Bayesian and machine learning-based fault detection and diagnostics for marine applications

Michail Cheliotis, Iraklis Lazakis, Angelos Cheliotis

Research output: Contribution to journalArticlepeer-review

18 Citations (Scopus)
24 Downloads (Pure)

Abstract

Marine maintenance can improve ship performance by leveraging predictive maintenance, Machine Learning and Data Analytics. This paper aims to enrich the literature, by developing a novel framework for ship diagnostics based on operational data and the probability of faults. Moreover, the framework can identify the root cause of developing faults avoiding black-box Neural Networks, and complex physics-based models. This research integrates Machine Learning-based Fault Detection, Exponentially Weighted Moving Average control charts, and Bayesian diagnostic networks which allow the examination of the rate of development (fault profile) of faults and failure modes. For validation, the case study of a marine Main Engine is used to examine faults in the engine’s Air Cooler and Air and Gas Handling System. It is concluded that any simultaneous abnormal deviations in the Main Engine’s Exhaust Gas Temperature are more likely to be caused by a fault in the Air and Gas Handling System.

Original languageEnglish
Pages (from-to)2686-2698
Number of pages13
JournalShips and Offshore Structures
Volume17
Issue number12
Early online date9 Jan 2022
DOIs
Publication statusPublished - 2 Dec 2022

Keywords

  • condition monitoring
  • ship system diagnostics
  • Bayesian networks
  • fault detection
  • machine learning
  • ship operation

Fingerprint

Dive into the research topics of 'Bayesian and machine learning-based fault detection and diagnostics for marine applications'. Together they form a unique fingerprint.

Cite this