E-raamat: Intelligent Beam Control in Accelerators

1. Introduction

2. Beam Feedback Control

2.1. Beam Feedback Overview

2.2. Problem Formulation

2.3. Beam Response Matrix Identification

2.4. Static Linear Feedback Controller Design

2.4.1. Difficulties in Inversing a Response Matrix

2.4.2. Singular Value Decomposition

2.4.3. Response Matrix Inverse with SVD

2.4.4. Response Matrix Inverse with Least-square Method

2.4.5. Robust Control Design

2.5. Summary

3. Beam Optimization

3.1. Beam Optimization Overview

3.1.1. Optimizations in Beam Control

3.1.2. Formulation of Optimization Problems

3.1.3. Considerations of Online Optimization

3.2. Optimization Algorithms

3.2.1. A Test Function

3.2.2. Spontaneous Correlation Optimization

3.2.3. Random Walk Optimization

3.2.4. Robust Conjugate Direction Search

3.2.5. Genetic Algorithm

3.2.6. Particle Swarm Optimization

3.2.7. Summary of Optimization Algorithms

3.3. Beam Optimization Examples and Tools

3.3.1. Practical Considerations

2

3.3.2. FEL Optimization with SCO

3.3.3. Operating Point Switching

3.3.4. Optimization Tools

3.4. Summary

4. Machine Learning for Beam Control

4.1. Overview of Machine Learning

4.1.1. Introduction

4.1.2. Machine Learning Models

4.1.3. Typical Applications of Machine Learning

4.2. Accelerator Modeling with Machine Learning

4.2.1. Artificial Neural Network Regression Model

4.2.2. Gaussian Process Regression Model

4.3. Applications of Machine-learning Models in Beam Control

4.3.1. Surrogate Models of Beam Response

4.3.2. Response Matrix Estimation with NN Surrogate Model

4.3.3. Beam Optimization with NN Surrogate Model

4.3.4. Feedforward Control with NN Surrogate Model

4.3.5. Beam Optimization with GP Surrogate Model

4.4. Feedback Control with Reinforcement Learning

4.4.1. Introduction to Reinforcement Learning

4.4.2. Feedback Controller Design with Natural Actor-critic Algorithm

4.4.3. Example: RF Cavity Controller Design

4.4.4. Example: Static Feedback Controller Design

4.4.5. Further Reading

4.5. Other Applications of Machine-learning in Accelerators

4.5.1. Virtual Diagnostics

4.5.2. Fault Prediction

4.5.3. Classification

4.6. Summary

Zheqiao Geng is a senior electronic engineer at the Paul Scherrer Institute in Switzerland. He graduated with a bachelors degree from Tsinghua University in Beijing, China, in 2002. In 2007 he received his Ph.D. degree in nuclear engineering from the Graduate School of Chinese Academy of Sciences. For more than ten years, he worked on accelerator RF and beam control systems in different labs, including IHEP (China), DESY (Germany), SLAC (USA) and PSI (Switzerland). He was the key developer of critical aspects of the LLRF systems for various accelerator projects such as the European XFEL, LCLS and SwissFEL. At SLAC, he led the system-level design of the LCLS-II LLRF system. At PSI, he developed the control strategy for setting up and stabilizing the two-bunch operation of SwissFEL. As an internationally acclaimed RF and beam control expert, he was appointed as a PSI Senior Expert in 2021.



Stefan Simrock is a control system coordinator atthe ITER Organization located in southern France. He studied physics and microwave engineering at the Technical University of Darmstadt where he received his Ph.D. in engineering physics in 1988. From 1988 to 1996, he worked at the Thomas Jefferson National Accelerator Facility as a RF controls group leader and the deputy for the technical performance of the accelerator. He joined DESY in 1996 as the leader of a multidisciplinary team responsible for the design, construction, and commissioning of the control system for the superconducting linac at the TESLA Test Facility. In 2004, he was an appointed group leader of beam controls group responsible for the timing, synchronization, and beam feedback systems of all 10 accelerators at DESY. At the same time, he was the project leader for the RF Control System for FLASH and the European XFEL. Since 2010, he is responsible for the integration of ITER diagnostics with the central control system, machine protection system, safety system, andplasma control system.   Together with Dr. Zheqiao Geng, he published a book Low-Level Radio Frequency Systems (978-3-030-94418-6) in the series of Particle Acceleration and Detection in 2022.