STAGE 01
Satellite DC Bus
A nominal approximately 70-V DC source represents the spacecraft power bus feeding the conversion system.
Research Profile
High-frequency power conversion for demanding energy systems.
I am a PhD Research Scholar in Electrical Engineering at IIT Roorkee working in power electronics, with a particular interest in high-frequency GaN-based converters, isolated DC–DC conversion, high-gain architectures, and power conditioning systems for spaceborne applications.
My research focuses on understanding the complete converter rather than treating topology, semiconductor devices, magnetics, gate drivers, control, and thermal behaviour as independent problems. The objective is to develop power conversion systems that are efficient, compact, high-frequency, and physically realizable.
Current PhD Research
High-frequency GaN-based power conversion.
Primary Research Direction
Design and Analysis of High-Frequency Isolated GaN-Based DC–DC Converters for Spaceborne Electronic Power Conditioning
My current research is directed toward high-frequency power conversion using gallium nitride (GaN) power semiconductor devices. The work focuses on converter architectures capable of achieving high efficiency, high power density, and reliable operation at elevated switching frequencies.
A key application context is spaceborne electronic power conditioning, where converter size, mass, efficiency, thermal constraints, reliability, and electromagnetic compatibility are particularly important.
The research brings together topology selection, semiconductor characterisation, switching-loss analysis, magnetic-component design, gate-drive optimisation, soft-switching techniques, control, thermal behaviour, parasitic effects, and experimental validation.
Research Architecture
From the DC bus to the regulated output.
One of the system architectures being investigated considers a two-stage power conversion path for a spaceborne electronic power conditioning system.
STAGE 02
Pre-Regulation / Boost Stage
A regulated high-voltage DC link is generated from the lower-voltage spacecraft bus, providing a suitable input for the subsequent isolated stage.
STAGE 03
High-Frequency GaN Conversion
A high-frequency isolated GaN-based converter performs the required voltage transformation and power conditioning.
Technical Themes
Problems I am interested in solving.
Converter Topology & Architecture
Comparative analysis and design of isolated and high-gain converter architectures.
- Resonant and soft-switching converters
- High-gain DC–DC architectures
- Half-bridge and full-bridge configurations
- Topology selection based on application constraints
GaN & Wide-Bandgap Devices
Understanding how wide-bandgap semiconductor properties enable higher-frequency power conversion.
- GaN switching characteristics
- Switching and conduction losses
- Device voltage and current stress
- Parasitic-induced switching behaviour
High-Speed Gate Drivers
Gate-drive design becomes increasingly important as switching transitions become faster.
- Gate resistance and switching speed
- Dead-time optimisation
- dv/dt and di/dt effects
- Protection and reliable turn-off
High-Frequency Magnetic Components
High switching frequency creates new constraints in transformer and inductor design.
- High-frequency transformer design
- Core and winding losses
- Leakage and magnetising inductance
- Parasitic capacitance
Soft Switching & Loss Reduction
Exploring switching strategies that reduce semiconductor losses while maintaining practical operating ranges.
- ZVS and ZCS operation
- Switching transition analysis
- Loss modelling
- Efficiency optimisation
Thermal, EMI & Reliability
A converter must remain reliable outside ideal simulation conditions.
- Thermal management
- Electromagnetic interference
- Device stress and protection
- Non-ideal component behaviour
Research Questions
Questions behind the converter.
QUESTION 01
How far can switching frequency be increased?
Increasing switching frequency can reduce magnetic component size, but eventually switching losses, parasitics, EMI, gate-drive limitations, and thermal constraints dominate the design.
QUESTION 02
Which topology provides the best system-level trade-off?
The highest efficiency topology under one operating condition may not provide the best solution across the full load and input-voltage range.
QUESTION 03
How should GaN devices actually be driven?
The benefits of GaN depend heavily on gate-drive implementation, layout, parasitic inductance, dead-time, switching transitions, and protection.
QUESTION 04
Can high power density coexist with reliability?
Increasing power density introduces thermal, electrical, mechanical, and electromagnetic constraints that must be considered at the system level.
Previous Research
Intelligent battery state estimation.
Deep Learning Architectures for Precise SoC and SoH Estimation in Lithium-Ion Batteries using Bio-Inspired Optimization
My M.Tech research at VNIT Nagpur focused on intelligent estimation of State of Charge (SoC) and State of Health (SoH) in lithium-ion batteries.
The work investigated deep-learning architectures including LSTM, GRU, and CNN-LSTM models for battery state estimation and examined bio-inspired optimisation techniques for improving model performance.
The research gave me experience at the intersection of battery modelling, machine learning, time-series prediction, optimisation, and battery management systems.
This work now forms part of the broader foundation from which I approach power-electronics research: understanding both the energy-conversion hardware and the intelligent systems that interact with it.
Publications & Research Work
Selected research.
Battery Systems · Deep Learning
Comparative Analysis of Deep Learning Models for State-of-Charge Estimation in Li-ion Batteries
Comparative investigation of recurrent and hybrid deep-learning architectures for accurate battery State-of-Charge estimation.
Publication / PDF — Coming Soon
Battery Degradation · Optimization
State-of-Health Estimation and Degradation Tracking using CNN-LSTM and Enhanced Genetic Optimization
Research on intelligent State-of-Health estimation, degradation tracking, and optimisation-based tuning of deep-learning models.
View Paper PDF
Battery Management Systems
Active Cell Balancing Strategies in Battery Management Systems
Analysis of active cell-balancing approaches, energy redistribution, efficiency, cell equalisation, and implications for battery life and system reliability.
Technical Presentation — Coming SoonResearch Toolkit
Tools I use to turn ideas into models and experiments.
MATLAB
Simulink
Python
NumPy
C
MATLAB Control Systems
Power Electronics Simulation
DSP-Based Control
TMS320F28335
TMS320F28379D
PWM
Digital Control
Future Directions
Where I want the research to go.
Higher Power Density
Pushing switching frequency and converter power density while maintaining acceptable thermal and reliability margins.
Intelligent Power Electronics
Exploring data-driven and machine-learning approaches for monitoring, optimisation, prediction, and fault-aware operation.
Spaceborne Power Systems
Developing compact, efficient, and reliable power-conversion architectures for demanding aerospace and spaceborne applications.
Research Philosophy
Simulation is the beginning, not the destination.
A converter that works in simulation is only a hypothesis. The real engineering problem begins when parasitic inductance, device capacitance, magnetic losses, thermal limitations, measurement uncertainty, EMI, gate-drive behaviour, and component tolerances enter the system. My goal is to bridge that gap between mathematical models, simulation, hardware, and experimentally validated engineering.