Genetic Algorithm-Based Optimization Approach and Generic Tool for Calibrating Traffic Microscopic Simulation Parameters
- 1 January 2002
- journal article
- research article
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 1800 (1), 6-15
- https://doi.org/10.3141/1800-02
Abstract
GENOSIM is a generic traffic microsimulation parameter optimization tool that uses genetic algorithms and was implemented in the Port Area network in downtown Toronto, Canada. GENOSIM was developed as a pilot software as part of the pursuit of a fast, systematic, and robust calibration process. It employs the state of the art in combinatorial parametric optimization to automate the tedious task of hand calibrating traffic microsimulation models. The employed global search technique, genetic algorithms, can be integrated with any dynamic traffic microscopic simulation tool. In this research, Paramics, the microscopic traffic simulation platform currently adopted at the University of Toronto Intelligent Transportation Systems Centre, was used. Paramics consists of high-performance, cross-linked traffic models that have multiple user adjustable parameters. Genetic algorithms in GENOSIM manipulate the values of those control parameters and search for an optimal set of values that minimize the discrepancy between simulation output and real field data. Results obtained by replicating observed vehicle counts are promising.Keywords
This publication has 2 references indexed in Scilit:
- System identification and linearisation using genetic algorithms with simulated annealingPublished by Institution of Engineering and Technology (IET) ,1995
- System identification and control using genetic algorithmsIEEE Transactions on Systems, Man, and Cybernetics, 1992