基于优化广义回归神经网络的变电站设备温度预测

doi:10.11930/j.issn.1004-9649.2016.07.054.06

摘要/Abstract

摘要： 变电站设备温度情况是反映设备是否正常运行的一个重要指标,对设备温度进行及时的预测能够保证电力系统高效、稳定运行。针对高压设备温度存在非线性、随机性等特点,应用广义回归神经网络（GRNN）对其进行预测。运用K-近邻思想和多轮投票机制确定最优平滑因子,对GRNN进行优化。结合某变电站的设备温度历史数据,以高压开关柜为例,划分训练样本和预测样本,建立了基于优化GRNN的高压开关柜温度预测模型。Matlab仿真实验表明：优化的GRNN预测结果优于BP神经网络的预测结果,在训练速度和预测精度上都有显著提高。

关键词: 变电站, 温度预测, 优化的GRNN算法, K-近邻, 平滑因子

Abstract: The temperature of substation equipments is an important indicator of device operation condition. Timely equipment temperature prediction guarantees power system operation efficiency and stability. Because of nonlinear and stochastic characteristics of equipment temperature, a generalized regression neural network(GRNN) model is applied to temperature prediction. The optimal soothing factor of GRNN is determined by K-nearest neighbor and multiple rounds of voting. Based on temperature history data of one substation equipment, taken a high voltage switch cabinet as example, a high voltage switch cabinet temperature forecast model is built based on optimized GRNN. The MATLAB simulation results show that proposed model generates better results compare to BP neural network with significant improvement on training speed and prediction accuracy.

Key words: substation, temperature forecast, optimized GRNN algorithm, K-nearest neighbor, smoothing factor

中图分类号:

TM771

孔雪卉, 张慧芬. 基于优化广义回归神经网络的变电站设备温度预测[J]. 中国电力, 2016, 49(7): 54-59.

KONG Xuehui, ZHANG Huifen. Temperature Forecast for Substation Equipment Based on Optimized General Regression Neural Network[J]. Electric Power, 2016, 49(7): 54-59.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.2016.07.054.06

https://www.electricpower.com.cn/CN/Y2016/V49/I7/54

参考文献

[1] 辛毅. 电力设备温度在线监测及预警系统的设计与实现[D]. 济南：山东大学,2013.
[2] 孟晓,胡炎,邰能灵,等. 变电站无线测温告警算法[J]. 电力系统自动化,2013,37（5）：81-86.
MENG Xiao, HU Yan, TAI Nengling, et al . Algorithm of wireless temperature measurement and alarm in substations [J]. Automation of Electric Power Systems, 2013, 37(5): 81-86.
[3] 刘琦,程春,吴健,等. 智能变电站温度监测主站系统的设计与实现[J]. 电力系统保护与控制,2013,41（4）：130-135.
LIU Qi, CHENG Chun, WU Jian, et al . Design and implementation of temperature monitoring master station system for intelligent substation [J]. Power System Protection and Control, 2013, 41(4): 130-135.
[4] 郎茂祥. 预测理论与方法[M]. 北京：清华大学出版社,2011.
[5] PRADHAN M K, RAMU T S. Prediction of hottest spot temperature(HST) in power and station transform[J]. IEEE Trans on Power Delivery, 2004, 18(4): 1275-1283.
[6] 管火志. 基于混沌时间序列的变电站无线测温装置预测预警研究[D]. 天津：天津大学,2008.
[7] 刘振亮. 基于BP神经网络的机电设备温度监测预警管理系统硏究[D]. 太原：太原理工大学,2012.
[8] 张旭东,李运泽.基于BP神经网络的纳卫星轨道温度预测[J].北京航空航天大学学报,2008,34（12）：1424-1427.
ZHANG Xudong, LI Yunze. Temperature prediction for nano satellite on orbit based on BP neural network[J]. Journal of Beijing University of Aeronautics and Astronautics, 2008, 34(12): 1424-1427.
[9] 李孟励. 基于 BP 神经网络的变压器绕组热点温度预测方法研究[D]. 重庆：重庆大学,2010.
[10] KUMAR P, NIGAM SP, KUMAR N. Vehicular traffic noise modeling using artificial neural network approach[J]. Transportation Research Part C, 2014, 40(1): 1561-1569.
[11] 潘睿,刘俊勇,倪雅琦,等. 基于负荷预测及广义回归神经网络的短路电流超短期预测[J]. 电力系统保护与控制,2010,38（18）：94-99.
PAN Rui, LIU Junyong, NI Yaqi, et al . Ultra-short term forecasting for short circuit current based on load forecasting and general regression neural network [J]. Power System Protection and Control, 2010, 38(18): 94-99.
[12] 陈伟根,奚红娟,苏小平,等. 广义回归神经网络在变压器绕组热点温度预测中的应用[J]. 高电压技术,2012,38（1）：16-21.
CHEN Weigen, XI Hongjuan, SU Xiaoping, et al . Application of generalized regression neural network to transformer winding hot spot temperature forecastting [J]. High Voltage Engineering, 2012, 38(1): 16-21.
[13] MIRZAEI M, JASNI J, HIZAM H, et al . Static voltage stability analysis using generalized regression neural network[C]//International Power Engineering and Optimization Conference.2013: 391-396.
[14] YANG Zhengxiang, XU Guimin, WANG Jinwen. Transport volume forecast based on GRNN network[C]//2010 2nd International Conference on Future Computer and Communication(ICFCC). Wuhan, China: IEEE, 2010: 629-632.
[15] 雷雨,赵丹宁. 广义回归神经网络在卫星钟差短期预报中的应用[J]. 宇航计测技术,2013,43（1）：39-44.
LEI Yu, ZHAO Danning. Application of generalized regression neural network in short-time prediction for satellite clock error [J]. Journal of Astronautic Metrology and Measurement, 2013, 43(1): 39-44.
[16] 肖微,韩钦,朱文滔,等. 基于广义回归神经网络的电缆接头温度预测[J]. 电气应用,2013,32（13）：34-37.
XIAO Wei, HAN Qin, ZHU Wentao, et al . Application of generalized regression neural network[J]. Electrotechnical Application, 2013, 32(13): 34-37.
[17] 伊良忠,章超,裴峥. 广义回归神经网络的改进及在交通预测中的应用[J]. 山东大学学报（工学版）,2013,43（1）：9-14.
YI Liangzhong, ZHANG Chao, PEI Zheng. A modified general regression neural network with its application in traffic prediction[J]. Journal of Shandong University (Engineering Science), 2013, 43(1): 9-14.
[18] 徐晓艳. 基于 K 近邻算法的中文文本分类研究[D]. 合肥：安徽大学,2012.
[19] 罗毅. 基于灰色理论与广义回归神经网络的客运量预测模型研究[D]. 成都：西南交通大学,2007.