Analysis of Key Factors in Heat Demand Prediction with Neural Networks

The development of heat metering has promoted the development of statistic models for the prediction of heat demand, due to the large amount of available data, or big data. Weather data have been commonly used as input in such statistic models. In order to understand the impacts of direct solar radiance and wind speed on the model performance comprehensively, a model based on Elman neural networks (ENN) was adopted, of which the results can help heat producers to optimize their production and thus mitigate costs. Compared with the measured heat demand, the introduction of wind speed and direct solar radiation has opposite impacts on the performance of ENN and the inclusion of wind speed can improve the prediction accuracy of ENN. However, ENN cannot benefit from the introduction of both wind speed and direct solar radiation simultaneously.