Control, Data and Fault Analysis for the Optical Synchronization System at European XFEL

Supervisors: Prof. Herbert Werner (TUHH), Prof. Gerwald Lichtenberg (HAW), Dr. Annika Eichler (DESY)

The free electron lasers European XFEL and FLASH at DESY provide ultra-short XUV and X-ray photon pulses used to study dynamics in matter with femtosecond precision in pump-probe configurations. To stabilize the accelerator and to carry out pump-probe experiments with high precision in both facilities femtosecond stable optical synchronization systems have been added to traditional coaxial cable based RF distributions. The optical synchronization technology is based on distributing short laser pulses from mode-locked laser oscillator in dispersion and drift compensated optical fibers. This technology is rather new, still under development, and the system complexity is significant. To ensure synchronization performance in the entire facility, beside the stable optical reference distribution, the underlying subsystem (optical laser oscillators, diagnostics, RF regulations) need to be carefully taken into account to minimize timing drifts and fast timing jitters.
For this project, the optical synchronization system for European XFEL and FLASH accelerators shall be analysed from a data point of view. For this purpose, an extensive data acquisition system is required, recording all relevant data from the system for mid- and long-term analysis. Internal changes in the system dynamics due to component aging and environmental changes are to be investigated and prototype algorithms for health monitoring and fault detection developed based on the gathered data. Additionally, an integrated model of optical synchronization and accelerator chain for arrival time jitter of the X-ray photon pulse respective to the experiment trigger (in particular a pump-probe laser pulse) shall be developed as a basis for a subsequent synthesis of a distributed control solution for the optical synchronization system, optimizing arrival time jitter on a system wide level rather than on each individual control loop in the signal chain. Finally, the insights from data analysis shall be integrated with the control solution to ensure optimal performance under all operating conditions and achieve automatic adaption of the controller to changing parameters.