Abstract: To address the shortcomings of traditional operation and maintenance (O&M) practices for the Three Gorges ship lift, such as insufficient data mining and reliance on subjective experience in fault diagnosis, this paper employed digital twin theory integrated with key technologies including fault diagnosis analysis methods, matter-element entropy weight method, convolutional neural network (CNN) algorithm, SSA-ICEEMDAN algorithm to develop an intelligent O&M platform for the ship lift. Through platform implementation and data analysis, the system successfully achieved dynamic monitoring, fault diagnosis, and predictive maintenance of the ship lift's operational status, giving a comprehensive health evaluation on the equipment. Application results demonstrated that the platform not only ensures the safe and stable operation of the Three Gorges ship lift, but significantly enhanced O&M efficiency as well. This successful application of digital twin technology and associated methods in the intelligent O&M of large-scale ship lift navigation facilities provides an innovative solution for analogous engineering challenges.