中压配电网中，弧光接地现象是最为常见的过电压故障之一，也是危害性较大的故障类型。然而弧光接地故障的信号特征非常微弱，使用传统的故障辨识方法难以检测出该故障，而快速准确的辨识出弧光接地故障及其位置可为继电保护装置动作提供重要依据。有鉴于此，论文提出了基于特征融合的弧光接地故障辨识方法以及相应故障的定位方法，主要工作如下：

首先，在中低压配电网弧光接地故障信息中提取出了12种数据特征。在分析弧光接地故障机理的基础上，研究电弧运动过程，并以动态电弧模型为基础，搭建中低压配电网弧光接地故障仿真模型，模拟不同工况、不同接地条件下的弧光接地故障的发展过程，最终从三相电压、三相电流、零序电压、零序电流中成功提取出了均值、峰间值、标准差等多种可能包含弧光接地故障信息的数据特征。

其次，对所提故障特征进行融合，在此基础上，提出一种改进D-S证据理论的弧光接地故障辨识方法。结合加权欧式距离，在D-S证据理论的基本概率分配函数计算中增加证据相似度、证据差异度和证据可信度权重，解决了多个数据特征之间的相关性以及冲突性问题，降低可信度冲突对故障辨识结果的干扰，实现弧光接地故障的精确辨识。利用仿真数据和实验数据验证所提方法的可行性，并与改进前D-S证据理论以及深度学习算法进行了对比分析。

最后，利用故障信息中包含的零序电流峰值、峭度等数据特征，提出了基于最大平均差（Max Mean Discrepancy，MMD）的断层断面弧光接地故障定位方法，解决了因存在位置盲区带来的定位误差问题。所提方法在动态时间弯曲方法（Dynamic Time Warping，DTW）的基础上，通过改变特征划分算法来消除位置盲区，准确区分故障区段与正常区段，实现弧光接地故障的精确定位研究。利用IEEE34算例，模拟不同中性点接地方式、故障初始相角、过渡电阻和故障位置等工况对论文所提定位方法的准确性进行了验证。为了验证方法的鲁棒性，还对加入了测量噪声的情况进行了定位验证。

In the medium-voltage distribution network, arc-flash grounding phenomenon is one of the most common overvoltage faults, and it is also a more harmful type of fault. However, the signal characteristics of arc ground fault are very weak, and it is difficult to detect the fault by using traditional fault identification methods, and quickly and accurately identify arc ground fault and its location can provide an important basis for the operation of the relay protection device. In view of this, this paper proposes an arc-flash ground fault identification method based on feature fusion and a corresponding fault location method, the main work is as follows:

Firstly, 12 kinds of data features are extracted from arc grounding fault information of medium and low voltage distribution network. Based on the analysis of arc grounding fault mechanism, arc motion process is studied. Based on dynamic arc model, arc grounding fault simulation model of medium and low voltage distribution network is built to simulate the development process of arc grounding fault under different working conditions and different grounding conditions. Finally, from the three phase voltage, three phase current, zero sequence voltage, zero sequence current, the mean value, peak to peak value, standard difference and other data features that may contain arc grounding fault information are successfully extracted.

Secondly, the proposed fault characteristics are fused, and on this basis, an arc-flash ground fault identification method that improves the D-S evidence theory is proposed. Combined with the weighted Euclidean distance, the weight of evidence similarity, evidence difference and evidence credibility is added to the calculation of the basic probability distribution function of D-S evidence theory, which solves the correlation and conflict problems between multiple data features, reduces the interference of credibility conflict on fault identification results, and realizes the accurate identification of arc-flash ground faults. The feasibility of the proposed method is verified by simulation data and experimental data, and compared with the improved D-S evidence theory and deep learning algorithm.

Finally, using the data characteristics such as zero-sequence current peak and steepness contained in the fault information, a fault fault location method based on Max Mean Discrepancy (MMD) is proposed, which solves the positioning error problem caused by the existence of location blind zone. Based on the Dynamic Time Warfare (DTW) method, the proposed method eliminates the position blind zone by changing the feature division algorithm, accurately distinguishes the fault section from the normal section, and realizes the accurate location research of arc-flash ground fault. Using IEEE34 examples, the accuracy of the proposed positioning method is verified by simulating different neutral point grounding methods, fault initial phase angle, transition resistance and fault location. In order to verify the robustness of the method, the positioning verification was also performed when measurement noise was added.

[1] 张婷.一流配电网让城市更智慧——国家电网公司世界一流城市配电网建设成效调查[J].国家电网,2020(12):64-67.

[2] 国家发展改革委.关于加快配电网建设改造的指导意见[EB/OL].[2015-08-20].

[3] 国家能源局.2021年电力安全监管重点任务[EB/OL].[2021-01-30].

[4] 陈雨婷,赵毅,吴俊达,孙文瑶,夏世威.考虑碳排放指标的配电网经济调度方法[J/OL].上海交通大学学报,2023,1:1-10.

[5] 谢李为,李勇，罗隆福等.基于距离矩阵与分支系数的配电网弧光接地故障定位方法[J].中国电机工程学报，2020，40（07）：2180-2191+2397.

[6] 要焕年,曹月梅.电力系统谐振接地[M].北京：中国电力出版社，2000.

[7] 李振兴，王朋飞，王新等.基于幅值特征和Hausdorff距离的配电网弧光接地故障定位方法[J].电力系统自动化，2020，44（07）：169-177.

[8] 刘科研,盛万兴,董伟杰.配电网弧光接地故障建模仿真与实验研究综述[J].高电压技术，2021，47（01）：12-22.

[9] 杨柳林，李德奎，陈延明等.基于RTDS的弧光接地故障自定义建模及仿真分析[J].电力系统保护与控制，2016，44（16）：137-142.

[10] Fan B,Zheng N,Zeng X,et al.A comparative investigation of flexible arc suppression devices in distribution network[J]. IET Power Electronics, 2022, 5:1-23.

[11] 郭凤仪,王鑫,刘艳丽.基于Mayr模型改进的适用于串联故障电弧的新模型[J].电工电能新技术,2015,34(02):67-71.

[12] 苏伟龙,许志红.高压直流继电器磁吹系统的建模与设计[J].电工技术学报,2022,37(06):1583-1594.

[13] 曹启纯,刘向军.高压直流继电器电弧运动仿真分析与实验研究[J].电工技术学报,2019,34(22):4699-4707.

[14] 王立军,贾申利,杨泽,史宗谦.纵磁作用下真空电弧的建模与仿真研究进展[J].高压电器,2017,53(03):17-25+34.

[15] 荣命哲,吴翊,杨飞,孙昊,纽春萍,段嘉炜.开关电弧电流零区非平衡态等离子体仿真研究现状[J].电工技术学报,2017,32(02):1-12+23.

[16] 朱冠南,杨晓艺,高宇航,钟昱铭,熊兰,周林.小电流直流电弧的动态模拟及其稳态特性研究[J].电测与仪表,2020,57(01):42-48.

[17] Li J,Yuan H,Xu M,et al.Research on Improved Schwarz Arc Model Considering Dynamic Arc Length[J]. Journal of Electrical Engineering & Technology, 2022: 1-12.

[18] 周虹屹,刘志刚,邓云川,杨云涵,高仕斌.针对电气化铁路电分段电弧的建模分析与抑制方法研究[J/OL].电网技术:1-11.

[19] Guo M F,Cai W Q, Zheng Z Y, et al. Fault phase selection method based on single-phase flexible arc suppression device for asymmetric distribution networks[J]. IEEE Transactions on Power Delivery, 2022, 37(6): 4548-4558.

[20] Li FW, Hao B, Ke Z, et al. Distribution Network Voltage Arc Suppression Method Based on Flexible Regulation of Neutral Point Potential of the New Grounding Transformer[J]. Frontiers in Energy Research, 2022, 10: 1-15.

[21] 马祖涛,李淑蓉.小电流接地故障电弧建模分析[J].电力系统及其自动化学报,2021,33(09):103-108.

[22] Jalil M,Samet H, Ghanbari T, et al. Development of Nottingham Arc Model for DC Series Arc Modeling in Photovoltaic Panels[J]. IEEE Transactions on Industrial Electronics, 2021, 69(12): 13647-13655.

[23] 杨帆,任伟,沈煜,雷杨,薛永端,徐丙垠.谐振接地配电网电弧接地故障暂态分析方法与辨识[J].电力系统及其自动化学报,2021,33(04):23-31.

[24] Cassie A.M. Theorie nouvelle des arcs de rupture et de la rigiditedes circuits[J].CICGRE,1939, 588-608.

[25] Sawicki A. About using of modified Cassie models to simulate impact of arc column length disturbances on the work of electrical devices[J]. Przeglad elektrotechniczny, 2012, 88(9A): 107-110.

[26] 刘宗杰,王成全,徐国强,李继强,王巨丰,王嬿蕾.对于Cassie电弧模型爆炸气体灭弧时间的研究[J].电子设计工程,2020,28(20):162-166.

[27] 曲娜,王建辉,刘金海,吴丰成.基于Cassie模型和L3/4范数的串联电弧故障检测方法[J].电网技术,2018,42(12):3992-3997.

[28] Mayr O.Beitrage zur Theorie des statischen und des dynamischen lichtbogens[J].Archiv fur Elektrotechnik,1943,37(12):588-608.

[29] 霍柳行,董爱华.基于Mayr模型的低压配电线路故障电弧研究[J].工矿自动化,2011,37(10):44-47.

[30] Christen T, Ranjan N. Generalized Mayr model for arcs in MV switches with splitter plates[J]. IEEE Transactions on Power Delivery, 2021, 37(1): 359-364.

[31] Liu Y J, Chang G W, Huang H M. Mayr's equation-based model for pantograph arc of high-speed railway traction system[J]. IEEE Transactions on Power Delivery, 2010, 25(3): 2025-2027.

[32] Wang J, Gleeson M, Smith W D, et al. The factors controlling along-arc and across-arc variations of primitive arc magma compositions: A global perspective[J]. Frontiers in Earth Science, 2022, 10:1-25.

[33] 晏坤,马尚,王伟,杜鹏,徐泽宝.基于小波分析和Cassie模型的低压串联电弧放电检测及故障保护仿真研究[J].电器与能效管理技术,2019,No.579(18):48-52+67.

[34] He J, Wang K, Li J. Application of an Improved mayr-type arc model in pyro-breakers utilized in superconducting fusion facilities[J]. Energies, 2021, 14(14): 4383-4394.

[35] 聂云良,赵宇,李瑞.基于改进Mayr模型与Comsol的串联故障电弧研究与火灾风险评估[J].电器与能效管理技术,2022,No.620(11):74-80.

[36] Habedank U.On the Mathematical Description of Arc Behaviour in the Vicinity of Current Zero[J].eztArchiv,1988,11(10):339-343.

[37] Smeets R P P，Kertsz V.Evaluation of high-voltage circuit breaker performance with a new validated arc model[J].IEE Proceedings-Generation，Transmission and Distribution，2000，147(2):121-125.

[38] 胡文广,王师奇.基于Mayr-Cassie的配电网络单相接地电弧建模及仿真研究[J].电工技术,2020,519(09):48-51.

[39] Jalil M, Samet H, Ghanbari T, et al. An enhanced Cassie–Mayr-based approach for DC series Arc modeling in PV systems[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-10.

[40] Yang F, Tang Z, Shen Y, et al. Parameter Determination Method of Cassie-Mayr Hybrid Arc Model Based on Magnetohydrodynamics Plasma Theory[J]. Frontiers in Energy Research, 2022, 10: 1-15.

[41] Hochrainer A.Eine Regelurgstechnische Betrachtng des Elektrischen Lichtbogens[J].ETZ-A，1971，92：367-371.

[42] Li S, Xue Y, Feng G, et al. Simulation analysis of intermittent arc grounding fault applying with improved cybernetic arc model[J]. The Journal of Engineering, 2019, 2019(16): 3196-3201.

[43] 郭威.配电线路故障类型辨识及故障选线定位方法的研究[D].华北电力大学(北京),2022.

[44] 郑聪,彭庆忠,周海峰,黄金满.基于人工智能专家系统的船舶电力系统故障诊断研究[J].广州航海学院学报,2021,29(03):5-8+24.

[45] 熊年昀,孟宪磊.基于专家系统的泵站在线故障诊断系统研究与开发[J].工业控制计算机,2022,35(09):37-39.

[46] Zhou Z, Ma Z, Jiang Y, et al. Fault Diagnosis Using Bond Graphs in an Expert System[J]. Energies, 2022, 15(15): 5703-5710.

[47] 王建军,郝素兰.液压机故障智能诊断专家系统研究[J].江苏科技信息,2022,39(15):47-49.

[48] 刘静,赵苗苗,刘瑶秋.基于专家系统的纯电动客车制动系统故障诊断[J].汽车电器,2021,397(09):29-31.

[49] Zhang Y, Chung C Y, Wen F, et al. An analytic model for fault diagnosis in power systems utilizing redundancy and temporal information of alarm messages[J]. IEEE Transactions on Power Systems, 2016, 31(6): 4877-4886.

[50] Jiang H, Dai X, Gao D W, et al. Spatial-temporal synchrophasor data characterization and analytics in smart grid fault detection, identification, and impact causal analysis[J]. IEEE Transactions on Smart Grid, 2016, 7(5): 2525-2536.

[51] 丁剑,白晓民,赵伟,方竹,李再华,仲悟之.基于复杂事件处理技术的电网故障信息分析及诊断方法[J].中国电机工程学报,2007,7(28):40-45.

[52] Liu Y,Chen Y,Paul P,et al.A review of power system fault diagnosis with spiking neural P systems[J]. Applied Sciences, 2021, 11(10): 4376-4395.

[53] Wu H, Ma X, Wen C. Multilevel fine fault diagnosis method for motors based on feature extraction of fractional fourier transform[J]. Sensors, 2022, 22(4): 1310-1321.

[54] Liu C, Zhuo F, Wang F. Fault diagnosis of commutation failure using wavelet transform and wavelet neural network in HVDC transmission system[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-8.

[55] Shafiullah M, AlShumayri K A, Alam M S. Machine learning tools for active distribution grid fault diagnosis[J]. Advances in Engineering Software, 2022, 173(1): 1-18.

[56] Huo D, Kang Y, Wang B, et al. Gear Fault Diagnosis Method Based on Multi-Sensor Information Fusion and VGG[J]. Entropy, 2022, 24(11): 1618-1625.

[57] Zhang Z, Xu X, Gong W, et al. Efficient federated convolutional neural network with information fusion for rolling bearing fault diagnosis[J]. Control Engineering Practice, 2021, 116: 1-13.

[58] 李腾飞,郝玉杰,袁方,孙锴,马晶,刘俊,彭鑫.基于多源特征信息融合的油浸式变压器故障智能诊断模型[J].电工电能新技术,2023,42(01):48-57.

[59] Thiviyanathan V A, Ker P J, Leong Y S, et al. Power transformer insulation system: A review on the reactions, fault detection, challenges and future prospects[J]. Alexandria Engineering Journal, 2022,61(10):7697-7713.

[60] 赵熙临,周建中,付波,刘辉.基于信息融合技术的电网故障诊断方法[J].华中科技大学学报(自然科学版),2009,37(03):98-101.

[61] Wu Z, Zhang Z, Wang W, et al. Fault Diagnosis Algorithm of Transformer and Circuit Breaker in Traction Power Supply System Based on IoT[J]. Energies, 2022, 15(23): 8812-8823.

[62] Fang J, Yan Y, Zhang H, et al. Research on fault section location method of distribution networks with arc suppression coil grounding based on energy leakage function[J]. IET Generation, Transmission & Distribution, 2020, 14(25): 6097-6106.

[63] Li Z, Qiao J, Wang Y, et al. A faulty section location method for distribution grid based on grounding fault transfer device[J]. Energy Reports, 2022, 81.

[64] 王楠,张利,杨以涵.10kV配电网单相接地故障交直流信号注入综合定位法[J].电网技术,2008,32(24):88-92.

[65] Sun J. Feature extraction of arc high impedance grounding fault of low‐voltage distribution lines based on Bayesian network optimisation algorithm[J]. IET Cyber‐Physical Systems: Theory & Applications, 2023, 23:1-12.

[66] 韩炜炜,黄鑫,李启鹏,杨海华.一起35kV不接地系统单相弧光接地事故分析[J].电工电气,2018(06):46-49.

[67] Borgwardt Karsten M, Gretton Arthur J. Integrating structured biological data by kernel maximum mean discrepancy. In: ISMB. 2006, p. 49–57.