Harnessing greylag goose optimization for efficient MPPT and seven-level inverter in renewable energy systems

Abstract

Owing to the significant increase in energy consumption, contemporary power systems are transitioning to a new standard characterized by enhanced access to renewable energy sources (RESs). RESs require interfaces to regulate the power generation. Maximum power point tracking (MPPT) is a technique employed in solar photovoltaic (PV) systems to modify operational parameters to ensureoptimal extraction of power from solar panels. MPPT operates under fluctuating conditions such as sunlight intensity and temperature. An inverter is a device that transforms a direct current into a sinusoidal alternating current. A multilevel inverter (MLI) can be utilized for RESs in two distinct modes: power-generating mode (stand-alone mode) and compensator mode (STATCOM). Limited research has been conducted on the optimization of controller gains in response to variations in a single phase load, particularly reactive load variations, across several scenarios. This load may exhibit an imbalance; hence, a more robust optimization approach must be used to address this problem. This study presents a control system that incorporates an optimized auxiliary MPPT controller for a seven-level inverter. The system uses a sophisticated greylag goose optimization (GGO) random search algorithm combined with the MPPT technique. The main objective is to create a system that enhances performance under diverse and imbalanced loading scenarios by utilizing sophisticated optimization techniques that determine the optimal switching angles for a seven-level inverter. This approach aims to eliminate specific harmonics and achieve a low total harmonic distortion (THD). The inverter THD output voltage was used as the objective function, and the proposed method is particularly beneficial in agricultural settings. The proposed MPPT-based seven-level invertersystem was simulated using MATLAB. The proposed GGO algorithm achieved a minimal THD of 1.95 %, surpassing methods such as salp swarm optimization (6.14 %), artificial neural networks with fuzzy logic (5.9 %), hybrid global selective algorithm (GSA) selective harmonic elimination (7.7 %), and genetic algorithms with particle swarm optimization (10.84 %), demonstrating its exceptional efficacy in improving power quality.