基于Volterra级数的提升小波变换边界处理及应用.pdfVIP

学兔兔 第33卷 第1期 仪 器 仪 表 学 报 V0l_33 No．1 2012年1月 Chinese Journal of Scientific Instrument Jan．2012 基于Volterra级数的提升小波变换边界处理及应用术 段礼祥，张来斌，陈敬龙 (中国石油大学(北京)机械与储运工程学院 北京 102249) 摘 要：针对现有的提升小波变换容易产生边界振荡和频率混叠的不足，提出了一种将Volterra级数模型和抗混叠提升小波包 相结合的信号处理方法。首先对信号两端进行数据延拓，用二阶Voherra级数预测模型对延拓信号进行预测；然后用抗混叠提 升算法对信号进行小波包分解。对仿真信号进行边界处理后，信号在边界不会出现振荡现象；用抗混叠提升小波包对信号进行 分解不会引起频率混叠现象。工程应用中，从强大的背景噪声中提取出了往复泵柱塞与缸套碰磨产生的微弱振动冲击信号．诊 断出了密封盘根过度磨损的故障。 关键词：信号处理；提升小波变换；Volterra级数；边界处理；故障诊断 中图分类号：TP306 文献标识码：A 国家标准学科分类代码：510．4030 Boundary treatment of lifting wavelet transform based on Volterra series model and its application Duan Lixiang，Zhang Laibin，Chen Jinglong (College ofMechanical and Transportation Engineering，China University ofPetroleum(Beijing)，Beijing 102249，China) Abstract：Considering that traditional lifting wavelet transform may lead to end distortions and frequency aliasing，a new signal processing method combining Voherra series model and anti--aliasing lifting scheme wavelet packet is pres-- ented．Firstly，the two ends of the original signal are extended and predicted using the second-order Volterra series mode1．Then the extended signal is decomposed with the method of anti-aliasing lifting scheme wavelet packet．The effectiveness of the proposed method is validated with simulation data．In the processing of an engineering signal，we acquired the extremely weak vibration signal generated by the collision between plunger and cylinder of a reciproca- ting water-flooding pump from strong interference background and precisely forecast the excessive wearing of the packing in the cylinder． Key words：signal processing；lifting wavelet transform；Volterra series；bounda