Identification of a railway wheelset system parameters
Wheelset is the basic component of a railway vehicle system. The wheel set has two wheels rigidly connected by an axle. There are mainly two non-constant values related to the wheelset that need to be estimated. They are the wheelset’s conicity and creep coefficients. Conicity (X) is a term related...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2010
|
| Subjects: |
| _version_ | 1796856290334474240 |
|---|---|
| author | Nazari, Ain |
| author_facet | Nazari, Ain |
| author_sort | Nazari, Ain |
| collection | ePrints |
| description | Wheelset is the basic component of a railway vehicle system. The wheel set has two wheels rigidly connected by an axle. There are mainly two non-constant values related to the wheelset that need to be estimated. They are the wheelset’s conicity and creep coefficients. Conicity (X) is a term related to the coning of the wheel tread and its value depends on the track and wheel profiles. For wheelset with profiled wheels, the conicity values is non-linear and depends on the relationship between the wheelset’s rolling radius difference and its lateral displacement Meanwhile, the creep coefficients (fit and fn) are related to the relative velocities of the wheel and the rail. This creep coefficient is non-linearly dependent on the normal force between the wheel and rail, and the value of this force is changing especially during curving. The values of the conicity and creep coefficients as well can be changed by other factors such as surface contamination, varying surface condition and railhead shapes. This project outlines the parameter estimator that can effectively estimate and track the time-varying conicity and creep coefficients. The performances of the Recursive Least Squares (RLS) and Recursive Instrumental Variable (RIV) used to estimate the conicity and creep coefficients values are compared. Via simulations, it is shown that the conicity and creep coefficients estimation using RLS algorithm takes longer time to converge to the true values and at the same time contains large amount of bias and oscillation. The RIV method overcome RLS problem in term of convergence and bias but the overshoot produced by this method is larger than RLS method. |
| first_indexed | 2024-03-05T18:40:51Z |
| format | Thesis |
| id | utm.eprints-26780 |
| institution | Universiti Teknologi Malaysia - ePrints |
| last_indexed | 2024-03-05T18:40:51Z |
| publishDate | 2010 |
| record_format | dspace |
| spelling | utm.eprints-267802017-08-21T01:28:28Z http://eprints.utm.my/26780/ Identification of a railway wheelset system parameters Nazari, Ain TK Electrical engineering. Electronics Nuclear engineering Wheelset is the basic component of a railway vehicle system. The wheel set has two wheels rigidly connected by an axle. There are mainly two non-constant values related to the wheelset that need to be estimated. They are the wheelset’s conicity and creep coefficients. Conicity (X) is a term related to the coning of the wheel tread and its value depends on the track and wheel profiles. For wheelset with profiled wheels, the conicity values is non-linear and depends on the relationship between the wheelset’s rolling radius difference and its lateral displacement Meanwhile, the creep coefficients (fit and fn) are related to the relative velocities of the wheel and the rail. This creep coefficient is non-linearly dependent on the normal force between the wheel and rail, and the value of this force is changing especially during curving. The values of the conicity and creep coefficients as well can be changed by other factors such as surface contamination, varying surface condition and railhead shapes. This project outlines the parameter estimator that can effectively estimate and track the time-varying conicity and creep coefficients. The performances of the Recursive Least Squares (RLS) and Recursive Instrumental Variable (RIV) used to estimate the conicity and creep coefficients values are compared. Via simulations, it is shown that the conicity and creep coefficients estimation using RLS algorithm takes longer time to converge to the true values and at the same time contains large amount of bias and oscillation. The RIV method overcome RLS problem in term of convergence and bias but the overshoot produced by this method is larger than RLS method. 2010 Thesis NonPeerReviewed Nazari, Ain (2010) Identification of a railway wheelset system parameters. Masters thesis, Universiti Teknologi Malaysia, Faculty of Electrical Engineering. http://libraryopac.utm.my/client/en_AU/main/search/results?qu=Identification+of+a+railway+wheelset+system+parameters&te= |
| spellingShingle | TK Electrical engineering. Electronics Nuclear engineering Nazari, Ain Identification of a railway wheelset system parameters |
| title | Identification of a railway wheelset system parameters |
| title_full | Identification of a railway wheelset system parameters |
| title_fullStr | Identification of a railway wheelset system parameters |
| title_full_unstemmed | Identification of a railway wheelset system parameters |
| title_short | Identification of a railway wheelset system parameters |
| title_sort | identification of a railway wheelset system parameters |
| topic | TK Electrical engineering. Electronics Nuclear engineering |
| work_keys_str_mv | AT nazariain identificationofarailwaywheelsetsystemparameters |