模糊聚类分析在色木孔洞缺陷位置声波检测中的应用  

Application of Fuzzy Cluster Analysis on Diagnosing the Hole Defects in Acer mono Wood Using Acoustic Testing

作　　者：郭瑞[1] 马肖祥 邢涛[1] 孟宪静[1] 

机构地区：[1]东北林业大学工程技术学院,黑龙江哈尔滨150040

出　　处：《西北林学院学报》2015年第4期251-256,共6页Journal of Northwest Forestry University

基　　金：中央高校基本科研业务费专项资金支撑项目(DL11CB02);国家自然科学基金支撑项目(41171274);中国博士后科学基金支撑项目(2011M500036)

摘　　要：为了实现木材孔洞缺陷位置的检测,提出了一种基于模糊聚类分析的新型木材声波无损检测方法。针对端部孔、无孔洞和中间孔木材试件,运用敲击法采集木材声波信号,提取时频特征向量作为样本数据,运用基于传递闭包的模糊相似矩阵对训练样本进行聚类分析,建立不同类别的模糊模式库,采用最大隶属度原则对待测样本进行识别。结果表明:此方法克服了模糊聚类单一分析方法的不确定性,实现了多指标定量化的检测;该方法能够有效地对色木孔洞缺陷位置进行检测,且准确率较高,检测端部孔样本的准确率为84%,无孔洞样本准确率为94%,中间孔样本准确率为92%。In order to achieve the non-destructive detection of wood hole defects,a new method based on fuzzy clustering analysis of wood acoustic was developed.Acoustic signals of the wood specimen with the hole defects at the end and in the middle as well as without hole defects were collected by percussion method.Time-frequency feature vectors were extracted as the sample data,and the fuzzy similar matrix based on the transitive closure was imposed to the cluster analysis for the training samples to create different classes of fuzzy pattern a maximum degree principle to identify the test samples was adopted.Experimental results showed that this method overcame the uncertainty of the fuzzy clustering single analysis,achieved a multi-index quantitative detection.The method could effectively detect the position of the hole defects in Acer mono wood with high accuracy,84% for the wood the hole defects at the end,94% for the wood without defects,92%for the wood with hole defects in the middle.

关 键 词：模糊模式识别 木材声波 无损检测 模糊聚类 

分 类 号：S781.5[农业科学—木材科学与技术]