Application of Artificial Neural Network in predicting building's energy consumption
Rahim Zahedi, Alireza Aslani, Arash Gitifar, Omid Noudeh Farahani, Hossein Yousefi
Abstract
The energy consumption of a residential building is mostly attended due to energy use and efficiency. The rate of building's energy consumption in developed countries is about one third of the total amount of energy consumption and this amount of energy consumption in developing countries is about 40 percent of the total amount of energy consumption. Accordingly, forecasting the energy consumption of buildings has been raised as a challenge in recent decades. Modeling energy consumption in residential buildings has become possible with the advances made in computing and simulation, and one of these significant advances is the emergence of artificial intelligence in the development of statistical models. Studies have shown that the Artificial Neural Network method can be used to predict the nonlinear behavior of building's energy consumption. In this model, climatic variables are the input and building energy consumption is the output variable. The network was built in MATLAB software and trained with Levenberg-Marquardt algorithm. The results show that the neural network has a 93.7 percent ability to estimate the energy consumption of buildings.