| Title |
A Study Energy Efficiency Prediction Model with AI-Based in Healthcare Building |
| Authors |
Yun Kyoung Yang ; Jin Chul Park |
| DOI |
https://doi.org/10.6110/KJACR.2022.34.7.336 |
| Keywords |
인공지능; 인공신경망; 에너지효율; 의료건물 Artificial intelligence; Artificial Neural Network; Energy efficiency; Healthcare Building |
| Abstract |
Because of global warming, interest in building energy reduction measures and eco-friendly energy use is increasing. Currently, interest in artificial intelligence is increasing as a way to efficiently save energy. The purpose of this paper was to study the building energy efficiency prediction model, by comparing the prediction accuracy of the artificial intelligence model, and using the optimized model. An optimized model was presented by comparing the predicted energy use with the actual energy use of the building, by analyzing the daily energy consumption value and the energy consumption effect, according to the facility field components using the target building. First, the correlation analysis between the data variables confirmed the highest correlation with energy consumption, which shows that the energy efficiency prediction model using the environmental variables of the data could be very limited because the building's cooling and heating consumption is proportional to the size of the building. As a result of model performance comparison and analysis, the ANN model showed the best performance with an average RMSE value of 153.2. It shows the difference between the predicted energy consumption and the actual energy consumption through the optimized hyperparameter ANN model, and the MLR analysis shows a very high prediction accuracy compared to RMES 197.0. |