Real vehicle test method and device for autonomous vehicle in multi-vehicle interaction scenario

Abstract

The present invention relates to the technical field of autonomous vehicles, and specifically discloses a method for real vehicle testing of autonomous vehicles for multi-vehicle interaction scenarios, including: calculating fusion data from the lidar point cloud data, millimeter wave radar millimeter wave point cloud data, and camera video data obtained from the test vehicle; calculating the position and speed data of the vehicles around the test vehicle based on the fusion data, and at the same time calculating the precise positioning data of the test vehicle based on the fusion data and the pose data of the integrated navigation sensor; calculating the relative position and relative speed data between the test vehicle and the adjacent vehicles among the surrounding vehicles based on the position and speed data of the surrounding vehicles and the precise positioning data of the test vehicle, and outputting the test results of the vehicle under test. The present invention also discloses a real vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios. The present invention can expand and complete more complex multi-vehicle interaction scenario tests.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous vehicles, and more specifically, to a method and device for real vehicle testing of autonomous vehicles for multi-vehicle interaction scenarios. Background Art

[0002] Autonomous driving is an important trend in the current development of the automotive industry. Although significant progress has been made in autonomous driving technology in recent years, there is still a certain gap from commercial operation, and it needs to be iterated and improved through public road testing. To meet the growing demand for public road testing of autonomous vehicles and ensure basic road traffic safety, in 2018, the Ministry of Industry and Information Technology, the Ministry of Public Security, and the Ministry of Transport jointly issued the "Administrative Regulations on Road Testing of Intelligent Connected Vehicles (Trial)", and revised and issued the "Administrative Regulations on Road Testing and Demonstration Applications of Intelligent Connected Vehicles (Trial)" (hereinafter referred to as the "Administrative Regulations") in 2021 to guide public road testing activities of intelligent connected vehicles.

[0003] The Administrative Regulations stipulate 8 general inspection items for autonomous driving functions of intelligent connected vehicles. The fourth item is "recognition and response to the driving status of surrounding vehicles", that is, real vehicle testing is carried out on the interaction between autonomous vehicles and other surrounding vehicles. In the actual operation process, the current real vehicle testing mainly relies on installing combined navigation devices on the test vehicle and surrounding target vehicles, and evaluating the autonomous driving ability of autonomous vehicles by measuring data such as the positions, speeds, and accelerations of the two vehicles.

[0004] Currently, similar testing devices such as VBOX and RT have been widely used in real vehicle testing, and the testing methods have been relatively mature. However, since equipment needs to be installed on each vehicle, the installation and debugging time is long, and at the same time, due to the high cost of the equipment, it is difficult to expand to multi-vehicle interaction scenarios. Summary of the Invention

[0005] To solve the deficiencies in the prior art, the present invention provides a method and device for real vehicle testing of autonomous vehicles for multi-vehicle interaction scenarios, which are used to conveniently carry out testing of multi-vehicle interaction scenarios of autonomous vehicles.

[0006] As the first aspect of the present invention, a method for real vehicle testing of autonomous vehicles for multi-vehicle interaction scenarios is provided, including the following steps:

[0007] Step S1: Obtain the laser point cloud data of the lidar, the millimeter wave point cloud data of the millimeter wave radar, the video data of the camera, and the pose data of the combined navigation sensor on the test vehicle respectively;

[0008] Step S2: Perform fusion calculation on the lidar point cloud data, millimeter-wave radar point cloud data, and camera video data of the test vehicle to obtain fusion data;

[0009] Step S3: Calculate the position and speed data of the vehicles around the test vehicle based on the fusion data, and simultaneously perform synchronous positioning calculation based on the fusion data and the pose data of the integrated navigation sensor to obtain the precise positioning data of the test vehicle;

[0010] Step S4: Perform relative data calculation based on the position and speed data of the vehicles around the test vehicle and the precise positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and the adjacent vehicles among the vehicles around the test vehicle;

[0011] Step S5: Output the test result of the vehicle under test according to the relative position and relative speed data between adjacent vehicles, where the vehicle under test is one of the vehicles around the test vehicle.

[0012] Further, the separately obtaining the lidar point cloud data, millimeter-wave radar point cloud data, camera video data, and pose data of the integrated navigation sensor of the test vehicle further includes:

[0013] Before the test starts, the RTK base station completes initialization to achieve differential positioning;

[0014] After the lidar, millimeter-wave radar, camera, and integrated navigation sensor are started, the warm-up process is completed, and the collected data is output respectively.

[0015] Further, it further includes:

[0016] Save data before the test starts, monitor the test process in a multi-vehicle interaction scenario, and output relevant test results after the test ends;

[0017] Evaluate the performance of the vehicle under test in the test scenario according to the output relevant test results.

[0018] Further, the monitoring of the test process in a multi-vehicle interaction scenario and the output of relevant test results after the test ends further includes:

[0019] (1) Test whether the vehicle under test can respond effectively when other vehicles suddenly appear in the same lane after the leading test vehicle cuts out during the following process of the vehicle under test. Among them, the vehicles involved in the test scenario are divided into the vehicle under test, the test vehicle, and other vehicles;

[0020] (2) Select a long straight three-lane road as the test road;

[0021] At the start of the test, the three vehicles are in the same lane. Both the vehicle under test and the test vehicle are traveling at a speed of 60 km / h. There is another vehicle traveling at 30 km / h 200 m in front of the test vehicle.

[0022] During the test, when the test vehicle cuts into the adjacent right lane at a distance of 30 m from the vehicle in front, the vehicle under test should recognize the other vehicle in the same lane ahead and decelerate.

[0023] (3) At the start of the test, record the initial driving state data of the three vehicles. During the test, transmit the position and speed data of the three vehicles in real time, and at the same time calculate the relative data between the three vehicles.

[0024] (4) After the test is completed, export the relative data between the three vehicles.

[0025] Furthermore, evaluating the performance of the vehicle under test in the test scenario according to the relevant test results output further includes:

[0026] Judging the test results of the vehicle under test according to the exported relative data between the three vehicles;

[0027] Among them, the vehicle under test can recognize and respond to other vehicles after the test vehicle cuts out, and the deceleration does not exceed 4 m / s 2 , and at the same time, the vehicle under test does not collide with other vehicles.

[0028] As the second aspect of the present invention, there is provided a real vehicle test device for an autonomous vehicle for multi-vehicle interaction scenarios, which is installed on a test vehicle and is used to implement the real vehicle test method for an autonomous vehicle for multi-vehicle interaction scenarios described above. The real vehicle test device for an autonomous vehicle for multi-vehicle interaction scenarios includes:

[0029] A perception module for respectively collecting the laser point cloud data, millimeter wave point cloud data and video data of the vehicles around the test vehicle;

[0030] A positioning module for collecting the pose data of the test vehicle;

[0031] A calculation module, configured to perform fusion calculation on the lidar point cloud data, millimeter wave point cloud data, and video data of the vehicles around the test vehicle to obtain fusion data; calculate the position and speed data of the vehicles around the test vehicle based on the fusion data, and simultaneously perform synchronous positioning calculation based on the fusion data and the pose data of the test vehicle to obtain the accurate positioning data of the test vehicle; perform relative data calculation based on the position and speed data of the vehicles around the test vehicle and the accurate positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and an adjacent vehicle among the vehicles around the test vehicle; output the test result of the vehicle under test according to the relative position and relative speed data between adjacent vehicles, where the vehicle under test is one of the vehicles around the test vehicle.

[0032] Further, the perception module further includes a lidar, a millimeter wave radar, and a camera;

[0033] The lidar is installed on the roof of the test vehicle and is configured to collect the lidar point cloud data of the vehicles around the test vehicle;

[0034] The millimeter wave radar is installed above the license plate frame of the test vehicle and is configured to collect the millimeter wave point cloud data of the vehicles around the test vehicle;

[0035] The camera is installed at the front windshield of the test vehicle and is configured to collect the video data of the vehicles around the test vehicle.

[0036] Further, the positioning module further includes a combined navigation sensor, and the combined navigation sensor is installed inside the test vehicle and is configured to collect the pose data of the test vehicle.

[0037] Further, the combined navigation sensor includes a GNSS receiver and an inertial measurement unit, and is capable of receiving RTK base station signals to improve the positioning accuracy of the test vehicle;

[0038] Wherein, the GNSS receiver performs combined navigation on the received satellite navigation information to obtain the pitch, roll, heading, position, speed, and time data of the test vehicle, and outputs the position, speed, and attitude information of the test vehicle after inertial calculation.

[0039] Further, the calculation module includes an industrial computer, and the industrial computer is installed inside the test vehicle and is configured to evaluate the performance of the vehicle under test in the test scenario according to the test result of the vehicle under test.

[0040] The real vehicle test method for autonomous vehicles provided by the present invention for multi-vehicle interaction scenarios has the following advantages:

[0041] (1) There is no need to install and debug the test device on the vehicle to be tested because the test device itself is also a test vehicle in the test scenario. After the initial debugging and setting are completed, only simple self-positioning needs to be done, and it can be used for subsequent vehicle-to-vehicle interaction scenario tests.

[0042] (2) Provide a multi-vehicle interaction in-vehicle test device that is easy to expand the number of participating vehicles in the scenario. Since the test device can sense multiple surrounding vehicles, for multi-vehicle interaction scenarios with more than two vehicles, it can also be completed through this test device without having to install corresponding test devices on the additional vehicles, and it can further expand to complete more complex multi-vehicle interaction scenario tests. Description of the Drawings

[0043] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention, but do not constitute a limitation to the present invention.

[0044] Figure 1 It is a flowchart of the in-vehicle test method for autonomous vehicles facing multi-vehicle interaction scenarios provided by the present invention.

[0045] Figure 2 It is a flowchart of the specific implementation manner of the in-vehicle test method for autonomous vehicles facing multi-vehicle interaction scenarios provided by the present invention.

[0046] Figure 3 It is an installation schematic diagram of the in-vehicle test device for autonomous vehicles facing multi-vehicle interaction scenarios provided by the present invention.

[0047] Figure 4 It is a working principle diagram of the in-vehicle test device for autonomous vehicles facing multi-vehicle interaction scenarios provided by the present invention.

[0048] Figure 5 It is a schematic diagram of the vehicle exiting the scenario provided by the present invention. Specific Embodiments

[0049] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0050] To further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following will, in conjunction with the accompanying drawings and preferred embodiments, elaborate in detail on the specific implementation manner, structure, features, and effects of the real vehicle test method for an autonomous vehicle facing a multi-vehicle interaction scenario proposed according to the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention.

[0051] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances for the embodiments of the present invention described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily need to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0052] In this embodiment, a real vehicle test method for an autonomous vehicle facing a multi-vehicle interaction scenario is provided, as Figure 1 shown. The real vehicle test method for an autonomous vehicle facing a multi-vehicle interaction scenario includes:

[0053] Step S1: Obtain the laser point cloud data of the lidar, the millimeter-wave point cloud data of the millimeter-wave radar, the video data of the camera, and the pose data of the integrated navigation sensor on the test vehicle respectively;

[0054] Step S2: Perform fusion calculation on the laser point cloud data of the lidar, the millimeter-wave point cloud data of the millimeter-wave radar, and the video data of the camera on the test vehicle to obtain fusion data;

[0055] Step S3: Calculate the position and speed data of the vehicles around the test vehicle based on the fusion data, and at the same time perform synchronous positioning calculation based on the fusion data and the pose data of the integrated navigation sensor to obtain the precise positioning data of the test vehicle;

[0056] Step S4: Perform relative data calculation based on the position and speed data of the vehicles around the test vehicle and the precise positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and the adjacent vehicles among the vehicles around the test vehicle;

[0057] Step S5: Output the test result of the vehicle under test according to the relative position and relative speed data between adjacent vehicles, where the vehicle under test is one of the vehicles surrounding the test vehicle.

[0058] As Figure 2 shown, the specific process of the in-vehicle test method for autonomous vehicles facing multi-vehicle interaction scenarios is as follows:

[0059] First, before the test starts, the RTK base station completes initialization to achieve differential positioning; after the lidar, millimeter-wave radar, camera, and integrated navigation sensor are started, the warm-up process is completed, and the collected data is output respectively.

[0060] Second, select the vehicle cut-out scenario as a typical multi-vehicle interaction scenario. The scenario schematic diagram is as Figure 5 shown. Save the data before the test starts, monitor the test process in the vehicle cut-out scenario, and output the relevant test results after the test ends.

[0061] Specifically, it also includes:

[0062] (1) As Figure 5 shown, test whether the vehicle under test can effectively respond when other vehicles suddenly appear in the same lane after the test vehicle in front cuts out during the following process. Among them, according to the multi-vehicle interaction scenario, the vehicles involved in the test scenario are divided into the vehicle under test ①, the test vehicle ②, and other vehicles ③.

[0063] (2) According to the test purpose, select the test area and design the route trajectories and interaction timings of the three types of vehicles. For example, select a long straight three-lane road as the test road.

[0064] At the start of the test, the three vehicles are in the same lane. The vehicle under test and the test vehicle are both driving at a speed of 60 km / h, and 200 m in front of the test vehicle is another vehicle driving at a speed of 30 km / h.

[0065] During the test, when the test vehicle cuts out to the adjacent right lane at a distance of 30 m from the vehicle in front, the vehicle under test should recognize the other vehicle in the same lane in front and decelerate.

[0066] (3) After confirming that all vehicles are in the state to be tested, start the test. At the start of the test, record the initial driving state data of the three vehicles. The other vehicles except the vehicle under test complete the predetermined actions according to the interaction state, and the vehicle under test completes the recognition and response to the surrounding vehicles according to the interaction state. During the test, the positions and speed data of the three vehicles are transmitted in real time, and the relative data between the three vehicles are calculated at the same time.

[0067] (4) After the test is completed, export the relative position, relative speed and other data between the three vehicles.

[0068] Third, according to the relevant test results output, evaluate the performance of the vehicle under test in the test scenario according to the rules for each scenario.

[0069] Preferably, evaluating the performance of the vehicle under test in the test scenario according to the relevant test results output further includes:

[0070] Judging the test results of the vehicle under test according to the relative data exported among the three vehicles;

[0071] Among them, the vehicle under test can identify and respond to other vehicles after the test vehicle cuts out, and the deceleration does not exceed 4 m / s 2 , and at the same time, the vehicle under test does not collide with other vehicles. There is a safety officer on the vehicle under test. If there is a risk of collision, the safety officer takes over the vehicle under test.

[0072] As another embodiment of the present invention, as Figure 4 shown, a real vehicle test device for an autonomous vehicle for a multi-vehicle interaction scenario is provided. The test device is installed on the test vehicle. The real vehicle test device for an autonomous vehicle for a multi-vehicle interaction scenario includes:

[0073] A perception module, configured to respectively collect the laser point cloud data, millimeter wave point cloud data, and video data of the vehicles around the test vehicle;

[0074] A positioning module, configured to collect the pose data of the test vehicle;

[0075] A calculation module, configured to perform fusion calculation on the laser point cloud data, millimeter wave point cloud data, and video data of the vehicles around the test vehicle to obtain fusion data; calculate the position and speed data of the vehicles around the test vehicle according to the fusion data, and at the same time perform synchronous positioning calculation according to the fusion data and the pose data of the test vehicle to obtain the accurate positioning data of the test vehicle; perform relative data calculation according to the position and speed data of the vehicles around the test vehicle and the accurate positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and the adjacent vehicles among the vehicles around the test vehicle; output the test results of the vehicle under test according to the relative position and relative speed data between the adjacent vehicles, where the vehicle under test is one of the vehicles around the test vehicle.

[0076] Preferably, as Figure 4 shown, the perception module further includes a lidar, a millimeter wave radar, and a camera; the lidar, the millimeter wave radar, and the camera are respectively installed on the test vehicle, and the test vehicle is operated by a driver;

[0077] The lidar is installed on the roof of the test vehicle and is used to collect the lidar point cloud data of the vehicles around the test vehicle;

[0078] The millimeter-wave radar is installed above the license plate holder of the test vehicle and is used to collect the millimeter-wave point cloud data of the vehicles around the test vehicle;

[0079] The camera is installed at the front windshield of the test vehicle and is used to collect the video data of the vehicles around the test vehicle.

[0080] In the embodiment of the present invention, as Figure 3 shown, the test vehicle is a mass-produced passenger vehicle, and the energy type is pure electric. The pure electric vehicle is convenient for modules such as the perception module, the positioning module, and the calculation module to take points. The perception module mainly consists of a lidar, a millimeter-wave radar, and a camera, and the installation positions are the roof, above the license plate holder, and the front windshield respectively. The three sensors can output lidar point cloud, millimeter-wave point cloud, and video image respectively.

[0081] Specifically, the test vehicle is a vehicle that can be normally maneuvered and driven. There are installation interfaces for the perception module on the vehicle body, and there are installation positions for the positioning module and the calculation module inside the vehicle, and it can supply power to the additionally installed electronic devices.

[0082] Specifically, the perception module includes sensors such as lidar and camera, and can sense the surrounding environment information, including obstacles, vehicles, infrastructure, etc.

[0083] Specifically, the positioning module includes components such as integrated navigation, antenna, and RTK radio station, and can achieve its own precise positioning.

[0084] Specifically, the calculation module is a calculation unit that can perform tasks such as integrated positioning and data processing. A high-precision map of the test device operation site is stored in the calculation module for integrated positioning. The calculation module can receive the position, speed, etc. data of all participating vehicles in the test, and calculate the relative distance, speed, etc. between vehicles for test evaluation.

[0085] Preferably, the positioning module further includes an integrated navigation sensor, and the integrated navigation sensor is installed inside the test vehicle and is used to collect the pose data of the test vehicle.

[0086] Preferably, the integrated navigation sensor includes a GNSS receiver and an inertial measurement unit, and can receive the RTK base station signal to improve the positioning accuracy of the test vehicle;

[0087] Among them, the GNSS receiver performs integrated navigation on the received satellite navigation information to obtain the pitch, roll, heading, position, speed, and time data of the test vehicle, and outputs the position, speed, and attitude information of the test vehicle after inertial calculation, with a certain navigation accuracy maintaining function within a short period of time.

[0088] Preferably, the calculation module includes an industrial personal computer installed inside the test vehicle, which is used to evaluate the performance of the vehicle under test in the test scenario according to the test results of the vehicle under test.

[0089] The embodiment of the present invention also provides a real vehicle test system for autonomous vehicles facing multi-vehicle interaction scenarios, which is installed on the test vehicle where the test device is located and has functions such as parameter display, data saving, data calculation, and result output.

[0090] The parameter display function includes parameter value display and parameter visualization, and the status data of the vehicles involved in the scenario, such as vehicle speed, position, etc., are displayed in real time on the system interface; at the same time, based on the vehicle position data and the high-precision map, the relative positions between all vehicles and their relative positions to lanes, infrastructure, and obstacles are displayed in three dimensions, and the three-dimensional display interface should be able to be updated in real time.

[0091] The data saving function mainly saves the selected parameters to be saved together with the time stamp in a certain format.

[0092] The data calculation function mainly calculates the data that cannot be directly obtained by sensors but is required for test evaluation.

[0093] The result data function mainly outputs the test data and evaluation data in a certain format.

[0094] The specific description of the test system functions is as follows.

[0095] 1. Overall function description:

[0096] The main function is to display and be able to output and save the status data of the test vehicle itself, the status data of surrounding vehicles, and the process data.

[0097] 2. Data types

[0098] The status data of the test vehicle itself are mainly position, speed, acceleration, heading angle, etc.

[0099] The status data of surrounding vehicles are mainly the identity ID, speed, relative distance, position, etc. of surrounding vehicles.

[0100] The process data are mainly the flag bits in the positioning calculation process of the test vehicle itself, such as position update flag, positioning calculation completion flag, etc.

[0101] 3. Data storage

[0102] The system can store and export three types of data, and the export format can be csv, xlsx or other tabular formats, etc. The exported content should include two types of data and the corresponding timestamps. The save and export functions should have conventional functions such as selecting the save path and the export path.

[0103] 4. Human-computer interaction

[0104] The main body of the test system should be installed on the industrial control computer of the test device, and after running, it can be interacted with through a WEB browser or other independent windows.

[0105] 5. Software interface

[0106] The interface includes three parts. One is the 3D display; the second is the real-time display of 3 types of data; the third is the menu bar, which has function buttons such as saving data and exporting data.

[0107] The embodiment of the present invention also provides a non-transitory computer storage medium, which stores computer-executable instructions that can execute the real vehicle test of the autonomous vehicle for the multi-vehicle interaction scenario in any of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (abbreviation: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0108] The above is only a preferred embodiment of the present invention, and it does not impose any form of limitation on the present invention. Although the present invention has been disclosed above with a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments by using the above-disclosed technical content within the scope of the technical solution of the present invention. However, as long as it does not depart from the content of the technical solution of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A real vehicle test method for autonomous vehicles in a multi-vehicle interaction scenario, characterized in that, Including the following steps: Step S1: Obtain the lidar point cloud data, millimeter-wave radar millimeter-wave point cloud data, camera video data, and pose data of the integrated navigation sensor on the test vehicle respectively; Step S2: Perform fusion calculation on the lidar point cloud data, millimeter-wave radar millimeter-wave point cloud data, and camera video data on the test vehicle to obtain fusion data; Step S3: Calculate the position and speed data of the vehicles around the test vehicle based on the fusion data, and at the same time perform synchronous positioning calculation based on the fusion data and the pose data of the integrated navigation sensor to obtain the precise positioning data of the test vehicle; Step S4: Perform relative data calculation based on the position and speed data of the vehicles around the test vehicle and the precise positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and the adjacent vehicles among the vehicles around the test vehicle; Step S5: Output the test result of the vehicle under test according to the relative position and relative speed data between adjacent vehicles, where the vehicle under test is one of the vehicles around the test vehicle; Among them, it also includes: Start saving data before the test starts, monitor the test process in a multi-vehicle interaction scenario, and output relevant test results after the test ends; Evaluate the performance of the vehicle under test in the test scenario according to the output relevant test results; Among them, the monitoring of the test process in a multi-vehicle interaction scenario and the output of relevant test results after the test ends also include: (1) Test whether the vehicle under test can respond effectively when other vehicles suddenly appear in the same lane after the leading test vehicle cuts out during the following process. Among them, the vehicles involved in the test scenario are divided into the vehicle under test, the test vehicle, and other vehicles; (2) Select a long straight three-lane road as the test road; At the start of the test, the three vehicles are in the same lane, both the vehicle under test and the test vehicle are driving at a speed of 60 km / h, and 200 m in front of the test vehicle is another vehicle driving at a speed of 30 km / h; During the test, when the test vehicle cuts out to the adjacent lane on the right at a distance of 30 m from the vehicle in front, the vehicle under test should recognize the other vehicle in front in the same lane and decelerate; (3) At the start of the test, record the initial driving state data of the three vehicles; during the test, transmit the position and speed data of the three vehicles in real time, and at the same time calculate the relative data between the three vehicles; (4) After the test is completed, export the relative data between the three vehicles; Among them, the evaluation of the performance of the vehicle under test in the test scenario according to the output relevant test results also includes: Judge the test result of the vehicle under test according to the exported relative data between the three vehicles; Among them, the vehicle under test can identify and respond to other vehicles after the test vehicle cuts out, with a deceleration not exceeding 4 m / s 2 , and at the same time, the vehicle under test does not collide with other vehicles.

2. The real vehicle test method for an autonomous vehicle for a multi-vehicle interaction scenario according to claim 1, wherein, The obtaining of the lidar point cloud data, millimeter-wave radar millimeter-wave point cloud data, camera video data, and pose data of the integrated navigation sensor on the test vehicle respectively also includes: Before the test starts, the RTK base station completes initialization to achieve differential positioning; After the lidar, millimeter-wave radar, camera, and integrated navigation sensor are started, they complete the warm-up process and respectively output the collected data.

3. A real vehicle test device for an autonomous vehicle in a multi-vehicle interaction scenario, installed on a test vehicle, for implementing the real vehicle test method for an autonomous vehicle in a multi-vehicle interaction scenario described in any one of claims 1 to 2, characterized in that, The in-vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios includes: A perception module, configured to respectively collect lidar point cloud data, millimeter-wave point cloud data, and video data of the vehicles around the test vehicle; A positioning module, configured to collect the pose data of the test vehicle; A calculation module, configured to perform fusion calculation on the lidar point cloud data, millimeter-wave point cloud data, and video data of the vehicles around the test vehicle to obtain fusion data; calculate the position and speed data of the vehicles around the test vehicle based on the fusion data, and at the same time perform synchronous positioning calculation based on the fusion data and the pose data of the test vehicle to obtain the precise positioning data of the test vehicle; perform relative data calculation based on the position and speed data of the vehicles around the test vehicle and the precise positioning data of the test vehicle to obtain the relative position and relative speed data between the test vehicle and an adjacent vehicle among the vehicles around the test vehicle; output the test result of the measured vehicle according to the relative position and relative speed data between adjacent vehicles, where the measured vehicle is one of the vehicles around the test vehicle.

4. The in-vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios according to claim 3, characterized in that The perception module further includes a lidar, a millimeter-wave radar, and a camera; The lidar is installed on the roof of the test vehicle and is configured to collect lidar point cloud data of the vehicles around the test vehicle; The millimeter-wave radar is installed above the license plate holder of the test vehicle and is configured to collect millimeter-wave point cloud data of the vehicles around the test vehicle; The camera is installed at the front windshield of the test vehicle and is configured to collect video data of the vehicles around the test vehicle.

5. The on-vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios according to claim 3, wherein The positioning module further includes an integrated navigation sensor, and the integrated navigation sensor is installed inside the test vehicle and is configured to collect the pose data of the test vehicle.

6. The on-vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios according to claim 5, wherein The integrated navigation sensor includes a GNSS receiver and an inertial measurement unit, and is capable of receiving RTK base station signals to improve the positioning accuracy of the test vehicle; Wherein, the GNSS receiver performs integrated navigation on the received satellite navigation information to obtain the pitch, roll, heading, position, speed, and time data of the test vehicle, and outputs the position, speed, and attitude information of the test vehicle after inertial solution.

7. The on-vehicle test device for autonomous vehicles for multi-vehicle interaction scenarios according to claim 3, characterized in that The calculation module includes an industrial computer, and the industrial computer is installed inside the test vehicle and is configured to evaluate the performance of the measured vehicle in the test scenario according to the test result of the measured vehicle.

Images