Performance enhancement of photovoltaic systems using hybrid LSTM-CNN solar forecasting integrated with P&O MPPT

Abstract

The increasing penetration of photovoltaic (PV) systems in smart grids highlights the need for reliable solutions to mitigate the inherent intermittency of solar energy. Short-term variability in solar irradiance remains a critical challenge for stable grid operation and efficient PV energy management. This paper proposes an integrated forecasting-control framework that combines short-term global horizontal irradiance (GHI) prediction with a conventional P&O MPPT strategy to enhance PV system performance. A hybrid LSTM-CNN architecture is developed to forecast one-step-ahead GHI under the semi-arid climatic conditions of Dakhla, Morocco, a region characterized by high solar potential and pronounced irradiance fluctuations. The forecasting model is validated using measured irradiance data from the National Renewable Energy Laboratory (NREL) via the National Solar Radiation Database (NSRDB). Predicted irradiance is then used to improve PV power estimation and support predictive maximum power point tracking (MPPT) operation. Simulation results obtained in MATLAB/Simulink demonstrate that the proposed framework achieves accurate GHI forecasting, faster MPPT convergence, reduced steady-state oscillations, and improved PV power stability under rapidly changing irradiance. The proposed approach provides a practical and computationally efficient solution for enhancing the dynamic response and energy extraction efficiency of PV systems in smart grid applications.