FAULT ANALYSIS OF SHIP MACHINERY USING MACHINE LEARNING TECHNIQUES

AK A., Inceisci F. K.

INTERNATIONAL JOURNAL OF MARITIME ENGINEERING, cilt.164, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 164
  • Basım Tarihi: 2022
  • Doi Numarası: 10.5750/ijme.v164i1.769
  • Dergi Adı: INTERNATIONAL JOURNAL OF MARITIME ENGINEERING
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Applied Science & Technology Source, Compendex
  • Anahtar Kelimeler: Fault analysis, Ship machines, Machine Learning, ARTIFICIAL NEURAL-NETWORK
  • Marmara Üniversitesi Adresli: Evet

Özet

Estimating the probability of failure of ship systems has become an important issue. The rapid development of information technologies in recent years has made it possible to integrate machine learning techniques into the design of ship systems. Predicting and preventing the failures that may occur in the ship machinery systems can extend the life of ship machinery. The study presented aimed to predict the turbine failure of a ship engine. For this purpose, the data obtained from an LM-2500 type ship engine were analysed using Artificial Neural Networks (ANN) algorithms. These results were compared with the following techniques: linear regression; decision tree regression; "k" nearest neighbours' regression; random forest regression; Bayesian ridge regression; extra tree regression; and linear Support Vector Regression (SVR). The study showed that the ANN method determined the failure prediction of ship machinery with a higher accuracy than the regression methods.