STI Publications - View Publication Form #16819

Back to Search Results Print
Publication Information
Title Developing Robust Digital Twins and Reinforcement Learning for Accelerator Control Systems at the Fermilab Booster
Abstract We describe the offline machine learning (ML) development for an effort to precisely regulate the Gradient Magnet Power Supply (GMPS) at the Fermilab Booster accelerator complex via a Field-Programmable Gate Array (FPGA). As part of this effort, we created a digital twin of the Booster-GMPS control system by training a Long Short-Term Memory (LSTM) to capture its full dynamics. We outline the path we took to carefully validate our digital twin before deploying it as a reinforcement learning (RL) environment. Additionally, we demonstrate the use of a Deep Q-Network (DQN) policy model with the capability to regulate the GMPS against realistic time-varying perturbations.
Author(s) D. Kafkes, Malachi Schram
Publication Date August 2021
Document Type Meeting, Proceedings
Primary Institution Fermi National Accelerator Laboratory, Batavia, IL
Affiliation Comp Sci&Tech (CST) Div / Data Science / Data Science
Funding Source Nuclear Physics (NP), FNAL-LDRD-2019-027: Accelerator Control with Artif
Proprietary? No
This publication conveys Technical Science Results
Document Numbers
JLAB Number: JLAB-CST-21-3502 OSTI Number: 1828299
LANL Number: Other Number: DOE/OR/23177-5351
Associated with an experiment No
Associated with EIC No
Supported by Jefferson Lab LDRD Funding No
Meeting / Conference
Meeting Name IPAC21
Meeting Date 5/24/2021
Document Subtype Paper (proceedings only)
Proceedings
Title Proceedings of IPAC 2021
Editor(s) Liu Lin, John M. Byrd, Regis Neuenschwander, Renan Picoreti, Volker R. W. Schaa
Publisher JACOW
Refereed No
Page Number 2268-2271
Year 2021
Attachments/Datasets/DOI Link
Document(s)
tupab327.pdf (STI Document)
tupab327.pdf (Accepted Manuscript)
DOI Link
Dataset(s) (none)
Back to Search Results Print