In this thesis a validation methodology for vehicle dynamics simulation models and its application is presented. The developed validation paradigm has a top-down approach to the problem. First, the term "validation" is explored and defined. It is established that a simulation model can never be absolutely valid, since a simulation is only an imitation of the reality, but can only be "not invalid". According to this axiom, it is concluded that a simulation model's validity is dependent on the application for which it is aimed. Next, it is ascertained that vehicle dynamics simulation models can only be validated using test maneuvers although they are aimed for real life maneuvers. Consequently the target real life events should be analyzed in terms of requirements and simulation targets should be determined. According to these requirements, a group of test maneuvers which reveal the dynamic characteristics of the vehicle and exhibit similar characteristics to those of the real event should be selected at the start of the model development project. The selected maneuvers should separately be analyzed, objective techniques to handle the data be explicitly defined and validation metrics and criteria be declared. The experimental data and simulation data are handled according to these guidelines, and the results are compared according to the defined validity criteria. If the simulation results and metrics satisfy the criteria, then the simulation is deemed to be "not invalid" and is corroborated to be used in the planned task. If the simulation model fails to meet one or more of the defined criteria, the model is deemed invalid, and model iteration should be performed. The results are analyzed to determine if the results indicate a modeling error or a modeling inadequacy; and if a conditional validity in terms of system variables can be defined. Three test cases are presented to demonstrate the application of the paradigm. Each maneuver is separately handled, data analysis methodologies are explained, metrics and accuracy criteria are defined, implemented and results are presented. New definitions and calculation techniques for reference points, such as the alignment point for the step response maneuver, or the partitioning point for the double lane change maneuver are introduced, which can help automating the assessment procedure. The developed methodology successfully identified the shortcomings of the tested simulation model, and defined the limits of application. Several insights for the deficiencies of the model are reported in the analysis but the iteration step of the methodology is not demonstrated. All in all, the approach offers a step-by-step procedural methodology for the simulation engineer. Utilizing the proposed methodology will help to achieve more time and cost efficient simulation projects with increased model confidence by enhancing the traceability of the validation process.


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Development and demonstration of a validation methodology for vehicle lateral dynamics simulation models


    Beteiligte:
    Kutluay, Emir (Autor:in)


    Erscheinungsdatum :

    2012


    Format / Umfang :

    126 Seiten, 42 Bilder, 22 Tabellen, 101 Quellen




    Medientyp :

    Hochschulschrift


    Format :

    Print


    Sprache :

    Englisch






    Validation of Railway Vehicle Lateral Dynamics Models

    GOSTLING, R. J. / COOPERRIDER, N. | Taylor & Francis Verlag | 1983


    Methodology for Development and Validation of Multipurpose Simulation Models

    Carlsson, Magnus / Andersson, Henric / Gavel, Hampus et al. | AIAA | 2012


    Methodology for Development and Validation of Multipurpose Simulation Models

    Carlsson, M. / Andersson, H. / Gavel, H. et al. | British Library Conference Proceedings | 2012