Optimizing PV String Performance with a Minimal Data Sample Framework for Shade Detection

This article presents a new shade detection method to discriminate shade occurrences from uniform irradiated conditions. The method uses only two measurements of the system's characteristics, taken at constant current and constant voltage, to accurately detect the duty ratio of uniform irradiated conditions. Based on the data retrieved at predicted duty value, the discrimination between uniform irradiated conditions and occurrence of shade events is accurately identified. The algorithm also predicts duty value of right power peak (RPP) in the fourth sample to determine the occurrence of less complex with minimal shade involvement. For uniform irradiated and less complex shade events, the method predicts the Maximum Power Point (MPP) zone and uses a simple Perturb and Observe (P&O) algorithm to quickly converge to the optimal operating point. For complex shade events, a new twin-phase P&O method is proposed with two different step sizes to locate the Global Maximum Power Point (GMPP) zone. Based on the information on multiple MPPs, the GMPP algorithm then explores the wide search space around the GMPP zone to find the accurate power. The proposed method is evaluated using nine different shade events, and the results of real-time experiments on a lab-scale prototype model are analyzed in detail.