作者：安天瑜1, 马煜2, 高阳3, 杨博文3, 魏家和3
作者单位：1. 国家电网东北电力调度控制中心，辽宁 沈阳 110179;
2. 国网辽宁沈阳供电公司调度控制中心，辽宁 沈阳 110052;
3. 沈阳工程学院，辽宁 沈阳 110015
关键词：故障检测;故障定位;融合深度特征;小波变换;深度神经网络
摘要：
为解决含分布式电源配电网发生故障时无法快速准确获取故障类型、故障相序和故障位置的问题，该文基于深度学习理论和小波变换的思想，提出一种基于融合深度特征的含分布式电源配电网智能故障检测新方法，达到对含分布式电源配电网故障实时识别并准确定位的效果。算法仿真验证结果表明，该方法在故障类型和故障相序识别的准确度方面有一定提高，且对故障位置定位的误差显著降低。
Intelligent fault detection of distribution network with distributed generation based on integration depth feature
AN Tianyu1, MA Yu2, GAO Yang3, YANG Bowen3, WEI Jiahe3
1. State Grid Northeast Power Dispatch Control Center, Shenyang 110179, China;
2. State Grid Shenyang Electric Power Supply Company Dispatch Control Center, Shenyang 110052, China;
3. Shenyang Institute of Engineering, Shenyang 110015, China
Abstract: In order to solve the problem that the fault type, fault phase sequence and fault location cannot be obtained quickly and accurately when the distributed power distribution network has faults, based on the deep learning theory and the idea of wavelet transform, this paper proposes a new intelligent fault detection method for the distributed power distribution network based on the integration of deep features, so as to achieve the effect of real-time identification and accurate location of the faults in the distributed power distribution network. The simulation results of the algorithm show that this method has a certain improvement in the accuracy of fault type and fault phase sequence identification, and the error of fault location is significantly reduced.
Keywords: fault detection;fault location;integration depth feature;wavelet transform;deep neural network
2023, 49(2):58-65 收稿日期: 2022-04-07;收到修改稿日期: 2022-05-19
基金项目: 国家重点研发计划08专项270项目资助（2020YFC0827003）
作者简介: 安天瑜(1976-),男,山东聊城市人,高级工程师,博士,研究方向为电力系统分析、电力调度自动化
参考文献
[1] 陈辉, 周羽生, 刘诗涵, 等. 基于改进和声算法的含DG配电网故障定位[J]. 电力科学与技术学报, 2018, 33(2): 123-128+134
[2] 梁梦可, 滕欢, 李雪松, 等. 考虑非平衡特性的辐射状配电网故障定位方法[J]. 电力科学与技术学报, 2020, 35(2): 61-68
[3] BENKERCHA R, MOULAHOUM S. Fault detection and diagnosis based on C4.5 decision tree algorithm for grid connected PV system[J]. Solar Energy, 2018, 173: 610-634
[4] OKUMUS H, NUROGLU F M. A random forest-based approach for fault location detection in distribution systems[J]. Electrical Engineering, 2021, 103(1): 257-264
[5] 彭成, 王松松, 贺婧, 等. 基于离散小波变换和随机森林的轴承故障诊断研究[J]. 计算机应用研究, 2021, 38(1): 101-105
[6] 李亮, 范瑾, 闫林, 等. 基于混合采样和支持向量机的变压器故障诊断[J]. 中国电力, 2021, 54(12): 150-155
[7] 李雅欣, 侯慧娟, 张立静, 等. 近邻成分分析和k近邻学习融合的变压器不平衡样本故障诊断[J]. 高电压技术, 2021, 47(2): 472-479
[8] 徐先峰, 徐晨杰, 张艳波, 等. 基于行波特征分类的有源配电网故障定位[J]. 重庆大学学报, 2022, 45(11): 59-68
[9] 徐瑞东, 常仲学, 宋国兵, 等. 注入探测信号的直流配电网接地故障识别方法[J]. 电网技术, 2021, 45(11): 4269-4277
[10] 曾祥君, 陈磊, 喻锟, 等. 基于配电网双端信息融合的单相断线故障实时监测方法[J]. 电力科学与技术学报, 2020, 35(3): 12-18
[11] 陶彩霞, 杜雪, 高锋阳, 等. 基于经验小波变换的混合输电线路单相接地故障测距[J]. 电力系统保护与控制, 2021, 49(10): 105-112
[12] JAMES J Q, HOU Y, LAM A Y S, et al. Intelligent fault detection scheme for microgrids with wavelet-based deep neural networks[J]. IEEE Transactions on Smart Grid, 2017, 10(2): 1694-1703
[13] 史加荣, 王丹, 尚凡华, 等. 随机梯度下降算法研究进展[J]. 自动化学报, 2021, 47(9): 2103-2119
[14] 杨观赐, 杨静, 李少波, 等. 基于Dopout与ADAM优化器的改进CNN算法[J]. 华中科技大学学报(自然科学版), 2018, 46(7): 122-127
