...
首页> 外文期刊>Electric Power Components and Systems >A New Algorithm for Busbar Fault Zone Identification Using Relevance Vector Machine
【24h】

A New Algorithm for Busbar Fault Zone Identification Using Relevance Vector Machine

机译:基于相关向量机的母线故障区域识别新算法。

获取原文
获取原文并翻译 | 示例

摘要

This article presents relevance vector machine (RVM) based relaying scheme for busbar protection which correctly differentiates between internal faults on busbar and external faults. To validate the proposed scheme, numerous computer simulations have been carried out on an existing 220 kV Indian power generating station having different types of bays such as line, transformer, reactor & generator with generator transformer (GT). Various fault conditions (test data set of 23,760) have been simulating using the power system computer aided design (PSCAD)/ electromagnetic transient direct current (EMTDC) software package (Winnipeg, MB, Canada) by varying fault & system parameters. The proposed RVM based fault discrimination scheme is executed in MATLAB software (The Math-Works, Natick, Massachusetts, USA) by loading the simulation data. Comparative evaluation of the proposed RVM based scheme with the existing support vector machine (SVM) based scheme clearly indicates the superiority of the proposed scheme in terms of decision speed (faster than SVM based scheme) and classification accuracy (more than 99%). Moreover, it does not operate under different types of external faults and system disturbances. Subsequently, the proposed scheme remains stable during severe current transformer (CT) saturation condition.
机译:本文介绍了基于相关矢量机(RVM)的母线保护继电器方案,该方案可正确区分母线内部故障和外部故障。为了验证所提出的方案,已经对现有的220 kV印度发电站进行了许多计算机模拟,这些发电站具有不同类型的机架,例如线路,变压器,电抗器和带有发电机变压器的发电机(GT)。通过改变故障和系统参数,使用电力系统计算机辅助设计(PSCAD)/电磁暂态直流电(EMTDC)软件包(加拿大温尼伯,加拿大)已经模拟了各种故障条件(测试数据集为23,760)。通过加载仿真数据,在MATLAB软件(美国马萨诸塞州纳蒂克的Math-Works公司)中执行基于RVM的故障判别方案。与基于支持向量机(SVM)的现有方案相比,对基于RVM的方案的比较评估清楚地表明了该方案在决策速度(比基于SVM的方案更快)和分类准确性(超过99%)方面的优势。而且,它不能在不同类型的外部故障和系统干扰下运行。随后,提出的方案在严重的电流互感器(CT)饱和条件下保持稳定。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号