Great Circle Path Planning for USV Voyages Across Oceans with Irregul ar Static Obstacles

On the surface of the earth's sphere, the shortest path of an unmanned surface vehicle's (USV's) voyage across an ocean is a great circle arc rather than a rhumb line segment. To plan the great circle paths fitting to USV dynamics, this paper develops a great-circle-based nonlinear programming (GCNLP) approach for USV voyages across oceans with irregular static obstacles. We formulate the path planning problem as an optimization problem whose feasible solutions are the USV control vector consisting of force in surge and moment in yaw, which generates a USV planned path. The objective function of the optimization problem is the sum of all the great circle arc lengths between the adjacent waypoints in a USV planned path, such that an optimal USV planned path is the shortest. The constraints of the optimization problem are formulated using a new irregular static obstacle geometric model and the dynamic model of USVs, such that a USV planned path bypasses irregular static obstacles and fits to the USV dynamics. Using a nonlinear programming technique to search for the optimal solution of the optimization problem, we obtain the USV great circle planned path. Path planning and comparison results with the A* path planning method demonstrate that the USV great circle planned path under our proposed GCNLP approach is shorter. Our proposed GCNLP approach can be applied to the USV sailing in different waters using the corresponding irregular static obstacle data.