电子测量与仪器学报

周媛,左洪福,刘鹏鹏.基于相关时间规整的航空发动机缓变故障诊断[J].电子测量与仪器学报,2017,31(9):1358-1364

基于相关时间规整的航空发动机缓变故障诊断

Aero engine gradual changing fault diagnosis based on canonical time warping algorithm

DOI：10.13382/j.jemi.2017.09.002

英文关键词:aero engine gradual changing fault fault diagnosis the canonical time warping algorithm concurrent fault

基金项目:江苏高校品牌专业建设工程（1181081501003）、国家自然科学基金（61403198）、江苏省自然科学基金（BK20140827）资助项目

作者	单位
周媛	南京信息工程大学电子与信息工程学院南京210044
左洪福	南京航空航天大学民航学院南京210016
刘鹏鹏	中国船舶工业系统工程研究院北京100094

Author	Institution
Zhou Yuan	College of Electronic and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
Zuo Hongfu	College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
Liu Pengpeng	Systems Engineering Research Institute, China State Shipbuilding Corporation, Beijing 100094, China

摘要点击次数: 2205

全文下载次数: 7189

中文摘要:

针对航空发动机退化导致的缓变故障诊断问题，提出了一种基于相关时间规整算法的诊断模型，该模型通过挖掘发动机退化过程中过渡信息，根据退化数据中发动机状态变化特征来识别故障模式。通过仿真数据和实际案例数据实验证明，在监测数据满足1~2个飞行循环间隔前提下，该模型能够区分发动机正常状态和故障状态，对发动机本体的缓变故障能够定位到部件级，平均G－mean值为0.948 7，拥有较好的鲁棒性和准确度，为民航发动机健康管理提供了一种可行的工程方法。

英文摘要:

Aiming at the degradation caused gradual changing fault diagnosis of aero engine, this paper proposed a diagnosis model based on the canonical time warping (CTW) algorithm, which can discriminate the fault pattern based on the transition features from degradation data via mining the transition information of degradation. We conducted the proposed model on both simulated data and real data. The experimental results show that the proposed model can recognize the normal state and fault state and locate the gradual changing fault in component level, whoseG－meanvalue is 0.948 7 in the premise that the flight cycle interval is 1 or 2. The proposed model provides a feasible engineering method for civil aviation aero engine health management, and has good robustness and high accuracy.

查看全文查看/发表评论下载PDF阅读器