广义局部特征尺度分解方法及其应用  

The method of generalized local characteristic-scale decomposition and its application

作　　者：吴占涛[1,2] 程军圣[1] 李宝庆[1] 郑近德[3] 

机构地区：[1]湖南大学汽车车身先进设计制造国家重点实验室,湖南长沙410082 [2]湖南大学现代工程训练中心,湖南长沙410082 [3]安徽工业大学机械工程学院,安徽马鞍山243032

出　　处：《振动工程学报》2016年第2期331-339,共9页Journal of Vibration Engineering

基　　金：国家自然科学基金资助项目(51375152;51575168);国家科技支撑计划资助项目(2015BAF32B03);湖南省科技计划资助项目(2014WK3005);智能型新能源汽车国家2011协同创新中心;湖南省绿色汽车2011协同创新中心资助项目

摘　　要：针对自适应时频分析方法局部特征尺度分解(Local Characteristic-scale Decomposition,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,提出了广义局部特征尺度分解(Generalized Local Characteristic-scale Decomposition,GLCD)方法。该方法在借鉴EMD和LCD等方法思路的基础上,采用典型的数值计算方法定义了5种不同的均值曲线,并定义了瞬时频率具有物理意义的广义内禀尺度分量(Generalized Intrinsic Scale Component,GISC)和GISC判据。GLCD从不同均值曲线筛分得到的每一阶分量中选取最优分量作为该阶最终的GISC,确保了最终结果为最优。研究了GLCD方法的原理及GISC判据,通过仿真信号将GLCD方法与LCD方法进行了对比,结果表明GLCD在提高分量精确性和正交性等方面具有一定的优越性,并将GLCD方法应用于转子不对中和碰摩故障的诊断,结果表明了方法的有效性。A non-stationary signal method Generalized Local Characteristic-scale Decomposition （GLCD） was proposed for improving the self-adaptive time-frequency analysis method Local Characteristic-scale Decomposition （LCD）, of which, the property of mean curve interpolation points was mainly decided by adjacent similar extremum points which can＇t properly reflect the overall trends of signal, thereby the decomposition precision was lowered. To improve the LCD method, the ideas of EMD and LCD were referenced, typical numerical methods were used to define five different mean curves, meanwhile Generalized Intrinsic Scale Component（GISC） with physical instantaneous frequency and the criterion of GISC were defined, To ensure the final result was optimal, GLCD selected the best component from different components obtained by sifting with different mean cures in each rank as the final GISC. The paper firstly studied the theory of GLCD and the criterion of GISC, then simulation experiments were used to compare GLCD with LCD. The results indicate that GLCD is more efficient in improving the veracity and orthogonality in components than LCD. Finally, the proposed method was applied to diagnose the rotor with misalignment and rub-impact fault successfully which indicated the effectiveness of GLCD.

关 键 词：故障诊断 局部特征尺度分解 广义局部特征尺度分解 广义内禀尺度分量 时频分析 

分 类 号：TH165.3[机械工程—机械制造及自动化] TH911.7