Case Study: Vibration trip and post-event Analysis with Auto-Associative Neural Networks on a Large Steam Turbine.

L. Fromaigeat and G. Nicchiotti
Submission Type: 
Full Paper
phmec_16_029.pdf1.69 MBJune 23, 2016 - 3:29am

The paper presents the case study of a vibration trip event on a 350 MW steam turbine at a coal-based thermal power plant.
The plant is equipped with a Meggitt VM600 protection system, condition monitoring analysis software and an automatic diagnostic tool. The Machine Protection System (MPS) and CMS configuration combines sensors, electronic hardware, firmware and software specific to this application.
The vibration protection system initiated a trip having identified high vibration. The trip prevented further damage. Subsequent analysis of the data using the condition monitoring software established the bearings most affected and pin pointed the source of high vibration.
The data was post processed using an Auto-Associative Neural Networks (AANN) that had been trained with healthy data recorded several hours prior to the trip.
AANN are methodologies widely used for novelty and anomaly detection. AANNs, also called auto-encoders are of particular interest when only one class of data, usually associated to a normal or a reference class, is available. This is typically the situation when the problem of fault detection on rotating machinery is addressed.
The AANN results indicated that such approach would have been capable to detect the failure event in advance compared to the automatic diagnostic system based on rules, demonstrating the validity of the approach in this context.
Various aspects related to vibration: sensing, protection, condition monitoring, analysis, automatic diagnostics using rules and using the AANN will be presented and their results discussed.

Publication Year: 
Publication Volume: 
Publication Control Number: 
Page Count: 
Submission Keywords: 
Condition Based Maintenance
applications: industrial
Auto-Associative Neural Network
Experience Feedback
Submission Topic Areas: 
Data-driven methods for fault detection, diagnosis, and prognosis
Industrial applications
Systems and platform applications
Submitted by: 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed