电工技术学报  2015, Vol. 30 Issue (12): 320-329    DOI:
高电压与绝缘技术 |
基于改进量子粒子群优化稀疏分解的局放信号去噪方法
王永强,谢军,律方成
华北电力大学河北省输变电设备安全防御重点实验室 保定 071001
PD Signal Denoising Method Based on Improved Quantum-Behaved Particle Swarm Optimization Sparse Decomposition
Wang Yongqiang,Xie Jun,Lü Fangcheng
Hebei Provincial Key Laboratory of Power Transmission Equipment Security Defense North China Electric Power University Baoding 071003 China
全文: PDF (756 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 噪声抑制是局放在线监测的关键环节之一.针对局放信号噪声抑制问题,提出一种基于改进量子粒子群优化稀疏分解的局放信号去噪方法.该方法基于信号的稀疏分解思想,构建了仅与局放信号时频特性相匹配的匹配局放信号过完备原子库;基于匹配追踪(MP)算法在该原子库中对染噪局放信号进行最佳匹配原子搜索,并通过改进量子粒子群算法加速搜索进程,同时以残差比阈值作为MP迭代终止条件;基于各次MP迭代搜索得到最佳匹配原子仅可对原始无噪局放信号分量进行稀疏表示,而难以对噪声分量进行表示的原理,实现局放信号稀疏分解去噪目的.运用本文介绍方法对局放仿真信号及实测信号进行了去噪处理,并与基于形态学-小波的局放去噪结果作对比.结果表明,本文介绍方法能有效对局放信号进行去噪处理,去噪结果准确性高且波形无畸变,较好保留局放信号原始特征.
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
关键词 改进量子粒子群稀疏分解匹配追踪局部放电信号去噪    
Abstract:Noise suppression is one of the key links for partial discharge(PD) online monitoring. Focusing on the noise suppression, a PD signal denoising method based on improved quantum-behaved particle swarm optimization sparse decomposition was given in this paper. More specifically, the principle of this method is signal sparse decomposition. The partial discharge signal matching overcomplete dictionary which only matches the features of PD signals was built. Based upon these, polluted PD signal was sparse decomposed by matching pursuit(MP) algorithm in this dictionary to search the best matching atoms. Meanwhile, the improved quantum-behaved particle swarm optimization(IQPSO) was presented to accelerate the searching process, and the residual ratio was chosen to be the terminating condition of the iteration as well. Since no noise PD signal component can be sparse represented by the best atoms while the noise cannot, the goal of denoising was finally achieved. The denoising method presented in this article is applied on the simulated and measuring signals, the results are critical compared with the effect of the PD denoising method based on the morphological wavelet, the results show that the denoising method of this paper is available to precisely suppress the noise interference of PD signal with high accuracy results; the distortion of waveform after denoising cannot be found and the features of PD signal were well kept.
Key wordsImproved quantum-behaved particle swarm optimization(IQPSO)    sparse decom- position    matching pursuit(MP)    partial discharge(PD)    denoising   
收稿日期: 2014-04-14      出版日期: 2015-09-14
PACS: TM835  
基金资助:国家高技术研究发展计划(863计划)(2012AA050802),国家电网资助项目(GY17201200047)和中央高校基金(13MS73)资助项目
作者简介: 王永强 男,1975年生,博士,副教授,研究方向电气设备在线监测与故障诊断.谢 军 男,1988年生,博士研究生,研究方向为电气设备在线监测与故障诊断.
引用本文:   
王永强,谢军,律方成. 基于改进量子粒子群优化稀疏分解的局放信号去噪方法[J]. 电工技术学报, 2015, 30(12): 320-329. Wang Yongqiang,Xie Jun,Lü Fangcheng. PD Signal Denoising Method Based on Improved Quantum-Behaved Particle Swarm Optimization Sparse Decomposition. Transactions of China Electrotechnical Society, 2015, 30(12): 320-329.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/Y2015/V30/I12/320