应用经验模态分解的箱型结构裂纹诊断方法  

A new crack diagnosis method for box structures based on empirical mode decomposition

作　　者：谢一村 樊世超[1] 杨艳静[1] 

机构地区：[1]北京卫星环境工程研究所,北京100094

出　　处：《航天器环境工程》2015年第5期489-495,共7页Spacecraft Environment Engineering

摘　　要：为获得航天器结构的更多故障信息,文章提出了一种应用经验模态分解对一个两层箱型结构进行裂纹故障诊断的方法。先对该结构进行有限元建模和模态分析,发现结构有无故障均对模态的影响非常小;再对实际结构进行正弦振动试验,对振动响应信号进行经验模态分解,利用固有模态函数构造标识向量,提取用标识向量间的标识夹角作为诊断结构故障的指标。分析结果表明,结合施加给结构的主振方向,利用标识夹角能识别出裂纹的位置和方向。该方法为航天器结构故障的诊断提供了一种新的思路。A new crack diagnosis method for a two-story box structure based on the empirical mode decompositionis proposed in this paper. According to the FEM simulation analysis, it is shown that the modal of the structure can bebarely influenced by the crack. The response signals of the swept sine vibration test are empirically decomposed into aset of intrinsic mode functions, from which tag vectors are constructed, then tag angles are defined to diagnose thefailure of the board. Combined with the load direction to the structure, the position and the direction of the crack can beidentified by the tag angles.

关 键 词：箱型结构 经验模态分解 故障诊断 标识向量 

分 类 号：V415.6[航空宇航科学与技术—航空宇航推进理论与工程]