Research on target recognition algorithm and control strategy of AEB system in curved roads

A target recognition method based on curve coordinate transformation was proposed to address the challenge of misidentification in existing automatic emergency braking (AEB) systems under curved conditions. The geometric information of the road model could be reflects through sensors, the curve coordinate transformation method was used to locate the position of the main vehicle and the target vehicle, the relative distance between vehicles were calculated, and the dangerous target vehicle was determined by using the logic gate restriction method. A longitudinal collision avoidance control strategy combining the Honda algorithm and TTC algorithm were proposed to address the traditional collision avoidance algorithms based on the safe distance model or time-to-collision (TTC) algorithm, which cannot balance the safety and comfort issues during the braking process. Using the TTC algorithm as a forward collision warning strategy, and the autonomous emergency braking strategy was designed based on the optimized Honda algorithm. The simulation results show that the proposed method based on curve coordinate transformation can accurately calculate the distance between the main vehicle and the target vehicle, accurately and efficiently identify dangerous target vehicles. The fusion algorithm based on the collaborative control of the safety distance algorithm and TTC algorithm can effectively avoid longitudinal car following collisions, and take into account the safety and comfort of the emergency braking process.