A Research on the Test and Evaluation Method of Autonomous Vehicle Emergency Braking System Based on Typical Test Scenarios

With the emergence of self-driving vehicles, the Autonomous Emergency Braking System (AEB) has become a crucial element in preventing traffic collisions. When comes to public trust and road safety, AEB needs to be really reliable and working well. In this paper, a series of testing and evaluation methods are introduced about AEB system using some typical and key scenarios that reflect the real world driving scenario. Do research starting from having the basic structure for all kind of test scenario like CCR (car-to-car rear stationary), CCRm(car-to-car rear moving), CCRb(car-to-car rear braking), VRU(Vulnerable Road User) and more. The test protocol details, test environment, instruments in vehicle, target, recording data. The fundamental aspect of the evaluation method comprises of 4 KPIs: Collision_avoidance, Speed_reduction_on_impact, Ttc_at_ttc_intervention and Peak_braking_deceleration. In the paper, it takes hypothetrical test data on these situations to prove how this works. The results show that AEB performance depends much on the situation, whether it’s initial velocity, relative velocity, or target type. That means we should test out different situations to see how far we can push an AEB. The proposed methodology gives automakers, regulatory bodies and research institutions a regular, iterative method to do strong evaluations and comparisons of AE B performance, which is good for making safer AD Ts.