부산대학교 도서관

This research presents a technique for SWH called Modified Shallow Swarm Optimisation with Deep Learning-Driven Energy Use Prediction (MSSODL-EUP). For energy use prediction, the MSSODL-EUP technique employs data normalisation and a Restricted Boltzmann Machine (RBM) model. The MSSO method, which combines quasi-oppositional based learning (QOBL) with the SSO algorithm, is used to optimise the hyperparameters of the RBM model. Comprehensive simulations show that the MSSODL-EUP technique outperforms competing algorithms in terms of prediction performance. The performance of the MSSODL-EUP technique is compared to that of existing algorithms in the comparative study. For each approach, the mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) values are computed. MSSODL-EUP has MSE, RMSE, and MAE values of 0.0066, 0.0811, and 0.0616, respectively. These findings illustrate the MSSODL-EUP technique's superiority in properly predicting energy utilisation in solar water heating systems.