Towards Optimal Motion Planning of Surface Vehicles in Confined Waters

Abstract

Abstract This paper presents a method for motion planning of fully actuated autonomous surface vehicles in confined waters. The goal is to generate a dynamically feasible trajectory from a given initial state to a given final state without collisions of the vehicle with obstacles. Both the own vehicle and static obstacles are represented as convex polygons. A sampling-based approach is used, where multiple samples in a state space are connected by third-order polynomials. A method is proposed to compute optimal polynomials that both minimize a cost function and account for input variable constraints. Within the cost function, time is weighted together with the control effort. The obtained objective function is a scalar nonlinear function that is solved using numerical minimization techniques. Embedded in a sampling-based motion planning algorithm, the major motion planning problem is solved by many small optimizations. The result is an optimal trajectory from an initial state to a final state, taking into account dynamics and input variable constraints, and avoiding collisions of the vehicle with static obstacles. A simulation is used to demonstrate that the proposed method is basically applicable.