Obstacle Avoidance for autonomous agricultural vehicles

By Merel Stevens

Abstract
This thesis addresses the avoidance of static obstacles in autonomous agricultural vehicles (AAV) through the design of a path planning algorithm. Existing path planning methods most often do not take the kinematic constraints of non-holonomic vehicles into account. In this research, multiple local- and global path planning methods and field- and obstacle representations are investigated. The proposed methodology builds upon the Ariadne’s Clew algorithm with key features of the A* algorithm. Key objectives include the avoidance of static obstacles on the headland of an agricultural field and planning a path with respect to the magnitude of the maximum curvature and maximum curvature rate of the AAV. The established algorithm is capable of finding solution paths in scenarios with cluttered obstacles and narrow spaces.