Carbon emission prediction of low-carbon regional office buildings based on improved RNN model

Accurate prediction of building carbon emissions is the foundation of building energy conservation and emission reduction. This study proposes four factors that have a significant impact on carbon emissions during the operation phase of office buildings: personnel characteristics, building characteristics, electrical equipment characteristics, and temperature characteristics. A predictive model for carbon emissions in office buildings is constructed using a recurrent neural network. The model is trained and tested using time-series data of electricity carbon emissions from three office buildings, and the experimental results validate the prediction error within a range of 4.6%. This indicates that the hidden patterns and fluctuation trends of electricity carbon emissions during the operation phase of these three office buildings have been successfully captured by the predictive model, and accurate prediction of building carbon emissions has been achieved within the error range, providing a method reference for carbon emission prediction of similar buildings.