基于正交优化时频原子分解算法的电能质量扰动匹配特征分析  

Analysis on Matching Features of Power Quality Disturbances Based on Orthogonal Optimization of Time-Frequency Atom Decomposition

作　　者：侯世英[1] 张文玉[1] 孙韬[1] 罗书豪[1] 

机构地区：[1]输配电装备及系统安全与新技术国家重点实验室(重庆大学),重庆市沙坪坝区400044

出　　处：《电网技术》2013年第3期647-652,共6页Power System Technology

基　　金：高等学校学科创新引智计划项目(B08036);重庆大学输配电装备及系统安全与新技术国家重点实验室自主研究项目(2007DA10512711204)的资助~~

摘　　要：针对电能质量扰动信号时频局部化信息量较广难以简洁、灵活提取有效细微特征以及匹配追踪收敛速度较差的问题,提出一种应用于电能质量扰动分解重构及扰动特征参量提取的匹配时频原子框架及其改进方法。在Gabor时频原子库离散优化基础上,通过匹配追踪算法对扰动信号进行自适应分解,同时对搜寻的最佳时频原子进行正交变换,减小冗余分量,设定迭代次数或残差阈值作为终止条件,从而获得一系列匹配扰动信号波形特征的正交时频原子及其参量化形式。仿真结果表明,该框架能有效分解提取电能质量扰动信号时频特征参量,相对基于匹配追踪的稀疏分解,改进算法单一扰动匹配特征重构信噪比高达55dB,多重扰动达35dB,均方误差数量级为0.001,匹配扰动特征精度及收敛性能进一步提高,满足电能质量分析要求。In allusion to the problems that power quality disturbance signals possess a lot of localized information and it is difficult to extract effective and fine features of the signals flexibly and sententiously as well as the convergence speed of matching pursuit is insufficient,an improved method using time-frequency atom framework is proposed to decompose and reconstruct power quality disturbances and extract feature parameters of disturbances.Based on the discrete optimization of redundant Gabor dictionary of time-frequency atoms,the disturbance signal is adaptively decomposed by matching pursuit algorithm,meanwhile the Schimidt orthogonalization is applied to the optimal time-frequency atoms being sought to reduce the redundant components and the iteration times or the residual error threshold are set as the termination condition to attain a series of orthogonal time-frequency atoms and their parametrization forms.Simulation results show that such a framework can decompose and extract time-frequency feature parameters of power quality disturbance signals effectively.Relative to the sparse decomposition based on matching pursuit,using the improved algorithm the signal to noise ratio（SNR） of reconstruction of matching feature of single disturbance can reach up to 55 dB,the SNR of multiple disturbances is higher than 35 dB,and the order of magnitude of mean square error is 0.001,thus the accuracy of matched disturbance feature and convergence performance are improved further to meet the demand of power quality analysis.

关 键 词：匹配追踪 特征提取 分解重构 时频原子分解算法 电能质量扰动 

分 类 号：TM93[电气工程—电力电子与电力传动]