Research on Critical Test Scenarios of Automated Vehicle Overtaking on Highway 2022-01-7018

The testing and verification of automated vehicles is of vital importance to improving the safety of automated vehicles. There are relatively few researches on the functional scenarios and critical test scenarios of the automated vehicle overtaking conditions on highway. Based on the complex scenario group composed of automated vehicles and surrounding vehicles, we adopted a full combination strategy to arrange and combine the relative positions and relative motions of the automated ego-vehicle and surrounding vehicles, and combined the principles of scenario screening to extract scenario library of overtaking functions with test value. Aiming at the problem that the number of test cases after the discretization combination of the dynamic scenario parameters of the ego-vehicle and the surrounding vehicles in the overtaking scenario was too large and the calculation could not be achieved, we adopted a three-way combination strategy to greatly reduce the number of test cases. According to the overtaking decision of automated vehicles on the highway and the motion planning and control algorithm based on model predictive control, we established a Matlab/Simulink and CarSim joint simulation platform to simulate the overtaking scenarios of automated ego-vehicle. We used the collision time, the corner distance between vehicles and the maximum deceleration of the ego-vehicle as the safety indicators of the overtaking scenario to evaluate the criticality of the scenario, obtained the critical test cases, and used the weighted Euclidean distance clustering method to reduce critical test cases for overtaking., and 11 sets of critical test cases for highway overtaking condition have been obtained, which can be used for overtaking safety test verification in closed test site.