Execution of synthetic Bayesian model average for solar energy forecasting

Accurate photovoltaic (PV) forecasting is quite crucial in planning and the regular operation of power system. Stochastic habit along with high risks PV signal uncertainty a probabilistic model required to address numerical weather prediction (NWP) underdispersion. In this study, new synthetic process based on Bayesian averaging (BMA) Ensemble Learning developed. The proposed initiated by improved self-organizing map (ISOM) clustering K-fold cross-validation for training process. To provide desirable learning different input samples, three learners including long short-term memory (LSTM) network, general regression neural network (GRNN), non-linear auto-regressive eXogenous NN (NARXNN) are employed. BMA approach combined output obtain accurate outcomes. Different models precisely compared obtained results over real-world engineering test site, that is, Arta-Solar case study. analysis recorded validate performance superiority model.