Condition-based maintenance method for multi-component systems under discrete-state condition: Subsea production system as a case

Components serve as the fundamental units of complex systems when implementing maintenance activities. Predicting the remaining useful life of components, evaluating their condition, and selecting maintenance measures are important factors that affect the accuracy and rationality of maintenance plans. The utilization of simplistic maintenance measures that aim for achieving a perfect state and the reliance on expert experience to obtain failure thresholds have limited the development of condition-based maintenance methods. To address these limitations, a condition-based maintenance method for multi-component systems under discrete-state conditions is proposed. A maintenance list is established based on the historical maintenance data of the system. Additionally, a Markov chain is employed to construct a single-component state transition model, which calculates the component state transition matrix based on historical operational data, and obtains the component state recovery matrix based on historical maintenance data. A single-step predictive model for components is established to study the probability of state transitions for components under different states and at different times. The maintenance optimization model is then applied to determine the maintenance plan of the system. The feasibility of the proposed method is demonstrated through an example involving critical equipment in a subsea production system for offshore oil.