| 引用本文: | 汪 斌,王 年,蒋云志,程志友,王 继,鲍文霞.改进FastICA算法在谐波检测中的应用[J].电力自动化设备,2011,31(3): |
| WANG Bin,WANG Nian,JIANG Yunzhi,CHENG Zhiyou,WANG Ji,BAO Wenxia.Application of improved FastICA in harmonic detection[J].Electric Power Automation Equipment,2011,31(3): |
|
| 摘要: |
| 介绍了一种基于改进的快速独立分量分析（FastICA）算法，并将其引入到谐波检测中。该算法在FastICA算法的基础上对牛顿迭代法进行了改进，使其满足三阶收敛，同时依据负熵极大的独立性准则实现了谐波信号的盲分离。为了更好地逼近真实信号，对分离后的信号进行幅值修正，从而完成谐波的检测。仿真实验结果表明了该算法相对于FastICA算法的优点在于减少了迭代次数和加快了收敛速度，同时在谐波分离准确性方面也明显优于FastICA算法。 |
| 关键词: 电力系统 独立分量分析 谐波检测 FastICA 牛顿迭代法 |
| DOI： |
| 分类号: |
| 基金项目:安徽省科技攻关计划重大科技专项项目（080102020 34）；安徽省高校青年教师资助项目（2008jq1023）；安徽省教育厅自然科学研究重点项目（KJ2010 A007） |
|
| Application of improved FastICA in harmonic detection |
|
WANG Bin, WANG Nian, JIANG Yunzhi, CHENG Zhiyou, WANG Ji, BAO Wenxia
|
|
Power Quality Engineering Research Center，Ministry of Education，Anhui University，Hefei 230039，China
|
| Abstract: |
| The improved FastICA（Fast Independent Component Analysis） algorithm is introduced to harmonic detection，which ameliorates the Newton iteration method based on FastICA algorithm to satisfy the third-order convergence and applies the independence criterion of negentropy maximization to realize the blind separation of harmonic signals. The amplitude of separated signal is amended to better approximate the true signal for completing the harmonic detection. Simulative results show that，compared with FastICA algorithm，it has less iterative time，quicker convergence and much higher harmonic separation accuracy. |
| Key words: power system independent component analysis harmonic detection FastICA Newton iteration method |
