基于声发射技术的钢丝绳断丝模式识别  

Pattern recognition for wire breakage detection of a steel rope by acoustic emission techniques

作　　者：邵永波[1] 裴珍[1] 朱祖铭[2] 于大安[1] 王师[1] 

机构地区：[1]东北大学信息科学与工程学院,沈阳110006 [2]中国科学院金属研究所,沈阳110015

出　　处：《应用声学》1999年第1期1-5,共5页Journal of Applied Acoustics

基　　金：原中国石油天然气总公司资助

摘　　要：基于大量实验及波形分析，给出钢丝绳断丝事件的声发射（ＡＥ）定量化表征形式．采用类内类间距离和二维图示方法综合选择ＡＥ特征，并根据其分布情况建立起基干Ｆｉｓｈｅｒ准则的分段线性分类器，以实现断丝事件的自动检测．Based on large quantity of experiments and waveform analyses, quantitative acoustic emission(AE) signature is achieved for wire breakage in a steel rope. By concur- rently using the measure of within-class and between-class variances and the method of two-dimensional plot, AE features are selected for classification. According to distribution of AE parameters,a piecewise linear classifier is designed with the Fisher discriminate analysis, such that wire breakage can be detected automatically.

关 键 词：声发射 钢丝绳 模式识别 断丝检测 

分 类 号：TG356.45[金属学及工艺—金属压力加工]