Novel hybrid computational intelligence approaches for predicting daily solar radiation

Accurate estimation of solar radiation is crucial for harnessing this abundant natural resource effectively. Measuring solar radiation directly requires ground station networks, which are either unavailable or very limited in many regions of the world, including Vietnam, particularly in remote areas due to resource constraints. Therefore, this study was carried out with the objective to develop hybrid artificial intelligence (AI) models to predict solar radiations correctly using other meteorological data such as wind speed, relative humidity, maximum and minimum temperature and rainfall which can be measured at site easily. In this study, we have proposed three novel hybrid AI models, namely ANFIS-GA, ANFIS-BBO and ANFIS-SA, which combine the adaptive neuro-fuzzy inference system (ANFIS) technique with genetic algorithm (GA), biogeography base optimization (BBO) and simulated annealing (SA), respectively, for predicting daily solar radiation in Hoa Binh province, Vietnam. The performance of these hybrid models was evaluated using statistical indicators, including correlation coefficient ( R ), root-mean-squared error (RMSE) and mean absolute error (MAE). The results demonstrate that all three optimized models outperform the single ANFIS model. Among them, the ANFIS-BBO model exhibits the highest predictive capability (RMSE = 3.141 MJ/m 2 , MAE = 2.439 MJ/m 2 , R = 0.874). Sensitivity analysis reveals that maximum temperature is the most influential factor for predicting daily solar radiation. The findings of this study have significant implications for predicting solar radiation using AI methods, particularly ANFIS-BBO, with minimal meteorological data in remote locations not only in Vietnam but also globally.