Forecasting the Demand of Short-Term Electric Power Load with Large-Scale LP-SVR

Pablo Rivas-Perea, Juan Cota-Ruiz, David Garcia Chaparro, Abel Quezada Carreón, Francisco J. Enríquez Aguilera, Jose-Gerardo Rosiles

Department of Computer Science, School of Engineering and Computer Science, Baylor University, Waco, USA.
Department of Electrical and Computer Engineering, Autonomous University of Ciudad Juárez (UACJ), Ciudad Juárez, México.
Science Applica- tions International Corporation, El Paso, USA..

DOI: 10.4236/sgre.2013.46051                

Abstract

This research studies short-term electricity load prediction with a large-scalelinear programming support vector regression (LP-SVR) model. The LP-SVR is compared with other three non-linear regression models: Collobert’s SVR, Feed-Forward Neural Networks (FFNN), and Bagged Regression Trees (BRT). The four models are trained to predict hourly day-ahead loads given temperature predictions, holiday information and historical loads. The models are trained on-hourly data from the New England Power Pool (NEPOOL) region from 2004 to 2007 and tested on out-of-sample data from 2008. Experimental results indicate that the proposed LP-SVR method gives the smallest error when compared against the other approaches. The LP-SVR shows a mean absolute percent error of 1.58% while the FFNN approach has a 1.61%. Similarly, the FFNN method shows a 330 MWh (Megawatts-hour) mean absolute error, whereas the LP-SVR approach gives a 238 MWh mean absolute error. This is a significant difference in terms of the extra power that would need to be produced if FFNN was used. The proposed LP-SVR model can be utilized for predicting power loads to a very low error, and it is comparable to FFNN and over-performs other state of the art methods such as: Bagged Regression Trees, and Large-Scale SVRs.

Keywords

Power Load Prediction; Linear Programming Support Vector Regression; Neural Networks for Regression; Bagged Regression Trees

Conflicts of Interest

The authors declare no conflicts of interest.

References

[1]	J. Valenzuela and M. Mazumdar, “On the Computation of the Probability Distribution of the Spot Market Price in a Deregulated Electricity Market,” 22nd IEEE Power En gineering Society International Conference on Power In dustry Computer Applications, 2001. Innovative Comput ing for Power—Electric Energy Meets the Market, Syd ney, 20 May 2001-24 May 2001, pp. 268-271.
[2]	Y. Yuan, Y. Wu, G. Yang and W. Zheng, “Adaptive Hy Brid Model for Long Term Load Prediction in Computa tional Grid,” 8th IEEE International Symposium on Clus ter Computing and the Grid, CCGRID ’08, Lyon, 19-22 May 2008, pp. 340-347. doi:10.1109/CCGRID.2008.60
[3]	P. Pai and W. Hong, “Support Vector Machines with Si mulated Annealing Algorithms in Electricity Load Fore casting,” Energy Conversion and Management, Vol. 46, No. 17, 2005, pp. 2669-2688. doi:10.1016/j.enconman.2005.02.004
[4]	M. Mohandes, “Support Vector Machines for Short-Term Electrical Load Forecasting,” International Journal of Energy Research, Vol. 26, No. 4, 2002, pp. 335-345. doi:10.1002/er.787
[5]	A. Jain and B. Satish, “Clustering Based Short Term Load Forecasting Using Support Vector Machines,” PowerTech, IEEE Bucharest, Bucharest, 28 June-2 July 2009, pp. 1-8. doi:10.1109/PTC.2009.5282144
[6]	N. E. ISO, “Nepol Zonal Load Data,” 2011. http://www.iso-ne.com
[7]	L. Hoogerheide, J. Kaashoek and H. Dijk, “Neural Net Work Approximations to Posterior Densities: An Ana lytical Approach,” Econometric Institute Report, Erasmus School of Economics (ESE), Rotterdam, 2010.
[8]	A. Gallant and H. White, “There Exists a Neural Network That Does Not Make Avoidable Mistakes,” IEEE Inter national Conference on Neural Networks, San Diego, 24-27 July 1988, pp. 657-664. doi:10.1109/ICNN.1988.23903
[9]	C. Bishop, “Neural Networks for Pattern Recognition,” Oxford University Press, Oxford, 1995.
[10]	M. Canty, “Image Analysis, Classification and Change Detection in Remote Sensing: With Algorithms for ENVI/ IDL,” CRC Press, Boca Raton, 2007.
[11]	R. Berk, “Bagging,” Statistical Learning from a Regres sion Perspective, Springer, New York, 2008, pp. 1-24. doi:10.1007/978-0-387-77501-2_4
[12]	A. Cutler, D. Cutler and J. Stevens, “Tree-Based Meth ods,” High-Dimensional Data Analysis in Cancer Re search, Springer, New York, 2009, pp. 1-19.
[13]	R. Collobert and S. Bengio, “Svmtorch: Support Vector Machines for Large-Scale Regression Problems,” Journal of Machine Learning Research, Vol. 1, No. 2, 2001, pp. 143-160.
[14]	E. Osuna, R. Freund and F. Girosi, “An Improved Train ing Algorithm for Support Vector Machines,” Proceed ings of the IEEE Workshop, Neural Networks for Signal Processing, 24-26 September 1997, pp. 276-285. doi:10.1109/NNSP.1997.622408
[15]	T. Joachims, “Making Large Scalesvm Learning Practi cal,” Advances in Kernel Methods—Support Vector Learn ing, MIT Press, Cambridge, 1999.
[16]	P. R. Perea, “Algorithms for Training Large-Scale Linear Pro gramming Support Vector Regression and Classification,” Ph.D. Dissertation, The University of Texas, El Paso, 2011.
[17]	P. Rivas-Perea, J. Cota-Ruiz and J. G. Rosiles, “An Algo rithm for Training a Large Scale Support Vector Machine for Regression Based Onlinear Programming and De composition Methods,” Pattern Recognition Letters, Vol. 34, No. 4, 2012, pp. 439-451. doi:10.1016/j.patrec.2012.10.026