Highly Efficient MPP Tracker based on Adaptive Neuro-fuzzy Inference System for Stand-Alone Photovoltaic Generator System

Abstract

Recently, AI-based maximum power point tracking (MPPT) controllers for photovoltaic (PV) generator systems have become widespread. Compared to traditional maximum power point (MPP) trackers, the AI-based trackers have lower oscillations around the MPP, high tracking speed, and the least calculation time. Among the various AI methods, the Adaptive Neural Fuzzy Inference System (ANFIS) is widely used for PV systems. Nevertheless, obtaining precise training data and tuning the ANFIS model presents significant difficulties for developing an effective ANFIS-MPPT technique. This article proposes a highly efficient MPP tracker based on adaptive ANFIS with direct control for stand-alone PV generators. The proposed ANFIS-MPPT technique can trace the MPP under rapidly changing solar radiation and cell temperatures. The duty cycle of the boost converter is directly adjusted; hence the PI control loop is eliminated in this technique. The training data for the suggested technique are extracted with the aid of a multi-variable step perturbation and observation (MV-P&O-MPPT) algorithm for avoiding the errors that are usually included in an experimental dataset. The proposed ANFIS-MPPT technique is simulated and compared to other AI-based MPPT techniques such as Artificial Neural Networks (ANN) and Fuzzy Logic Controller (FLC). Simulation results confirm that the presented technique precisely tracks the MPP and achieves higher efficiency under different climatic conditions.