期刊文献+

基于矩阵算法和BP神经网络的智能站二次系统故障定位方法 被引量:5

Fault Location Method of Secondary System in Smart Substation Based on Matrix Algorithm and BP Neural Network
下载PDF
导出
摘要 针对智能变电站二次系统故障分析及定位面临数据规模庞大、故障机理复杂等问题,提出一种矩阵算法和BP神经网络相结合的二次系统故障定位方法。首先,从信息物理系统视角出发,对智能变电站二次系统的交互逻辑进行梳理抽象,以逻辑节点为最小单元建立二次系统连通状态矩阵和逻辑节点描述模型。然后,在获取二次系统拓扑关联和交互信息后,建立二次系统状态解析式,利用矩阵算法快速确定可疑故障范围。在此基础上,将基于矩阵算法得到的故障定位结果进行编码作为BP神经网络的输入数据,利用BP神经网络优良的拟合能力提高故障定位精度。最后以具体故障事件为例,阐述了所提方法的故障定位过程。算例结果表明,较之结合诸如循环神经网络和卷积神经网络等方案,所提方法具有更高的故障定位准确率。 To address issues such as large data scale and complex fault mechanism in secondary-system fault analysis and location of smart substations, a fault location method for secondary systems based on matrix algorithm and BP neural network is proposed. First, from the perspective of cyber-physical systems, we combed and abstracted the interaction logic of the secondary system of smart substation. On this basis, we took the logical node as the smallest unit to construct the logical node description model and communication state matrix of substation secondary systems. Then, analytical formula of the secondary system state was established and the matrix algorithm was used to quickly determine the suspicious fault zones after obtaining the secondary system topology association and interaction information. We encoded the fault location results obtained from matrix algorithm, made them as the input data of the BP neural network, and improved the fault location accuracy with the excellent fitting ability of BP neural network. Finally, a secondary system failure event was studied to demonstrate the analysis process of the proposed approach. The results show that compared with schemes based on combined recurrent neural network and convolutional neural network, the proposed method has higher location accuracy.
作者 戴志辉 耿宏贤 韩健硕 李金铄 方伟 DAI Zhihui;GENG Hongxian;HAN Jianshuo;LI Jinshuo;FANG Wei(School of Electrical and Electronic Engineering,North China Electric Power University,Baoding 071003,China;State Grid Hebei Cangzhou Power Supply Company,Cangzhou 061000,China)
出处 《华北电力大学学报(自然科学版)》 CAS 北大核心 2022年第6期1-10,共10页 Journal of North China Electric Power University:Natural Science Edition
基金 国家自然科学基金资助项目(51877084) 国家重点研发计划专项课题(2016YFB0900203) 河北省自然科学基金项目(E2018502063)。
关键词 智能变电站 二次系统 故障定位 矩阵判据 神经网络 smart substation secondary system fault location matrix criterion neural network
  • 相关文献

参考文献20

二级参考文献322

共引文献547

同被引文献53

引证文献5

二级引证文献4

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部