In order to improve the safety of AEB (autonomous emergency braking) pedestrian protection, a hierarchical pedestrian collision avoidance strategy based on upper fuzzy control and lower PID control was proposed. Taking an E-class SUV vehicle as the research object, its dynamic model was established. Based on the real pedestrian test scenarios at home and abroad, the TTC (time to collision) risk assessment model was established. Then the control strategies were simulated and verified through the joint simulation of Matlab and CarSim. The simulation results show that the strategies of the pedestrian collision avoidance system can satisfy the domestic pedestrian test condition standards, and the minimum safe distance from vehicle to pedestrians is 0.9 m. The fuzzy control can automatically adjust the braking strength, the output deceleration range is controlled within 4.8 m/s² to 6.1 m/s² for better comfort on the premise of ensuring safety. The risk assessment model can correctly send out pedestrian collision warning signals without missed alarm and false alarm.