Motion planning for fixed-wing aircraft to harvest wind energy using motion primitives

Abstract

In this paper, I propose a path (motion) planner for fixed-wing aircraft to exploit wind energy using motion primitives. The planner operates in three main steps: the reduced-order path planning, the full- state autopilot, and the selection step. Initially, the planner constructs a tree structure to explore paths, using the cost to prune or select branches. Upon sufficient growth, the tree reveals candidate paths. The planner generates guidance commands for the full-state autopilots by utilizing elements from the nodes on the candidate paths and motion primitives. After the autopilots generate paths, the planner in the last step selects the path with the lowest control surface efforts as a solution among the paths in the second step. Two simulations demonstrate that this solution path offers about 10% greater efficiency than a direct, straight path to the goal. It highlights the effectiveness of the proposed path-planning method in exploiting wind energy.