To address the problems of target allocation conflicts, trajectory oscillations in underactuated systems, and degraded control performance under disturbance conditions during the cooperative docking of multiple unmanned surface vessel (USV) clusters, this study proposes a hierarchical, distributed cooperative docking navigation method that integrates potential and differential game theories. First, at the decision-making layer, a distributed optimization framework based on potential games is formulated to resolve conflicts of interest in target allocation among unmanned surface vessels. By designing a global potential function that reflects the total cost of the cluster, self-organization of each unmanned surface vessel is induced, and convergence to a pure-strategy Nash equilibrium is achieved. Second, at the guidance layer, in order to solve the problem of trajectory oscillations caused by the motion coupling characteristics of underactuated systems, a continuously variable weighted guidance strategy with dynamic adjustment based on an S-shaped function is designed, by which a smooth transition from the far-field approach mode to the near-field alignment mode is achieved. Finally, at the control layer, an integral term of the tracking error is introduced into the augmented model, and a nine-dimensional augmented state space is formulated. The docking and collision avoidance requirements are modeled as performance indices and nonlinear potential field constraints. In combination with specific docking requirements, a quadratic cost function is formulated, and the algebraic Riccati equation is solved based on differential game theory and optimal control theory, through which a globally optimal state-feedback control law is obtained. Simulation results verify that the proposed method can successfully accomplish docking tasks in both typical longitudinal and lateral docking scenarios. Disturbances can be effectively suppressed by the integral augmented optimal control structure, the steady-state error can be kept within 0.1 meters, the relative distance among all unmanned surface vessels can be maintained above the safety threshold at all times, the total energy consumption cost can be reduced by approximately 46 percent, and the docking time can be shortened by approximately 20 percent, demonstrating excellent robustness and high efficiency.