Artificial neural network-based nonlinear control and modeling of a DC microgrid incorporating regenerative FC/HPEV and energy storage system

Abstract

This study addresses the challenge of mitigating global warming by focusing on DC microgrids integrating renewable energy sources. The research specifically explores the modeling and nonlinear control design of DC microgrids featuring a novel renewable source called hybrid photoelectrochemical and voltaic cells (HPEV), alongside fuel cells and an energy storage system. The HPEV and fuel cells serve as primary sources, while the energy storage system includes a battery bank and ultracapacitor as secondary power sources. The primary objective is to derive a mathematical model for the considered DC microgrid, ensuring each power source maximizes output despite disturbances and varying climatic conditions. To optimize power extraction from HPEV, an artificial neural network is implemented. Subsequently, a nonlinear sliding mode control is applied to manage and stabilize the DC bus voltage, with global asymptotic stability confirmed through Lyapunov stability criteria. Additionally, the study introduces an energy management algorithm for effective power management within the microgrid. The system’s efficiency is validated through MATLAB Simulink simulations under variable load demands, comparing the results with those obtained from a Lyapunov redesign controller. The study concludes with real-time hardware-in-loop experiments, further validating the system’s performance and comparing experimental results with simulated outcomes.