The road environments of the future smart cities will be more digitized and connected. Many entities and applications within this highly digitized urban environment are to communicate with one another, in order to realize intelligent decision-making processes. The fact that most of the smart city applications and devices are connected that provisions uninterrupted communication bit-pipes places Information and Communication Technologies (ICT) and network service providers in a very important position. Hence, many decision instances, which to-date did not need to take the ICT into account, now strongly depend on the availability of communication services. In the face of recent technological developments, it becomes evident that autonomous and connected vehicles will be an integral part of this smart city ecosystem and Cooperative Intelligent Transport Systems (C-ITS). When widely adopted and operated as an on-demand mobility service, the envisioned fully autonomous and connected vehicles will bring benefits such as increased road safety, transport efficiency, and passenger comfort. Like many other smart city services, the seamless operation of automated and connected driving applications, i.e., vehicle platooning, advanced driving, extended sensors, and remote driving, will require reliable and uninterrupted connectivity. One way to realize this would be to capture the interrelation between the route planning component of autonomous and connected vehicles and the network management system of communication service providers. The planning and control instances deployed at both ends of this interrelation would mutually benefit each other. By using the advantage of the hierarchical and deterministic route and motion planning mechanism of autonomous vehicles, the network resources can be proactively allocated along the vehicles’ planned trajectories. This preemptively initiated resource allocation process based on the programmatically determined routes of autonomous and connected vehicles would definitely contribute to achieving connectivity with the required level of Quality of Service (QoS). Complementing the trajectory-based network management mechanism, the autonomous vehicles’ route planning component can take into account the communication network status information. In this way, the routes of vehicles are to be optimized and adjusted by considering both traffic-related and communication network-related conditions, which is addressed in this work. The network-aware route planning mechanism would enable vehicles to travel along roads with better network availability and higher signal strength. In this direction, we design and implement a modular framework that allows users to model an urban scenario comprising the C-ITS elements such as autonomous and connected vehicles, wireless access points, dynamic data storage entity. By means of the established simulation environment, (near) real-time traffic and network status information can be collected from the modeled road environment. Using this dynamic data collected from the urban scenario, a group of vehicle instances representing autonomous and connected vehicles are rerouted based on the Multi-Objective Evolutionary Algorithm (MOEA) approach. The developed modular framework enables to import and remove different objectives and constraints easily to/from the route optimization problem model. In this way, different route optimization problem instances can be defined and solved depending on the modeled scenario. The problem objectives include conventional metrics such as traveled distance, travel time, and congestion level on the roads. Additionally, communication network quality information can be incorporated into the optimization problem model as an unconventional metric. An adjustable mix of different objectives and constraints can be considered in the dynamic route planning process. A dynamic data storage entity, representing the Local Dynamic Map (LDM) component of the C-ITS, is implemented and extended with Network Context (NC) object. The provided network data layer enables the maintenance of (near) real-time network status information on the road environment. In this manner, the network conditions are taken into account by the route planning methodology, which is highly critical for the seamless operation of many automated vehicle applications such as vehicle platooning, advanced driving, extended sensors, and remote driving. A mechanism to calculate the projected congestion contribution on each road segment is implemented, which reflects the future congestion levels on the roads. Considering this additional metric as an objective or constraint would contribute to achieving global traffic optimization rather than routing the vehicles in a greedy way. The Multi-Objective Evolutionary Algorithms (MOEAs) used to reroute the vehicles, such as NSGA-2 and NSGA-3, efficiently generate a set of optimal and trade-off solutions for the vehicles.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    A multi-objective route planning framework for automated and connected vehicles


    Weitere Titelangaben:

    Ein Multi-Objective-Routenplanungs-Framework für automatisierte und vernetzte Fahrzeuge


    Beteiligte:
    Bila, Cem (Autor:in) / Technische Universität Berlin (Gastgebende Institution)

    Erscheinungsdatum :

    2022



    Medientyp :

    Sonstige


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Klassifikation :

    DDC:    000





    Route planning of automated guided vehicles

    Hsu Wen Jing / Huang Shell Ying | IEEE | 1994


    Eco-Driving System for Connected Automated Vehicles: Multi-Objective Trajectory Optimization

    Yang, Xianfeng Terry / Huang, Ke / Zhang, Zhehao et al. | IEEE | 2021


    Route planning of automated guided vehicles

    Jing Hsu Wen / Ying, Huang-Shell | Tema Archiv | 1994


    Route Planning of Automated Guided Vehicles

    Jing, H. W. / Ying, H. S. / IEEE; Industrial Electronics Society | British Library Conference Proceedings | 1994