For motion planning and control of autonomous vehicles to be proactive and safe, pedestrians' and other road users' motions must be considered. In this paper, we present a vehicle motion planning and control framework, based on Model Predictive Control, accounting for moving obstacles. Measured pedestrian states are fed into a prediction layer which translates each pedestrians' predicted motion into constraints for the MPC problem. Simulations and experimental validation were performed with simulated crossing pedestrians to show the performance of the framework. Experimental results show that the controller is stable even under significant input delays, while still maintaining very low computational times. In addition, real pedestrian data was used to further validate the developed framework in simulations.


    Access

    Download


    Export, share and cite



    Title :

    Real-Time Constrained Trajectory Planning and Vehicle Control for Proactive Autonomous Driving With Road Users


    Contributors:

    Publication date :

    2019


    Remarks:

    7 pages, 9 figures, accepted to European Control Conference (ECC) 2019


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English




    FOCUSED TRAJECTORY PLANNING FOR AUTONOMOUS ON-ROAD DRIVING

    Gu, T. / Snider, J. / Dolan, J. et al. | British Library Conference Proceedings | 2013


    Driving-behavior-oriented trajectory planning for autonomous vehicle driving on urban structural road

    Zeng, Dequan / Yu, Zhuoping / Xiong, Lu et al. | SAGE Publications | 2021


    Trajectory Planning of Autonomous Driving Vehicles Based on Road-Vehicle Fusion

    Li, Han / Yu, Guizhen / Zhou, Bin et al. | ASCE | 2020


    Focused Trajectory Planning for autonomous on-road driving

    Gu, Tianyu / Snider, Jarrod / Dolan, John M. et al. | IEEE | 2013


    Dynamic Trajectory Planning for Vehicle Autonomous Driving

    Zhang, Sumin / Deng, Weiwen / Zhao, Qingrong et al. | IEEE | 2013