Genetic algorithm optimizations of traffic signal timings have been shown to be effective, continually outperforming traditional optimization tools such as Synchro and TRANSYT-7F. However, their application has been limited to scholarly research and evaluations. Only one tool has matured to a commercial deployment: direct CorSim optimization, a feature of TRANSYT-7F. A genetic algorithm formulation, VisSim-based genetic algorithm optimization of signal timings (VISGAOST), is presented; it builds on the best of the recorded methods by extending their capabilities. It optimizes four basic signal timing parameters with VisSim microsimulation as an evaluation environment. The program brings new optimization features not available in the direct CorSim optimization, such as the optimization of phasing sequences, multiple coordinated systems and uncoordinated intersections, fully actuated isolated intersections, and multiple time periods. The formulation has two features that enhance and reduce computational time: optimization resumption and parallel computing. The program has been tested on two VisSim networks: a hypothetical grid network and a real-world arterial of actuated–coordinated intersections in Park City, Utah. The results show that timing plans optimized by the genetic algorithm outperformed the best Synchro plans in both cases, reducing delay and stops by at least 5%.


    Access

    Download

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    VisSim-Based Genetic Algorithm Optimization of Signal Timings


    Additional title:

    Transportation Research Record


    Contributors:


    Publication date :

    2007-01-01




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English



    VisSim-Based Genetic Algorithm Optimization of Signal Timings

    Stevanovic, Aleksandar | Online Contents | 2007




    Genetic algorithm calibration method based on VISSIM and C #

    WANG JING / REN JIABO | European Patent Office | 2023

    Free access