航空发动机快变信号稀疏时频诊断方法研究

The aero-engine is the heart of the airplane. The safety and reliable operation is the first priority of the aero-engine, which always attracts the global concern. However, the aero-engine services in extreme environment, such as high speed, high temperature, heavy load, and strong disturbance, which may result in serious accident. Thus health monitoring and machine fault diagnosis play a significant role in guaranteeing the safety and reliable operation of aero-engine. Because of the factors such as the frequently varying condition, the sharp speed-up or speed-down, and high-speed varying-stiffness, the instantaneous frequency of the vibration of the aero-engine is always highly oscillated and changed with time. The vibration signals with fast time-varying instantaneous frequency is strong nonstationary. This project will analyze aero-engine vibration signals with fast time-varying instantaneous frequency, and research the sparse time-frequency method for aero-engine fault diagnosis. Firstly, the dynamic modeling of aero-engine dual-rotor system and fault mechanism will be studied to reveal the vibration mechanism of the high-oscillated time-varying signals. Secondly, matching synchrosqueezing transform will be studied to analyze the high-oscillated time-varying signals, and thus to improve the energy concentration of the time-frequency representation for the high-oscillated time-varying signals. Thirdly, the sparse time-representation dictionary can be constructed by the proposed matching synchrosqueezing transform, and thus to transform forward and backward between the physical space and the sparse domain. Then the object function for sparse time-frequency representation of the high-oscillated time-varying signals can be constructed and eventually solved fast by advanced optimization algorithm. As the result, the sparsity of the high-oscillated time-varying signals in the time-frequency domain and the extraction accuracy can be improved. Lastly, taking a center aero-engine dual-rotor system as the research object, the experimental and engineering study will be implemented. The research results of this project will provide theoretical basis and technical support for aero-engine fault diagnosis, and thus will provide a significant role in guaranteeing the safety and reliable operation of mechanical equipment.

航空发动机是飞机的“心脏”，常常工作于高速、高温、重载、强扰动等极端服役环境，导致灾难性事故时有发生，健康监测与故障诊断对保障航空发动机运行安全至关重要。本项目以航空发动机频繁变工况、大幅升降速、高转速变刚度运行等引起的瞬时频率快速变化、且具有强时变非平稳特性的快变信号为分析对象，研究快变信号稀疏时频诊断方法。具体包括：研究航空发动机双转子系统动力学建模与故障机理，揭示快变信号产生机理；研究快变信号匹配同步压缩变换方法，实现快变信号高聚集性时频表征；构建快变信号稀疏时频表示字典，研究快变信号稀疏时频表示的目标函数构造与快速求解方法，提高快变信号在时频域的稀疏性与提取精度；以某型航空发动机双转子系统为对象，开展实验与工程验证研究。本项目在上述研究基础上，提出航空发动机快变信号稀疏时频诊断方法，有望为航空发动机等机械装备故障诊断提供理论支撑和技术支持，对保障机械装备运行安全具有重要意义。

航空发动机是飞机的“心脏”，常常工作于高速、高温、重载、强扰动等极端服役环境，导致灾难性事故时有发生，健康监测与故障诊断对保障航空发动机运行安全至关重要。本项目以航空发动机频繁变工况、大幅升降速、高转速变刚度运行等引起的瞬时频率快速变化、且具有强时变非平稳特性的快变信号为分析对象，研究快变信号稀疏时频诊断方法。.（1）提出了航空发动机转子系统建模与典型故障（碰摩和转轴裂纹）建模方法，数学本质上分别揭示了转子系统碰摩故障和转轴裂纹故障诱发快变振动的物理机理；（2）提出了匹配同步压缩小波变换方法，采用“匹配重排”策略的匹配时频分析思想，构造了能够匹配信号调频本质的匹配瞬时频率重排算子，解决了传统时频重排方法和同步压缩小波变换在时频聚集性与重构性质之间不可兼得的问题；（3）提出了非凸稀疏正则化与保凸优化、组内组间稀疏正则化、脊线加权的快变信号稀疏建模等方法，实现了快变信号稀疏时频表示；（4）搭建了航空发动机双转子系统等试验平台，开展了实验研究与工程应用。.基于本项目相关研究成果，发表论文22篇，其中SCI论文17篇（2篇ESI高被引论文），《机械工程学报》1篇，会议论文4篇；申请发明专利3项，授权3项。相关研究成果得到麻省理工学院（MIT）、IEEE Fellow Steven B. Leeb教授，英国利物浦大学、《Journal of Sound and Vibration》杂志副主编H. Ouyang教授，IEEE/ASME会士、美国凯斯西储大学Robert X. Gao教授，香港城市大学Kwok Leung TSUI教授，东北大学、中科院闻邦椿院士，上海交通大学、杰青/长江/万人计划学者彭志科教授，浙江大学、长江学者徐兵教授等著名专家的引用与正面评价。基于本项目的资助与相关研究，项目负责人于2018年晋升副教授，获批国家自然科学基金重大研究计划培育项目1项，作为骨干成员参与国家自然科学基金重点项目和国家重点研发计划项目各一项。

内容获取失败，请点击重试