Optimization-Based Motion Planning for Autonomous Parking Considering Dynamic Obstacle: A Hierarchical Framework
Xuemin Chi, Zhitao Liu, Jihao Huang, Feng Hong, Hongye Su · 2022 34th Chinese Control and Decision Conference (CCDC) · 2022
We present a hierarchical framework based on graph search and model predictive control (MPC) for electric autonomous vehicle (EAV) parking maneuver in tight-environment. At high-level, only static obstacles are considered, and the scenario-based hybrid A* (SHA*), which is faster than the traditional hybrid A*, is designed to provide an initial guess (also known as a global path) for the parking task. To extract the velocity and acceleration profile from an initial guess, an optimal control problem (OCP) is built. In low-level, a NMPC-based strategy is used to avoid dynamic obstacles (also konwn as local planning). The efficacy of SHA* is evaluated through 148 different simulation schemes and the proposed hierarchical parking framework is demonstrated through a real-time parallel parking simulation.