A spatial load forecasting method based on load regularity analysis

A new method, based on regularity analysis of cellular load, is presented to determine the reasonable maximum of the load density index method of spatial load forecasting (SLF). Through analysis of cellular historical load data, each cellular load is decomposed into regular and random components. The regular part is used to reveal the regularity of the cellular load in different cycles, using random components to describe and characterize the intrinsic error of the cellular load. Adverse effects that occur by eliminating random components to suppress the intrinsic error allow extraction of the maximum of the regular part as the Class I cellular reasonable load maximum. Using the reasonable maximum and the land use information of Class I cells, load density of I is obtained. The load density is used to determine the coordination coefficient of the classi cation load density. According to the constraint relationship of the historical load, we established the power supply area, land information, classi cation load density, relationship equation between Class I cellular load, and classi cation load density. Classi cation load density index is obtained by using the least square method and is used to predict its value in the target year. Then the load of Class II cell can be obtained, so that the spatial load forecasting of the urban power system is realized. The analysis of an example indicates that the method is effective in improving the prediction accuracy of spatial load forecasting.

PDF

___

[1] Willis HL, Northcote-Green JED. Spatial electric load forecasting: A tutorial review. Proc IEEE 1983; 71: 232-253.
[2] Willis HL, Engel MV, Buri MJ. Spatial load forecasting. Comput Appl Power IEEE 1995; 8: 40-43.
[3] Wu HC, Lu CN. A data mining approach for spatial modeling in small area load forecast. IEEE T Power Syst 2002; 17: 516-521.
[4] Carreno EM, Rocha RM, Padilha-Feltrin A. A cellular automaton approach to spatial electric load forecasting. IEEE T Power Syst 2011; 26: 532-540.
[5] Hwang KJ, Kim GW. Spatial load forecasting model for electrical distribution planning. In: 2004 Russian{Korean International Symposium on Science and Technology; 26 June{3 July 2004; Tomsk, Russia. New York, NY, USA: IEEE. pp. 237-241.
[6] Bai X, Chao Z, Gang MU. Review and prospect of the spatial load forecasting methods. Zhongguo Dianji Gongcheng Xuebao/Proc CSEE 2013; 33: 78-92 (article in Chinese).
[7] Tong S, Wen F, Liang C. A two-dimension wavelet threshold de-noising method for electric load data pre-processing. Dianli Xitong Zidonghua/Autom Elect Power Syst 2012; 36: 101-105 (article in Chinese).
[8] Kang C, Xia Q. Review of power system load forecasting and its development. Dianli Xitong Zidonghua/Autom Electr Power Syst 2004; 28: 1-11 (article in Chinese).
[9] Chen YH, Gang MU, Duan FL. Identi cation and management to anomalous data in short-term load forecasting. J NEIEP 2002; 22: 1-5 (article in Chinese).
[10] Ye LI, Kang C, Chen X. Joint adaptive training of parameters in integrated models and related single forecasting methods. Dianli Xitong Zidonghua/Autom Electr Power Syst 2010; 34: 36-40 (article in Chinese).
[11] Gang MU, Hou KY, Yang YH, Hui YJJiang KZ. Studies on load regularity evaluating method for load forecasting. Zhongguo Dianji Gongcheng Xuebao/Proc CSEE 2001; 21: 96-101 (article in Chinese).
[12] Karayiannis NB. An axiomatic approach to soft learning vector quantization and clustering. IEEE T Neur Netw 1999; 10: 1153-1165.
[13] Jiang P, Ma X. A hybrid forecasting approach applied in the electrical power system based on data preprocessing, optimization and arti cial intelligence algorithms. Appl Math Model 2016; 40: 10631-10649.
[14] Pijoan A, Kamara-Esteban O, Borges CE. Environment modelling for spatial load forecasting. In: ACM 2014 International Workshop on Agent Environments for Multi-Agent Systems; 6 May 2014; Paris, France. New York, NY, USA: IEEE. pp. 188-206.
[15] Melo JD, Carreno EM, Padilha-Feltrin A, Minussi CR. Grid-based simulation method for spatial electric load forecasting using power-law distribution with fractal exponent. Int T Electr Energy 2015; 26: 1339-1357.
[16] Melo JD, Carreno EM, Padilha-Feltrin A. Estimation of a preference map of new consumers for spatial load forecasting simulation methods using a spatial analysis of points. Int J Elec Power 2015; 67: 299-305.
[17] Xiao B, Ping LI. Load regularity analysis on spatial load forecasting of urban power system. Power Syst Technol 2009; 33: 113-119 (article in Chinese).
[18] Xiao B, Mu G, Li P, Yang YH, Yan H, Yang CJ. A time series mismatch corrective method for spatial load forecasting. Dianli Xitong Zidonghua/Autom Electr Power Syst 2010; 34: 50-54 (article in Chinese).
[19] Chat eld C. Guide of Time Series Analysis. Beijing, China: Yuhang Press, 1986.
[20] Xiao B, Yang X, Mu G, Song K. A load density index method based on historical data of cell load. Power Syst Technol 2014; 38: 1014-1019 (article in Chinese).
[21] Li GZ, Liu WY, Yun HZ, Gao YH. A new data preprocessing method for bus load forecasting. Power Syst Technol 2010; 34: 149-154 (article in Chinese).