On land transportation vehicles play major role in governmental, nongovernmental organization and individuals for moving goods and peoples. Which used to increase, economy development on the individuals and nations, quality of education, healthcare and other benefit’s. But due to vehicles accident countries loss big amount of capital, manpower, and additionally influence negative impact on the healthcare. Vehicle accident injures all age groups. This research is done to provide way to solve this existing problem through sensor fusion. Information fusion provides an important role of any system electrical, electromechanical, biomedical, and any other filed of science and technology. In this research, the main parameters are heartbeat, body temperature, speed and load of the vehicle. The mathematical models of heartbeat are selected by comparing different models from past research [1], [2], [3], [4] and [5]. And the model of body temperature is also selected from past research [6], [7] and [8]. The speed is investigated based on Newton second law of motion on upward, down ward, and on the strait line. Since most of Loads of the vehicle are big, four active metallic strain gauges are used to measure the load of the vehicle. Body temperature of the driver, speed and load of the vehicle are filtered by extended kalman filter. Due to the hard nonlinearity property of heartbeat, it did not filter by extended kalman filter. Then dynamic and linear artificial neural network are designed to track the biological status of the driver and vehicle dynamics respectively. Finally, hierarchical structure, of Bayesian sensor fusion technique is used. These are checked by three Coues that are low, medium and high risk of accident depending on the average traffic accident rule and regulation


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    SENSOR FUSION TO CONTROL VEHICLE DRIVER THROUGH A RECURRENT NEURAL NETWORK


    Beteiligte:
    TALEGETA MOGES (Autor:in)

    Erscheinungsdatum :

    2019-05-17


    Anmerkungen:

    doi:10.20372/nadre/4004



    Medientyp :

    Hochschulschrift


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Klassifikation :

    DDC:    629




    ENHANCED MULTI-SENSOR DATA FUSION METHOD USING RECURRENT NEURAL NETWORK

    He, Jing / Guo, Chengjun / Tao, Chao | TIBKAT | 2021





    Vehicle motion prediction control method based on recurrent neural network

    WU XIALAI / LIN LING / YI JIAN et al. | Europäisches Patentamt | 2023

    Freier Zugriff