基于线性预测分析MIG焊电弧声信号特征提取  

Feature extraction of arc sound signal in MIG welding based on linear prediction analysis

作　　者：刘立君[1,2] 王奇[1] 兰虎[2] 郑红艳[2] 李峰 

机构地区：[1]浙江大学宁波理工学院,浙江宁波315100 [2]哈尔滨理工大学,哈尔滨150080 [3]宁波双林汽车部件股份有限公司,浙江宁波315153

出　　处：《焊接学报》2010年第7期35-38,共4页Transactions of The China Welding Institution

基　　金：哈尔滨工业大学现代焊接国家重点实验室基金资助项目(W08003);黑龙江省自然科学基金资助项目(E2007-01);黑龙江省青年骨干教师基金资助项目(1153G009);哈尔滨市科技创新基金资助项目(2007RFQXG055)

摘　　要：以不同焊接状态电弧声信号识别为目的,从MIG焊平板对接射流过渡电弧声信号入手,探寻了电弧声的产生机理和传播特性.提出电弧声是气体放电的伴生物,是等离子体集体振荡以波形方式传播的结果,也是声源和声道共同作用的产物,其频谱特性主要取决于声道的频率响应.在此基础上,建立了电弧声线性产生模型,采用Levin-son-Durbin递推算法提取了电弧声信号的线性预测系数(LPC)、反射系数和对数面积比系数,并构建了电弧声信号的联合特征向量.Aiming at recognition of the arc sound signal in different welding status,the mechanism of arc sound generation and propagation were investigated to MIG butt welding with spray transfer. It is proposed that arc sound is an associated thing of gas discharge,is the result of transmitting by the wave when the plasma collectively vibrates,is also the outcome of their combined action,including the sound source and the sound track. Among them,the spectrum characteristic depends mainly on the frequency respond of sound track. Based on the above analysis,the linear generation model of arc sound was set up. The linear prediction coefficients (LPC){ai},reflection coefficientsand logarithm area ratio coefficientswere extracted by Levinson-Durbin algorithm. In addition,the combined characteristic vector of arc sound signal was constructed.

关 键 词：线性预测分析 MIG焊 电弧声 特征提取 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]