At present, the path planning algorithm of area coverage exhibits certain shortcomings, such as overreliance on the starting point, a lack of consideration for ocean currents, and the existence of excessively sharp turns in the planned path. While considering the starting point of USVs and the influence of ocean currents, path planning using multiple USVs to cover designated sea areas was studied. When the obstacle information is known, an improved multi-USV optimal coverage path planning partitioning algorithm is proposed. Based on environmental modeling, the roulette selection method is used to optimize the starting point. Taking into account the constraints of ocean current, the generation cost influenced by the current is incorporated into the mobile generation cost. This integration facilitates the adaptive improvement of the algorithm flow field. Additionally, the resulting original path is smoothed through the utilization of second-order and third-order Bezier curves. The optimal area coverage path is derived, significantly enhancing the practicality of the algorithm. Experimental results verify the reliability and effectiveness of the algorithm.