The paper presents a strategy for cooperative collision avoidance and obstacle avoidance for underactuated unmanned ship formation based on an improved artificial potential field method and input quantization. The strategy involves designing kinematic guidance and dynamic control laws hierarchically, using a distributed guidance law and incorporating a repulsion function from the artificial potential field method to achieve cooperative collision and obstacle avoidance at the kinematics level. Additionally, a radial basis function neural network is utilized to approximate external disturbances and unmodeled dynamics, while an adaptive neural network quantization controller is designed to handle quantization of input variables. The stability of the closed-loop system is proven using input-to-state stability theory. Finally,simulation results using Matlab validate the effectiveness of the proposed strategy