太阳能学报2024,Vol.45Issue(3) :326-336.DOI:10.19912/j.0254-0096.tynxb.2022-1745

基于改进差动电流函数的柔直配电线路保护原理

PROTECTION SCHEME FOR FLEXIBLE DC DISTRIBUTION LINES BASED ON IMPROVED DIFFERENTIAL CURRENT FUNCTION

戴志辉 李毅然 焦彦军 刘珊珊 史琛 周鑫泽
太阳能学报2024,Vol.45Issue(3) :326-336.DOI:10.19912/j.0254-0096.tynxb.2022-1745

基于改进差动电流函数的柔直配电线路保护原理

PROTECTION SCHEME FOR FLEXIBLE DC DISTRIBUTION LINES BASED ON IMPROVED DIFFERENTIAL CURRENT FUNCTION

戴志辉 1李毅然 1焦彦军 1刘珊珊 1史琛 1周鑫泽1
扫码查看

作者信息

  • 1. 新能源电力系统全国重点实验室华北电力大学,保定 071003
  • 折叠

摘要

针对纵差保护辨识高阻故障困难,且易受噪声、区外短路和异常数据等因素干扰的问题,提出基于改进差动电流函数的柔直配电线路保护原理.改进差动电流函数放大了区内外短路暂态特征差异,有利于高阻故障识别,同时也避免了区外故障对保护可靠性的影响;基于Same卷积的Hausdorff距离算法可有效抑制异常数据干扰,最终通过两者配合实现故障区域和故障类型辨识.PSCAD/EMTDC仿真验证表明,所提保护能有效识别直流线路短路,不受过渡电阻、噪声和异常数据等因素影响,适用于含多类型换流器和分布式可再生能源电源的柔直配电网.

Abstract

Aiming at the difficulty of differential protection in high-impedance-fault identification and easily disturbed by interferences,such as noise,external short-circuit faults and abnormal data,a protection scheme for flexible DC distribution lines based on improved differential current function is proposed.The improved differential current function enhances the difference of inner and external short-circuit transient characteristics,which is conducive to the identification of high-impedance fault,and avoiding the influence of external fault on the reliability.The improved Hausdorff distance algorithm based on the Same convolution can effectively suppress the interference of abnormal data.Finally,the fault zone and fault type are identified by the cooperation of improved differential current function and its convolutional Hausdorff distance.Simulation results by the PSCAD/EMTDC show that the proposed protection can effectively identify DC line short-circuit faults,and is not disturbed by interferences,such as the fault resistance,noise and abnormal data.The proposed protection is suitable for flexible DC distribution networks with multi-type converters and distributed renewable-energy sources.

关键词

直流配电系统/故障电流/故障检测/改进差动电流函数/卷积Hausdorff距离

Key words

DC distribution system/fault currents/fault detection/improved differential current function/convolutional Hausdorff distance

引用本文复制引用

基金项目

国家自然科学基金(51877084)

出版年

2024
太阳能学报
中国可再生能源学会

太阳能学报

CSTPCDCSCD北大核心
影响因子:0.392
ISSN:0254-0096
参考文献量27
段落导航相关论文