Analysis of Regression-based MPPT Algorithms in PV System

This paper aims to evaluate the potential of the regression method as an MPPT (Maximum Power Point Tracking) control strategy for PV panels. To this end, three MPPT algorithms were designed, simulated and tested in MATLAB / Simulink. These are a regression algorithm (R), a "perturb and observe" algorithm (P&O), and a hybrid algorithm (H). Of these, two are based on conventional methods (P&O is an online method and R is an offline method) and one is based on a hybrid method (R - P&O). A comparative analysis of MPPT control on the PV systems was performed based on the results obtained from the MATLAB simulations. The parameters taken into account in this review are design complexity, tracking speed, tracking accuracy, sensitivity to environmental change and cost. The advantages and disadvantages of these methods are presented in detail.