Subsea trees are typical marine equipment subject to internal degradation and external impacts. Resilience is used to evaluate the ability to resist performance degradation and restore functionality from disruption. Resilience assessment can provide quantitative support for on-site production and maintenance decisions. However, existing resilience assessment methods often ignore the random process of natural degradation, resulting in distorted results. To address this, a resilience assessment method for subsea trees based on probabilistic interval models is proposed. It enables probabilistic resilience outputs while accounting for natural degradation randomness. A natural degradation model is established by integrating a semi-Markov process with a dynamic Bayesian network to characterize the random evolution of equipment states. Disaster shock and maintenance models are constructed using a Bayesian network and a logistic function, respectively. System reliability and normalized mean residual life are quantified by a weighted-area method. They are then combined into a two-dimensional resilience vector. Monte Carlo simulation is used to derive the probability distribution of system resilience. The method is validated on the control system of a subsea tree in the South China Sea. Results show that resilience fluctuations under random shocks can be characterized, providing a more reliable probabilistic basis for resilience assessment under uncertainty.
原文链接
分享
引用
