Parameter estimation of photovoltaic single-diode model using an integral form and hybrid objective function

Based on an integral form of the I–V relation of photovoltaic (PV) single-diode model, a non-iterative linear method is proposed for parameter estimation from a single experimental I–V curve of PV modules/cells. After appending a parameter refinement process using deterministic optimization algorithms, best accuracy has been achieved in four commonly-used case studies. Meanwhile, the computational cost in objective function evaluation is reduced by an order of magnitude compared to other state-of-the-art methods. In some case studies, negative parameter estimates are obtained along with large uncertainties, which are attributed to defective or low-quality input data. The framework of uncertainty analysis corresponds to interval estimation in mathematical statistics, and is further validated through a statistical assessment of parameter values reported by 80 methods in literature. Moreover, a hybrid objective function is conceived to account for both current and power errors, resulting in improved prediction of the maximum-power point compared to the conventional current-error-based function. To facilitate the application of this work, a MATLAB app has been developed and is freely provided to the research community. • Propose a non-iterative method for PV single-diode model parameter estimation. • Obtain high-accuracy PV parameters at a substantially reduced computation cost. • Propose a hybrid objective function for better maximum power point prediction. • Develop a framework of uncertainty analysis for PV model parameters.