Utilization of ensemble random vector functional link network for freshwater prediction of active solar stills with nanoparticles

This study introduces an Ensemble Random Vector Functional Link Networks (EnsRVFL) to predict the yield of active solar stills with nanoparticles. The active solar still system composed of a conventional solar still (RSD) integrated with a suction fan and an external condensation system. The addition of two different nanoparticles of Cu2O and Al2O3 into the basin of the solar still was investigated. The yield of the active solar still (MSD) was enhanced by about 140% and 100% when using Cu2O and Al2O3 respectively compared with that of RSD. In addition, using the fan increased the efficiency of the active solar still with Cu2O and Al2O3 to be 36.02% and 32.82%, respectively. Besides, the thermal efficacy of the passive solar still was kept almost constant at around 20.5%. Then, the proposed EndRVFL method was used to anticipate the hourly yield of the solar stills. Real experimental data were used to train and test the developed model. To assess the accuracy of the proposed model, the predicted results by EnsRVFL as well as standalone RVFL were compared with the experimental ones using different statistical assessment criteria. The coefficient of determination (R-2) ranges between 0.942 and 0.978 for RVFL and 0.982 and 0.991 for EnsRVFL for all investigated cases, which reveals the outperformance of EnsRVFL over standalone RVFL.