Application of System Identification Techniques to an RF Cavity Tuning Loop

Modern system identification is applied to rf cavity tuning on the ISIS synchrotron. Four types of test signals are investigated to assess their suitability for real time measurement in an accelerator environment. The Pseudo Random Binary Signal (PRBS) appears to be the most advantageous. Measurements under normal operating conditions allow automatic identification for a self-adapting loop. The interactive software MATLAB is used to process the data and the identified model is represented in pole-zero form. The model shows good correlation with system performance.