Maximum power extraction in partial shaded grid-connected PV system using hybrid fuzzy logic/neural network-based variable step size MPPT

Abstract

The photovoltaic (PV) system’s output power varies owing to solar radiation’s irregularity, which confines their usage for various applications. Implementation of maximum power tracking (MPT) algorithms increases the efficiency and power generated from solar cells. When the array is partially obscured by clouds or structures, several local maximum power peaks (LMPPs) appear in the solar cell characteristics. Traditional MPPT algorithms, rather than following the global peak power point (GPPP), are preferable to following the local peak power point. If partial shading causes numerous LPPPs, it is necessary to look into how the MPPT technique can keep track of GPPP. Employing soft computing approaches such as the hybrid neural network/fuzzy method with variable step size perturb and observing MPPT, it is possible to trace the GPPP and also augment solar energy extraction. The present research paper focuses on hybrid fuzzy/neural network MPPT integrated with a high-step-up DC-DC converter to harvest the utmost power from the solar PV array. The voltage transients are reduced by controlling the DC link voltage along with solar radiation and temperature variations. The proposed MPPT technique is shown to be effective under both uniform and partial shade conditions in a series of simulations. From the test results, the efficiency of the overall system has increased from 91 to 98% for partial shading and uniform operating conditions.