点击次数：
所属单位：航空学院
发表刊物：Eur. Workshop Struct. Heal. Monit., EWSHM
摘要：For the practical aerospace application of structural health monitoring (SHM), one of the key challenges is the reliability of damage evaluation under time varying conditions. Recently, Gaussian Mixture Model (GMM) based SHM methods have been studied to deal with this problem. However, most of the mixture model in these methods could not change with the time varying conditions adaptively, which results in unreliable damage monitoring especially for complex structures. In this paper, an improved dynamic probability mixture model for damage monitoring is proposed. An adaptive probability density peaks search clustering based Expectation Maximization (EM) algorithm is proposed to construct GMM efficiently, and GMM is used to model the probability characteristic of guided wave features introduced by time varying conditions. The damage evaluation is realized by observing the cumulative migration trend of GMMs. The method is validated by the hole-edge cracks monitoring of an aircraft wing spar, which is a main load-bearing structure of an aircraft structure. The results indicate that the method efficiently realize adaptive and reliable damage monitoring under varying structural boundary conditions. © 2018 NDT.net. All rights reserved.
是否译文：否
发表时间：2018-01-01
合写作者：Fang, Fang,邱雷,袁慎芳,王一平
通讯作者：Fang, Fang,顾芳芳