English

应用科学学报 ›› 2012, Vol. 30 ›› Issue (4): 363-368.doi: 10.3969/j.issn.0255-8297.2012.04.006

• 电子技术 • 上一篇    下一篇

基于FastICA的电流传感器相位差测量方法

龚国良， 鲁华祥， 刘沛华， 陈天翔   

  1. 中国科学院半导体研究所神经网络实验室，北京100083
  • 收稿日期:2011-03-03 修回日期:2011-09-01 出版日期:2012-07-23 发布日期:2012-07-25
  • 基金资助:

    国家自然科学基金(No.60576033)；国家“863”高技术研究发展计划基金(No.2007AA04Z423, No.2006AA01Z106)；福建省自
    然科学基金(No.2008J04001)；厦门市科技项目基金(No.3502Z20083031)资助

Phase Difference Measurement for Current Sensor Based on FastICA

GONG Guo-liang, LU Hua-xiang, LIU Pei-hua, CHEN Tian-xiang   

  1. Laboratory of Artificial Neural Networks, Institute of Semiconductors,
    Chinese Academy of Sciences, Beijing 100083, China
  • Received:2011-03-03 Revised:2011-09-01 Online:2012-07-23 Published:2012-07-25
  • Supported by:

    龚国良，博士生，研究方向：电力系统绝缘在线监测，E-mail: gongmianjie@163.com；鲁华祥，研究员，博导，研究方向：智能控制及算法，E-mail: luhx@semi.ac.cn

PDF

1220

被引次数

1

摘要：

电流传感器的相位差易受环境的影响，为提高电力绝缘在线监测系统的可靠性和准确度，文中提出了一种在线监测电流传感器相位差的测量方法. 该方法引入独立分量分析（independent component analysis,ICA）对电流传感器的输出信号进行分离. 给出一种补偿观测信号与源分量数目的方法，建立ICA 的数学模型. 针对FastICA 算法每次分离结果误差不同的局限性，用一个关于混合矩阵的评价函数选取多次分离结果中相位测量误差较小的结果. 实验结果显示：对于信噪比为10.9 dB 的信号，评价函数能使相位测量误差小于0.06± 的样本接受率从51.4% 提高到81%.

关键词: 相位差, 在线监测, 介质损耗, 独立分量分析, 负熵

Abstract:

Phase difference of a current sensor is susceptible to environmental impact. In order to improve
reliability and accuracy of the on-line insulation monitoring system, a phase difference measurement method
for current sensor on-line monitoring is proposed. In the method, independent component analysis (ICA) is
used to extract the test signal from the output of current sensor. An ICA model is established, and a method
given to compensate the number of observed signals and its components. To solve the problem that phase
errors are different in different test runs, an evaluation index of the mixing matrix is proposed to find the most
accurate result. Experimental results indicate that for a signal with SNR=10.9 dB, the percentage of phase
errors below 0.06± is raised form 51.4% to 81% by using the evaluation index.

Key words: phase difference, on-line monitoring, dielectric loss, independent component analysis (ICA), negentropy

中图分类号: