Wing-assisted technology is an effective way to reduce emissions and promote the decarbonization of the shipping industry. The lifting and lowering of wing-sail is usually driven by hydraulic system. Leakage, as an important failure form, directly affects the safety as well as the functioning of hydraulic system. To increase the system reliability and improve the wing-assisted effect, it is essential to conduct leakage fault diagnosis of lifting and lowering hydraulic system. In this paper, an AMESim simulation model of lifting and lowering hydraulic system of a Very Large Crude Carrier (VLCC) is established to analyze the operation characteristics of the hydraulic system. The effectiveness of the model is verified by the operation data of the actual hydraulic system. On this basis, a wavelet packet transform (WPT)-based sensitive feature extracting method of leakage fault for the hydraulic system is proposed. Subsequently, a support vector machine (SVM)-based multi-classification model and diagnosis method of leakage fault are proposed. The study results show that the proposed method has an accuracy of as high as 97.5% for six leakage fault modes. It is of great significance for ensuring the reliability of the wing-sail operation and improving the utilization rate of the offshore wind resources.