| Author: | Lam, Chun-man |
| Title: | Train wheel profile wearing analysis by machine vision |
| Degree: | M.Sc. |
| Year: | 1998 |
| Subject: | Railroads -- Trains -- Wheels -- Maintenance and repair Computer vision -- Industrial applications Image processing -- Digital techniques Hong Kong Polytechnic University -- Dissertations |
| Department: | Multi-disciplinary Studies Department of Electrical Engineering |
| Pages: | [86] leaves : ill. ; 31 cm |
| Language: | English |
| Abstract: | The maintenance schedules for electric train, in particular commuter train such as the EMUs that operated in Hong Kong KCRC, call for frequent examination and measurement of the underframe components and systems. The condition of train wheel profile is one critical measurement item because it would affect the train operational safety especially when it is running through curves. At present, wheel profile measurement is done by style type drawing instrument where the measured profile is drawn on paper. It is not easy to establish a long term monitoring system to store the past measurement records and keep track of the trend of the wheel profile wearing. The analyzing work of the profile wearing condition is mainly based on human judgment where inconsistency result will be obtained. If both the analyzing and measurement works of train wheel profile are automated at a suitably chosen site, for example, the train wash plant where the train visits every few days, it will lead potential improvement on the maintenance cost, train availability and reliability, and the risk of derailment would also be minimized. In this project, the main objective is to explore the feasibility of using machine vision techniques to monitor and analyze the wheel profile. Image processing / description technologies will be applied for analyzing the wheel profile wearing condition and identifing the critical wearing point. Steps on analyze the wheel profile were recommended. It involves the work of obtaining wheel profile image by either scanning or camera capturing, application of image processing techniques to separate the profile image from the background of the "raw" image and the application of curve fitting and image description techniques to identify the critical wearing point. The performance of the quadratic least square and B-spline techniques in curve fitting for the wheel profile application are almost the same. The quadratic least square method is concluded to be the best one because of its speed and acceptable accuracy. Bending energy is found to be a means of reflecting the severity of wheel wear but cannot help to identify the critical wearing point from the profile. The two critical wheel profile wearing point, the point of tread wear and point of flange wear, are found to be able to identify by detecting the first zero crossing point and the minimum point of the slope curve. Demonstration on using the camera to capture the image from the real scene with the application of geometric transformation to generate an image for wheel profile analysis has been carried out. It successfully demonstrates the image that captured by a camera at some restricted location could be used for further analysis after appropriate geometric transformation. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b1441868x.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.76 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/3245