Passenger comfort is a key factor in railway vehicles as it significantly influences passengers' feelings and experiences during travel. For evaluating and measuring passenger comfort when travelling by train, the vibration of a railway vehicle, which is transmitted via a car body to passengers, is known as whole-body vibration (WBV) and plays a vital role as a criterion for assessing the comfort of passengers. The railway system, including the vehicles (rolling stock) and infrastructure (track), is made up of key components which can affect the vibrations of railway vehicles. The key components of rolling stock and track quality deteriorate over time due to train operations, usage, and ageing which may affect the vibrations of railway vehicles and passenger comfort. When the key components of rolling stock and track quality are degraded, maintenance activities are needed to improve passenger comfort. The research aims to improve the understanding of the effect of railway maintenance on passenger comfort and to consider the possibility of linking the degradations of selected key components of rolling stock and track quality to passenger comfort and to investigate the potential of machine learning in identifying the effect of specific maintenance activities. The main achievements and contributions from the research include:1. A model linking the degradation of primary and secondary suspension components of the railway vehicle and track quality to passenger comfort was developed. Regarding the developed model, the simulated carbody accelerations representing vibrations of the railway vehicle transferred to passenger’s body were collected and analysed for passenger comfort evaluation according to the standards (EN12299 and ISO 2631-1). 2. An improved understanding of the effect of railway maintenance on passenger comfort was obtained regarding the analysis of accelerations in time domain and frequency domain relating to degradation of suspension components (springs and dampers) of the railway vehicle and track quality. This work revealed that degradation of secondary dampers running on the poorest track quality affect passenger comfort the most. 3.The machine learning technique was developed using carbody accelerations and proposed to predict and prioritise the fault of suspension components of the railway vehicle linking passenger comfort for specific maintenance activities. This machine learning technique could be implemented as an onboard condition-based monitoring system in trains to alert for specific maintenance activities such as inspection and replacement.
| Date of Award | 3 Jul 2026 |
|---|
| Original language | English |
|---|
| Supervisor | Adam Bevan (Main Supervisor) & Hassna Louadah (Co-Supervisor) |
|---|