A New Hybrid BAT-ANFIS-Based Power Tracking Technique for Partial Shaded Photovoltaic Systems

Abstract

Photovoltaic (PV) power has proved to be the most reliable and sustainable technology as a primary source of power. The intermittent and fragmentary nature of solar energy has its own shortcomings due to which the PV system fails to meet the continuous demand of power structure set-up. Hence, optimization algorithm such as maximum power point tracking (MPPT) has been implemented to augment and improve the power efficiency of the PV system. Conventional techniques such as Incremental conductance and perturb and observe based MPPT fail to tackle the non-linearity and oscillations issue in tracking the maximum power especially when the array is shaded. MPPT is incorporated to overcome these limitations. Modern techniques such as artificial neural network, evolutionary algorithms (EA) and Fuzzy logic, can be additionally integrated into the system to select the desired algorithm to obtain maximum optimized output. In this research work, a BAT EA trained Adaptive Neuro-Fuzzy Inference System (ANFIS) based MPPT is implemented for a partially shaded PV array. A modified SEPIC converter controlled by BAT ANFIS ensures maximum power delivery to the load. The suggested technique is tested for various patterns of shading and the results reveals that the BAT ANFIS MPPT is advantageous.