The modeling of solar photovoltaic (PV) cells is complex due to their current-voltage characteristics being heavily influenced by factors like irradiance and temperature. There are many difficulties that can arise from traditional analytical and numerical techniques used to model them. As a result, they often experience premature convergence and inaccurate results when the weather changes and/or when partial shadows are on them. In recent years, new adaptive hybrid optimization frameworks have been developed to utilize evolutionary computations for better parameter estimation and extraction from PV cells. This  paper proposes an Adaptive Hybrid Evolutionary Algorithm (AHEA), which combines Differential Evolution (DE), the Whale Optimization Algorithm (WOA) and a mutation-adaptive Particle Swarm Optimization (PSO) in an efficient manner to provide the most accurate parameter estimation of solar cells. Furthermore, also provides scenario-based analytical models using a depth-augmented reinforcement learning framework that consider how the changes in the irradiance and temperature will occur before estimating the parameters. The proposed algorithm has been validated based on similar operating conditions to both single-diode and double-diode PV models. Simulation results indicate that the AHEA outperforms traditional algorithms such as Genetic Algorithms (GA), PSO and standalone WOA for root mean square error (RMSE) and convergence speed.In addition, the results from empirical studies with varied sunlight levels provide evidence supporting the utility of this combination approach for modeling of solar cells in real time.  This is an optimal model for intelligent adaptive maximum power point tracking (MPPT) system and smart grid integration, as well as for forecasting future usages of solar energy.

Keywords: Solar cell modelling, Hybrid evolutionary algorithm, Adaptive optimization, Differential evolution, Whale optimization, Photovoltaic systems, Dynamic environmental conditions