Performance analysis of machine learning for fault diagnosis of ballast water treatment system
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The demand for ecofriendly ships in shipbuilding and maritime industries is growing steadily. In response to strict environ-mental regulations on shipping by the International Maritime Organization, the demand for ship machinery and equipment that carry ecofriendly labels, including ballast water treatment systems (BWTS), is increasing. The BWTS involves early-stage equipment error and fault diagnostics, and its malfunction can have major cost and time consequences. This study expands on previous research on fault diagnosis using the SVM, a machine learning model. The two aspects of this expansion are an increased window size range used to generate the features and the introduction of several machine learning models. We used 47,435 sensor data points to compare and analyze the results and evaluate the classification accuracy by increasing the window size range to 10. We demonstrate that the previous model can be easily applied to other machine learning models and that the SVM model improves performance through feature gener-ation. The F1 score of the random forest model with the highest performance score of 99.73% indicates potential for industrial appli-cations if accompanied by expert monitoring and verification.
Abstract
The demand for ecofriendly ships in shipbuilding and maritime industries is growing steadily. In response to strict environ-mental regulations on shipping by the International Maritime Organization, the demand for ship machinery and equipment that carry ecofriendly labels, including ballast water treatment systems (BWTS), is increasing. The BWTS involves early-stage equipment error and fault diagnostics, and its malfunction can have major cost and time consequences. This study expands on previous research on fault diagnosis using the SVM, a machine learning model. The two aspects of this expansion are an increased window size range used to generate the features and the introduction of several machine learning models. We used 47,435 sensor data points to compare and analyze the results and evaluate the classification accuracy by increasing the window size range to 10. We demonstrate that the previous model can be easily applied to other machine learning models and that the SVM model improves performance through feature gener-ation. The F1 score of the random forest model with the highest performance score of 99.73% indicates potential for industrial appli-cations if accompanied by expert monitoring and verification.
참고문헌 (20)
- "International Maritime Organization"[URL]
- 김은찬, "IMO 선박평형수 처리장치 기술과 발전 방향에 대한 고찰", 한국해양환경•에너지학회지, 15(4), pp.349~356 (2012), 한국해양환경·에너지학회[DOI]
- 손상호, "신개념 기계식 선박평형수 처리장치의 디스크 형상에 따른 유동특성에 관한 수치해석 연구", 한국마린엔지니어링학회지, 39(1), pp.19~27 (2015), 한국마린엔지니어링학회[DOI]
- A. Mouzakitis, "Classification of fault diagnosis methods for control systems", Measurement and Control, 46(10), pp.303~308 (2013)
- 서지노, "선박 평형수 처리 시스템의 고장 예측 및 e-Navigation을 이용한 고장 수리 시스템", 한국인터넷방송통신학회 논문지, 17(1), pp.145~151 (2017), 한국인터넷방송통신학회[DOI]
- A. Géron, "Hands-On Machine Learning with Scikit-Learn, Keras and Tensorflow: Concepts, Tools, and Techniques to Build Intelligence Systems" (2019), O’Relliy
- D. E. Rumelhart, "Learn-ing representations by back-propagating errors", Nature, 323, pp.533~536 (1986)
- C. Cortes, "Support-vector networks", Machine Learning, 20(3), pp.273~297 (1995)
- J. K. Kim, "Diagnosing faults in BWTS based on machine learning", International Journal of Computer Science and Network Security, 20(8), pp.104~111 (2020)
- B. D. Yoon, "Diagnosis and prognostic of engineering system sta-tus using artificial intelligence", Journal of the Korean So-ciety of Mechanical Engineers, 57(3), pp.38~41 (2017)
- C. M. Bishop, "Pattern Recognition and Machine Learning" (2006), Springer
- D. R. Cox, "The regression analysis of binary sequences", Journal of the Royal Statistical Society : Series B(Method-ological), 20(2), pp.215~232 (1958)
- N. S. Altman, "An introduction to kernel and nearest neigh-bor nonparametric regression", The American Statistician, 46(3), pp.175~185 (1992)
- F. Rosenblatt, "The perceptron : A probabilistic model for information storage and organization in the brain", Psycho-logical Review, 65(6), pp.386~408 (1958)
- L. Breiman, "Random forests", Machine Learning, 45, pp.5~32 (2001)
- C. Tianqi, "XGBoost: A scalable tree boost-ing system", Proceedings of the 22nd ACM SIGKDD Inter-national Conference on Knowledge Discovery and Data Mining, pp.785~794 (2016)
- "Scikit-learn: machine learning in Python & mdash; scikit-learn 0.24.2 documentation"[URL]
- "NumPy"[URL]
- "Pandas - Python Data Analysis Library"[URL]
- 박호민, "Deep learning–based drone detection with SWIR cameras", 한국마린엔지니어링학회지, 44(6), pp.500~506 (2020), 한국마린엔지니어링학회[DOI]