电力信号的压缩感知重构算法电测与仪表.DOCVIP

基于压缩感知的电力信号压缩与重构研究* 陈雷1,2，（1.，；2. 燕山大学 电气工程学院，） 摘要：对几种典型的压缩感知重构算法的实现方法和特性SAMP算法可以在稀疏度未知的情况下，使用更少的采样值精确重构原始信号。 关键词：； 中图分类号： 文献标识码： 文章编号：1001-1390（201）00-0000-00 on power signal compression and reconstruction based on compressed sensing Chen Lei1,2, Zheng Dezhong2, Wang Zhongdong1 (1. Northeast Petroleum University, Qinhuangdao 066004, Hebei, China. 2. Key Laboratory of Measurement Technology and Instrumentation, Yanshan University, Qinhuangdao 066004, Hebei, China) Abstract: A method for compressive sampling and reconstruction of power signal based on compressed sensing theory is proposed to solve the problem of acquisition and compression of the electrical signal, and thus avoiding the redundancy sampling. Firstly, the implementation steps and the characteristic of the compressed sensing reconstruction algorithm such as IRLS, OMP, SP and SAMP are analyzed. Secondly, one-dimensional sparse signal and the power signal containing harmonics and inter-harmonics components, which are sparse under the Fourier transform are reconstructed under the four algorithms. The simulation results show that the greedy algorithm is more suitable for the electrical signal reconstruction due to its fast speed and low computational complexity. Compared with other greedy algorithms, the SAMP algorithm can reconstruct the original electrical signal with fewer sampling in the situation of unknown sparsity. Keywords: compressed sensing, power signal, greedy reconstruction algorithm, sparsity 0　引 言 随着信息技术的高速发展和人们对电能信息需求量的增加，需要不间断地对电网电能信号进行采集，以便进行相关的参数检测和数据分析。 为了降低存储和传输成本，必须对电能数据进行压缩处理。为此，很多学者提出了针对电能信号的压缩和解压缩方法[1-5]，这些压缩方法都是在?Nyquist采样定理的基础上先进行高速采样，然后再压缩处理，在压缩过程中丢弃了大量已采集的非重要数据，这些数据之所以可以丢弃，是因为其并不包含原始数据的关键信息，是冗余的。因此这些方法虽然取得了较好的压缩效果，但前期的采样和存储过程却消耗了大量的采样和存储资源。 近几年提出的压缩感知(Compressed Sensing)理论指出，如果信号本身是稀疏的或可压缩的，或者在某种变换域下是稀疏的或可压缩的，则可以利用一种与变换域不相关的测量矩阵，将高维信号投影到低维空间上，实现压缩采样。同时，由于这种在低维空间上的投影包含了重构原始高维信号所需要的关键信息，因此可以用非线性优化算法从低维空间上以高概率精确恢复原始信号[6-7]。可见只要能找到电能信号的合适的稀疏表示空间，即可利用