Talk by Vishnu Varadan

Abstract: This talk delves into the intricacies surrounding the control of a prototype Beam Wire Scanner (BWS) featuring a Tubular Linear Magnetic Coupling (TLMC), a pivotal instrument in particle accelerators and beam physics experiments. The focus is on utilizing nonlinear system identification techniques, particularly neural networks, to characterize the TLMC system. This characterization reveals its suitability for precisely determining wire position without significantly affecting beam size accuracy. The discussion further explores optimal reference trajectory generation methods, showcasing their effectiveness in minimizing wire vibration and augmenting the repeatability and reliability of the BWS. Acknowledging the influence of evolving mechanical conditions on the Permanent Magnet Synchronous Motor (PMSM), the focus moves on towards innovative data-driven motor drive control strategies. Simulations presented during the talk highlight the proposed Deep Predictive Control (DeePC) architecture's ability to adapt seamlessly to changing motor parameters. This adaptation ensures high performance and adaptability in BWS operations, emphasizing the potential impact of these advancements on the field of particle acceleration and beam physics experimentation.

Bio: Vishnu Varadan completed his bachelor's in Electronics and Instrumentation Engineering in 2019 from Anna University, India graduating at the top of his class and earning the University Gold Medal. Later, he went on to work for Caterpillar for a couple of years on the development of control systems for diesel-electric locomotives. He started his master's in Autumn 2021 at ETH Zurich in Electrical Engineering and Information Technology, specializing in Control Systems. During this period he was involved in several projects including a the development of a receding horizon game theoretical framework for a smart grid and the design of a sequential control scheme for a liquid rocket engine. Currently, he is working on his Master's Thesis at CERN, Geneva with the Beam Instrumentation Group. His major research interests include direct and indirect data-driven control, optimization and optimal control, game theory and receding horizon control.