Abstract: Aiming at the system composed of the same and degraded independent components,an optimal maintenance decision optimization model was proposed.The degradation process of components in the system was described as a Markov process with discrete time continuous states.Based on the analysis of system's transition characteristics,the Markov process theory was used to solve the stationary probability expression of each state of the individual components in the system.An optimal decision model for the average cost rate of a multi-component system with multi-state degradation characteristics was established.The optimal detection period,opportunity and preventive maintenance threshold of the system were obtained to ensure the minimal long-term average cost rate.The wind turbine was used as an example to analyze and verify the correctness and applicability of the maintenance decision model.The results showed that the model could help the actual multi-component system to give a more economical and reasonable maintenance strategy.

Key words: multi-unit systems, discrete state modeling, deterioration state space partitioning method, maintenance decision, wind turbine

摘要： 针对劣化独立的同类型部件组成的系统,提出了最优的维修决策优化模型。将系统中部件的劣化过程描述为具有时间连续状态离散的马尔可夫过程,在分析系统转移特性的基础上,运用马尔可夫过程理论求解了系统中单个部件各状态的平稳概率表达式,建立了具有多状态劣化特性的多部件系统的平均费用率最优决策模型,在系统长期平均费用率最小的情况下,得到系统最优的检测周期、机会和预防维修阈值。以风电机组为例,验证了维修决策模型的正确性和适用性。结果表明,该模型可以为实际的多部件系统给出较为经济且合理的维修策略。

关键词: 多部件系统, 离散状态建模, 劣化状态空间划分, 维修决策, 风电机组