Adaptive Neuro-Fuzzy Inference-Based Control of Parallel Connected DC-DC Converters for DC Microgrid Application

DC microgrid has been popular in recent years, because of its excellent features such as decentralized design, environment friendly, and easy integration with renewable energy sources. In a typical DC microgrid, multiple DC-DC converters are connected in parallel for reliable operation. But in this parallel connection, accurate current sharing among converters and DC bus voltage regulation is not as desired. To overcome the inaccurate current sharing and DC voltage regulation problem, a novel Adaptive Neuro-Fuzzy Inference System (ANFIS) based control using the backpropagation training algorithm is proposed. The training efficiency of the proposed system is around ≈ 99.8%. The simulation results demonstrate that the ANFIS-based strategy has lesser oscillations during the sudden change in load conditions and is damped out in less time in comparison to conventional proportional-integral control.