Anomaly Detection at the European XFEL using a Parity Space based Method

TL;DR

This paper introduces a parity space based anomaly detection method for superconducting RF cavities at EuXFEL, capable of identifying and distinguishing quench events from other anomalies through residual signal analysis.

Contribution

The paper presents a novel model-based parity space method for anomaly detection that generates unique signatures for different fault types, improving fault diagnosis in RF cavities.

Findings

Successfully verified at EuXFEL with experimental data.

Effectively distinguishes quenches from other anomalies.

Provides a fault signature catalog for better diagnosis.

Abstract

A novel approach to detect anomalies in superconducting radio-frequency cavities is presented, based on the parity space method with the goal to detect quenches and distinguish them from other anomalies. The model-based parity space method relies on analytical redundancy and generates a residual signal computed from measurable RF waveforms. The residual is a sensitive indicator of deviation from the model and provides different signatures for different types of anomalies. This new method not only helps with detecting faults, but also provides a catalogue of unique signatures, based on the detected fault. The method was experimentally verified at the European X-ray Free Electron Laser (EuXFEL). Various types of anomalies incorrectly detected as quenches by the current quench detection system are analysed using this new approach.