欢迎来到天天文库
浏览记录
ID:52425461
大小:1.47 MB
页数:5页
时间:2020-03-27
《油液磨粒超声回波信号双树复小波自适应降噪最优分解层数的研究.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、2015年10月机床与液压Oct.2015第43卷第19期MACHINETOOL&HYDRAULICSVol.43No.19DOI10.3969/j.issn.1001-3881.2015.19.050油液磨粒超声回波信号双树复小波自适应降噪最优分解层数的研究李一宁2!张培林2!徐超2!杨玉栋2!张云强2!吕纯1(1.军械工程学院七系,河北石家庄050003$2.武汉军械士官学校四系,湖北武汉430075)摘要:磨粒超声回波信号受到各种因素的影响,从而存在噪声,分解层数的选取对降噪效果影响很大。为此,提出了-种基于粒子群优化算法(PS0)的最优分解层数选取方
2、法,将得到的最优分解层数代人双树复小波域,采用-种渐近半软阈值函数与-种自适应阈值选择方法相结合,对含噪磨粒回波信号进行双树复小波阈值降噪,选取信噪比(SNR)和均方根误差(RMSE)两个参数评价降噪结果。仿真与实验结果表明:通过粒子群优化算法选取的分解层数得到的信噪比最高,油液磨粒超声回波信号自适应降噪方法对磨粒超声回波信号具有显著的降噪效果,明显提高了信噪比,降低了均方根误差,还原了信号的波形特征,为后续的特征提取与智能识别打下了良好的基础。关键词:在线磨粒检测;分解层数;自适应降噪;双树复小波变换;粒子群优化算法中图分类号:TP391.42文献标志码:
3、A文章编号:1001-3881(2015)19-205-5ResearchofOptimalDecompositionLevelinAdaptiveDe-noisingUltrasonicEchoSignalofWearDebrisinOilBasedonDual-treeComplexWaveletTransformLIYining1,ZHANGPeilin1,XUChao1,YANGYudong2,ZHANGYunqiang1,LVChun1(1.Department7st,0rdnanceEngineeringCollege,ShijiazhuangH
4、ebei050003,China;2.DepartmentFour,Wuhan0rdnance0fficerSchool,WuhanHubei430075,China)Abstract:Theultrasonicechosignalofweardebrisisinfluencedbymanymatters,therebynoisescompositionlevelhasgreatinfluenceontheeffectofnoisede-noising.Therefore,akindofalgoritlimlevelbasedonparticleswarmo
5、ptimization(PS0)methodwasputforward.Theobtainedoptimalthedual-treecomplexwaveletfield,andamethodwasusedwhichcombinedtheasymptoticsemi-softthresholdselection.Dual-treecomplexwavelettransform(DTCWT)thresholdwasutilizedtode-noiseforthenoisyultrasonicechosignal.Accordingtotheparameters
6、ofsignal-to-noiseratio(SNR)androotmeansquareerror(RMSE),theresultofde-noisingwasevaluated.SimulatedandexperimentalresultsshowthattheSNRisthehighestwhichPS0,theadaptivede-noisingmethiodforultrasonicechosignalofweardebrisinoilhasobviouseffectonsignalde-noising,anditobviouslyimprovest
7、heSNRandreducestheRMSEaswellasrestorethewaveformfeaturesofthiesignal,whichlaysagoodfoundationforthefurtherfeatureextractionandintelligentrecognition.Keywords:0nlineweardebrisdetecting;Decompositionlevel;Adaptivede-noising;Dual-treecomplexwaveletticleswarmoptimizationn〇前言液磨粒在线检测技术越来
8、越受到人们的重视[3]。机械设备在运行过程中,各零部
此文档下载收益归作者所有