Vehicle-in-loop automatic driving vehicle testing method, system and device in mixed scene

Abstract

The invention discloses a vehicle-in-the-loop automatic driving vehicle testing method, system and device in a mixed scene. The method comprises the steps that a 3D scene model is generated based on asemantic map; a real vehicle obtains vehicle-mounted sensor data in a real scene configured with obstacles, vehicle-mounted sensor data are preprocessed, pedestrian, vehicle and dynamic and static obstacle information is configured in a simulation scene, vehicle obstacle avoidance and multi-vehicle interaction based on laser radar ranging are used, and a virtual traffic flow is formed; obstacle information configured in the simulation scene is processed into a sensing result, and the sensing result is fused to serve as a sensing result in simulation; a sensing result in the simulation scene is fused with sensor data preprocessed in a real scene to obtain a control instruction of the vehicle, and the control instruction is executed by the real vehicle to move in the simulation scene and the real scene, and the sensing result in the simulation scene and the multi-sensor sensing result in the real scene are updated and fed back to form a test loop.

Description

technical field [0001] The invention belongs to the technical field of automatic driving vehicle testing, and in particular relates to a vehicle-in-the-loop automatic driving vehicle testing method, system and device in a mixed scene. Background technique [0002] The stability and safety of self-driving vehicles must be guaranteed before they are put into practical use, which requires comprehensive and repeated testing of software, hardware, and all possible unexpected situations. An autonomous vehicle includes at least multiple components such as vehicles, sensors, computing units, operating systems, and algorithm software. For traditional vehicles, manufacturers find and correct vehicle problems through millions of kilometers of actual road tests. However, this method is dangerous and time-consuming for self-driving vehicles, a multi-component system that combines hardware and software. Therefore, everyone is looking for A safe, reliable, low-cost, repeatable method for ...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the problems of high risk factor, high test cost, difficult to reproduce the scene or highly dependent on the test site in the prior art for the simulation test of autonomous driving vehicles, the present invention provides a hybrid scene that supports the vehicle in the virtual and real world. Flexible self-driving vehicle testing method, system and device for the loop

Images