Motion planning by adding geometric constraint of roadside to beam curvature method

Motion planning is a fundamental research field in autonomous navigation for intelligent vehicle. In order to realize stable and comfortable driving like human driver does, we present a novel motion planning method based on the compound grid and geometry representation of unstructured road environment. By adding geometric constraint of roadside as well as the hidden traffic rules to the improved beam curvature method, the proposed method not only inherited the performance of sensitivity to obstacle avoidance, but also can realize quick, smooth and stable and comfortable driving in unstructured environments.