This paper proposes an enhanced sine-cosine algorithm (ESCA) for parameter identification of photovoltaic (PV) cells. The ESCA improves the original sine-cosine algorithm (SCA) by introducing a population average position concept, a personal destination agent mutation mechanism, and a competitive selection mechanism. These modifications enhance the exploration ability, diversity maintenance, and search accuracy of the algorithm. The ESCA is evaluated using single-diode (SDM), double-diode (DDM), three-diode (TDM), and photovoltaic module (PVM) models, and compared with eight existing methods. Experimental results show that ESCA outperforms other methods in terms of diversity maintenance, efficiency, and stability. It achieves lower standard deviation values in parameter estimation for three PVM models, indicating higher accuracy. The ESCA is able to accurately extract parameters for six different PV models, demonstrating its effectiveness in parameter identification. The algorithm's performance is further validated through statistical analysis, showing that ESCA has better convergence accuracy and more reliable parameter evaluation results compared to other methods. The ESCA also has a reasonable computational time, making it a promising tool for parameter extraction in complex PV models. The study highlights the importance of considering environmental factors in PV modeling and suggests that future work should focus on improving the stability and accuracy of parameter identification in dynamic environments.This paper proposes an enhanced sine-cosine algorithm (ESCA) for parameter identification of photovoltaic (PV) cells. The ESCA improves the original sine-cosine algorithm (SCA) by introducing a population average position concept, a personal destination agent mutation mechanism, and a competitive selection mechanism. These modifications enhance the exploration ability, diversity maintenance, and search accuracy of the algorithm. The ESCA is evaluated using single-diode (SDM), double-diode (DDM), three-diode (TDM), and photovoltaic module (PVM) models, and compared with eight existing methods. Experimental results show that ESCA outperforms other methods in terms of diversity maintenance, efficiency, and stability. It achieves lower standard deviation values in parameter estimation for three PVM models, indicating higher accuracy. The ESCA is able to accurately extract parameters for six different PV models, demonstrating its effectiveness in parameter identification. The algorithm's performance is further validated through statistical analysis, showing that ESCA has better convergence accuracy and more reliable parameter evaluation results compared to other methods. The ESCA also has a reasonable computational time, making it a promising tool for parameter extraction in complex PV models. The study highlights the importance of considering environmental factors in PV modeling and suggests that future work should focus on improving the stability and accuracy of parameter identification in dynamic environments.