耦合时变不确定性量化的水电机组运行可靠性建模与实时评估

Hydroelectric energy is clean and renewable energy, hydroelectric generator units is the key equipment of hydropower production process. And it is generally characterized as complex structure and functional units coupled with each other under dynamic environmental and operational conditions. To solve these problems, this project focuses on the deep investigation of the reliability modelling and assessment considering the time-varying uncertainty, and the hydroelectric generator units is chosen as the engineering background. The characteristics and property of the time-varying uncertainty for the hydroelectric generator units will be researched, and the methods of describing and modelling for the uncertainty variables will be studied. Meanwhile, the Copula theory and the multi-source information integration will be utilized to analyze the reliability modelling and assessment, respectively. Major attention will be laid on the following research contents: (1) Develop an integrated reliability modelling method based on the multiple degradation processes considering dynamic environmental and operational conditions; (2) Introduce a Bayesian dynamic model for reliability assessment. Finally, a coherent methodology of the reliability modelling and assessment for the complex equipment and systems is constructed based on the techniques and methods developed above. The research can expand uncertainty quantification theory, enrich and perfect the reliability modelling and assessment methods, as well as improve the maintenance decision-making and the health management for the complex equipment and systems.

水电能源是清洁可再生能源，水电机组是水电能源生产过程中关键的基础设备。针对水电机组系统结构复杂、功能部件相互耦合、服役环境及工况多变等关键问题，本项目以水电机组为研究对象，以时变不确定性下的可靠性建模与评估为切入点，研究水电机组运行过程中所面临的时变不确定性因素的特征和性质，提出时变不确定性的量化、表征及耦合方法；将Copula理论与多源信息融合方法相结合，研究基于多性能退化分析的水电机组可靠性建模与评估，着重研究考虑多性能退化、动态环境和变工况应力的水电机组时变可靠性的建模方法，提出基于贝叶斯动态模型的水电机组运行可靠性评估方法，形成面向复杂装备和系统的可靠性实时评估理论和方法。该项目对拓展不确定性量化理论，丰富和完善现有运行可靠性建模与评估方法，提高复杂装备系统的运行维护决策和健康管理水平具有实际应用价值。

研究了水电机组运行过程中面临的时变不确定性量化和分析问题，基于损伤累积理论建立了与时变不确定性因素相关的失效物理模型。识别了水电机组关键功能特性和关键功能部件，确认了关键功能部件对应的性能指标，基于性能退化实验数据建立了多性能退化分析的水电机组可靠性分析模型。分别将环境条件对应的影响因子和工况应力对应的影响因子引入到了所建立的可靠性模型中，研究了考虑多性能退化和动态环境及变工况作用下的水电机组时变可靠性建模方法。应用经典贝叶斯方法的两步近似法，采用序贯蒙特卡洛方法中的数据仿真方法对建立的可靠性模型进行了动态评估，并采用近似贝叶斯计算方法进行了模型算法的效率与精度的检验，有效建立了融合性能退化试验、运行状态监测数据和任务工况信息的水电机组动态可靠性评估方法。