Method of UHV Transformer Top Oil Temperature Forecasting Based on Decoupling Analysis and Euclidean Distance

被引：0

作者：

Tan F. ^{[1
]}

Zhu C. ^{[1
]}

Xu G. ^{[1
]}

Chen H. ^{[1
]}

He J. ^{[2
]}

机构：

[1] State Grid Jiangsu Electric Power Co., Ltd. Maintenance Branch Company, Nanjing

[2] Department of Electrical Engineering, Southeast University, Nanjing

来源：

Gaodianya Jishu/High Voltage Engineering | 2022年 / 48卷 / 01期

基金：

中国国家自然科学基金;

关键词：

Decoupling analysis; Direct path coefficient; Euclidean distance; Path analysis; Similar hour; Support vector machines; Top oil temperature; UHV transformer;

D O I：

10.13336/j.1003-6520.hve.20201525

中图分类号：

学科分类号：

摘要：

To solve the problems of low accuracy in forecasting ultra-high-voltage (UHV) transformer top oil temperature, a method of forecasting UHV transformer top oil temperature based on decoupling analysis and Euclidean distance is proposed, in which the path analysis is adopted to realize the decoupling analysis of various factors. Firstly, the direct path coefficients of 9 factors, including temperature, humidity, wind speed, air pressure, rainfall, light intensity, horizontal date, vertical time, and load, are calculated by horizontal analysis and vertical analysis, and the main factors affecting UHV transformer top oil temperature are selected. Then, the correlation of three factors, including weather, time, and load, are analyzed by using Euclidean distance based on weighted optimization of direct path coefficients, and the comprehensive correlation is obtained by using linear weighting principle. Finally, on the basis of fully researching the steps of selecting similar hour based on comprehensive correlation, the method of UHV transformer top oil temperature forecasting is realized by using the support vector machine and linear weighting method. The results show that the average error of the method for forecasting UHV main transformer top oil temperature is 2.19%, and forecasting UHV regulator transformer top oil temperature is 2.50%. The method has high forecasting accuracy, which verifies its feasibility and validity. © 2022, High Voltage Engineering Editorial Department of CEPRI. All right reserved.

引用

页码：298 / 306

页数：8

共 22 条

[1]

ZHANG Jiangong, WANG Yanzhao, CHEN Yuchao, Et al., Measurement for sound power of a UHV transformer based on a combined vibration and sound pressure method, High Voltage Engineering, 45, 6, pp. 1843-1850, (2019)

[2]

SUN Jiaan, LI Lin, Time-periodic FEM using parareal for transformer electromagnetic characteristics analysis with sub-synchronous components, Proceedings of the CSEE, 40, 13, pp. 4348-4357, (2020)

[3]

ZHOU Lijun, TANG Haolong, WANG Lujia, Et al., Modeling and analysis of transformer overload based on top oil temperature rise, High Voltage Engineering, 45, 8, pp. 2502-2508, (2019)

[4]

QI Bo, WANG Yiming, ZHANG Peng, Et al., Deep recurrent belief network model for trend prediction of transformer oil chromatography data, Power System Technology, 43, 6, pp. 1892-1899, (2019)

[5]

LI Kejun, XU Yanshun, WEI Bengang, Et al., Prediction model for top oil temperature of transformer based on hybrid kernel extreme learning machine trained and optimized by particle swarm optimization, High Voltage Engineering, 44, 8, pp. 2501-2508, (2018)

[6]

CUI Yu, HOU Huijuan, XU Mingkai, Et al., A prediction method for dissolved gas in power transformer oil based on dual-stage attention mechanism, Proceedings of the CSEE, 40, 1, pp. 338-347, (2020)

[7]

QI Xiaowu, LI Kejun, YU Xiaoyan, Et al., Transformer top oil temperature interval prediction based on kernel extreme learning machine and Bootstrap method, Proceedings of the CSEE, 37, 19, pp. 5821-5828, (2017)

[8]

LI Lin, XIE Yuqing, LIU Gang, Et al., Influencing factor analysis for disc-type winding temperature rise of off-immersed power transformer, Electric Power Automation Equipment, 36, 12, pp. 83-88, (2016)

[9]

CHEN Weigen, SU Xiaoping, CHEN Xi, Et al., Influence factor analysis and improvement of the thermal model for predicting transformer top oil temperature, High Voltage Engineering, 37, 6, pp. 1329-1335, (2011)

[10]

WANG Yongqiang, YUE Guoliang, HE Jie, Et al., Study on prediction of top oil temperature for power transformer based on Kalman filter algorithm, High Voltage Apparatus, 50, 8, pp. 74-79, (2014)

← 1 2 3 →