Swedish district heating systems. October The stepwise regression model SR in this study is The architecture of the NN with one hidden layer based on the initial selection of regressors provided by of neurons with sigmoidal activation function and a ARX models and is constructed by iteratively adding linear output layer is shown in Fig. Other influential inputs include solar radiation and extracted features denoting population activities such as day of the week. Additional weather related represented by a direct linear link with a variables temperature, solar radiation, wind speed, nonlinear feature extractor hidden nonlinear relative humidity and extracted features days of neurons.
Top Selected Products and Reviews
Development of an urban enconman. In this paper, short-term forecasting of heat load in a district heating system of Ljubljana is presented. The results RW model and 3.