Robust Parameter Estimation for Photovoltaic Array Model under Partial Shading Condition

Abstract

Modeling the photovoltaic (PV) array can analyze the influence of temperature, irradiance, and other factors on the I‐V characteristic curve. The model can be used to replace the actual PV array to implement various PV experiments for reducing experimental costs and saving experimental time. Parameter estimation can make the parameters of the PV array model more accurate and make the outputs of model consistent with the outputs of the actual device. This work focuses on PV array under partial shading condition problems. The main contribution is to use different types of robust estimators for parameter estimation and present a comparative analysis of the traditional weighted least squares (WLS) estimator and eight robust estimators, including the Quasi‐weighted least squares (QWLS) estimator, Correntropy, etc. The performance of these estimators is analyzed in the simulation, where both random errors and gross errors are considered. The results showed that the robust estimators have better performance than the traditional WLS estimator especially when the measurement data contain gross errors. The robust estimators can be used in the parameter estimation for PV array model to estimate the accurate parameters. © 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.