Parameter identification and generality analysis of photovoltaic module dual-diode model based on artificial hummingbird algorithm

The aim of this study is to propose a photovoltaic (PV) module simulation model with high accuracy under practical working conditions and strong applicability in the engineering field to meet various PV system simulation needs. Unlike previous model-building methods, this study combines the advantages of analytical and metaheuristic algorithms. First, the applicability of various metaheuristic algorithms is comprehensively compared and the seven parameters of the PV cell under standard test conditions are extracted using the double diode model, which verifies that the artificial hummingbird algorithm has higher accuracy than other algorithms. Then, the seven parameters under different conditions are corrected using the analytical method. In terms of the correction method, the ideal factor correction is added on the basis of previous methods to solve the deviation between simulated data and measured data in the non-linear section. Finally, the root mean squared error between the simulated current data and the measured current data of the proposed model under three different temperatures and irradiance is 0.0697, 0.0570 and 0.0289 A, respectively. This paper uses existing parameter extraction methods and the latest metaheuristic algorithms to solve the problem of the nonlinearity, multivariable and multimode of photovoltaic cell dual diode model parameters. Seventeen algorithms are compared against experimental data for predicting cell parameters. Graphical Abstract

Automated summary

The aim of this study is to propose a high-accuracy simulation model for photovoltaic modules by combining analytical and metaheuristic algorithms. The artificial hummingbird algorithm is found to have higher accuracy in extracting parameters compared to other algorithms. The proposed model corrects the parameters using the analytical method and adds an ideal factor correction to address nonlinear deviations. The model achieves a low root mean squared error between simulated and measured current data.