Improved MGO method

The MGO method is improved by utilizing several hybrid strategies to augment local search capabilities and better exploration of interesting

This study introduces a new predictive output current for the photovoltaic (PV) system using actual experimental data. There are three primary contributions proposed in this research: First, the MGO method is improved by utilizing several hybrid strategies to augment local search capabilities and better exploration of interesting areas inside the search space. Afterward, the architecture of the multilayer feedforward artificial neural networks is refined. The IMGO is implemented to select optimal hyperparameters of the model, including number of neurons in the hidden layers and learning rate using Bayesian regularization backpropagation algorithm. Finally, the proposed IMGOMFFNN is integrated with the polynomial regression model for enhancing the prediction accuracy of the PV system. The experimental findings indicated that the proposed IMGO method is extremely perfect for dealing with difficult problems with a high degree of precision, stability, and rapid convergence. Furthermore, the proposed hybrid IMGOPRFANNmodel offered a high degree with actual measured PV current using various statistical assessments and outperformed ant lion optimizer based on random forest (ALORF) model, two stages of ANN (ALO2ANN) model, long short-term memory (LSTM), gated recurrent unit (GRU), extreme learning machine (ELM), least square support vector machine (LSSVM), and convolutional neural network (CNN) model.