电能质量信号的KSVD-NRAMP归一化自适应稀疏重构算法  

KSVD-NRAMP normalized adaptive sparse reconstruction algorithm of power quality signals

作　　者：肖儿良[1,2] 冯杰 简献忠 王如志[3] XIAO Er-liang;FENG Jie;JIAN Xian-zhong;WANG Ru-zhi(School of Optical-Electrical & Computer Engineering, University of Shanghai for Science & Technology,Shanghai 200093, China;Shanghai Key Laboratory of Modern Optical System, Shanghai 200093, China;College of Materials Science and Engineering, Beijing University of Technology,Beijing 100124,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院 [2]上海市现代光学系统实验室,上海200093 [3]北京工业大学材料科学与工程学院,北京100124

出　　处：《软件导刊》2018年第12期87-91,共5页Software Guide

基　　金：国家自然科学基金项目(11774017)

摘　　要：针对稀疏重构算法在电能质量重构中存在实时性差、重构精度低的问题,提出一种基于特征向量归一化的K奇异值分解(KSVD-NRAMP)自适应稀疏重构算法。算法针对电能质量信号的非线性非稳态特征,采用迭代式匹配追踪得到信号稀疏特征矩阵,然后对矩阵进行归一化处理,量化特征向量,加快函数收敛速度。接着对得到的矩阵原子进行奇异值分解,改善迭代步长波动造成信号重构精度低的问题,最后构建信号的高斯随机矩阵并重构信号。当信号压缩率在50%～90%时,该算法重构信噪比其它重构算法的重构信噪比高出26dB～28dB。实验结果表明,该算法重构精度更高且计算时间短,为电能质量信号的研究提供了一种新思路。Due to the poor real-time performance and low reconstruction accuracy of sparse reconstruction algorithm in power quality reconstruction,the author proposed an adaptive sparse reconstruction algorithm based on the normalized K singular value decomposition of eigenvector(KSVD-NRAMP).Firstly aiming at the nonlinear and unstable features of power quality signals,algorithm can obtain the sparse feature matrix of signals by adopting the iterative matching pursuit.Then the algorithm adopted the normalization processing of sparse feature matrix,quantizing eigenvector,accelerating the convergence speed of the function.Afterwards,aiming at the obtained matrix atoms,the algorithm can realize the decomposition of the singular value and improve the problem of low signal reconstruction accuracy caused by the fluctuation of iterative step.At last,the algorithm can construct the Gaussian random signals matrix and reconstruct the sparse signals.When the signal compression ratio is between50%and90%,the reconstruction SNR of the proposed reconstruction algorithm is26dB^28dB higher than that of other reconstruction algorithms.Experimental results show that the proposed algorithm has higher reconstruction accuracy and shorter computation time,and provides a new idea for the study of power quality signals.

关 键 词：归一化 奇异值分解 压缩感知 电能质量信号重构 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]