多小波对风机故障信号降噪处理的比较研究  

Comparative Study of Multi-wavelet in Fan De-noising

作　　者：王红君[1] 贺鹏[1] 赵辉[1,2] 岳有军[1] 刘明明 

机构地区：[1]天津理工大学天津市复杂系统控制理论与应用重点实验室,天津300384 [2]天津农学院,天津300384 [3]河北钢铁集团邯钢气体厂,河北邯郸056015

出　　处：《化工自动化及仪表》2013年第2期163-166,共4页Control and Instruments in Chemical Industry

基　　金：天津市自然科学基金项目(09JCZDJC23900;10JCZDJC23100);天津市科技支撑计划项目(10ZCECJD43080)

摘　　要：在风机故障诊断系统中,因为现场采集的故障信号含有噪声,对故障诊断的准确性产生很大影响,所以有必要进行降噪处理。采用"平衡预处理的SA4多小波"和"正交插值的SA4多小波"对现场采集的含有噪声的故障信号进行处理,并与已有的"系数重复行的GHM多小波"、"平衡预处理的CL多小波"降噪方法进行比较,结果显示平衡预处理的SA4多小波的降噪方法比GHM和CL多小波的降噪方法更适合风机的故障诊断。The noise in collected fault signals can influence the accuracy of fan＇ s fault diagnosis system and the fan de-noising becomes necessary. Both SA4 multi-wavelet with balanced-pretreatment and CL multi-wavelet were employed to process field＇ s noise-contained fault signals. Comparing de-nosing results shows that SA4 multi-wavelet with balanced-pretreatment outperforms both GHM multi-wavelet which boasting of duplicate coefficient rows and CL multi-wavelet,and it suits fan＇s fault diagnosis most.

关 键 词：SA4 多小波 预处理 降噪 风机 

分 类 号：TH861[机械工程—仪器科学与技术]