BIOGRAPHICAL SKETCH
I am a Ph.D. candidate in Mechanical Engineering at the University of Alberta, specializing in the development of advanced machine learning models for energy systems. My research focuses on the state-of-health (SOH) estimation and remaining useful life (RUL) prediction of solid oxide fuel cells (SOFCs), which are promising alternatives to fossil fuels in power generation. Under supervision of Dr. M. Shahbakhti and Dr. C. R. Koch, in a collaborative project with Cummins Inc., I employ physics-informed machine learning (PIML) techniques to advance predictive maintenance methods that can mitigate SOFC degradation, expanding their lifespan and reducing costs.
As the first stage of my doctoral work, I developed a steady-state performance prediction model for diverse tubular and planar SOFCs under healthy conditions. Using a deep neural network (DNN) model, I successfully reduced the training time and data requirements for developing predictive models by up to 85% and 90%, respectively, through transfer learning. Building on this foundation, my current focus is on integrating fuel cell voltage and impedance time-series data with knowledge of the underlying physics into real-time PIML models for faulty conditions. By training a temporal graph convolutional network (TGCN) under essential fault conditions such as Redox cycling, I achieved a root mean square error (RMSE) of 0.084 in predicting SOH over a long horizon during accelerated degradation tests.
With an M.Sc. in Mechatronics Engineering and a B.Sc. in Aerospace Engineering, my academic journey has provided me with a diverse technical background in model-based and data-driven control systems, as well as advanced programming and machine learning methodologies. During my master’s thesis, I developed a dynamic simulation for internal combustion engines and vehicle dynamics for real-time hardware-in-the-loop applications. My bachelor’s thesis focused on the design, fabrication, and control of a reaction wheel inverted pendulum using a sliding mode controller. These experiences have shaped my passion for addressing real-world challenges at the intersection of machine learning, control and diagnostics systems, and sustainable energy technologies.
Through my interdisciplinary expertise, I strive to contribute to the advancement of energy solutions. My research bridges academic rigor with industrial applicability, positioning me to make meaningful contributions to the global transition toward cleaner and more sustainable energy systems.
(November, 2024)