Energy simulation through design builder and temperature forecasting using multilayer perceptron and Gaussian regression algorithm

Building design under different climate change scenarios should increase their thermal energy performance and consecutively reduce their impact on the environment. To compare the performance of a building temperature analysis, thermal comfort, CO2 emission and energy consumption design characteristics to its traditional construction, a comparative simulation analysis was performed on a 3RC building. Compared to the thermal comfort scenario, the proposed 3R cement design features include the wall, roof and plastering. The change in temperature between winter and summer was less than 5 °C, and the relative humidity dropped with temperature, which nearly matched the data collected. To do this, Design Builder energy simulations were run utilizing meteorological data for the location. Simulations validate the advantages of energy and greenhouse gas execution, which resulted in a 12% reduction in annual energy consumption and a 5% reduction in CO2 emissions. To accomplish this, two machine learning techniques were evaluated in this study that could be applied to forecasting temperature in a building. To compare their accuracy in terms of the R-coefficient and root mean square error as well as their performance in terms of the Gaussian process and multilayer perceptron, the models execute the experiment using real data. The findings show that over all horizons, it has the highest average accuracy of 0.95%. © 2023, The Author(s), under exclusive licence to Springer Nature Switzerland AG.