摘要
传统的单端行波测距方法无法有效识别特殊波头,由于受零模波速不稳定的影响,导致基于模量波速差的行波测距结果误差较大。本文对上述问题提出改进,分析得到随着传输距离的增加,零模检测波速单调递减,而两模量传输时差则单调递增,故考虑基于BP神经网络的训练学习对零模波速进行估算,而后再利用模量波速差进行故障测距,仿真表明,该方法对各种故障情况均适用且具有较高的精度。
The special wave head cannot be effectively identified for the conventional single ended traveling wave fault location and because the destabilizing effect of the zero mode phase velocity bring about big error of trave ling wave fault location result based on the modulus wave velocity difference. Aimed at these problems, this paper an alyzed the zero mode detection wave velocity monotonically decreasing with transmission distance increasing, mean while the transit time difference of the two modulus are monotonically increasing. Thus estimated zero mode phase velocity based on BP neural network, and then used modulus wave velocity difference for fault location. Simulation shows that the method is applicable and has high precision for various fault conditions.
出处
《山西焦煤科技》
2013年第10期4-8,共5页
Shanxi Coking Coal Science & Technology
关键词
煤矿电网
单相接地
故障测距
相模变换
Coal mine power grid
Single phase grounding
Fault location
Phase -model transformation
