小波包最优基分解树的降噪滤波方法研究

Research on Denoising and Filtering Method based on Wavelet Packet Optimal Base Decomposition Tree

装载机的传动系统工作时传感器所采集到的信号受环境干扰大,不利于数据分析,因此在研究其工作状态时需对数据信号进行降噪滤波。本文以5t装载机的前轴扭矩信号为研究对象,剥离其趋势项后采用小波包变换法进行db9-6尺度分解,进而通过小波包最优分解树得到重构信号。通过与巴特沃斯去噪、小波变换去噪法对比后得出:巴特沃斯滤波处理法相较于原始信号存在相位偏移;小波包最优基分解树去噪后的信噪比、均方根误差分别为16.38与74.71,与小波变换去噪法相比结果近似,更适用于工况识别、人工智能算法等领域。研究结果可为其他同类型工程机械数据信号处理提供方法依据。

Signals of loader′s transmission system collected by sensors are always disturbed because of the harsh environment,which is not conducive to data analysis.Therefore,the signal should first be denoised and filtered when studying its working condition.Taking a front axle torque signal of a 5 t loader as the research object,the wavelet packet transform method was used to decompose the db9-6 scale after eliminating the trend,and then the reconstructed signal was obtained by the wavelet packet optimal decomposition tree.Compared with the Butterworth denoising and wavelet denoising methods,it is concluded that Butterworth filtering method has phase deviation compared with the original signal,and the signal-to-noise ratio and root mean square error of optimal base decomposition tree of wavelet packet after de-noising are 16.38 and 74.71 respectively,which are similar to the wavelet transform denoising results and more suitable for the fields of working condition identification and artificial intelligence algorithms,etc.The research results could provide a method basis for data signal processing of other similar construction machinery.