【摘要】 基于燃气涡轮发动机由计划维修模式向视情维修(Condition Based Maintenance, CBM)模式转变、由被动保障向主动保障发展的实际需求，针对目前燃气涡轮发动机在故障预测与健康管理(Prognostics and Health Management, PHM)系统设计上存在的整体性与系统性不强的问题，通过开展燃气涡轮发动机的PHM技术研究，围绕故障诊断、寿命预测、保障决策等PHM系统的核心功能要素，提出了一种新型燃气涡轮发动机PHM系统架构。该架构包括实时监测模块与保障决策模块，实时监测模块主要基于测试性建模实现故障诊断逻辑设计以满足涡轮发动机故障检测与健康管理的实时性要求，保障决策模块为燃气涡轮发动机的CBM保障提供了一种数据驱动的决策生成方法，通过长短时记忆网络(Long Short-Term Memory, LSTM)算法预测关键部件寿命，以匹配最佳的维修保障方案，提升燃气涡轮发动机的保障效能，最后给出了该系统架构的部分核心功能的设计。更多还原

【Abstract】 Based on the actual demand of the transformation of gas turbine engines from planned maintenance mode to condition based maintenance(CBM) mode, and from passive guarantee to active guarantee, in view of the problem of the weak integrity and systematization of the current prognostics and health management(PHM) system design of gas turbine engines, throgh the research of PHM technology of gas turbine engine, a new gas turbine engine PHM system structure is proposed, focusing on the core functional elements of gas turbine PHM system such as fault diagnosis, life prediction and maintance decision-making.The system structure includes the real-time monitoring module and the maintance decision-making module.The real-time monitoring module mainly realizes fault diagnosis logic design based on testability modeling to meet the real-time requirements of turbine engine fault detection and health management.The maintance decision-making module provides a data-driven decision-making method for gas turbine engine CBM guarantee, and predicts the life of key components of gas turbin engine through long short-term memory(LSTM) algorithm to match the best maintenance and guarantee plan, and improves the efficiency of gas turbine engine maintance work.Finally, the design of some of the core functions of the system architecture is provided.更多还原