摘要
小波多分辨率分解能够将信号在不同的尺度上展开 ,因而具有对信号按频带进行处理的能力 ,这对于建立表征识别故障信号的特征以及清除信号的干扰与噪声等方面具有十分重要的意义。目前在钢丝绳LF检测中对断丝信号的处理仍比较困难 ,将它应用于断丝信号的消噪平滑和奇异性检测 ,使得信号更加光滑 ,避免断丝误判 ,同时使奇异点更加明显 ,有效地提高内外部断丝识别和定量检测的准确率 。
The multi-resolution analysis of wavelet can decompose signal on different scales. So it has the ability of processing for signal in frequency-band, which is very useful to build up the identification features of signal and to cancel the signal noise. Nowadays it is still difficult to process the signals of broken steel ropes in LF testing. This paper applies it for signal denoising and smoothing and singularity detection. The method makes signals more smoothly, avoids mistakes and makes singular points outstanding. It can effectively improve accuracy of recognizing internal and external broken wire ropes and quantitative inspection. The experiment indicates it is evidently effective.
出处
《煤矿机械》
北大核心
2005年第2期117-118,共2页
Coal Mine Machinery
关键词
多分辨分析
消噪
奇异性检测
LF检测
multi-resolution analysis
denoising
singularity detection
LF (Localized Fault) testing
