Considering non-periodic inspection and imperfect maintenance comprehensively, and integrating the (s, S) inventory policy, this paper proposes a novel joint optimization strategy for condition-based maintenance and spare parts inventory. The system degradation process is described based on the Wiener process, and residual degradation and degradation rate are introduced to establish an imperfect condition-based maintenance model. Based on the concept of first-passage time, the probability distribution of remaining useful life (RUL) is derived, and the interval of non-periodic performance inspection is determined according to the predicted value of RUL. A joint optimization model integrating the (s, S) spare parts inventory policy is established, and the model is optimized and solved by combining Monte Carlo simulation with genetic algorithm, with the objective of minimizing the long-term average maintenance cost rate of components. The proposed method is validated using gyroscope drift degradation data, and compares it with optimization models based on different assumptions. The results show that the proposed method can effectively reduce operation and maintenance costs, significantly optimize the spare parts inventory level while improving system reliability, providing new technical support for maintenance decision-making of industrial equipment.