A self-adaptive particle swarm speed optimization method for unmanned sailboats considering roll constraint

The unmanned sailboat is increasingly used for marine data collection and environmental monitoring. However, it faces significant challenges in optimizing speed under complex sea conditions while ensuring navigation safety. Notably, large roll angle during navigation can greatly increase the risk of capsizing. To address this problem, a 6-degree-of-freedom dynamic model of the unmanned sailboat is established, and a self-adaptive particle swarm speed optimization algorithm with roll constraint is proposed. This algorithm is used to optimize the sail angle to achieve maximum speed while minimizing roll angle. It adopts dynamic-objective constraint-handling method to convert the single-objective constraint into a bi-objective optimization problem, and maps the roll angle constraint onto the input parameters of the wing sail. The optimized sail angle with maximum output speed and minimum roll angle is applied to the unmanned sailboat to form a feedback control. The sea trial has been performed and the results demonstrate the superiority, effectiveness, and robustness of the proposed algorithm.