Deep learning -Based Control and Management of grid connected hybrid renewable energy system

Abstract

Hybrid renewable energy sources (HRESs), including wind and photovoltaic systems (PV), are gaining popularity as alternatives to traditional sources for power in distributed generation, with the integration of energy storage systems(BAT) enhancing their effectiveness, such as battery, are important for ensuring equilibrium between the variable nature of energy production and the changing conditions of dynamic RL load. The situation demands sophisticated power control and management strategies to address difficult circumstances. This study examines and assesses an effective method for regulating DC bus voltage and enhancing power quality under conditions of dynamic load variation. The proposed power management strategy for hybrid renewable energy systems uses a deep learning intelligent controller, particle swarm optimization (PSO), Recurrent Neural Networks (RNN) and Convolution Neural Networks (CNN), creating an innovative an oversight power management structure designed for photovoltaic systems equipped with battery storage. The goal is to maintain consistent power flow, ensure quality through Total Harmonic Distortion, and ensure uninterrupted service by preventing system components from exceeding operational limits, enhancing DC link bus voltage regulation in renewable hybrid systems.