振动攻丝扭矩信号处理中应用小波包消噪

Application of Wavelet Pocket to Noise-Killing in Processing Torque Signal during Vibration Tapping

简述小波包分析的基本理论及其主要算法 .讨论用该理论进行信号消噪的基本原理 ,并应用于振动攻丝扭矩的信号处理 .通过对所采样的扭矩信号进行小波包分解 ,获得信号精细的低频和高频两个部分 .利用阈值量化处理的方法去除噪声 ,再将剩余部分进行重构 ,从而达到消噪声的目的 .实验结果表明 ,该方法能够有效地消除采样信号中的噪声 ,进而提高扭矩值的计算精度 .

Beginning with the basic theory and the main algorithm of wavelet pocket analysis, the authors discuss the basic principle of using this theory to signal denoising; and apply it to torque signal processing during vibration tapping. With regard to the fineness of signals, the part of low frequency and the part of high frequency as two parts are obtained through decomposing sampled torque signals with wavelet pocket. After denoising by applying quantized processing of threshold value, the rest are reconstructed and the objective of denoising is attained. Experimental results show that the method is able to kill the noise effectively in sampled signals and then to improve the computational accuracy of torque value.