Research on Short-Term Gas Load Forecasting Based on Support Vector Machine Model

The short-term gas load forecasting model is developed based on the SVM regression. Gas supply data from a North-China city are taken for model validation. The forecasting error is less than 5% when the through-year data is used in the SVM model training. The un-update model which fits the realistic situation gives only slight error level increase. The data preprocessing, including the grouping and normalization, is effective for increasing accuracy of the regression analysis. In the normalization process, wide data range may be useful to get more accurate forecasting results. The SVM forecasting model developed in this paper may be effective for practical using, especially for the un-updated model.

[1]  F. Girosi,et al.  Nonlinear prediction of chaotic time series using support vector machines , 1997, Neural Networks for Signal Processing VII. Proceedings of the 1997 IEEE Signal Processing Society Workshop.

[2]  Ming-Wei Chang,et al.  Load forecasting using support vector Machines: a study on EUNITE competition 2001 , 2004, IEEE Transactions on Power Systems.

[3]  Carlos E. Pedreira,et al.  Neural networks for short-term load forecasting: a review and evaluation , 2001 .

[4]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[5]  Peng Shi-ni.  Methods of City Gas Load Forecasting and Their Analysis , 2005 .

[6]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.