The Application of the Optimal GM (1,1) Model for Heating Load Forecasting

Heating load forecasting of municipal steam heating system is an important part of the smart city. According to the forecast results, heating companies can arrange production reasonably and provide heating uniform and stable to improve energy efficiency. For the features of the time series of heating monitor data, we present the optimal GM (1, 1) model based on the metabolic GM (1, 1) model. Results from theoretical analysis, simulations, and experiments show that the performance of the optimal GM (1, 1) model forecasting is effective with smaller error, robustness against outliers and able to meet forecast demand.