Adaptive production and maintenance optimization of a hybrid solar–wind system under nigerian operating conditions

This paper presents an integrated framework for production and maintenance optimization in a hybrid solar–wind energy system under Nigerian climatic conditions. The approach couples power generation forecasting with reliability-driven maintenance scheduling to minimize life-cycle cost while ensuring operational reliability. Power output is predicted using an Artificial Neural Network (ANN), while component degradation is modeled through the Weibull reliability function incorporating a virtual age formulation. A restoration impact factor (α) is introduced to represent imperfect maintenance, quantifying partial reliability recovery after each intervention. Remaining Useful Life (RUL) is predicted using an LSTM-based model trained on normalized operational covariates and updated online after each maintenance event. Results over a 60-month horizon identify the optimal strategy as five preventive and twenty-one predictive maintenance actions, achieving a total cost of €194,000 and maintaining system reliability above a threshold. Sensitivity analysis shows that increasing corrective maintenance cost shifts the optimal plan toward preventive dominance, confirming the model’s adaptive behavior. The proposed α-based framework provides a practical, data-driven tool for reliability-centered maintenance planning in hybrid renewable systems, offering improved cost efficiency and resilience in variable operating environments.