摘要
随着系统容量日趋增大、电压等级逐渐提高,直流系统故障电弧威胁日益提升,新型源荷设备的接入更是增加了系统干扰的复杂程度,传统保护技术无法满足直流故障电弧及时准确的检测需求。文中从电弧仿真、特征构建、算法设计、硬件实现4个方面总结了直流故障电弧检测技术的研究进展,讨论了不同电弧仿真模型的缺陷,分析了复合电弧模型的针对性改进;论述了基于物理、电磁、电气信号的特征构建方法,对比了不同时频检测特征优势;并从故障电弧特征融合的角度,基于树莓派平台比较了机器学习算法与阈值比较算法的检测应用效果;介绍了国内外直流故障电弧测试标准与测试平台,已研制的故障电弧检测样机在体积、成本方面更具优势。未来故障电弧检测设备可集成其他故障检测功能以综合提升直流系统安全保护效率,同时结合机器学习、5G网络等先进技术实现智能化检测应用。
With the gradual increase of the capacity and the voltage level,the threat of DC arc faults has undoubtedly increased.New source-load equipment has increased the complexity of system interference in DC systems.Traditional protection technologies cannot meet the requirements of timely and accurate detection of DC arc fault.In this paper,the research progress of the low-voltage DC arc fault detection technology has been summarized from the aspect of arc simulation,feature construction,algorithm design,hardware implementation.The defects of different arc fault simulation models are discussed and the composite arc fault model is proposed.This paper discusses the construction method of feature based on physical,electromagnetic and electrical signals and lists the advantages of different frequency detection features.From the prospects of the arc fault feature fusion method,the detection performances of machine learning algorithm and threshold comparison algorithm are compared based on the Raspberry pi platform.The DC arc fault test standards and platforms are introduced and the arc fault detection prototype has advantages in volume and cost.In the future,arc fault detection devices can integrate other fault detection functions to improve the efficiency of DC system security protection.Machine learning,5G network and other advanced technologies can be combined to achieve intelligent detection applications.
作者
汪倩
陈思磊
孟羽
杨淇
李兴文
WANG Qian;CHEN Silei;MENG Yu;YANG Qi;LI Xingwen(School of Science,Xi’an University of Technology,Xi’an 710048,China;School of Electrical Engineering,Xi’an University of Technology,Xi’an 710048,China;State Key Laboratory of Electrical Insulation and Power Equipment,Xi’an Jiaotong University,Xi’an 710049,China)
出处
《高压电器》
CAS
CSCD
北大核心
2023年第5期93-103,共11页
High Voltage Apparatus
基金
陕西省科技计划项目(2022TD-59)。
关键词
直流场景
故障电弧
模型仿真
检测特征
机器学习
DC system
arc fault
model simulation
detection variable
machine learning
