A Three-DOF Wave Compensation Control Method for Ships Based on the DMAPPO Algorithm

When engineering ships perform servicing operations on offshore wind turbines, wave action induces six-degrees-of-freedom (6-DOF) motion, severely compromising operational safety. Among these, pitch, roll, and yaw motions exert the greatest impact on ship stability. Therefore, a decision multi-agent proximal policy optimization (DMAPPO) method for ship motion compensation control is proposed to achieve stable ship operations. In this method, a MAPPO algorithm is proposed to capture the motion characteristics of ships under complex and variable sea conditions. Furthermore, integrating MAPPO with PID control through decision strategies to enhance stability requirements in compensation control, while applying local and global rewards to enhance multi-agent coordination capabilities. Additionally, the decision strategy effectively suppresses oscillations caused by control strategy switching through the decision function and action selection strategy. Finally, the effectiveness of the proposed method was validated through simulation and experimentation, achieving compensation efficiency exceeding 94%.