Electric Heating Load Forecasting Method Based on Improved Thermal Comfort Model and LSTM

The accuracy of the electric heating load forecast in a new has close relationship with safety and stability distribution network normal operation. It also enormous implications on architecture network. Firstly, thermal comfort model human body was established to analyze comfortable temperature main crowd under different temperatures levels humidity. Secondly, it analyzed influence factors load, from perspective meteorological factors, selected difference between actual humidity by gray correlation analysis. Finally, attention mechanism utilized promote precision combined adjunction model, then data results predicted were obtained. In verification, measured certain area eastern Inner Mongolia used. showed that after considering input vector most relative such as comfort, LSTM can realize accurate prediction load.