Parameter extraction for photovoltaic device performance characterization

Photovoltaic systems for solar energy harvesting have seen accelerated growth over the past few decades. Crystalline silicon based photovoltaic systems are one of the most widespread photovoltaic solar cell technologies in use today. Such increased penetration of the photovoltaic systems creates new opportunities but also raises challenges, for photovoltaic manufacturers that are tasked with making their product more reliable and of good quality. As a final step during the fabrication of photovoltaic solar cells or modules, the current-voltage characteristics corresponding to each finished product, are measured under standard testing conditions. Measurement of electrical characteristics of the finished product is important for qualification purposes. For quality assessment purposes, certain metrics associated with the current-voltage characteristics, such as open-circuit voltage, short-circuit current, maximum current, maximum voltage, etc., are used based on a chi-by-the-eye method in adjudicating as to whether a given photovoltaic device is of “good” quality. In doing so, only a few of the salient points in the current voltage characteristic are utilized, while detailed information, inherent in the current-voltage characteristic, is not considered. In this thesis, means whereby empirical models of photovoltaic solar cells or modules may be used to fit the experimentally measured current-voltage characteristics is proposed. Accordingly, model parameters that can be used to characterize photovoltaic solar cells or modules are extracted from their measured current-voltage characteristics. These model parameters include photo-generated current, reverse saturation current, diode ideality factor, series resistance, and shunt resistance. For the realization of this objective, one of the most commonly used models to predict the current-voltage characteristic of a photovoltaic solar cell or module is adopted. A curve fitting approach is proposed based on a non-linear optimization technique to extract model parameters associated with given current-voltage characteristics. In this thesis, computational solutions are developed for accurately extracting model parameters. The presented research spans from performance modeling, parameter estimation methods, to application of an optimization algorithm. Both experimental and standard data are used to investigate the accuracy of the extracted model parameter results obtained to validate the optimal performance of the proposed technique. The possible applications of this approach are discussed.Applied Science, Faculty ofEngineering, School of (Okanagan)Graduat