基于小波变换的扭矩传感器瞬时信号预处理

Transient signal preprocessing of torque sensor based on wavelet transformation

为了使扭矩信号的变化情况不受干扰信号的影响,通过采集扭矩传感器的瞬时信号,利用Haar小波的信号分解和重构算法对采样信号进行了分析和处理,将采样信号分解成多个层次上的高频和低频部分,针对实际需求,采用了强制滤波的方法,对高频部分信号的小波系数设定合适的阈值进行修正,再通过信号重构来达到对瞬时信号消噪的目的。实验结果表明,利用小波变换的方法对扭矩传感器瞬时信号进行预处理是可行和有效的。

In order to make the changing of fatigue signal do not influenced by jamming signal, the signal decomposing and recovery algorithm based on Haar wavelet was used to analyze and process the signal. By sampling, the transient torque signal was decomposed into low frequency part and high frequency part in each decomposing levels. According to the actual requirements, wavelet coefficient in high frequency signal was modified by setting a suitable threshold value, then de-noising could be reached by signal reconstruction. The result shows that, using wavelet transformation to preprocess the transient signal of torque sensor is useful and effective.