Contribution to a conference proceedings

PUBDB-2023-00272

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Condition Monitoring and Fault Detection of a Laser Oscillator Feedback System

Gruenhagen, A. (Corresponding author)DESY* ; Eichler, A.XFEL.EU*DESY* ; Tropmann-Frick, M. ; Fey, G.Extern*

2023

Abstract: The successful operation of industrial plants like the European X-ray free electron laser relies on the correct functioning of many dynamic systems that operate in a closed loop with controllers. In this paper, we present how data-based machine learning methods can monitor and detect disturbances of such dynamic systems based on the controller output signal. We implement four feature extraction methods based on statistics from the time domain, statistics from the frequency domain, characteristics of spectral peaks, and the autoencoder latent space representation of the frequency domain. These extracted features require no system understanding and can easily be transferred to other dynamic systems. We systematically compare the performance of 19 state-of-the-art fault detection methods to decide which combination of feature extraction and fault detection is most appropriate to model the condition of an actively controlled phase-locked laser oscillator. Our experimental evaluation shows that especially clustering algorithms are very well suited for detecting disturbed system conditions.

---

Contributing Institute(s):

1. Strahlkontrollen (MSK)

Research Program(s):

1. 621 - Accelerator Research and Development (POF4-621) (POF4-621)
2. HIDSS-0002 - DASHH: Data Science in Hamburg - Helmholtz Graduate School for the Structure of Matter (2019_IVF-HIDSS-0002) (2019_IVF-HIDSS-0002)

Experiment(s):

1. No specific instrument

Appears in the scientific report 2023

---