With advancements in modern technology and the increasing complexity of various dependable systems, rolling stock system reliability, risk, and maintenance targets are becoming increasingly performance-critical indicators for train operating companies. Rolling stock punctuality, in terms of public performance measured by the United Kingdom Office of Rail and Road, shows an increase in the number of train delays due to technical failures. In addition, rapid changes in technology increase the risk of human error, which could lead to an accident if a holistic view of risk management is not considered. Additionally, the individualised application of reliability, risk, and maintenance techniques for rolling stock system analyses increases the likelihood of the desired performance targets not being achieved. Therefore, this research is intended to develop a composite hybrid model as an integrated framework for optimising rolling stock reliability and maintenance strategies, using both classical steady-state techniques, dynamic probability-based models, and machine-learning techniques. Although the individual tools in the proposed framework are employed widely within the rolling stock sector and in various industries, the existing literature, combined with industry experience, suggests that the current applications of these tools are largely independent of each other. In addition to evaluating risk and reliability, the proposed hybrid model simultaneously allocates maintenance strategies to all subsystems based on the specified life cycle phase. In addition, hybrid probability-based models and machine learning techniques allow for comprehensive risk assessment and optimisation of the maintenance strategy for a railway rolling stock subsystem, which can mitigate the likelihood of rail accidents, such as derailments and collisions. All the proposed innovative hybrid models are illustrated through various case studies using critical rolling stock subsystems from the rail industry. The various cases studies indicate the hybrid framework performs significantly better than conventional reliability, risk, and maintenance methods for railway rolling stock subsystems.
Date of Award | 1 Aug 2022 |
---|
Original language | English |
---|
Awarding Institution | - The University of Manchester
|
---|
Supervisor | Moray Kidd (Supervisor) & Akilu Yunusa-Kaltungo (Supervisor) |
---|
- Railways
- Rolling Stock
- Reliability
- Safety
- Maintenance
Development of a composite hybrid framework for optimising the reliability of rolling stock subsystems
Appoh, F. (Author). 1 Aug 2022
Student thesis: Phd