基于隐马尔可夫和逆推法的水力发电机部件动态检修策略研究  

Research on Dynamic Maintenance Strategy of Hydraulic Generator Components Based on Hidden Markov and Inverse Method

作　　者：郝珈萱 Hao Jiaxuan(Shaanxi Provincial Institute of Water Resources and Electric Power Investigation and Design,Xi'an 710000,Shaanxi)

机构地区：[1]陕西省水利电力勘测设计研究院,陕西西安710000

出　　处：《陕西水利》2024年第9期164-167,共4页Shaanxi Water Resources

摘　　要：为提升现有水力发电机部件检修策略制定水平,提出一种基于隐马尔科夫和逆推法的水力发电机部件动态检修策略制定方法。以水电站水力发电机轮转叶片为研究对象,基于隐马尔科夫模型预测轮转叶片不同裂纹状态的可能停留时间,以动态更新检修周期;根据状态预测结果,基于逆推法量化轮转叶片不同裂纹状态的风险贡献率,以制定差异化的个体最优检修策略。结果表明:该方法可以有效降低风险概率,减少检修成本。研究结果可为工程实践提供理论基础。IIn order to improve the existing maintenance strategy formulation level of hydraulic generator components,a dynamic maintenance strategy formulation method of hydraulic generator components based on hidden Markov and backstepping method is proposed.Taking the rotating blades of the hydropower station as the research object,the possible residence time of different crack states of the rotating blades is predicted based on the hidden Markov model to dynamically update the maintenance cycle.According to the state prediction results,the risk contribution rate of different crack states of the rotating blade is quantified based on the inverse method to formulate a differentiated individual optimal maintenance strategy.The results show that this method can effectively reduce the risk probability and reduce the maintenance cost.The research results can provide a theoretical basis for engineering practice.

关 键 词：水力发电机组 隐型马尔科夫模型 逆推法 检修策略 

分 类 号：TV734.21[水利工程—水利水电工程]