Optimized MPPT for Grid tied Transformer Less PV System: A Comparative Analysis

¹ Senior Assistant Professor, EEE Dept., CVR College of Engineering, Hyderabad, Telangana, India
² Associate Professor, EEE Dept., ANURG University, Hyderabad, Telangana, India.
³ Professor, EEE Dept., JNTU Hyderabad, Telangana, India

^* Corresponding author: janardhan.gu@gmil.com

Abstract

For grid-tied photovoltaic (PV) systems, Maximum Power Point Tracking (MPPT) algorithms based on artificial neural networks (ANNs) are prone to initialization issues, which could cause them to converge at local maxima rather than the global maximum power point (MPP). This means that, a regular retraining on big datasets is required. This paper presents a novel hybrid MPPT algorithm that combines Extreme Gradient Boosting (XGBoost) and Vascular Invasive Growth Optimization (VIGO) to address this challenge. The exploration-exploitation conundrum that traditional optimization algorithms have is addressed by VIGO, and the convergence speed and accuracy of MPPT are improved by XGBoost. To assess its performance, the suggested approach is compared with well-known methods such as Grasshopper Optimization Algorithm (GOA), Sparrow Search Algorithm (SSA), and Particle Swarm Optimization (PSO). This comparison study shows that the hybrid VIGO-XGBoost method produces improved maximum power.