基于lagrange插值的局部特征尺度分解方法及其应用

《基于lagrange插值的局部特征尺度分解方法及其应用》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、基于Lagrange插值的局部特征尺度分解方法及其应用摘要：针对局部特征尺度分解(LocalCharacteristic-scaleDecomposition,LCD)方法中均值曲线插值点的属性主要由相邻两同类极值点的属性决定，不能很好地体现数据的整体变化趋势，从而可能引起分解精度降低，提出了基于Lagrange插值的局部特征尺度分解(LagrangeInterpolationbasedLocalCharacteristic-scaleDecomposition,LILCD)方法.该方法采用Lagrange插值取

2、代LCD中的线性插值，且均值曲线的插值点是由相邻的3个同类极值点构成的Lagrange插值多项式计算产生.引入了对称系数的概念，并给出了最优对称系数评价准则.研究了LILCD方法的原理及最优对称系数评价准则，通过仿真信号将LILCD方法与LCD方法进行了对比，结果表明LILCD在提高分量精确性和正交性方面具有一定的优越性.将LILCD方法应用于转子不对中故障的诊断，结果表明了方法的有效性.关键词：局部特征尺度分解；Lagrange插值；故障诊断；不对中故障；时频分析中图分类号：TH165；TH911.7文献标识码

3、：ATheMethodofLagrangeInterpolationBasedLocalCharacteristic-scaleDecompositionandItsApplicationWUZhantaol，CHENGJunshengl,LIBaoqingl,ZHENGJinde2(1.StateKeyLaboratoryofAdvancedDesignandManufacturingforVehicleBody,HunanUniversity,Changsha410082，China；2.SchoolofMe

4、chanicalEngineering,AnhuiUnivetsityofTechnology,Ma’anshan243032，China)Abstract：Anewnon-stationarysignalmethodLagrangeInterpolationbasedLocalCharacteristic-scaleDecomposition(LILCD)wasproposedtoimprovetheLCDmethod，inwhichthepropertyofmeancurveinterpolationpoin

5、tswasmainlydecidedbyadjacentsimilarextremumpointswhichcannotproperlyreflecttheoveralltrendsofsignal,andthedecompositionprecisionwaslowered.ToimprovetheLCDmethod，LagrangeinterpolationwasusedinLILCDtoreplacethelinearinterpolationinLCD,andthemeancurveinterpolati

6、onpointswerecomputedwiththeLagrangeinterpolationpolynomialgeneratedbythreeadjacentsimilarextremumpoints.Symmetriccoefficientwasintroduced，andtheoptimalevaluationcriteriaofsymmetriccoefficientweregiven.ThepaperfirstlystudiedthetheoryofLILCD，andthen,simulatione

7、xperimentsweredonetocompareLILCDwithLCD.TheresultshaveindicatedthatLILCDismoreefficientinimprovingtheveracityandorthogonalityincomponentsthanLCD.Finally，theproposedmethodwasappliedtodiagnosetherotorwithmisalignmentfault，whichindicatestheeffectivenessofLILCD.K

8、eywords：localcharacteristic-scaledecomposition(LCD)；Lagrangeinterpolation；faultdiagnosis；misalignmentfault；time-frequencyanalysis??机械设备发生故障时，其振动信号一般是非平稳、非线性信号，由于时频分析方法能同时提供非平稳信号在时域和频域的局部化