Thesis

Accommodating maintenance in prognostics

Creator
Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2021
Thesis identifier
  • T15904
Person Identifier (Local)
  • 201261004
Qualification Level
Qualification Name
Department, School or Faculty
Abstract
  • Steam turbines are an important asset of nuclear power plants, and are required to operate reliably and efficiently. Unplanned outages have a significant impact on the ability of the plant to generate electricity. Therefore, condition-based maintenance (CBM) can be used for predictive and proactive maintenance to avoid unplanned outages while reducing operating costs and increasing the reliability and availability of the plant. In CBM, the information gathered can be interpreted for prognostics (the prediction of failure time or remaining useful life (RUL)). The aim of this project was to address two areas of challenges in prognostics, the selection of predictive technique and accommodation of post-maintenance effects, to improve the efficacy of prognostics. The selection of an appropriate predictive algorithm is a key activity for an effective development of prognostics. In this research, a formal approach for the evaluation and selection of predictive techniques is developed to facilitate a methodic selection process of predictive techniques by engineering experts. This approach is then implemented for a case study provided by the engineering experts. Therefore, as a result of formal evaluation, a probabilistic technique the Bayesian Linear Regression (BLR) and a non-probabilistic technique the Support Vector Regression (SVR) were selected for prognostics implementation. In this project, the knowledge of prognostics implementation is extended by including post maintenance affects into prognostics. Maintenance aims to restore a machine into a state where it is safe and reliable to operate while recovering the health of the machine. However, such activities result in introduction of uncertainties that are associated with predictions due to deviations in degradation model. Thus, affecting accuracy and efficacy of predictions. Therefore, such vulnerabilities must be addressed by incorporating the information from maintenance events for accurate and reliable predictions. This thesis presents two frameworks which are adapted for probabilistic and non-probabilistic prognostic techniques to accommodate maintenance. Two case studies: a real-world case study from a nuclear power plant in the UK and a synthetic case study which was generated based on the characteristics of a real-world case study are used for the implementation and validation of the frameworks. The results of the implementation hold a promise for predicting remaining useful life while accommodating maintenance repairs. Therefore, ensuring increased asset availability with higher reliability, maintenance cost effectiveness and operational safety.
Advisor / supervisor
  • McArthur, Stephen, 1971-
  • Catterson, Victoria
  • West, Graeme
Resource Type
Note
  • Error on title page - year of award is 2021
DOI
Date Created
  • 2020

Relations

Items