Virtual Validation Method for Automated Driving Vehicles Based on Traffic Accident

With the advancement in automatic driving technology, testing methods-such as road test for conventional automatic is difficult to verify the safety and reliability of automatic driving vehicles. Virtual testing method has become an important mean due to the technical superiority of test efficiency and time consumption. This paper selects a highway traffic accident from China In-Depth Accident Study (CIDAS) database and analyzes the accident cause. Builds the traffic accident scenario in virtual simulation software-CarMaker, including actual road environment information, ego vehicle, and other traffic participants maneuver and so on; then take a hardware-in-the-loop(HIL) testing, and the ACC and AEB control system based on radar simulator is used to test the ability of the ADAS system of respond to traffic accident scenario. Testing result shows that the ADAS system can reduce the likelihood of collision in some way. Testing process verifies the feasibility of virtual test that verify the functionality of the ADAS system, which will provide an method to large-scale virtual simulation test for AD and study test scenario.