Interval forecasting for heating load using support vector regression and error correcting Markov chains

As previously heating load forecasting methods are mostly deterministic, that is, point forecasting. In this paper, a new integrated interval forecasting approach based on support vector regression (SVR) and error correcting Markov chains is proposed to predict hourly heating load. Firstly, the architecture of the forecasting approach is presented. Then the forecasting system is applied to heating load collected from a certain heating supply station. Finally the forecast results are presented, and the simulation results illustrate that the forecasting approach can meet the demands of optimization control and operation for energy-saving.