Method and system for testing autonomous vehicles

Abstract

The application discloses a kind of test method and system of automatic driving vehicle, belong to the field of automatic driving test.In the method, simulation platform is based on the virtual test scene of test vehicle, virtual scene information and virtual sensor information are sent to test vehicle, so that automatic driving controller controls test vehicle to carry out automatic driving based on these information, and motion state information and related control signal in the process of automatic driving are sent to simulation platform, and test result is determined by simulation platform.Due to the above-mentioned virtual sensor information includes virtual noise signal, so the authenticity of virtual sensor information is improved, so that the test process is closer to real road conditions, and the effectiveness of the test method is improved.Moreover, in the above-mentioned test process, the AR display on the test vehicle can display the fusion picture of virtual test scene and real driving environment in real time, which is convenient for relevant personnel to intuitively observe the interaction between test vehicle and virtual driving environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automatic driving test, in particular to a test method and system of an automatic driving vehicle. BACKGROUND

[0002] Before the automatic driving vehicle is put into commercial use, it must be fully tested and verified in terms of functional safety and performance safety to ensure the safety of consumers and the public. Generally, the test content includes sensors, algorithms, actuators, human-machine interfaces, etc., and from the aspects of application functions, performance, stability and robustness, functional safety, intended functional safety, type certification, etc., to ensure that the vehicle can be on the road autonomously.

[0003] In the related art, XiL (X in the Loop) technology (a method of incorporating human feedback and decision-making into the system control loop) is used to test the automatic driving vehicle. For example, using XiL technology, a simulation platform is used to model the hardware of the automatic driving vehicle, digitize the automatic driving environment, and parameterize the automatic driving scene, etc., to test the functional safety and performance safety of the automatic driving vehicle.

[0004] However, the above test process has poor authenticity, resulting in low accuracy of the test results. SUMMARY

[0005] The embodiments of the present application provide a test method and system of an automatic driving vehicle, and the technical solutions are as follows:

[0006] In a first aspect, a test method of an automatic driving vehicle is provided, which is applied to a test system of the automatic driving vehicle, the system comprising a simulation platform and a test vehicle, the test vehicle having a plurality of sensors, an automatic driving controller, and an augmented reality (AR) display, and the method comprising:

[0007] The simulation platform constructs a virtual test scene of the test vehicle, and based on the virtual test scene, sends virtual scene information and virtual sensor information to the test vehicle, the virtual test scene being used to simulate an automatic driving environment of the test vehicle, the virtual scene information indicating the automatic driving environment of the test vehicle, and the virtual sensor information comprising virtual signals of the plurality of sensors, each virtual signal of a sensor comprising a virtual true value signal and a virtual noise signal;

[0008] The test vehicle controls the test vehicle to perform automatic driving through the automatic driving controller based on the virtual scene information and the virtual sensor information, and displays a fusion picture of the virtual test scene and a real driving environment of the test vehicle through the AR display;

[0009] The test vehicle sends motion state information and control signals in an automatic driving process to the simulation platform, the motion state information indicating a pose of the test vehicle in the automatic driving process, and the control signals including signals output by the automatic driving controller in the automatic driving process.

[0010] The simulation platform determines a test result of the test vehicle based on the motion state information, the control signals and a test target, and updates a motion state of the test vehicle in the virtual test scene, the test result indicating an automatic driving performance of the test vehicle under the test target.

[0011] In some embodiments, the method further includes:

[0012] The simulation platform generates a reference virtual signal for each sensor, the reference virtual signal including a reference virtual true value signal and a reference virtual noise signal.

[0013] The simulation platform acquires a true signal for each sensor, fuses the true signal for each sensor with the reference virtual signal to obtain a virtual signal for each sensor.

[0014] In some embodiments, the virtual scene information includes at least one of weather information, road information and traffic flow information.

[0015] In some embodiments, the test vehicle controls the test vehicle to perform automatic driving through the automatic driving controller based on the virtual scene information and the virtual sensor information, including:

[0016] The automatic driving controller performs noise reduction on the virtual signals of the plurality of sensors based on the virtual scene information and the virtual sensor information, and fuses the virtual signals of the plurality of sensors after noise reduction to determine an automatic driving environment of the test vehicle in the virtual test scene.

[0017] The automatic driving controller controls the test vehicle to perform automatic driving based on the automatic driving environment in the virtual test scene.

[0018] In some embodiments, the method further includes:

[0019] The simulation platform stores the virtual test scene information, the virtual sensor information, the motion state information, the control signals and the test result.

[0020] In a second aspect, a test system of an automatic driving vehicle is provided, the system including a simulation platform and a test vehicle, the test vehicle having a plurality of sensors, an automatic driving controller and an augmented reality (AR) display.

[0021] The simulation platform is configured to construct a virtual test scene of the test vehicle, and send virtual scene information and virtual sensor information to the test vehicle based on the virtual test scene, the virtual test scene being configured to simulate an automatic driving environment of the test vehicle, the virtual scene information being indicative of the automatic driving environment of the test vehicle, and the virtual sensor information including virtual signals of the plurality of sensors, each virtual signal of a sensor including a virtual true value signal and a virtual noise signal.

[0022] The test vehicle is configured to:

[0023] control the test vehicle to perform automatic driving based on the virtual scene information and the virtual sensor information by using the automatic driving controller, and display a fusion picture of the virtual test scene and a real driving environment of the test vehicle by using the AR display.

[0024] send motion state information and control signals in the automatic driving process to the simulation platform, the motion state information being indicative of a pose of the test vehicle in the automatic driving process, and the control signals including signals output by the automatic driving controller in the automatic driving process.

[0025] The simulation platform is further configured to determine a test result of the test vehicle based on the motion state information, the control signals and a test target, and update a motion state of the test vehicle in the virtual test scene, the test result being indicative of an automatic driving performance of the test vehicle under the test target.

[0026] In some embodiments, the simulation platform is further configured to:

[0027] generate a reference virtual signal of each sensor, the reference virtual signal including a reference virtual true value signal and a reference virtual noise signal.

[0028] obtain a real signal of each sensor, fuse the real signal of each sensor with the reference virtual signal to obtain the virtual signal of each sensor.

[0029] In some embodiments, the virtual scene information includes at least one of weather information, road information and traffic flow information.

[0030] In some embodiments, the automatic driving controller on the test vehicle is configured to perform noise reduction on the virtual signals of the plurality of sensors based on the virtual scene information and the virtual sensor information, fuse the virtual signals of the plurality of sensors after noise reduction, and determine an automatic driving environment of the test vehicle in the virtual test scene based on the virtual scene information and the virtual sensor information; and control the test vehicle to perform automatic driving based on the automatic driving environment in the virtual test scene.

[0031] In some embodiments, the simulation platform is further configured to store the virtual test scene information, the virtual sensor information, the motion state information, the control signal and the test result.

[0032] In a third aspect, an electronic device is provided, which includes a memory and a processor, and the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the steps performed by the simulation platform in the test method of the autonomous vehicle according to the first aspect.

[0033] In a fourth aspect, a computer readable storage medium is provided, which stores at least one computer program, and the at least one computer program is loaded and executed by a processor to implement the steps performed by the simulation platform in the test method of the autonomous vehicle according to the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort based on these drawings.

[0035] Figure 1 is a schematic diagram of an implementation environment provided by an embodiment of the present application;

[0036] Figure 2 is a flowchart of a test method of an autonomous vehicle provided by an embodiment of the present application;

[0037] Figure 3 is a schematic diagram of a test method of an autonomous vehicle provided by an embodiment of the present application;

[0038] Figure 4 is a schematic diagram of another test method of an autonomous vehicle provided by an embodiment of the present application;

[0039] Figure 5 is a schematic diagram of a test device of an autonomous vehicle provided by an embodiment of the present application;

[0040] Figure 6 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of the present application clearer, the embodiments of the present application will be further described in detail with reference to the drawings.

[0042] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.) and signals involved in the present application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions. For example, the vehicle sensor signals involved in the present application are obtained under sufficient authorization.

[0043] Figure 1 is a schematic diagram of an implementation environment provided by an embodiment of the present application. As shown in Figure 1 , the implementation environment includes a test system 100 of an autonomous vehicle, the test system 100 of the autonomous vehicle includes a simulation platform 101 and a test vehicle 102, the simulation platform 101 and the test vehicle 102 are communicatively connected, for example, through a wired network or a wireless network.

[0044] The simulation platform 101 is a tool for simulating and testing autonomous driving systems. It provides a safe and controllable test environment for autonomous driving systems by simulating real-world traffic environments, vehicle behaviors and sensor data, etc. The simulation platform 101 can help researchers to perform a large number of simulation tests on the autonomous driving system of the vehicle before actual road testing, so as to find and solve potential problems. In the embodiments of the present application, the simulation platform 101 is used to provide functions such as scene modeling, sensor simulation, vehicle dynamics simulation, and environment interaction. For example, the simulation platform 101 uses mathematical models and computer simulation technology to simulate the motion, environment and sensors of the test vehicle 102. Illustratively, the simulation platform 101 can be implemented by using cloud computing and distributed systems to improve computing power and efficiency, or by using electronic devices with computing power (such as servers, etc.) to implement the simulation platform 101, which is not limited in the present application.

[0045] Test vehicle 102 refers to the physical vehicle to be tested, which has multiple sensors, an autonomous driving controller, and an augmented reality (AR) display. These multiple sensors include cameras, LiDAR, millimeter-wave radar, etc., which are not limited in this application. For example, the multiple sensors may also include wheel speed sensors, steering angle sensors, acoustic sensors, oxygen monitors, fuel gauge sensors, engine oil pressure sensors, etc. An ADAS / AD algorithm platform is deployed on the autonomous driving controller. The ADAS / AD algorithm platform refers to the software platform used to develop and implement advanced driver assistance systems (ADAS) and autonomous driving (AD) functions. An ADAS / AD algorithm platform typically includes the following modules: Sensor data fusion: fusing data from different sensors (such as cameras, radar, LiDAR, etc.) to obtain more comprehensive and accurate environmental information. Target detection and classification: identifying and classifying various targets on the road, such as vehicles, pedestrians, obstacles, etc. Perception and modeling: perceiving and modeling the environment around the vehicle, including lane line detection and road sign recognition. Path planning and decision-making: Based on perceived environmental information and vehicle status, the system plans a suitable driving path and makes decisions such as acceleration, deceleration, and lane changing. Control module: Translates these decisions into specific vehicle control commands, such as steering and braking. Fault diagnosis and safety module: Monitors system status, diagnoses potential faults, and takes corresponding safety measures. Data recording and analysis: Records driving data for subsequent algorithm optimization and performance evaluation. These modules collaborate to achieve the functions of ADAS / AD. Specific module settings and functions vary depending on different algorithm platforms and application scenarios. Illustratively, the autonomous driving controller controls the test vehicle 102 for autonomous driving through the ADAS / AD algorithm platform. The AR display can be implemented through a smartphone, tablet, head-mounted device, or car windshield, etc., and this application does not limit this. The AR display can provide the driver in the test vehicle 102 with real-time navigation instructions, road information, obstacle warnings, etc., enhancing driving safety and convenience.

[0046] Based on the above Figure 1 The test system 100 for the autonomous vehicle shown in this application provides a test method for the autonomous vehicle. (See also...) Figure 2 , Figure 2 This is a flowchart of a testing method for an autonomous vehicle provided in an embodiment of this application, such as... Figure 2 As shown, the method includes the following steps 201 to 204.

[0047] 201. The simulation platform constructs a virtual test scenario for the test vehicle and sends virtual scenario information and virtual sensor information to the test vehicle based on the virtual test scenario.

[0048] In the embodiments of the present application, the virtual test scene is used to simulate the automatic driving environment of the test vehicle. The simulation platform runs a real-time simulation machine, which is used to simulate the behavior and environment of the automatic driving vehicle in real time. Generally, the real-time simulation machine needs to have high computing power, real-time data processing and transmission capability, and accurate simulation of the automatic driving system, and it can be integrated with other components and modules in the simulation platform to realize a more complete automatic driving simulation system. Illustratively, the simulation platform runs the real-time simulation machine based on the test vehicle and the test field (such as a closed test field) where the test vehicle is located to construct the virtual test scene of the test vehicle. This process is also the scene modeling process, that is, by establishing realistic road, vehicle, pedestrian, traffic signal and other models, a real traffic scene is simulated. For example, the simulation platform and the test vehicle are connected through network communication. The test vehicle sends real-time video pictures collected by the camera to the simulation platform, so that the simulation platform constructs the virtual test scene of the test vehicle according to the received real-time video pictures.

[0049] After the simulation platform constructs the virtual test scene, the simulation platform sends virtual scene information and virtual sensor information to the test vehicle based on the constructed virtual test scene. The virtual scene information indicates the automatic driving environment of the test vehicle. Illustratively, the virtual scene information includes at least one of weather information, road information, and traffic flow information. The weather information is, for example, ground temperature, day or night, humidity, etc., and the present application is not limited thereto. The road information is, for example, lane lines, guardrails, road steps, etc., and the present application is not limited thereto. The traffic flow information is, for example, the current road motor vehicle flow, non-motor vehicle movement, pedestrian movement, etc., and the present application is not limited thereto.

[0050] The virtual sensor information includes virtual signals of a plurality of sensors, and each virtual signal of the sensors includes a virtual true value signal and a virtual noise signal. The plurality of sensors include a camera, a laser radar, a millimeter wave radar, and the like, and the application is not limited thereto. Illustratively, the simulation platform simulates the true value signals and noise signals of the plurality of sensors on the test vehicle to obtain the virtual true value signals and virtual noise signals of each sensor. It should be understood that, since the simulation platform introduces noise signals when simulating the perception signals of the sensors, in this way, the authenticity of the virtual sensor information can be improved, the test process is closer to the real road conditions, and thus the effectiveness of the test method is improved. For example, taking a camera as an example, the simulation platform simulates that the video captured by the camera has a pedestrian 10 meters in front of the vehicle, i.e., the virtual true value signal, and on this basis, a telegraph pole is added beside the pedestrian in the video, i.e., the virtual noise signal is added. When the test vehicle receives the virtual signal of the camera, it needs to perform noise reduction processing to identify the target pedestrian in the video. As can be seen, in this way, the authenticity of the video captured by the camera can be improved, the test process is closer to the real road conditions, and thus the effectiveness of the test method is improved.

[0051] In some embodiments, the simulation platform generates a reference virtual signal of each sensor, the reference virtual signal including a reference virtual true value signal and a reference virtual noise signal; obtains a real signal of each sensor, and fuses the real signal of each sensor with the reference virtual signal to obtain a virtual signal of each sensor. The reference virtual signal can be understood as an initial virtual signal, i.e., the virtual data simulated by the simulation platform. For any sensor, the simulation platform fuses the real signal of the sensor with the reference virtual signal in a manner such as data-level fusion (directly merging the real data and the virtual data at the data level, which can be unified by data preprocessing, data standardization, and the like, and then merging them into a data set), feature-level fusion (fusing the real and virtual data at the feature level, i.e., extracting the key features of the real and virtual data, and then combining or fusing these features to form a comprehensive feature vector), spatio-temporal fusion (considering the correlation of the real and virtual data in time and space, which can be aligned and fused according to the time sequence or the spatial position to better capture the dynamic relationship between the data), and the like, and the application is not limited thereto.

[0052] The above process is that, for any sensor, the simulation platform fuses the real data of the sensor with the virtual data, so that the authenticity of the virtual sensor information is further improved, the test process is closer to the real road conditions, and the effectiveness of the test method is improved. In addition, the simulation platform can fuse the real data and the virtual data on the basis of single road data, ensure the authenticity of the virtual information of each sensor, make the test process as close as possible to the real road conditions, and greatly improve the effectiveness of the test method. For example, there are multiple cameras on the test vehicle, and for any camera, the simulation platform fuses the reference virtual signal and the real signal of the camera to obtain the virtual signal of the camera.

[0053] 202. The test vehicle controls the test vehicle to perform automatic driving through the automatic driving controller based on the virtual scene information and the virtual sensor information, and displays a fusion picture of the virtual test scene and the real driving environment of the test vehicle through the AR display.

[0054] In the embodiment of the application, the test vehicle receives virtual scene information and virtual sensor information, controls the test vehicle to perform automatic driving through the ADAS / AD algorithm platform deployed on the automatic driving controller, and displays a fusion picture of the virtual test scene and the real driving environment of the test vehicle through the AR display. The process in which the automatic driving controller controls the test vehicle to perform automatic driving includes: based on the virtual scene information and the virtual sensor information, denoising the virtual signals of multiple sensors, and fusing the denoised virtual signals of the multiple sensors to determine the automatic driving environment of the test vehicle in the virtual test scene; based on the automatic driving environment in the virtual test scene, controlling the test vehicle to perform automatic driving. For example, in the virtual sensor information, the virtual signal of the camera indicates that there are two targets in front of the test vehicle, and the virtual signal of the millimeter wave radar indicates that there are three targets in front of the test vehicle. The automatic driving controller fuses the virtual signals of the two sensors based on the ADAS / AD algorithm platform to determine the accurate automatic driving environment, such as two targets in front, both pedestrians, based on which the automatic driving controller controls the test vehicle to brake.

[0055] In addition, the test vehicle displays a fusion picture of the virtual test scene and the real driving environment of the test vehicle through the AR display, improves the authenticity of virtual-real combination, and enables the driver on the test vehicle to intuitively observe the interaction relationship between the test vehicle and the virtual driving environment and the virtual traffic flow, thereby improving the immersion of the subjective evaluation dimension.

[0056] 203. The test vehicle sends the motion state information and the control signal in the automatic driving process to the simulation platform.

[0057] In the embodiments of the present application, the motion state information indicates the pose of the test vehicle in the automatic driving process, i.e., the position, speed, acceleration, angular velocity, etc. of the test vehicle in the automatic driving process, which is not limited in the present application. For example, the motion state information is determined by an inertial measurement unit (IMU) and a global positioning system (GPS) on the test vehicle, which is not limited in the present application. The control signal includes the signal output by the automatic driving controller in the automatic driving process. In some embodiments, the control signal also includes the signal output by each actuator on the test vehicle in the automatic driving process. It should be understood that the automatic driving controller on the test vehicle can send signals to each actuator (such as a steering wheel, a clutch, an accelerator, a brake, etc.) on the test vehicle to control the actuators to perform corresponding actions, and the test vehicle sends the signal output by the automatic driving controller in the automatic driving process to the simulation platform, which provides technical support for determining the test result of the test vehicle for the simulation platform. Moreover, by sending the motion state information and the control signal of the test vehicle in the automatic driving process to the simulation platform, it is convenient for the simulation platform to update the motion state of the test vehicle in the virtual test scene, which provides technical support for simulating the automatic driving environment in the future time period.

[0058] In addition, the motion state information and the control signal can be sent to the simulation platform by the automatic driving controller, or can be sent to the simulation platform by the message forwarding device on the test vehicle, which is not limited in the present application.

[0059] 204、The simulation platform determines the test result of the test vehicle based on the motion state information, the control signal and the test target, and updates the motion state of the test vehicle in the virtual test scene.

[0060] In the embodiments of the present application, the test result indicates the automatic driving performance of the test vehicle under the test target. The test target is used to evaluate the automatic driving performance of the test vehicle. For example, the test target is the perception ability of the test vehicle. The simulation platform determines whether the sensor can accurately perceive the surrounding environment, including roads, vehicles, pedestrians, obstacles, etc., based on the motion state information and control signals of the test vehicle in the automatic driving process. For another example, the test target is the decision-making ability of the test vehicle in different scenarios. The simulation platform determines whether the test vehicle can perform lane changing, obstacle avoidance, etc. according to the traffic signal, and determines whether the test vehicle can accelerate, brake, turn, etc. in time, based on the motion state information and control signals of the test vehicle in the automatic driving process. It should be noted that the number and type of test targets are not limited in the present application. In actual application, the test targets can be set according to business needs, such as vehicle speed, energy efficiency, comfort, etc. The present application is not limited thereto. In this way, the simulation platform can automatically evaluate the automatic driving performance of the test vehicle according to the pre-set test target, realize the automatic evaluation of the test result, and improve the comprehensiveness and consistency of objective evaluation. In some embodiments, the simulation platform sends the test result to the test vehicle, and the test vehicle displays the test result through the AR display, so that the driving personnel on the test vehicle can know the test result of the test vehicle in real time, and improve the human-computer interaction efficiency. In other embodiments, the simulation platform displays the test result by digital twinning, so that the relevant personnel can intuitively know the test result of the current test vehicle, and improve the human-computer interaction efficiency.

[0061] In addition, the simulation platform updates the motion state of the test vehicle in the virtual test scene based on the motion state information and the control signal of the test vehicle in the automatic driving process, and in some embodiments, the simulation platform digitally twins the virtual test scene to intuitively understand the current situation of the virtual test scene by relevant personnel. Moreover, the simulation platform can simulate the automatic driving environment of the future time period based on the updated virtual test scene, or generate the virtual scene information and the virtual sensor information of the next cycle, so that the test vehicle continues to drive automatically. It should be understood that the test process of the test vehicle can be understood as evaluating the automatic driving performance of the vehicle according to the automatic driving situation of the test vehicle in multiple cycles, for example, a cycle is 50 milliseconds or 1 second, etc. The simulation platform does not limit the division method of the cycle, for example, a cycle can also be the display time length of a frame of picture, etc. After the simulation platform constructs the virtual test scene of the test vehicle, in each cycle, the simulation platform sends the virtual scene information and the virtual sensor information to the test vehicle, the test vehicle controls the test vehicle to drive automatically based on the virtual scene information and the virtual sensor information through the automatic driving controller, and displays the fusion picture of the virtual test scene and the real driving environment of the test vehicle through the AR display, and the test vehicle sends the motion state information and the control signal in the automatic driving process to the simulation platform, and the simulation platform determines the test result of the test vehicle based on the motion state information, the control signal and the test target, and updates the motion state of the test vehicle in the virtual test scene, generates the virtual scene information and the virtual sensor information of the next cycle.

[0062] Reference is made below to Figure 3 two cycles after the simulation platform constructs the virtual test scene, to illustrate the above-mentioned test method of the automatic driving vehicle. Figure 3 is a schematic diagram of a test method of an automatic driving vehicle provided by an embodiment of the present application. As shown in Figure 3 , the simulation platform is in communication connection with the test vehicle, and the test vehicle has an automatic driving controller, and the automatic driving controller has an ADAS / AD algorithm platform disposed thereon, for example, the ADAS / AD algorithm platform includes a perception fusion algorithm (for perceiving the environment around the vehicle and fusing data from different sensors), a decision planning algorithm (for planning a suitable driving path according to the perceived environment information and the vehicle state, and making decisions such as acceleration, deceleration, lane change, etc.), and a motion control algorithm (for converting the decisions into specific vehicle control instructions such as steering, braking, etc.). Illustratively, the test method of the automatic driving vehicle includes the following steps:

[0063] Step 1, the simulation platform constructs a virtual test scene of the test vehicle.

[0064] Step 2, the simulation platform sends virtual scene information and virtual sensor information to the automatic driving controller on the test vehicle based on the virtual test scene, wherein the virtual scene information includes weather information, road information and traffic flow information, and the virtual sensor information includes virtual signals of multiple sensors, which are virtual signals fused with real signals and noise signals. This process can also be understood as a scene injection process.

[0065] Step 3, the automatic driving controller controls the test vehicle to perform automatic driving in the first period based on the ADAS / AD algorithm platform, and obtains motion state information of the test vehicle in the first period, and sends the motion state information and control signals in the first period to the simulation platform.

[0066] Step 4, the simulation platform updates the motion state of the test vehicle in the virtual test scene based on the motion state information in the first period, and performs digital twin display to generate virtual scene information and virtual sensor information in the next period, i.e., the second period, and sends the virtual scene information and virtual sensor information in the second period to the automatic driving controller.

[0067] Step 5, the automatic driving controller controls the test vehicle to perform automatic driving in the second period based on the ADAS / AD algorithm platform, and obtains motion state information of the test vehicle in the second period, and sends the motion state information and control signals in the second period to the simulation platform. This process can also be understood as a vehicle state feedback process.

[0068] Step 6, the simulation platform determines the test result of the test vehicle based on the received motion state information, control signals and test target.

[0069] In addition, in the above process, the test vehicle displays a fusion picture of the virtual test scene and the real driving environment of the test vehicle in real time through the AR display. This process is described with reference to Figure 4 , Figure 4 is another schematic diagram of a test method of an automatic driving vehicle provided by the embodiment of the present application. As shown in Figure 4 , the camera on the test vehicle sends the real-time video picture collected to the simulation platform, and the pose device such as IMU and GPS on the test vehicle collects the motion state information in real time and sends it to the simulation platform, and the simulation platform fuses and draws the virtual test scene and the real driving environment of the test vehicle to obtain a fusion picture, which is displayed through the AR display, so that the driver on the test vehicle can intuitively observe the interaction relationship between the test vehicle and the virtual driving environment and the virtual traffic flow.

[0070] In some embodiments, the simulation platform provides information storage functions, which can store relevant information generated during the test process of the autonomous vehicle, such as virtual test scene information, virtual sensor information, motion state information, control signals, and test results, etc. By storing this information, it can provide technical support for subsequent fault reproduction and comparative analysis. Among them, fault reproduction refers to when a system fails or an abnormal situation occurs during testing or actual operation, trying to recreate or simulate the same fault condition to determine the root cause of the problem and the solution. Comparative analysis refers to comparing the system performance under normal conditions with the system performance under fault conditions to find differences and abnormalities. It should be understood that since the simulation platform stores relevant information generated during the test process, it can quickly implement fault reproduction, thereby improving fault analysis efficiency.

[0071] In summary, in the test method of the autonomous vehicle provided in the embodiments of the present application, after the simulation platform constructs the virtual test scene of the test vehicle, the virtual scene information and the virtual sensor information are sent to the test vehicle, so that the autonomous driving controller on the test vehicle controls the test vehicle to perform autonomous driving based on these information, and the motion state information and the control signals during the autonomous driving process are sent to the simulation platform to determine the test result of the test vehicle. In this process, since the virtual sensor information includes virtual noise signals, the authenticity of the virtual sensor information is improved, making the test process closer to the real road conditions, thereby improving the effectiveness of the test method, and in the test process, the test vehicle displays the fusion picture of the virtual test scene and the real driving environment of the test vehicle in real time through the AR display, which facilitates the driving personnel on the test vehicle to intuitively observe the interaction between the test vehicle and the virtual driving environment.

[0072] Referring to Figure 5 , the embodiments of the present application provide a test device for an autonomous vehicle, which is configured in a simulation platform in a test system of an autonomous vehicle, and the system further includes a test vehicle, which has a plurality of sensors, an autonomous driving controller, and an augmented reality (AR) display. The device can be realized as part of or all functions of the aforementioned simulation platform through software, hardware, or a combination of both. Illustratively, the device includes a scene construction module 501, a sending module 502, a receiving module 503, a determination module 504, and an updating module 505.

[0073] The scene construction module 501 is configured to construct a virtual test scene of the test vehicle, and the virtual test scene is used to simulate an autonomous driving environment of the test vehicle.

[0074] The sending module 502 is configured to send virtual scene information and virtual sensor information to the test vehicle based on the virtual test scene, so that the test vehicle controls the test vehicle to perform automatic driving through an automatic driving controller based on the virtual scene information and the virtual sensor information, and displays a fusion picture of the virtual test scene and a real driving environment of the test vehicle through an AR display; the virtual scene information indicates an automatic driving environment of the test vehicle, and the virtual sensor information includes virtual signals of a plurality of sensors, and the virtual signal of each sensor includes a virtual true value signal and a virtual noise signal.

[0075] The receiving module 503 is configured to receive motion state information and a control signal in an automatic driving process sent by the test vehicle, the motion state information indicating a pose of the test vehicle in the automatic driving process, and the control signal including a signal output by the automatic driving controller in the automatic driving process.

[0076] The determining module 504 is configured to determine a test result of the test vehicle based on the motion state information, the control signal and the test target, the test result indicating an automatic driving performance of the test vehicle under the test target.

[0077] The updating module 505 is configured to update the motion state of the test vehicle in the virtual test scene.

[0078] In some embodiments, the apparatus further includes a generating module configured to generate a reference virtual signal of each sensor, the reference virtual signal including a reference virtual true value signal and a reference virtual noise signal; and the simulation platform acquires a real signal of each sensor, and fuses the real signal of each sensor with the reference virtual signal to obtain the virtual signal of each sensor.

[0079] In some embodiments, the virtual scene information includes at least one of weather information, road information and traffic flow information.

[0080] In some embodiments, the apparatus further includes a storage module configured to store the virtual test scene information, the virtual sensor information, the motion state information, the control signal and the test result.

[0081] To sum up, the test device of the autonomous vehicle provided in the embodiments of the present application constructs a virtual test scene of the test vehicle, and then sends the virtual scene information and the virtual sensor information to the test vehicle, so that the autonomous driving controller on the test vehicle controls the test vehicle to perform autonomous driving based on the information, and sends the motion state information and the control signal in the autonomous driving process to the simulation platform, and the simulation platform determines the test result of the test vehicle. In this process, since the virtual sensor information includes a virtual noise signal, the authenticity of the virtual sensor information is improved, the test process is closer to the real road conditions, and the effectiveness of the test method is improved. Moreover, in the test process, the test vehicle displays a fusion picture of the virtual test scene and the real driving environment of the test vehicle in real time through the AR display, so that the driver on the test vehicle can intuitively observe the interaction between the test vehicle and the virtual driving environment.

[0082] It should be noted that the test device of the autonomous vehicle provided in the above embodiments is used to test the autonomous vehicle, and only the division of the above functional modules is used as an example for illustration. In actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure is divided into different functional modules to complete all or part of the functions described above. In addition, the test device of the autonomous vehicle and the test method of the autonomous vehicle provided in the above embodiments belong to the same concept, and the specific implementation process is described in the method embodiments, which will not be repeated here.

[0083] Reference Figure 6 The embodiments of the present application also provide an electronic device, Figure 6 FIG. 1 is a structural schematic diagram of an electronic device provided in the embodiments of the present application. The electronic device 600 can have great differences due to different configurations or performances, and can include one or more processors (Central Processing Units, CPUs) 601 and one or more memories 602, wherein the memory 602 stores at least one computer program, and the at least one computer program is loaded and executed by the processor 601 to realize the steps performed by the simulation platform in the test method of the autonomous vehicle provided in the above method embodiments. Of course, the electronic device can also have a wired or wireless network interface, a keyboard, an input and output interface, and other components for realizing the functions of the device, and details are not described here.

[0084] The serial numbers of the above embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0085] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or can be instructed by a program to complete the related hardware, and the program can be stored in a computer readable storage medium. The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0086] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A test method of an autonomous vehicle, characterized by, A test system applied to an autonomous vehicle, the system comprising a simulation platform and a test vehicle, the test vehicle having a plurality of sensors, an autonomous driving controller and an augmented reality (AR) display, the method comprising: the simulation platform constructing a virtual test scene of the test vehicle, and sending virtual scene information and virtual sensor information to the test vehicle based on the virtual test scene, the virtual test scene being used to simulate an autonomous driving environment of the test vehicle, the virtual scene information indicating the autonomous driving environment of the test vehicle, and the virtual sensor information comprising virtual signals of the plurality of sensors, each virtual signal of a sensor comprising a virtual true value signal and a virtual noise signal; the autonomous driving controller performing noise reduction on the virtual signals of the plurality of sensors based on the virtual scene information and the virtual sensor information, and fusing the virtual signals of the plurality of sensors after noise reduction to determine the autonomous driving environment of the test vehicle in the virtual test scene; the autonomous driving controller controlling the test vehicle to perform autonomous driving based on the autonomous driving environment in the virtual test scene, and displaying a fused picture of the virtual test scene and a real driving environment of the test vehicle through the AR display; the test vehicle sending motion state information and control signals in the autonomous driving process to the simulation platform, the motion state information indicating a pose of the test vehicle in the autonomous driving process, and the control signals comprising signals output by the autonomous driving controller in the autonomous driving process; the simulation platform determining a test result of the test vehicle based on the motion state information, the control signals and a test target, and updating a motion state of the test vehicle in the virtual test scene, the test result indicating an autonomous driving performance of the test vehicle under the test target; wherein the method further comprises: the simulation platform generating a reference virtual signal of each sensor, the reference virtual signal comprising a reference virtual true value signal and a reference virtual noise signal; the simulation platform obtaining a real signal of each sensor, and fusing the real signal of each sensor with the reference virtual signal to obtain the virtual signal of each sensor.

2. The method of claim 1, wherein, The virtual scene information comprises at least one of weather information, road information and traffic flow information.

3. The method of claim 1, wherein, The method further comprises: the simulation platform storing the virtual scene information, the virtual sensor information, the motion state information, the control signals and the test result.

4. A test system for an autonomous vehicle, characterized by The system comprises a simulation platform and a test vehicle, the test vehicle having a plurality of sensors, an autonomous driving controller and an augmented reality (AR) display; The simulation platform is configured to construct a virtual test scene of the test vehicle, and send virtual scene information and virtual sensor information to the test vehicle based on the virtual test scene, the virtual test scene being configured to simulate an automatic driving environment of the test vehicle, the virtual scene information being indicative of the automatic driving environment of the test vehicle, and the virtual sensor information including virtual signals of the plurality of sensors, each virtual signal of a sensor including a virtual true value signal and a virtual noise signal. The test vehicle is configured to: control the test vehicle to perform automatic driving through the automatic driving controller based on the virtual scene information and the virtual sensor information, and display a fusion picture of the virtual test scene and a real driving environment of the test vehicle through the AR display; send motion state information and control signals in the automatic driving process to the simulation platform, the motion state information being indicative of a pose of the test vehicle in the automatic driving process, and the control signals including signals output by the automatic driving controller in the automatic driving process; The simulation platform is further configured to determine a test result of the test vehicle based on the motion state information, the control signals and a test target, and update a motion state of the test vehicle in the virtual test scene, the test result being indicative of an automatic driving performance of the test vehicle under the test target. The simulation platform is further configured to generate a reference virtual signal of each sensor, the reference virtual signal including a reference virtual true value signal and a reference virtual noise signal. acquire a real signal of each sensor, and fuse the real signal of each sensor with the reference virtual signal to obtain the virtual signal of each sensor. The automatic driving controller on the test vehicle is configured to perform noise reduction on the virtual signals of the plurality of sensors based on the virtual scene information and the virtual sensor information, fuse the virtual signals of the plurality of sensors after noise reduction, and determine an automatic driving environment of the test vehicle in the virtual test scene; and control the test vehicle to perform automatic driving based on the automatic driving environment in the virtual test scene.

5. The system of claim 4, wherein, The virtual scene information includes at least one of weather information, road information and traffic flow information.

6. An electronic device, comprising: The electronic device includes a memory and a processor, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to implement the steps performed by the simulation platform in the test method of the automatic driving vehicle according to any one of claims 1-3.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores at least one computer program, the at least one computer program being loaded and executed by the processor to implement the steps performed by the simulation platform in the test method of the automatic driving vehicle according to any one of claims 1-3.

Images