基于樸素貝葉斯算法的避雷器缺陷識別方法研究
李亞錦1,劉英男1,張婉瑩1,于大洋1,張國新1,蘇寧2
(1 山東大學 電氣工程學院,山東 濟南 250061;
2 海南電網(wǎng)有限責任公司瓊海供電局,海南 瓊海 571400)
摘 要:高溫、高濕、高鹽特殊環(huán)境下,加速了避雷器劣化或潛伏性缺陷的發(fā)展。僅依靠避雷器監(jiān)測指標判斷缺陷,難以識別特殊環(huán)境下避雷器的異常狀態(tài)。提出一種基于樸素貝葉斯算法的避雷器缺陷識別技術,提取特殊環(huán)境下影響避雷器運行狀態(tài)的關鍵特征量,通過樸素貝葉斯算法計算訓練樣本的先驗概率和測試樣本的后驗概率,從而識別避雷器缺陷類型。利用實際監(jiān)測和檢測數(shù)據(jù)進行分析,驗證了所提方法的可行性和正確性。
關鍵詞:樸素貝葉斯算法;特殊環(huán)境;帶電檢測;避雷器;缺陷識別
中圖分類號:TM862 文獻標識碼:A 文章編號:1007-3175(2022)01-0020-04
Research on Classification of Arrester Defect Diagnosis Based on Naive
Bayes Algorithm Inference
LI Ya-jin1, LIU Ying-nan1, ZHANG Wan-ying1, YU Da-yang1, ZHANG Guo-xin1, SU Ning2
(1 School of Electrical Engineering, Shandong University, Jinan 250061, China;
2 Qionghai Power Supply Bureau of Hainan Electric Power Co., Ltd, Qionghai 571400, China)
Abstract: The special environment of high temperature, high humidity, and high salt could accelerate the development of the deterioration or latent defects of the arrester.It is difficult to identify the abnormal state of the arrester in the special environment just relying on the monitoring index of the arrester.This paper proposed arrester defect diagnosis technology based on Naive Bayes, which extracted the key features that affect the operation state of the arrester in a special environment. It calculated the prior probability of training samples and the posterior probability of test samples through a Naive Bayes algorithm to identify the type of arrester defects. The feasibility and correctness of the proposed method are analyzed and it verified by the actual monitoring and detection data.
Key words: Naive Bayes algorithm; special environment; live detection; arrester; defect diagnosis
參考文獻
[1] 張弛,曾杰,呂旺燕,等. 海島微電網(wǎng)高溫、高濕、高鹽環(huán)境條件監(jiān)測及防護研究[J] . 環(huán)境技術,2019,37(3) :89-94.
[2] 史志強,鄧維,羅日成,等. 氧化鋅避雷器受潮與電氣參數(shù)的關系[J] . 高壓電器,2019,55(4) :233-238.
[3] 王宏偉,張利民,姜建平,等. 特高壓站避雷器泄漏電流在線監(jiān)測和分析系統(tǒng)[J] . 電瓷避雷器,2019(6) :67-72.
[4] 田世杰,張亮,林楚喬,等. 基于帶電檢測技術金屬氧化物避雷器故障發(fā)現(xiàn)與分析[J] . 電工技術,2019(1) :91-92.
[5] 劉英男, 李亞錦, 蘇寧, 等. 基于帶電檢測的變電設備差異化運維策略研究[J] . 電工電氣,2021(9) :34-37.
[6] 張搏宇,李光范,張翠霞,等. 污穢條件下避雷器的內(nèi)部溫升研究[J] . 高電壓技術,2011,37(8) :2065-2072.
[7] 李海峰,劉俊勇,徐斌,等. 消除外部環(huán)境因素干擾的 MOA 在線監(jiān)測參數(shù)修正方法[J]. 高電壓技術,2018,44(8) :2580-2586.
[8] 陳景亮,姚學玲. MOV 阻性電流諧波補償法的研究[J]. 高電壓技術,2007,33(3) :53-57.
[9] 彭紅霞,文艷,王磊,等. 基于兩層知識架構的電力設備差異化運維技術[J] . 高壓電器,2019,55(7) :221-226.
[10] 方紅燕,王蕊,楊文志,等. 貝葉斯公式實例的深度挖掘[J] . 曲阜師范大學學報(自然科學版),2018,44(4) :7-10.
[11] WILLIAMS C K I , BARBER D . Bayesian Classification with Gaussian Processes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(12) :1342-1351.