Regresar

PV panel modeling and identification

Abstract:

In this chapter, the modelling techniques of PV panels from I-V characteristics are discussed. At the beginning, a necessary review on the variousmethods are presented, where difficulties in mathematics, drawbacks in accuracy, and challenges in implementation are highlighted. Next, a novel approach based on linear system identification is demonstrated in detail. Other than the prevailing methods of using approximation (analytical methods), iterative searching (classical optimization), or soft computing (artificial intelligence), the proposed method regards the PV diode model as the equivalent output of a dynamic system, so the diode model parameters can be linked to the transfer function coefficients of the same dynamic system. In this way, the problem of solving PV model parameters is equivalently converted to system identification in control theory, which can be perfectly solved by a simple integral-based linear least square method. Graphical meanings of the proposed method are illustrated to help readers understand the underlying principles. As compared to other methods, the proposed one has the following benefits: 1) unique solution, 2) no iterative or global searching, 3) easy to implement (linear least square), 4) accuracy, 5) extendable to multi-diode models. The effectiveness of the proposed method has been verified by indoor and outdoor PV module testing results. In addition, possible applications of the proposed method are discussed like online PV monitoring and diagnostics, non-contact measurement of POA irradiance and cell temperature, fast model identification for satellite PV panels, and etc.