节点文献

基于小波神经网络的小电流接地系统故障选线方法的研究

Research on the Approach of Faulty Line Detection in Small Current Grounding System Based on the Wavelet Neural Networks

【作者】 王念彬

【导师】 王彦文;

【作者基本信息】 中国矿业大学(北京) , 电力电子与电力传动, 2010, 博士

【摘要】 我国中压配电网普遍采用小电流接地方式,包括中性点不接地、经消弧线圈和电阻接地,对于我国煤矿6kV~10kV供电系统而言,普遍采用中性点经消弧线圈接地方式。本文对小电流接地系统发生单相接地时的稳态和暂态特征进行了深入的研究并阐述了其故障特征规律,总结了小电流接地系统故障选线的难点及存在的问题;并对小波理论及相关理论进行了详细的介绍,给出了在暂态分量选线中小波基函数的选取原则和方法。论文提出了基于工频滤波和小波域空域相关滤波的方法并首次在小电流选线中得以应用;定义了五次谐波、基于小波包分解的相关分析、基于行波检测的故障选线方法的故障测度,提出了基于小故障角和小波神经网络的智能融合算法相结合的故障选线方法。即,首先判断故障角的大小,如果是小故障角,则利用小故障角的方法;反之按稳态分量和暂态分量法相结合的基于小波神经网络智能融合算法进行故障选线。通过仿真验证了该方法的可行性。

【Abstract】 Voltage distribution network in China commonly used in small current grounding, including the neutral point grounded by the coil and ground resistance.6kV or lOkV power supply system in China coal mine are generally neutral with Resonant Grounding. The paper of small current grounding system ground fault occurs when the steady state and transient characteristics of in-depth research, summed up the small current grounding fault line of the difficulties and problems; and its associated wavelet theory was detail, especially for transient component Selection How to choose the appropriate wavelet packets are given the principles and methods of selection. Thesis, based on frequency filtering and wavelet domain spatial correlation filtering method; defines the fifth harmonic, based on wavelet packet decomposition of correlation analysis, fault detection based on traveling wave fault line selection measure, proposed based on a small fault angle and the combination of wavelet neural network fault line selection. That is, first of all determine the size of angle fault, if it is a small fault angle, the angle of approach to using small fault; the contrary by the steady component and transient component of the Combination of fusion algorithm based on wavelet neural network for fault line. The simulation verifies the feasibility and effectiveness.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络