Abstract: The severe issue of sediment redeposition under pile foundation wharves along China’s silty coastal shores leads to damage to the pile foundations and high silt dredging costs. By integrating the finite element analysis method of pile foundation stress with operational research methods of port management, and taking large coastal pile foundation wharf silt dredging projects as prototypes, this study undertakes an optimization research on desilting construction strategies for pile foundation wharves. It establishes a finite element numerical model for the wharf pile foundation-slope soil structure system, as well as an optimization mathematical model for port dredging construction strategies that comprehensively considers dredging costs, safety, and efficiency. The optimal dredging scheme was obtained using a genetic algorithm improved by reinforcement learning (DDQN-GA). The scheme effectively reduces dredging costs while fully considering structural safety, providing scientifically sound and rational dredging project recommendations for port management.