Laser-based Vehicle Control in Orchard

Abstract Daily work is very tough in orchard on slope land. Assistant robots are required to develop for workers. In this paper, an autonomous vehicle in orchard is provided as a base platform which is used for monitoring, pesticide spraying, harvesting and so on in the near future. The self-position identification is attained by using a laser range finder and map matching. The robust nonlinear control method which was proposed by authors for underactuated mobile robots is applied to the vehicle which is a by-wire all terrain one. A desired path to follow is generated at first. If the vehicle is a little far away from the desired path, another desired path from the current position is regenerated to cancel the current path error and smooth motion is achievable because in the case of underactuated systems sometimes a path error causes large motion of a vehicle. Experiments were performed in orchard. The experimental results were very fine to go through the orchard. The average path error in the lateral direction is about 20cm.