Trajectory Planning in Frenet Frame via Multi-Objective Optimization

Autonomous vehicles are an essential tool for promoting the development of intelligent transportation systems (ITS) and can effectively reduce traffic accidents caused by human errors. As important part automatic driving software system, path planning is responsible generating motion trajectory vehicle, which primary factor determining quality. However, solution space construction optimization problem formulation remain challenging research areas in field planning. In this paper, we propose a multi-objective algorithm static obstacle avoidance to improve comfort, safety anti-deviation planned trajectory. We decouple lateral longitudinal vehicle using Frenet frame discretize state generate target states vehicle. Based on initial states, set trajectories quintic quartic polynomials, respectively. addition, design cost function that comprehensively considers safety, deviation distance road center line combining acceleration check, curvature collision check. function, novel method quantify candidate considering size obstacles. The experimental results show proposed quantize paths comfort 13.47%, 32.19%, 59.36% 18.60% straight road, curvy intersection U-shaped Furthermore, 63.72%, 13.86%, 44.36%, 45.56%