OPTIMIZATION METHODS FOR REAL-TIME TRAFFIC CONTROL
In this dissertation, the author describes the development, testing and evaluation of three optimization techniques for real-time traffic control. These include: the Quasi-Newton methodology, multivariate genetic algorithms, and univariate genetic algorithms. These techniques are tested and evaluated in different traffic scenarios including varying traffic flow conditions and different phase sequencing alternatives. Total stopped delays is the criterion used as the measure of effectiveness (MOE. Minimization of total stopped delay is the criterion used for the optimization process, and the effectiveness of each optimization technique is compared on the basis of percent improvement in MOE.
- Publication Date: 1995 UMI Company, Ann Arbor MI Remarks: Thesis (Ph. D.)--University of Pittsburgh, 1995
University of PittsburghDepartment of Electrical and Computer Engineering
Pittsburgh, PA United States 15261
- Memon, Ghulam Qadir
- Publication Date: 1995
- Pagination: xxi, 154 p.
- TRT Terms: Computer algorithms; Real time control; Traffic control; Traffic signals
- Subject Areas: Operations and Traffic Management;
- Accession Number: 00793382
- Record Type: Publication
- Source Agency: UC Berkeley Transportation Library
- Report/Paper Numbers: 96-13812 (UMI Order #)
- Files: PATH
- Created Date: Jun 13 2000 12:00AM