基于相关性的钢丝绳断丝特征提取

断丝信号与正常信号的非相关性是断丝信号的重要时域特征。以相关分析理论为指导,研究钢丝绳检测信号的相关性,提出一种获取断丝相关特征的方法。利用检测信号的相关运算,通过选择时延,提取断丝信号与股波信号在特定时延点的相关特征。研究相关窗长度和时延对结果的影响,提出确定相关窗长度和最佳时延的方法。最后针对一段检测信号,验证提取断丝相关性特征的过程。相关性是时域信号的重要特征,基于该特征的相关性识别方法对于提高断丝自动识别的可靠性和准确性具有一定的应用价值。

CORRELATION ANALYSIS OF BROKEN WIRE SIGNAL IN WIRE ROPE TESTING

Correlation analysis is an important method to recognize broken wire from testing the signal of steel wire rope. Based on character of testing signal, by means of correlation function of testing signal, cross-correlation between signal with broken wire and standard signal extracting from testing signal is performed by choosing best time-relay. The length of correlation window has significant influence on correlation coefficient, and reasonable resolution is presented. As a key index for gaining stable correlation character index, time-relay is researched. At last, testing signal is analyzed with correlation analysis method. Reliability of automotive recognition can be increased with correlation analysis method, cooperating with other recognition methods in time domain.