This paper proposes a method that is based on ANN for monitoring of the vehicle behavior. Considering the control loop of driver-vehicle-environment a driver should perceive the environment and the vehicle behavior by processing received information from the environment and feedback from the vehicle. The precession of the driver's percipience is the critical element in such case. In this study, an ANN is applied for perception and prediction of the vehicle dynamic performance. Several relevant parameters from the vehicle and the environment, such as accelerator pedal travel and road grade, serve as information for the prediction. After training of the network with the measured data from a test vehicle, the network will be used for prediction of the driving speed. The comparison of the measured driving speed with the predicted speed can indicate the actual performance of the vehicle. The advantages of applying ANN for vehicle modeling are: 1) The complexity of vehicle modeling using physical models, which are base on structural mechanics and other physical processes, can be avoided. 2) The ANN with the real-time learning capability can model the system that has nonlinearly changing parameters. Thus, the model of the system can be updated in real-time. With this experimental study, it is proved that using the ANN for modeling of the vehicle longitudinal dynamics is feasible. This work has presented that this new approach is widely applicable for monitoring the vehicle dynamic performance. Further work will concentrate on optimization of the networks, investigation of the network real-time learning capability and extending the application area of this approach.


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Attempt of prediction of vehicle longitudinal dynamic performance using Artificial Neural Networks (ANN)


    Beteiligte:
    Albers, Albert (Autor:in) / Ott, Sascha (Autor:in) / Wang, Jiangang (Autor:in)


    Erscheinungsdatum :

    2007


    Format / Umfang :

    5 Seiten, 7 Bilder, 22 Quellen



    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Print


    Sprache :

    Englisch






    Attempt at Resilient Modulus Modeling Using Artificial Neural Networks

    Tutumluer, Erol / Meier, Roger W. | Transportation Research Record | 1996


    Attempt at Resilient Modulus Modeling Using Artificial Neural Networks

    Tutumluer, Erol / Meier, Roger | Transportation Research Record | 1996


    Attempt of monitoring vehicle longitudinal dynamic with simulation technique

    Albers, A. / Duser, T. / Seifermann, A. et al. | British Library Conference Proceedings | 2008