There are a thousand Hamlets in a thousand people's eyes. The conflicts induced by different measures also exist in mechanical health monitoring, especially for the ones operate in variable conditions like aero-engine and wind turbine. The fault features could be sensitive to the change of operation, and thus represented in different levels associated with the measuring scales. In addition, the situation maybe further compounded by the coupling of performance degradation and fault feature. These factors, mixed with the data sampling and multi-fault case, make it impossible to conduct a reasonable monitoring based on the single scale data. Thereafter, the present investigation plan to address above issues from the aspects of data reconstruction and association, diagnosis methodology and health evaluation: Construct the hybrid multiscale mode to support the data reconstruction and association, and then address the data deficiency and scale isolation problem; Construct the synergized model-data method to address the separation issues and realize fault tracking; Construct the frame of probabilistic evaluation of title problem to address the fault identifiability problem; Conduct the experimental validations on aero-engine and wind turbine at last. The present research may further develop the methodologies of data fusion, fault diagnosis and health evaluation, and thus supports the operational maintenance of aero-engine and wind turbine.
横看成岭侧成峰,远近高低各不同,观测角度和尺度导致了结论的差异,机械装备健康监测也是如此,特别是变工况运行状态下的重要装备,如航空发动机、风力发电机,其故障特征受变工况影响不稳定,在不同时空尺度上表现有所不同,加之正常性能退化与故障特征耦合,难以通过单一尺度数据进行有效监测,此外传统研究较少考虑数据抽取、多故障混合导致的置信度问题,因此本项目拟在变工况机械装备数据关联、诊断方法、健康评估三方面开展研究:建立宏-细观混合多尺度模型支持数据重构与关联,解决数据不完整与尺度间信息孤立问题,实现数据关联与特征解耦;建立模型-数据协同化健康监测方法,解决模型、数据驱动割裂问题,实现故障溯源;建立概率化健康监测评估方法,解决故障可辨识性问题,实现合理评估;最终在航发、风机等重要装备上开展实验和应用研究。本项目研究有望在数据融合、故障诊断、健康评估等方面有所创新突破,为航发、风机安全运行保障提供支持。
项目严格按照计划书执行,并进行了纵向深入扩展,形成了多尺度数据恢复与关联方法、模型-数据协同化健康监测方法,概率化健康监测评估方法,在此基础上,进一步打破了欧美在航空发动机转子叶片在线测振上的理论与技术垄断,借助与“余数定理”等问题提出了具有我国特色的极端低介入叶端定时测量理论,并研制了硬件样机,入选十三五科技成果展内展;提出了超声导波全聚焦相控阵列概念及其快速实现,有望进一步在大型机械结构的健康监测中发挥作用。项目累计发表SCI论文50篇,出版第一作者学术专著1部,发表EI论文7篇,授权国家发明专利41项,获得陕西省技术发明一等奖、陕西省高等学校自然科学一等奖、全国可再生能源学会科技创新二等奖各1项,相关成果在624、华电电科院等多家单位获得应用。
{{i.achievement_title}}
数据更新时间:2023-05-31
演化经济地理学视角下的产业结构演替与分叉研究评述
涡度相关技术及其在陆地生态系统通量研究中的应用
正交异性钢桥面板纵肋-面板疲劳开裂的CFRP加固研究
特斯拉涡轮机运行性能研究综述
栓接U肋钢箱梁考虑对接偏差的疲劳性能及改进方法研究
基于健康监测的桥梁结构多尺度损伤预后方法研究
基于多时间尺度模型的变工况大型旋转机械健康管理研究
面向变工况的复杂装备多域特征变关联设计关键技术
变工况下旋转机械健康状态识别的深度迁移学习方法研究