Mid-Long Term Load Forecasting Based on Combination Forecasting Model

Author(s): 
J. G. Zhou, W. Liu, & Q. Song

Affiliation(s): 
†School of Economics and Management, North China Electric Power University, Baoding, 071003, China, ‡State Grid Jining Power Supply Company, Jining, 272100, China

Cite this paper
J. G. Zhou, W. Liu,  Q. Song, “Mid-Long Term Load Forecasting Based on Combination Forecasting Model”, Journal of Mechanical Engineering Research and Developments, vol. 39, no. 2, pp. 262-269, 2016. DOI: 10.7508/jmerd.2016.02.001

ABSTRACT: The load forecasting of mid-long term power system has diversity, complexity and many uncertainties, which is a typical nonlinear system. In this paper, multiple linear regression analysis method will be used to select factors from all of the relevant factors, which can optimize the network structure, and reduce the input space dimensions of the combination forecasting. Then the combination forecasting model is used for forecasting, which is based on the RBF neural network and support vector machine. The empirical results show that the forecasting accuracy is higher after screening factors, and the forecasting accuracy of combination forecasting model is higher than the single forecasting models’ whether the factors are screened or not, which verifies the validity of the model.

Keywords : Power Load Forecasting; Multiple Linear Regression; Radial Basis Function Neural Network (RBFNN); Support Vector Machine (SVM), Combination Forecasting.

References
[1] H.Zhang, Y.J.Zhu, L.L.Fan, The long-term power load combination forecasting based on Markoff correction, J. East China Electr. Power, 41(1)(2013) 33-36.
[2] K.G.Chen, Z.W.Jin,W. Yan, Research and application of medium and long-term power load forecasting based on support vector machine, J. University Shanghai Sci. Technol. 30(2)(2008) 129-132.
[3] S.Zhang, R.Y.Zhang, D.W.Wang. Based on PCA-BPNN method for long-term power load forecasting, Control Engineering, 17 (6)(2010 ) 800-802.
[4] L.Li, X.R.Song, L.Q.Zhang. Study on long-term power load forecasting, Comput. Simul. 31(9)( 2014) 132-135.
[5] J.P.Zhu, J.Dai. The optimized influence factors of long-term power load forecasting, Comput. Simul. 25 (5)(2008) 226-229.
[6] D.S.Luo, X.Jin, G.Q.Sun, et al. Selection and combination method for medium and long term load forecasting model, Power Syst. Automation, 24 (4)(2012)50-156.
[7] S.Zhang, R.Y.Zhang, D.W.Wang. Based on the DPCA-BP neural network for long-term power load forecasting, J. Northeastern University (NATURAL SCIENCE EDITION), 31(4)(2010) 482-485.
[8] Y.Cui, C.Wang, X.L.Chen. Based on Grey Markov forecasting model of mid long term power load forecasting. Modern Power, 28 (3)(2011) 38-41.
[9] H.W.Li, W.J.Mao. Optimization of GM (1, 1) model and its application in medium term electric power load forecasting, Power Syst. Protection Control, 39 (13)(2011) 53-58.
[10] Y.G.Cheng,M.Lin, M.Y.Lin. The backbone of genetic algorithm and BP neural network in the city long-term power load forecast and analysis,Comput. Application, 1(2010)224-226.
[11] D.X.Niu, G.Cheng, B.Zhang, etc. Based on modified principal component analysis for radial basic function neural network in the medium and long term power load forecasting, J. North China Electr. Power University, 34 (3)(2007) 77-79.
[12] J.Ge, X.Y.Xiao, D.S.He. Prediction of medium and long-term power load combination forecasting based on support vector machine, Power Syst. Automation, 20 (1)(2008)84-88.