Abstract
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.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
