Parameter Estimation of Photovoltaic Systems Using Particle Swarm Optimization

Abstract

In the last years, solar photovoltaic generation has become a prominent renewable energy resource, due to its low environmental impact, continuously reducing costs and flexibility to be implemented as distributed, centralized, or isolated generation. The large-scale integration of this generation into electrical energy systems poses challenges to power system performance, planning, design, and operation, demanding rigorous representation of photovoltaic generation units. Considering the accurate gray box and white box models for solar photovoltaic systems already existing in the literature, proper models of photovoltaic generation systems can be obtained through parameter estimation. This work proposes a data-driven approach to estimate the electrical parameters associated with the single-diode model of a solar panel. A Particle Swarm Optimization algorithm is employed to determine the model parameters that best fits to a given set of measured data. Tests with measurement data of a real solar panel illustrate the proposed methodology.