My research will focus on using artificial neural networks to create surrogate models for physics-based battery models for parameter estimation, real-time control strategies, and educational tools. Specifically, I’ll be looking to improve upon the results I got this year using Decision Trees and Random Forests and get meaningful interpolation between data points. This would allow for much more effective surrogate model use in optimization applications. The educational aspect comes into play when considering the incredible speed benefits machine learning surrogate models provide, which can be used to create real-time graphing applications to help build intuition about the physics inside complex systems that cannot normally be interacted with in real time.
Advisor: Venkat Subramanian – Chemical Engineering