The invention relates to a central type dynamic path inducing method based on Sarsa learning. The method can calculate all Q value tables of probable destinations through Sarsa learning to provide route guidance for all vehicles. The step of establishing the Q value tables specifically comprises the following two steps: I, initializing the Q value table of each destination by means of a dynamic programming algorithm based on Q value by a communication center according to road network information and a previous road section transit time; II, updating the Q value tables by using a Sarsa learning method at an interval of T by the communication information center according to vehicle real-time information gathered by using a technique of Internet of vehicles; reading the Q value of corresponding road section in the Q value table by the communication information center according to position and destination information of the vehicle, and distributing road section for the vehicle by combining Boltzmann distribution of global and local parameters so as to realize path guidance. The method provided by the invention can calculate Q value tables for all probable destinations by using the Sarsa learning method, so that the optimum path guidance is provided for all vehicles.


    Access

    Download


    Export, share and cite



    Title :

    Central type dynamic path inducing method based on Sarsa learning


    Contributors:
    WEN FENG (author) / WANG XINGQIAO (author) / MIAO WEIPING (author) / YU YANG (author) / HUANG HAIXIN (author)

    Publication date :

    2015-05-27


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    English


    Classification :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS



    Cruise dynamic pricing based on SARSA algorithm

    Wang, Jing / Yang, Dong / Chen, Kaimin et al. | Taylor & Francis Verlag | 2021


    Hierarchical Sarsa Learning Based Route Guidance Algorithm

    Feng Wen / Xingqiao Wang / Xiaowei Xu | DOAJ | 2019

    Free access


    SARSA-based trunk line traffic control optimization method

    LIN JING / WEI PING / CHAI ZIHUI et al. | European Patent Office | 2021

    Free access

    Safe Reinforcement Learning for Single Train Trajectory Optimization via Shield SARSA

    Zhao, Zicong / Xun, Jing / Wen, Xuguang et al. | IEEE | 2023