Combination regression and neural network for short term load forecasting

The forecasting of gas demand has become one of the major research fields in gas engineering. Gas demand possesses dual property of increasement and seasonal fluctuation simultaneously, so it makes gas demand variation possess complex nonlinear character. Accurately forecast were essential part of an efficient gas system planning and operation. In this paper, a new forecasting model named regression combined neural network was put forward. We used regression to model gas demand trend, neural network was used for calculating predicted value and errors. Taking the advantages of regression analysis and artificial neural network, the model improves the forecasting accuracy of gas demand obviously. The results indicates that the model is effective and feasible for load forecasting.