Dynamic Model and Intelligent Maximum Power Point Tracking Approach for Large‐Scale Building‐Integrated Photovoltaic System

Abstract

Due to the complex surrounding environments, unpredictable occupants’ behaviors, and the photovoltaic (PV) module degradation after years of operation, the fast and accurate maximum power point tracking (MPPT) of building‐integrated photovoltaic (BIPV) system is very difficult. Herein, the MPPT of a huge BIPV system is modeled as a large‐scale global optimization (LSGO) problem, in which all the variables are proven to be separable. Then, a novel differential evolution (DE) algorithm with the “DE/Current‐to‐Tbest/Current‐to‐Gbest/1” mutation operation (DE‐TG1) is developed to optimize the aforementioned separable LSGO−MPPT model. In addition, the problem of PV module degradation and its effect on the LSGO‐MPPT model are analyzed in detail. To reduce the influence of PV degradation on MPPT accuracy, the dynamic LSGO−MPPT model is developed by introducing the annual drift rates into the proposed LSGO−MPPT model. Finally, the developed methodologies are comprehensively tested by simulation experiments. Experimental results show that the developed DE‐TG1 algorithm significantly outperforms the other state‐of‐the‐art algorithms on separable LSGO problems. In addition, result of the experiments also indicates that with the cooperation of the DE‐TG1 algorithm and the dynamic LSGO−MPPT model, the effective MPPT control of a large‐scale BIPV system can be obtained with accurate and robust performance.