A computational intelligence based maximum power point tracking for photovoltaic power generation system with small-signal analysis

There are multiple peak functions in its output power characteristic curve of a photovoltaic (PV) array under partial shading conditions (PSCs), the perturb and observe (P&O) may fail to track the global maximum power point (GMPP). Therefore, a reliable maximum power point tracking (MPPT) technique is essential to track the GMPP within an appropriate time. This article proposes a hybrid technique by combining an evolutionary optimization technique, namely the modified invasive weed optimization (MIWO) with the conventional P&O algorithm to enhance the search performance for the maximum power output of the PV system. MIWO executes in the initial stages of the tracking followed by the P&O at the final stages in the MPPT search process. The combined approach ensures faster convergence and better search to the GMPP under rapid climate change and PSCs. The search performance of the hybrid MIWO+P&O technique is examined on a standalone PV system through both MATLAB/Simulink environment and experimentally using dSPACE (DS1103)-based real-time microcontroller hardware setup. The performance of the proposed hybrid MPPT scheme is compared with the recent state-of-the-art MPPPT techniques. In addition, the small-signal analysis of the PV system is carried out to evaluate the loop robustness of the controller design. For a given set of system parameters, simulations for the small-signal model and robustness studies are analyzed to verify the results. The overall results justify the efficacy of the proposed hybrid MPPT algorithm.

Automated summary

This article proposes a hybrid technique by combining an evolutionary optimization technique with the conventional P&O algorithm to enhance the search performance for the maximum power output of the PV system. The combined approach ensures faster convergence and better search to the GMPP under rapid climate change and partial shading conditions.