基于快速独立分量分析算法的气体绝缘开关设备局部放电混合信号分离与缺陷类型辨识

发布时间：2018-05-01 00:38

本文选题：气体绝缘开关设备 + 局部放电　；参考：《高电压技术》2014年03期

【摘要】：为解决气体绝缘开关设备(GIS)内多种绝缘缺陷并存时的混合局部放电(简称局放)的信号分离和缺陷类型辨识问题,提出了基于快速独立分量分析(FastICA)算法的GIS局放混合信号分离与模式识别方法。该方法具体包括:采用FastICA算法从局放混合信号中分离出单一局放信号,针对各单一局放信号,提出的模式识别策略是通过增添相反符号和幅值归一化的单一局放信号用于分类器训练或者使用对信号符号不敏感的特征(如分形特征等)作为分类特征。算例分析结果表明,提出的方法能够实现GIS多绝缘缺陷局放信号的有效分离,并能对分离出的单一局放信号进行准确缺陷类型辨识,且具有抗噪能力强、计算速度快、robust性好等优点。
[Abstract]:In order to solve the problem of signal separation and defect type identification of mixed partial discharge (abbreviated partial discharge) in a gas insulated switchgear (GIS), a hybrid signal separation and pattern recognition method based on fast independent component analysis (FastICA) algorithm for GIS placement is proposed. This method includes: using the FastICA algorithm from the Bureau. A single local discharge signal is separated from the mixed signal. The proposed pattern recognition strategy is used to train the classifier by adding the opposite sign and amplitude normalization to the classifier, or to use the characteristics such as fractal characteristics, such as fractal characteristics, as classification characteristics. The proposed method can effectively separate the signal of the GIS multi insulation defect, and can identify the exact defect type for the isolated partial discharge signal, and it has the advantages of strong anti noise ability, fast calculation speed and good robust.

【作者单位】：国网山东省电力公司电力科学研究院;国网技术学院;山东大学电气工程学院;重庆大学输配电装备及系统安全与新技术国家重点实验室;
【基金】：国家重点基础研究发展计划(973项目)(2009CB724506)~~
【分类号】：TM564

【参考文献】