With the widespread application of unmanned surface vehicles in navigation, many unmanned surface vehicle clusters have become a key research direction. Firstly, using Leader-Follower for the unmanned surface vehicle structure formation, designed the unmanned surface vehicle controller based on a dis-tributed model predictive control algorithm to maintain the stability of the unmanned surface vehicle cluster formation. Secondly, to improve the obstacle avoidance performance of multiple unmanned surface vehicles, the structure of the traditional artificial potential field method has been enhanced, and dynamic regulatory factors are introduced to optimize the penetration of unmanned surface vehicles. And establish cost functions from security, gentleness, and power consumption, and comprehensively evaluate the optimization performance of the way to avoid the way. Finally, in response to changing the needs of the formation in the voyage of the unmanned surface vehicles, introduce stable formation functions, and achieve team form transformation in the state of the formation.