Accurate Parameter Estimation on Photovoltaic Modules Using Fisher Information Matrix and D-Optimal Design

Accurate estimation of the equivalent circuit parameters of photovoltaic modules enables comprehensive analysis of photovoltaic system’s electrical behaviour and facilitates the detection of faults. This paper proposes a rigorous method for optimizing the current-voltage curve measurements by minimizing the variance of the estimated parameters in the typical photovoltaic five-parameter equivalent circuit model. Leveraging the Fisher Information Matrix and the D-optimal experiment design technique, the proposed method aims to improve the accuracy of parameter estimation by optimizing the distribution of current or voltage sampling points, reducing the overall parameter uncertainty as quantified by the determinant of the inverse of Fisher Information Matrix. The study investigates both current and voltage-based sampling approaches, showing that current sampling is preferable under partial shading, while voltage sampling can be equivalently employed under uniform irradiance. The method is validated through both numerical simulations and experimental data from five real photovoltaic modules under both partial shading and uniform irradiance conditions. Experimental validation was conducted on five commercial photovoltaic modules of different technologies, including monocrystalline, polycrystalline, HIT, and TOPCon, confirming that the proposed optimization strategy consistently enhances estimation precision, with variance reductions up to two orders of magnitude