电能质量信号压缩采样稀疏基性能研究  

A Research on Performance of Compression Sampling Sparse Matrix of Power Quality Signal

作　　者：李雪梅[1] 苗桂君[2] 王学伟[2] 陆以彪 韩东 

机构地区：[1]北京电子科技职业学院,北京100016 [2]北京化工大学,北京100029 [3]黑龙江省电力科学研究院,哈尔滨150030

出　　处：《电测与仪表》2011年第9期14-17,共4页Electrical Measurement & Instrumentation

基　　金：黑龙江省电力有限公司重点科技项目(2010-222-3)

摘　　要：为解决电能质量信号传统采集压缩方法所面临的采样率高、采样资源浪费及硬件实现成本高的问题,本文研究并实现了对非稳态和稳态电能质量数据的压缩采样、采样数据空间稀疏基的选取和基于TV最小化共轭梯度法的信号重构。文中针对几类常见单一扰动和含有多重扰动的校准源实测信号进行算法的仿真分析和实验验证,并对比分析了傅立叶基和小波基对两种扰动信号的重构效果。仿真和实验表明,所提出的方法采样率比Nyquist采样率低73%,对于单一扰动信号,稀疏基采用傅里叶基时重构信噪比除暂态脉冲信号外均大于35dB;对于多重扰动信号,稀疏基采用小波基时重构信号信噪比大于22dB,满足电能质量分析的要求。To solve the problems of high sampling rate and sampling resource waste and high cost of hardware implementation faced by traditional collection and compression methods of power quality,the paper studies and implements acquisition of the measurements of unsteady and steady-state power quality,selection of sparse basis of the sample data space and TV-based conjugate gradient minimization method.In this paper,simulation analysis and experimental verification of the algorithms is performed for several kinds of single disturbances and measured signal of calibration source with multiple disturbances.Comparative analysis of reconstruction effects based on Fourier and wavelet-based is proceeded.Simulation and experiment show that the SNR of the reconstruction of the single disturbances using the Fourier-based is greater than 35dB except transient pulse and multiple disturbances using the wavelet-based is greater than 22dB when the sampling rate is lower than 73% of the Nyquist by the method proposed,which meets the requirements of power quality analysis.

关 键 词：电能质量 压缩传感 数据压缩 最小全变分 

分 类 号：TM933[电气工程—电力电子与电力传动] TP274[自动化与计算机技术—检测技术与自动化装置]