Feasible and Optimal Path Planning in Strong Current Fields

This paper addresses the problem of path planning in strong current fields. In such situations, existing approaches are subject to incorrectness and incompleteness issues. That is, they may return physically infeasible paths or no path at all, even if a feasible path exists. That is why we propose here a new approach called the sliding wavefront expansion. This algorithm, which combine an appropriate cost function and continuous optimization techniques, guarantees the existence of a path with an arbitrary precision. The validity and the global optimality of the path are theoretically proven. Simulation results on realistic environments, which is based on actual wind charts, are also provided.