电能质量扰动识别的不同时频分析方法研究  

Study on the Detection and Classification of Power Quality Disturbances Using Difference Time-frequency Methods

作　　者：张立国[1,2] 张淑清[1] 李莎莎[1] 乔永静[1] 张航飞[1] 李明星[1] 贺朋[1] 

机构地区：[1]燕山大学河北省测试计量技术及仪器重点实验室,河北秦皇岛066004 [2]河北省自动化研究所,河北石家庄050081

出　　处：《计量学报》2017年第3期345-350,共6页Acta Metrologica Sinica

基　　金：国家自然科学基金(61077071);河北省自然科学基金(F2015203413;F2015203392);河北省高层次人才资助项目(A2016002032)

摘　　要：分析了EEMD、LMD、ITD的算法、特点及分解不同扰动信号的实现步骤。经过实验模拟,对比分解所得效果图,得到适合各种电能质量扰动信号的最佳分解方法:对于电压暂升、电压暂降、电压中断幅值类扰动信号用EEMD方法分解效果最佳;脉冲暂态扰动EEMD和ITD均可,ITD方法更快、定位更准;暂态振荡信号用ITD方法效果较好;电压闪变扰动EEMD分解效果较好;谐波信号用ITD分解效果较好。EEMD,LMD,ITD are new time-frequency decomposition methods appear after EMD,which can apply to diagnose complex fault signal. The algorithm、characteristics and procedure of each methods are introduced firstly, then obtain the best suitable decomposition method to every power quality disturbance through the analysis of experiment effects. EEMD is suitable for power amplitude disturbances such as swells, sags, outages decomposition. EEMD and ITD methods are all suitable for impulsive transients but ITD is more quickly and more accurate positioning. ITD is suitable for oscillatory transients. The decomposition effect of voltage flicker using EEMD is better than others and ITD is suitable for the harmonic disturbance decomposition.

关 键 词：计量学 电能质量扰动信号 总体平均经验模式分解 局部均值分解 固有时间尺度分解 

分 类 号：TB971[一般工业技术—计量学]