抑制局部放電老化影響的XLPE電纜絕緣缺陷識別
符方達1，楊旭2，潘成3，姚雨杭3，江翼2，張靜2，王錄亮1
（1 海南電網有限責任公司電力科學研究院，海南 海口 570000；2 國網電力科學研究院武漢南瑞有限責任公司，湖北 武漢 430074；
3 武漢大學 電氣與自動化學院，湖北 武漢 430072）
摘 要：局部放電（PD）測量是檢測甚至識別交聯聚乙烯（XLPE）電纜絕緣缺陷的有效工具，設置了內半導電層破損、絕緣內部氣隙缺陷、絕緣表面劃痕缺陷和外半導電層爬電缺陷等四種絕緣缺陷，在直流條件下進行了各種缺陷的PD老化實驗，發現PD在不同老化階段表現出不同的特性，導致PD指紋參數隨著老化時間產生波動。為了提高識別效果，提出了基于BRNN 算法的缺陷識別模型，由局部放電特征將局放序列劃分為五個階段，分別提取每個階段下的指紋參數后再結合局部放電階段信息作為BRNN算法輸入。該方法將絕緣老化下局部放電的時序特性納入考慮，將缺陷識別效率由72.93%提升至93.71%。
關鍵詞：交聯聚乙烯（XLPE）電纜；絕緣缺陷；局部放電（PD）老化；指紋參數；BRNN模型
中圖分類號：TM726.4；TM855 文獻標識碼：A 文章編號：1007-3175(2020)04-0016-09
Identifying Insulation Defects of XLPE Cable with Suppressing the Influence of PD Aging
FU Fang-da1, YANG Xu2, PAN Cheng3, YAO Yu-hang3, JIANG Yi2, ZHANG Jing2, WANG Lu-liang1
（1 Hainan Electric Power Research Institute, Haikou 570000, China; 2 Wuhan Nari Electric Co., Ltd, Wuhan 430074, China;
3 School of Electrical Engineering and Automation, Wuhan University, Wuhan 430072, China）
Abstract: Partial discharge (PD) measurement is an effective tool for testing and even identifying the insulation defects of cross-linked polyethylene (XLPE) cables. Four kinds of defects were set, including internal semi-conductive layer damage, internal air gap of the insulation flaw, insulation surface scratch, and the outer semi-conductive layer creep.PD aging test with various defects were performed under DC conditions, and then found that PD showed different characteristics at different aging stages, which caused PD fingerprint parameters to fluctuate with aging time. In order to improve the recognition effect, a defect recognition model based on the BRNN algorithm is proposed. The partial discharge sequence is divided into five stages based on the partial discharge characteristics. The fingerprint parameters at each stage are extracted and combined with the partial discharge stage information as the BRNN algorithm input. This method takes into account the timing characteristics of partial discharge under insulation aging, and improves defect recognition efficiency from 72.93% to 93.71 %.
Key words: XLPE cable; insulation defects; PD aging; fingerprint parameters; BRNN model
參考文獻
[1] VAHEDY V.Polymer insulated high voltage cables[J]. IEEE Electrical Insulation Magazine，2006，22(3)：13-18.
[2] 何金良，黨斌，周垚，等. 擠壓型高壓直流電纜 研究進展及關鍵技術述評[J]. 高電壓技術，2015，41(5)：1417-1429.
[3] BAHDER G, KATZ C, EAGER G S, et al.Life expectancy of crosslinked polyethylene insulated cables rated 15 to 35 kV[J].IEEE Transactions Power Apparatus and Systems，1981，100(4)：1581-1590.
[4] MORSHUIS P H F, SMIT J J. Partial discharges at dc voltage：Their mechanism, detection and analysis[J].IEEE Transactions on Dielectrics and Electrical Insulation，2005，12(2)：328-340.
[5] CAVALLINI A, MONTANARI G C, TOZZI M, et al. Diagnostic of HVDC systems using partial discharges[J].IEEE Transactions on Dielectrics and Electrical Insulation，2011，18(1)：275-284.
[6] LI Junhao, HAN Xutao, LIU Zehui, et al.PD characteristics of oil-pressboard insulation under AC and DC mix voltage[J]. IEEE Transactions on Dielectrics and Electrical Insulation，2016，23(1)：444-450.
[7] FABIANI D, MONTANARI G C, DISSADO L A. Space charge accumulation due to ultra-fast charge packets in XLPE insulated cables：The effect of temperature and field[C]//IEEE 9th International Conference on Properties & Applications of Dielectric Materials，2009.
[8] 司文榮，李軍浩，袁鵬，等. 直流下局部放電序列信號檢測與特性分析[J]. 電工技術學報，2010，25(3)：164-171.
[9] SAHOO N C, SALAMA M M A, BARTNIKAS R.Trends in partial discharge pattern classification：A survey[J].IEEE Transactions on Dielectrics and Electrical Insulation，2005，12(2)：248-264.
[10] FROMM U.Interpretation of partial discharges at dc voltages[J].IEEE Transactions on Dielectrics and Electrical Insulation，1995，2(5)：761-770.
[11] GRAVES A.Hidden Markov Model Hybrids[M]. Toronto：Supervised Sequence Labelling with Recurrent Neural Networks，2012.
[12] SCHUSTER M, PALIWAL K K. Bidirectional recurrent neural networks[J]. IEEE Transactions on Signal Processing，1997，45(11)：2673-2681.
[13] KIM J Y, PARK D C, WOO D M.Application of Pruned Bilinear Recurrent Neural Network to Load Prediction[C]//The 8th ACS/IEEE International Conference on Computer Systems and Applications，2010.
[14] MOHANTY S, GHOSH S. Artificial neural networks modelling of breakdown voltage of solid insulating materials in the presence of void[J]. IET Science, Measurement & Technology，2010，4(5)：278-288.
