| Summary: | The purpose of this project is to determine the accurate energy consumption for heating and cooling of a building by using Artificial Neural Network (ANN) and to implement ANN for energy efficiency prediction. In this project, the data that has been obtained from Kaggle website was trained in ANN model using Levenberg-Marquardt (LM) method. The input variables are relative compactness, roof area, overall height, surface area, glazing area, wall area, glazing area distribution of a building, and orientation while output variables are the building's heating and ventilation demands. The training dataset consists of data from 768 residential buildings. The dataset was divided into 70 percents training samples, 15 percents validation samples, and 15 percents testing samples. The ANN model was able to predict heating and cooling loads with regression value (R) of 0.99174 and Mean Squared Error (MSE) of 1.6725 and able to show that surface area is the most effective factor in heating and cooling load. This process aids in the identification of prospective areas for enhancement, such as the modification of equipment schedules, the optimization of control settings, or the implementation of energy-conservation measures. © 2024 IEEE.
|
|---|