Optimized ANN-based MPPT for Solar Photovoltaic System

Photovoltaic systems are growing in popularity to generate clean electrical energy. Nevertheless, because of minimal efficiency, researchers have sought methods for enhancing its productivity. Maximum Power Point Trackers (MPPT) allow us to extract the most energy from PV panels that can be produced. It involves an algorithm to calculate the Maximum Power Point (MPP). The most commonly used MPPT methodologies, Incremental Conductance (I-C) and Perturb and Observe (P-O) provide long-term reliability for consistent environmental conditions, although tracking becomes complicated for varying environmental patterns. It was proved that ANN (Artificial Neural network) can track MPP with a small transient time and less ripple. Subsequently, to achieve the ideal MPP, the ANN method must be extensively adjusted. For desired performance, the number of hidden layers and neurons in every hidden layer must be properly chosen. Thus, the Seagull Optimisation (SO) algorithm is employed to optimise ANN in this work. Its capacity to perform is compared with that of P-O, ANN, and PSO. (Particle Swarm Optimization) for constant and varying climatic conditions.