A new parameter identification methodology for railway vehicle models based on stationary test results
- Autores: Miguel Aizpun Navarro
- Directores de la Tesis: Jordi Vinolas Prat (dir. tes.), Asier Alonso Pazos (codir. tes.)
- Lectura: En la Universidad de Navarra ( España ) en 2013
- Idioma: español
- Tribunal Calificador de la Tesis: José Germán Giménez Ortiz (presid.), Javier Eduardo Pérez Sarasola (secret.), Eduardo Gómez Martín (voc.), Sin Sin Hsu (voc.), Yann Bezin (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
The work carried out in this thesis is focused on the improvement of the validation process of the mathematical models of railway vehicles. Due to the increasing interest in including dynamic simulations in the acceptance process of new vehicles, it is essential to develop mathematical models which ensure the reliability of those simulations.
Therefore, the objective of this thesis is to develop a parameter identification methodology in order to obtain accurate estimations of the vehicle model parameters by means of the results of the stationary tests in the acceptance process.
Moreover, this methodology allows for the probabilistic calculation of the model parameters by estimating confidence intervals or uncertainty margins for those parameters. These parameters uncertainties are caused by the measurement uncertainties of the sensors used in the acceptance tests.
The method has been developed through a MATLAB Graphic User Interface, with the purpose of allowing the railway engineers to use it during the vehicle acceptance process.
In the first part of the methodology, the sensitivity analysis, several tools for assessing the importance of the model parameters in each of the tests are shown. The analysis results indicate which parameters could be identified in each test.
The main part of the methodology, the iterative method, is a process that efficiently estimates several model parameters based on the measured results of the acceptance tests.
Lastly, the methodology is completed with an uncertainty reduction procedure. This process significantly reduces the uncertainty margins of the identified parameters, provided that the results from several test runs are available.
Regarding the validation of the methodology, virtual verifications were performed and then followed by an experimental validation using experimental data from several tests.
The virtual validation was carried out by applying the methodology to the sway and wheel unloading tests, analysing the parameters that could be identified and their uncertainty margins.
The methodology was validated by using experimental data from two other tests. In the first one, the bogie rotational resistance test, several parameters were identified by means of a simplified version of the identification methodology. In the second test, the vehicle modal analysis test, a modified methodology was applied and satisfactory results for several vehicle model parameters were obtained.
