摘要
针对桩基缺陷信号特征提取困难的问题,对缺陷信号进行经验模态分解(EMD)并求取信息熵,构建了一种基于信息熵的均值特征向量;针对特征向量元素数较多,不能真实反映信号的特征以及过多的特征影响识别效率的问题,引入模糊聚类的相关技术方法,对均值特征向量进行相空间重构,对重构矩阵进行模糊聚类,根据聚类的结果构建新的特征向量。通过实验仿真证明了该特征向量取得了很好的识别效果,具有实际应用价值,可以作为一种有效的特征类型;同时验证了通过模糊聚类进行特征向量降维的有效性。
In this paper,according to the difficulty in feature extraction of pile foundation defect signals,the defect signals are firstly decomposed by empirical mode decomposition(EMD)and the information entropy is obtained.A mean value feature vector based on information entropy is constructed.Aiming at the problem that a large number of eigenvector elements cannot truly reflect the characteristics of the signal and too many features have affected the recognition efficiency,a theory of fuzzy clustering is introduced to phase space reconstruction of the mean value feature vector,and fuzzy clustering of the reconstruction matrix is performed,and further new eigenvectors are generated based on the clustering results.Through experimental simulation,it is found that the feature vector has a good recognition effect,and thus has practical application value and can be used as an effective feature type;at the same time,it validates effectiveness of the feature vector dimension reduction through fuzzy clustering.
作者
李衡
康维新
LI Heng;KANG Weixin(College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, China)
出处
《应用科技》
CAS
2019年第2期88-93,共6页
Applied Science and Technology
基金
国家自然科学基金项目(61371174)
关键词
桩基
经验模态分解
信息熵
模糊聚类
缺陷
无损检测
特征提取
识别
pile foundation
EMD
information entropy
fuzzy clustering
defect
non-destructive testing
feature extraction
identification