Real-time Artificial Intelligence for Accelerator Control: A Study at the Fermilab Booster

Abstract

We describe a method for precisely regulating the gradient magnet power supply at the Fermilab Booster accelerator complex using a neural network trained via reinforcement learning. We demonstrate preliminary results by training a surrogate machine-learning model on real accelerator data to emulate the Booster environment, and using this surrogate model in turn to train the neural network for its regulation task. We additionally show how the neural networks to be deployed for control purposes may be compiled to execute on field-programmable gate arrays. This capability is important for operational stability in complicated environments such as an accelerator facility.

16 pages, 10 figures. Phys. Rev. Accel. Beams vol 24, issue 10. Published 18 October 2021. For associated dataset and data sheet see http://doi.org/10.5281/zenodo.4088982

References in corpus (26)

• Continuous control with deep reinforcement learning
• Playing Atari with Deep Reinforcement Learning
• Deep Learning using Rectified Linear Units (ReLU)
• Machine learning and the physical sciences
• An Introduction to Deep Reinforcement Learning
• FPGA-based Accelerators of Deep Learning Networks for Learning and Classification: A Review
• Fast inference of deep neural networks in FPGAs for particle physics
• Automatic heterogeneous quantization of deep neural networks for low-latency inference on the edge for particle detectors
• Bayesian optimization of a free-electron laser
• Fast convolutional neural networks on FPGAs with hls4ml
• Compressing deep neural networks on FPGAs to binary and ternary precision with HLS4ML
• Neural Networks for Modeling and Control of Particle Accelerators
• Accelerating CNN inference on FPGAs: A Survey
• Machine Learning for Orders of Magnitude Speedup in Multi-Objective Optimization of Particle Accelerator Systems
• Fast inference of Boosted Decision Trees in FPGAs for particle physics
• Superconducting radio-frequency cavity fault classification using machine learning at Jefferson Laboratory
• Machine learning for beam dynamics studies at the CERN Large Hadron Collider
• Distance-Weighted Graph Neural Networks on FPGAs for Real-Time Particle Reconstruction in High Energy Physics
• Physics-informed Gaussian Process for Online Optimization of Particle Accelerators
• Accelerated Charged Particle Tracking with Graph Neural Networks on FPGAs
• First Steps Toward Incorporating Image Based Diagnostics Into Particle Accelerator Control Systems Using Convolutional Neural Networks
• Advanced Control Methods for Particle Accelerators (ACM4PA) 2019 Workshop Report
• Surrogate Modeling of the CLIC Final-Focus System using Artificial Neural Networks
• Model-free and Bayesian Ensembling Model-based Deep Reinforcement Learning for Particle Accelerator Control Demonstrated on the FERMI FEL
• A flexible FPGA accelerator for convolutional neural networks
• Quantized Guided Pruning for Efficient Hardware Implementations of Convolutional Neural Networks

Cited by in corpus (4)

• Applications and Techniques for Fast Machine Learning in Science
• Online-compatible Unsupervised Non-resonant Anomaly Detection
• Nanosecond machine learning event classification with boosted decision trees in FPGA for high energy physics
• Model-free and Bayesian Ensembling Model-based Deep Reinforcement Learning for Particle Accelerator Control Demonstrated on the FERMI FEL