Master's Thesis

Optimal Control Algorithms of Pulsed Accelerating Fields in Superconducting Spoke Cavities in ESS

Arnbjörn Einarsson, Babak Rajabian

Abstract

The aim of this master thesis is to compare two different control algorithms with respect to reference tracking and simulation time for a superconductive cavity in ESS. The two control algorithms are PID and MPC. The most common type of controller is the PID due to its simplicity. However, an increased interest in finding a control algorithm which can handle the reference tracking better than the PID controller, has been searched for better development of computers. The Model predictive control is a more advanced and requires more computer power. In this thesis the comparison is made using first a very simple cavity model and then extending it with Lorentz detuning of the cavity. Both these models are simulated with the two types of control algorithms. The result shows that both PID and MPC can manage the reference tracking well. In both cases, the controlled output is within its margins. However it can be seen that the PID has a faster response to the beam injection into the cavity. Due to the complexity of the MPC, this algorithm has a longer simulation time then PID. Furthermore the MPC calculates the control signal through an optimization problem.


Master's Thesis ISRN LUTFD2/TFRT--5905--SE, Department of Automatic Control, Lund University, Sweden, June 2012.