehicle under test and configure and calibrate the parameters;

[0053] (2) Based on the test requirements and test level proposed by the vehicle under test, the test task configuration module generates the test scenario task and route, delivers the generated test route to the vehicle under test, and completes the test preparation work.

[0054] (3) Power on the VBOX of the cloud-controlled simulated traffic participants and the vehicle under test and check whether each module is working properly; if it is normal, proceed to step (4); otherwise, power off the device and return to step (3).

[0055] (4) Based on the test scenario task and the driving route of the vehicle under test provided by the test task configuration module, the cloud control algorithm module determines the type and number of simulated traffic participants required for the test, and schedules the simulated traffic participants for this test to their respective test starting points and presets the routes;

[0056] (5) After the simulated traffic participants arrive at their respective designated locations, the cloud control platform issues a preset control command, and the simulated traffic participants begin to move, generating a test traffic environment;

[0057] (6) The cloud control platform enables test data storage, test video monitoring and visualization functions, and issues a start test command based on the on-site status of the test site;

[0058] (7) The vehicle under test is tested according to the preset route. During the test, the cloud control platform test data storage module collects and records test data in real time, and the test video monitoring collects and records video data in real time. Test management personnel remotely monitor the test site test status through the visualization module. The cloud control algorithm module continuously optimizes the trajectory of the cloud control simulated traffic participants according to the driving status of the vehicle under test.

[0059] (8) After the test ends or is terminated abnormally, the vehicle under test drives out of the test area, and the cloud control platform dispatches the cloud control simulated traffic participants back to the storage location.

[0060] (9) Based on the test data and video playback, design multi-objective evaluation indicators, combine expert scores, and use the analytic hierarchy process to complete the comprehensive performance evaluation of autonomous vehicles.

[0061] To verify the effectiveness of this solution, this embodiment conducts a typical urban scenario test for autonomous vehicles—a roundabout traffic scenario test (e.g., Figure 3 and Figure 4 As shown in the image, the test subject is an electric sedan equipped with Level 3 autonomous driving capabilities, possessing autonomous perception, decision-making, and control abilities. This test evaluates the intelligence level of the vehicle in a roundabout scenario. Specifically, it tests autonomous driving behaviors such as merging into a roundabout, following another vehicle in a roundabout, and merging out of a roundabout in a multi-vehicle environment. Based on the capabilities of the vehicle under test and the testing requirements, the scenario test level is set to a general intense conflict scenario. The specific process steps are as follows:

[0062] (1) Install the VBOX at the top center of the vehicle under test, and use a configuration laptop to connect to the VBOX via RJ45 to configure the vehicle parameter information, where the vehicle length L is 4.8 meters, the width W is 1.8 meters, the height H is 1.5 meters, and the wheelbase B is 2.7 meters.

[0063] (2) Generate a closed-loop test route of roundabout-transition section-roundabout. The roundabout scenario configuration test tasks mainly include traffic behaviors such as merging into the roundabout, following the roundabout, and merging out of the roundabout. According to the test level, this test is equipped with 3 cloud-controlled simulated cars in this scenario. Vehicle A and vehicle B take a roundabout route. Vehicle C enters the roundabout from the south side of the road and exits the roundabout from the south side of the road to the end of the road and makes a U-turn, and drives in a loop.

[0064] (3) Power on all cloud-controlled simulated traffic participants and VBOX, and check if the equipment is working properly. If it is, proceed to step (4); otherwise, power off and repeat step (3).

[0065] (4) The cloud control platform dispatches simulated traffic participants to the test starting point and sets the route;

[0066] (5) The cloud control platform issues control commands to simulate the operation of traffic participants and generate a test traffic environment;

[0067] (6) The cloud control platform enables test data storage, test video monitoring and visualization functions, and issues a start test command based on the on-site status of the test site;

[0068] (7) The test vehicle is tested according to a preset route. During the test, the cloud control platform's test data storage module collects and records test data in real time, and the test video monitoring collects and records video data in real time. Test management personnel remotely monitor the test site through the visualization module. The cloud control algorithm module continuously optimizes the trajectory of simulated traffic participants based on the driving status of the test vehicle. Specifically, taking the scenario of autonomous driving around an island and vehicle C quickly merging into the roundabout as an example, such as... Figure 4 As shown.

[0069] (7-1) When the autonomous vehicle under test is driving in the roundabout, the cloud control platform collects the real-time status information of the vehicle under test and vehicle C, and constructs a kinematic model as shown in equation (1).

[0070]

[0071] (7-2) Based on the preset path, find the potential collision point of the vehicle, denoted as P, and calculate the vehicle priority, as shown in Equation (2);

[0072]

[0073] Where, x ego(t p )=0,x1(t p ) < 0; PL C and PL ego These represent the primacy of vehicle C and the tested vehicle, respectively; t p The moment when the tested vehicle arrives at the potential collision initiation point; x C and x ego The positions of vehicle C and the vehicle under test; l C The length of vehicle C; w ego The width of the vehicle being tested.

[0074] (7-3) Based on the current state, calculate the priority condition when vehicle C takes the lead and the time difference between the two vehicles reaching the potential collision point is greater than the threshold ε under this test level, as shown in Equation (3).

[0075]

[0076] (7-4) Based on the preemptive condition obtained in (7-3), with minimizing the vehicle C acceleration as the objective function, an optimization model is constructed as follows, and the dynamic programming algorithm is used to solve it;

[0077]

[0078] st

[0079] 0≤v≤v max

[0080] 0≤a≤a max

[0081] Where N is the prediction optimization period, v max and a max These are the maximum values ​​of velocity and acceleration, respectively, and should also satisfy constraints (1) to (3).

[0082] (7-5) Adjust the trajectory of vehicle C based on the solution of (7-4) to generate a conflict scenario.

[0083] (8) If a collision occurs based on the actual operating performance of the vehicle under test, the test ends; otherwise, the test continues until the end.

[0084] (9) Based on the test data and video playback, a multi-objective evaluation method is adopted, using indicators such as safety, traffic efficiency, and comfort to construct a test evaluation system based on the analytic hierarchy process. Integrating expert scores, the final comprehensive performance evaluation of the tested vehicle under this scenario is obtained. Specific steps include:

[0085] (9-1) This test selected safety, traffic efficiency and comfort indicators. In order to determine the influence factor of each indicator on the objective function, a 3×3 judgment matrix A was constructed, which was established according to the importance of each element in the same layer. The specific expression is shown in equation (4).

[0086]

[0087] In the formula, a ij This represents the importance of indicator i to indicator j, where a ii =1, a ij >0 and a ij =1 / a ji The 1-9 scale method is used to quantify the relative importance of a certain indicator to other indicators within the same level, and to determine the elements a in the judgment matrix A. ij The value of .

[0088] (9-2) Calculate the eigenvector W and the largest eigenvalue λ max .

[0089] W i =(Πa ii ) 1 / n (5)

[0090] λ max =∑(AW) i / nW i (6)

[0091] (9-3) Consistency check

[0092] By checking whether the error is within a reasonable range, an average random consistency index is established, as shown in equation (7).

[0093]

[0094] In summary, this solution uses the test scenario and parametric conditions as system inputs. Based on the test requirements, it schedules a cluster of simulated traffic participants in the cloud to match the performance requirements of test vehicles in real traffic environments. It performs trajectory planning for the vehicle cluster and remotely controls the simulated traffic participants in the cloud. This enables continuous interaction between the simulated traffic participant cluster and the test vehicles, completing highly complete autonomous driving tests that closely resemble real traffic environments and cover high-risk, low-probability, and difficult-to-reproduce scenarios.

Claims

1. An autonomous driving cloud control testing system, applicable to autonomous driving scenarios involving rapid vehicle merging during roundabout driving, characterized in that, It includes a cloud control platform (1), a dedicated test network (2), simulated traffic participants (3), and a vehicle under test (4) equipped with a VBOX. The cloud control platform (1) is used to issue test tasks, schedule and plan the control of simulated traffic participants (3), and record and monitor test data in real time. The cloud control platform (1) includes a test task configuration module, a cloud control algorithm module, a test data storage module, and a test video monitoring module. The test task configuration module is used to set the test scenario, traffic flow, and intensity of conflict. The cloud control algorithm module schedules the simulated traffic participants (3) according to the test task and plans the global path; and generates a test condition that is challenging for the test vehicle (4) by optimizing the local trajectory according to the real-time position of the test vehicle, while avoiding the risk of collision between simulated traffic participants (3); The test data storage module is used to receive and store the real-time status information of the vehicle under test (4) and the trajectory data of the simulated traffic participants (3) during the test. The test video monitoring module is used to collect video data of the entire test process of the vehicle under test and record the entire test process; The process of generating test conditions by the cloud control algorithm module includes: Collect real-time status information of the tested vehicle (4) and simulated traffic participants (3) to construct a kinematic model; Based on the preset path, find the potential collision point of the vehicle, denoted as P, and calculate the vehicle's priority: in, , ; and The precession of the simulated traffic participants (3) and the tested vehicle (4) are respectively; The moment when the tested vehicle (4) arrives at the potential collision initiation point; and To simulate the positions of traffic participants (3) and the tested vehicle (4); To simulate the vehicle length of traffic participant (3); The width of the vehicle being tested (4); Based on the current state, it is calculated that the simulated traffic participant (3) is ahead, and the time difference between the arrival of the two vehicles at the potential collision point is greater than the threshold under this test level. The prior condition is used to construct an optimization model with the objective function of minimizing the acceleration of the simulated traffic participant (3), and the dynamic programming algorithm is used to solve it. The trajectory of the simulated traffic participant (3) is adjusted according to the solution to generate a conflict scenario. The dedicated test network (2) is used to upload vehicle status information and issue test instructions; The simulated traffic participants (3) consist of cloud-controlled simulated vehicles; The VBOX is used to acquire the status information of the vehicle under test (4) in real time.

2. The autonomous driving cloud control testing system according to claim 1, characterized in that, The status information of the vehicle under test (4) includes position, speed, heading, and acceleration.

3. The autonomous driving cloud control testing system according to claim 1, characterized in that, The test network includes a dedicated test base station and a 5G communication network. The dedicated test base station is deployed within the test site to achieve full coverage of the test site. The 5G communication network mentioned is specifically a 5G network provided by operators for autonomous driving testing.

4. The autonomous driving cloud control testing system according to claim 3, characterized in that, The cloud control platform (1) is deployed on the test server in the test site management room and is connected to the test video monitoring equipment and test base station in the test site through a fiber optic network.

5. The autonomous driving cloud control testing system according to claim 1, characterized in that, The cloud-controlled simulated car includes a fully steerable chassis and a detachable car cover. The fully steerable chassis is equipped with a second 5G communication module, a second chassis information acquisition module, a motion control module, and a second power supply module. The second chassis information acquisition module is used to collect the full steerable chassis position and posture information and the full steerable chassis status information, and uploads them to the cloud control platform (1) through the second 5G communication module. The motion control module receives real-time control commands through the second 5G communication module to realize chassis steering and drive / brake control.

6. The autonomous driving cloud control testing system according to claim 5, characterized in that, The removable car cover is made of polyethylene foam, which helps to cushion collisions.

7. The autonomous driving cloud control testing system according to claim 1, characterized in that, The VBOX includes a positioning module, a third 5G communication module, a test data acquisition module, a parameter configuration module, and a third power supply module. The positioning module is used to collect the positioning information of the vehicle under test (4) and upload it to the cloud control platform (1) through the third 5G communication module. The cloud control platform (1) generates the test conditions. The test data acquisition module is used to transmit the real-time acquired test data to the test data storage module for storage and recording via the third 5G communication module; The parameter configuration module is used to configure the size parameter information of the vehicle under test (4) in order to achieve the calibration of the vehicle under test (4).

8. An autonomous driving cloud control testing method using the autonomous driving cloud control testing system as described in claim 1, characterized in that, Includes the following steps: S1. Install VBOX on the vehicle under test and configure and calibrate the parameters; S2. Based on the test requirements and test level proposed by the vehicle under test, the cloud control platform generates the test scenario task and driving route, delivers the generated test route to the vehicle under test, and completes the test preparation work. S3. Simulate traffic participants and the vehicle under test. Power on the VBOX and check whether each module is working properly. If everything is normal, proceed to step S4; otherwise, power off the device and return to step S3 to execute again. S4. Based on the test scenario and driving route, the cloud control platform determines the type and number of simulated traffic participants required for the test, dispatches the simulated traffic participants to their respective test starting points, and presets routes. S5. After the simulated traffic participants arrive at their respective designated locations, the cloud control platform issues preset control commands, and the simulated traffic participants begin to move, generating a test traffic environment. S6, the cloud control platform enables test data storage, test video monitoring and visualization functions, and issues a start test command based on the on-site status of the test site; S7. The vehicle under test is tested according to the preset driving route. During the test, the cloud control platform collects and records test data and video data in real time. And based on the driving status of the tested vehicle, the cloud-controlled simulation of traffic participants' trajectories is continuously optimized; S8. After the test ends or is terminated abnormally, the tested vehicle drives out of the test area, and the cloud control platform dispatches the simulated traffic participants back to the storage location. S9. Based on test data and video playback, design multi-objective evaluation indicators, combine expert scores, and use the analytic hierarchy process to complete the comprehensive performance evaluation of autonomous vehicles.

Images