| PhD Seminar

Name of the Speaker: Mr. koushik Ghosh (EE21D750)
Guide: Dr. Kamalesh Hatua
Online meeting link: https://meet.google.com/moj-vzio-pjn
Date/Time: 14th March 2025 (Friday), 3:00 PM
Title: A Simplified Large-Signal Model-Based Control to Achieve Uniform Dynamic Response and a Frequency Dithering Technique to Attenuate Electromagnetic Interference (EMI) in Dual Active Bridge Series Resonant Converter.

Abstract :

A large-signal model-based control (LSMC) is proposed, which achieves a uniform transient response throughout the operating range, unlike the small-signal model-based linearized an d non-linear multi-variable control. The proposed LSMC method eliminates the operating point dependency through a lucid control methodology combining analog and digital blocks. The decoupling in multivariable control and system order reduction is achieved through the feedforward technique. This method also reduces the computational burden time of the controller as compared to the existing non-linear control methods. The achieved settling time for both output voltage and reactive power is 5 ms uniformly throughout the operating range of the converter with a computational burden of 5.6 us. Along with that, in order to attenuate EMI in DABSRC, a frequency dithering technique (FDT) combined with frequency-independent closed-loop control for dual active bridge series resonant converters (DABSRC). The proposed combined technique attenuates the EMI noise and simultaneously regulates the output variables, making it independent of switching frequency variation. The FDT considers 5 switching frequencies from 66-72 kHz in steps of 1.5 kHz. The proposed FDT considers the effect of resolution bandwidth (RBW) of 9 kHz as per EMI standards. The frequency dependency of the output variables is eliminated by a modifying the proposed LSMC. The proposed technique achieves a significant EMI attenuation of 8 dB along with decoupled output voltage and reactive power control. This achieves a filter size reduction by 34%. All the results are experimentally validated on a 2.4 kW DABSRC hardware prototype.