基于小波分析的交流CMT焊接电信号滤波

Electric signals filtering of AC CMT welding based on wavelet analysis

在传统阈值去噪法的基础上提出一种改进阈值去噪算法.该算法在阈值区间加入处理函数和调节系数,填补阈值区间内的空白,从而实现灵活调整信号衰减程度和平滑程度,提高重构后信号的连续性和逼近程度.试验利用coif小波对交流CMT焊接不锈钢电流电压信号进行3层分解,采用改进阈值法对分解后的小波系数进行处理.结果表明,通过灵活调整调节系数,改进算法能够很好地去除交流CMT焊接电信号中的高频噪声,保留有用突变成分和规则脉动信号,对有用信号的衰减比软阈值方法小,提高了与真实信号的逼近程度,适用于交流CMT焊接电信号的滤波.

An improved threshold de-noise algorithm based on the general de-noise threshold algorithm was proposed.A processing function and an adjustment coefficient were added to fill the gap within the threshold range. It is aimed at achieving the flexibility to adjust the degree of attenuation and smoothing and improving the continuity of the reconstructed signal and the degree of approximation of signals. Coif wavelet was used to decompose the electrical signal of stainless steel welding by AC CMT for 3-tier,then the decomposed wavelet coefficients were processed by improved threshold de-noise algorithm. The results showed that by adjusting the adjustment coefficient flexibly,high-frequency noise can be removed well as the useful mutation component and the rule pulsation signals were held through the processing of improved algorithm. The attenuation of the useful signals was lower than that the soft threshold method performed and improved the approximation to the real signals. The improved threshold de-noise algorithm is suitable for filtering of electrical signals of AC CMT.