Combination of reliability tools and artificial intelligence in a hybrid condition monitoring framework for ship machinery systems

Inadequate ship machinery maintenance can increase equipment failure posing a threat to the environment, affecting performance, having a great impact in terms of business losses by reducing ship availability, increasing downtime and moreover increasing the potential of major accidents occurring and endangering lives on-board. With high cost of ownership and overburdened crew, ship maintenance has become one of the major challenges in the marine industry. Though the industry is still predominantly reliant on a time-based, prescriptive approach to maintenance, technological advances, heightened expectation and competitive requirements as to ship availability and efficiency and the influence of the data revolution on vessel operations, have resulted in considerable interest in advanced maintenance techniques and favour a properly structured condition-based maintenance regime. In this respect, this thesis develops a hybrid framework oriented towards ship machinery condition monitoring utilising a combination of reliability tools (Fault Tree Analysis, Failure Modes & Effects Analysis, Reliability Block Diagrams) and data-driven approaches based on artificial neural networks (Self-Organising Maps, Nonlinear Autoregressive, Multilayer Perceptron). The above assist in identifying critical ship machinery systems and components and subsequently monitoring their condition through the employment of data clustering, time series forecasting, diagnostic and health assessment, leading to advisory generation of appropriate maintenance actions and recommendations. The above framework is applied to the case study of a Panamax container ship main engine for system, subsystem and component level and the results are validated with actual data recorded onboard. Sensitivity and cost benefit analysis are also presented. Key results include amongst others the identification of critical systems through a systematic approach, the ability of the Self-Organising Map to cluster data and monitor the status of the main engine and the forecasting capabilities of the Nonlinear Autoregressive time series neural networks to analyse available main engine data with high forecasting accuracy.Keywords: Artificial neural networks, data analysis, reliability tools, condition monitoring, predictive maintenance, maritime industry

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral-pg

Grantor dc:publisher.institution

University of Strathclyde

Year dc:date.issued

2018

Author dc:creator

• Raptodimos, Yiannis

Advisor dc:contributor.advisor

• Lazakis, Iraklis

DOI

10.48730/a7t1-7j63

Identifier

T15094

Author Identifier

201460844

OAI identifier oai:identifier

oai:strathclyde:w0892994f