摘要
针对单一选线无法实现适应各种单相接地故障的情况,提出基于极限学习机的融合选线方法。首先在MATLAB中建立110/35 kV的小电流系统模型,用小波包变换算法提取特征分量。采用零序导纳法、自适应陷波法、改进锁相环的5次谐波法分析在不同接地系统、不同接地电阻等故障类型发生故障时的暂态特征,利用极限学习机方法进行融合选线,并与PNN融合选线方法进行对比。实验证明本算法可实现100%正确选线,且选线时间短。
In view of the situation that single line selection could not adapt to all kinds of single phase to ground fault,this paper proposes a fusion method based on limit learning machine. First,the small current system model of 110/35 kV is established in MATLAB,respectively,the wavelet packet transform algorithm is used to extract the zero sequence admittance method,adaptive notch filter,PLL five harmonic method,wavelet analysis in different grounding system,grounding resistance,different fault initial angle and different fault type fault distance etc. when a fault occurs in the steady state and transient characteristics,extreme learning machine method is used for fusion line selection. And the PNN fusion method is compared with the line selection method. Experiments show that the algorithm can achieve 100% correction line selection,and the line selection time is short.
作者
梁喆
杨铁梅
LIANG Zhe;YANG Tie-mei(Taiyuan Universi ty of Science and Technology,Taiyuan 030024, Chin)
出处
《太原科技大学学报》
2018年第3期195-202,共8页
Journal of Taiyuan University of Science and Technology
基金
国家青年基金(F030507)
国家自然基金(U1510112)
山西重点实验室开放课题(MEI201504)
关键词
小电流系统
单相接地故障
融合选线
极限学习机
small current system,single phase to ground fault,fusion line selection,extreme learning machine
