Motion planning of mobile robot considering velocity-dependent cost and time
Sho Matsunaga, Yuichi Kobayashi, Kazuki Matsumaura, Chyon Hae Kim · 2016
Unmanned autonomous vehicles are desired to navigate in unstructured environment considering not only obstacle avoidance but also avoidance of unpreferred vibrations and impacts. Such unpreferred motion properties often depend on nature of terrain and velocity of vehicle, which have not been considered sufficiently. This paper presents an unmanned-vehicle's motion optimization method based on such velocity-dependent cost of traversing uneven terrains. The proposed method consists of two parts, prediction of velocity-dependent cost based on visual observation and planning using RASMO reflecting the predicted cost. Simulation and experiments were conducted to verify the proposed method.