Data-driven modeling of low pressure plasmas for semiconductor manufacturing

Professor: Venkattraman Ayyaswamy

Description: This project develops a data-driven modeling framework for plasmas used in materials processing by generating high-fidelity datasets from 2D simulations across a range of driving frequencies. Dynamic Mode Decomposition (DMD) is employed to extract dominant spatio-temporal features and construct a reduced-order model that captures the key dynamics of the system. The resulting model enables efficient prediction and analysis of plasma behavior over a wide frequency range, supporting improved design and control of plasma-based processes.

Preferred Qualifications: Willingness to learn our in-house code and/or experimental diagnostics, Python programming, Basic Linux or willingness to learn