Development of a parallel simulation algorithm for freeway traffic flows on distributed personal computer system,Phase 1

Principal Investigator:

Eil Kwon, Director, NATSRL, UMD-Civil Engineering

Co-Investigator

  • Haesun Park, Professor, Computer Science and Engineering

Project Summary:

Determining operational strategies for freeway management requires efficient computational algorithms that can generate control solutions in real time. While there has been substantial progress in developing traffic flow models during past decades, there still exists the lack of efficient, on-line computational environment that can be easily accessed by practicing engineers. To be sure, most existing freeway ramp metering systems use automatic rate selection strategies, which select appropriate metering rates in real time without employing rigorous calculation. Further, simulation is still being used mostly as a tool for off-line analysis of traffic flow. Although some parallel processing algorithms have been developed by various researchers for real time simulation, most of them require either main-frame computer systems equipped with multiprocessors or a network of workstations, whose costs and accessibility make the employment of parallel processing practically unfeasible to traffic engineers. The proposed research will investigate the feasibility of developing low-cost, personal computer-based parallel processing procedures that can be applicable to real time simulation of freeway flows using distributed computing technique. First, a distributed computing environment will be developed by linking PCs with network software. Secondly, the structure of existing freeway traffic models will be modified and algorithms will be developed for distributed computing. The scalability, of the algorithm and the reconfigurability of the connection topology will be addressed so that the system can be easily expanded in the future to handle larger problems with more processing units. A prototype version of the distributed simulation system will then be developed following the parallel, distributed algorithm. Finally, the performance of the prototype version will be extensively tested and its effectiveness/future potential will be analyzed.

Project Details:

  • Start date: 11/1994
  • Project Status: Completed
  • Research Area: Transportation Safety and Traffic Flow
  • Topics: Traffic Modeling and Data