Abstract: The decision on wind sail angle is crucial for effectively harnessing wind energy in windassisted ships. During ship navigation， the wind sail generates both thrust and lateral thrust， causing ship yawing and reduced energy efficiency. To address this problem， this paper introduced a lateral thrust weight factor and established a bilinear angle decision function composed of the thrust coefficient and the lateral thrust coefficient. The function aimed to minimize the deviation angle of the optimal course and utilized the particle swarm optimization algorithm to solve for the optimal weight factor applicable to the specific route. This approach enables achieving optimal sail angles under different relative wind directions， allowing for windassisted ship navigation without relying on steering and avoiding course deviation. Compared to the traditional angle strategy based on the maximum thrust method， the proposed approach results in slight improvement in heading longitudinal displacement within the same navigation time， substantial reduction in sideways displacement and course deviation angle， and smoother angle variations during fluctuations in relative wind direction. These findings are beneficial for wind sail manipulation. This research provides insights to enhance the energy efficiency of wind-assisted ships.

Key words: wind-assisted ship, course deviation, attack angle strategy, optimization