This project will create an end-to-end demonstration of the plasma control system toolkit being developed at MIT to control a small, levitated magnet system. This device provides a convenient, low cost, low risk proxy for the control system requirements of a tokamak. It provides a vehicle for exploring system modeling, algorithm development, supervisory control, and the training and real-time operation of machine learning based diagnostics and controllers. It is a proof-of-principle for the toolkit, exercising all the major architectural components of a tokamak control system apart from the specific physics and diagnostics.
Connect with the principal investigator to apply and delve into more of the role’s details.
