基于深度学习的柔性直流线路单端量波形特征保护

和敬涵; 张可欣; 李猛; 聂铭; 宋元伟

doi:10.16511/j.cnki.qhdxxb.2021.21.013

PDF(4211 KB)

清华大学学报（自然科学版） ›› 2021, Vol. 61 ›› Issue (5) : 478-486. DOI: 10.16511/j.cnki.qhdxxb.2021.21.013

保护

基于深度学习的柔性直流线路单端量波形特征保护

和敬涵, 张可欣, 李猛, 聂铭, 宋元伟

作者信息 +

Single terminal waveform characteristic protection of flexible DC lines based on deep learning

HE Jinghan, ZHANG Kexin, LI Meng, NIE Ming, SONG Yuanwei

Author information +

文章历史 +

摘要

多端柔性直流电网是未来智能电网的一种重要发展趋势，但对线路保护性能提出了更严苛的要求。针对现有柔性直流线路保护四性协调困难、阈值整定繁杂、耐受过渡电阻能力不足等问题，该文指出反行波波形特征蕴含丰富的故障位置信息，并以此为基础提出了基于深度学习的柔性直流线路单端量波形特征保护方案。该方案首先经极模变换、基于线路依频参数计算反行波；将归一化反行波作为输入量，通过所构建的深度学习模型自适应深入挖掘反行波波形特征，从而实现区内双极故障判别及区内单极故障选极。经仿真测试表明，所提方案可在2 ms内实现故障判别且可耐受200 Ω的过渡电阻。

Abstract

Multi-terminal flexible DC grids are a significant development trend for future smart grids, but they have significant line protection problems. Control systems must coordinate existing flexible DC line protection, complicated threshold settings and imperfect protection functions. This paper shows that the waveform characteristics of the reverse traveling wave contain abundant fault location information that can be used to develop a deep learning protection scheme. The reverse traveling wave is calculated based on a polar mode transformation and the line frequency parameters. Then, a normalized wave is used as the input to a deep learning model that adaptively mines the waveform characteristics. This then gives a bipolar fault diagnosis and single pole fault selection in the protected area. Simulations show that this scheme can provide a fault diagnosis within 2 ms and can withstand a 200 Ω transition resistance.

导出引用

和敬涵, 张可欣, 李猛, 聂铭, 宋元伟. 基于深度学习的柔性直流线路单端量波形特征保护[J]. 清华大学学报（自然科学版）. 2021, 61(5): 478-486 https://doi.org/10.16511/j.cnki.qhdxxb.2021.21.013

HE Jinghan, ZHANG Kexin, LI Meng, NIE Ming, SONG Yuanwei. Single terminal waveform characteristic protection of flexible DC lines based on deep learning[J]. Journal of Tsinghua University(Science and Technology). 2021, 61(5): 478-486 https://doi.org/10.16511/j.cnki.qhdxxb.2021.21.013

参考文献

[1] JUNG J J, CUI S H, LEE J H, et al.A new topology of multilevel VSC converter for a hybrid HVDC transmission system[J].IEEE Transactions on Power Electronics, 2017, 32(6):4199-4209.
[2] GEMMELL B, DORN J, RETZMANN D, et al.Prospects of multilevel VSC technologies for power transmission[C]//Proceedings of 2008 IEEE/PES Transmission and Distribution Conference and Exposition.Chicago, IL, USA:IEEE Press, 2008:1-16.
[3] DEBNATH S, QIN J C, BAHRANI B, et al.Operation, control, and applications of the modular multilevel converter:A review[J].IEEE Transactions on Power Electronics, 2015, 30(1):37-53.
[4] BERTHO R, LACERDA V A, MONARO R M, et al.Selective nonunit protection technique for multiterminal VSC-HVDC grids[J]. IEEE Transactions on Power Delivery, 2018, 33(5):2106-2114.
[5] 何佳伟, 李斌, 李晔, 等. 多端柔性直流电网快速方向纵联保护方案[J].中国电机工程学报, 2017, 37(23):6878-6887. HE J W, LI B, LI Y, et al.A fast directional pilot protection scheme for the MMC-based MTDC grid[J]. Proceedings of the CSEE, 2017, 37(23):6878-6887. (in Chinese)
[6] 薛士敏, 孙亚冰, 刘白冰, 等.基于Marti模型的柔性直流输电系统纵联行波差动保护[J].中国电机工程学报, 2019, 39(21):6288-6300.XUE S M, SUN Y B, LIU B B, et al.Longitudinal travelling wave differential protection for flexible HVDC system based on Marti model[J].Proceedings of the CSEE, 2019, 39(21):6288-6300.(in Chinese)
[7] 王帅, 毕天姝, 贾科.基于小波时间熵的MMC-HVDC架空线路单极接地故障检测方法[J].电网技术, 2016, 40(7):2179-2185. WANG S, BI T S, JIA K.Wavelet entropy based single pole grounding fault detection approach for MMC-HVDC overhead lines[J]. Power System Technology, 2016, 40(7):2179-2185. (in Chinese)
[8] 汤兰西, 董新洲, 施慎行, 等.柔性直流电网线路超高速行波保护原理与实现[J].电网技术, 2018, 42(10):3176-3186.TANG L X, DONG X Z, SHI S X, et al. Principle and implementation of ultra-high-speed travelling wave based protection for transmission line of flexible HVDC grid[J]. Power System Technology, 2018, 42(10):3176-3186. (in Chinese)
[9] 邓丰, 李欣然, 曾祥君, 等.基于波形唯一和时-频特征匹配的单端行波保护和故障定位方法[J].中国电机工程学报, 2018, 38(5):1475-1487.DENG F, LI X R, ZENG X J, et al. Research on single-end traveling wave based protection and fault location method based on waveform uniqueness and feature matching in the time and frequency domain[J]. Proceedings of the CSEE, 2018, 38(5):1475-1487. (in Chinese)
[10] LETERME W, BEERTEN J, VAN HERTEM D.Nonunit protection of HVDC grids with inductive DC cable termination[J].IEEE Transactions on Power Delivery, 2016, 31(2):820-828.
[11] RUI B J, LACERDA V A, MONARO R M, et al.Selective non-unit protection technique for multiterminal VSC-HVDC grids[J].IEEE Transactions on Power Delivery, 2018, 33(5):2106-2114.
[12] 高淑萍, 索南加乐, 宋国兵, 等.高压直流输电线路电流差动保护新原理[J].电力系统自动化, 2010, 34(17):45-49.GAO S P, SUO NAN J L, SONG G B, et al. A new current differential protection principle for HVDC transmission lines[J]. Automation of Electric Power Systems, 2010, 34(17):45-49. (in Chinese)
[13] TZELEPIS D, DY AŚG KO A, FUSIEK G, et al.Single-ended differential protection in MTDC networks using optical sensors[J].IEEE Transactions on Power Delivery, 2017, 32(3):1605-1615.
[14] BERTHO R, LACERDA V A, MONARO R M, et al.Selective nonunit protection technique for multiterminal VSC-HVDC grids[J]. IEEE Transactions on Power Delivery, 2018, 33(5):2106-2114.
[15] SANTOS R C, LE BLOND S, COURY D V, et al.A novel and comprehensive single terminal ANN based decision support for relaying of VSC based HVDC links[J]. Electric Power Systems Research, 2016, 141:333-343.
[16] YANG Q Q, LE BLOND S, AGGARWAL R, et al.New ANN method for multi-terminal HVDC protection relaying[J]. Electric Power Systems Research, 2017, 148:192-201.
[17] TZELEPIS D, DY AŚG KO A, FUSIEK G, et al. Advanced fault location in MTDC networks utilising optically multiplexed current measurements and machine learning approach[J].International Journal of Electrical Power & Energy Systems, 2018, 97:319-333.
[18] 陈先昌. 基于卷积神经网络的深度学习算法与应用研究[D]. 杭州:浙江工商大学,2014.CHEN X C. Research on algorithm and application of deep learning based on convolutional neural network[D]. Hangzhou:Zhejiang Gongshang University, 2014. (in Chinese)
[19] 杨赛昭, 向往, 张峻榤, 等. 基于人工神经网络的架空柔性直流电网故障检测方法[J]. 中国电机工程学报, 2019, 39(15):4416-4430. YANG S Z, XIANG W, ZHANG J J, et al. The artificial neural network based fault detection method for the overhead MMC based DC grid[J]. Proceedings of the CSEE, 2019, 39(15):4416-4430. (in Chinese)
[20] Aurelien G. Hands-on machine learning with Scikit-learn, Keras, and TensorFlow:Concepts, tools, and techniques to build intelligent systems[M]. 2nd ed. Sebastopol:O'Reilly Media,2019.