Citation

Hadi MA, Wallace CE. Transp. Res. Rec. 1994; 1456: 64-71.

Copyright

(Copyright © 1994, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Off-line optimization of signal timing plans involves the selection of cycle length, phase sequences, offsets, and green splits. None of the available signal timing optimization models explicitly optimizes all four parameters, and a combination of these models is usually used to obtain a total optimal design. One of the most effective programs used to optimize traffic signal timing is TRANSYT-7F. A major limitation of the program, however, is its inability to select signal phase sequences. Previous research applied genetic algorithms to phase sequence and timing optimization. This approach was functionally promising, but (at least in its experimental version) not very computationally efficient. The possibility of introducing a phase sequence optimization capability to TRANSYT-7F using the Cauchy simulated annealing algorithm was investigated. This is an optimization technique that makes an analogy between optimization problems and the annealing of physical solids. The simulated annealing algorithm is implemented to optimize cycle length, phase sequences, and offsets simultaneously on the basis of the progression opportunities calculated by TRANSYT-7F. The results suggest that the algorithm has potential for optimizing signal phasing and timing for arterial streets as well as multiarterial networks. The model is recommended for implementation in a future version of TRANSYT-7F, further advancing the utility of this important traffic signal timing tool.

Language: en

Keywords

Design; Mathematical models; Traffic signals; Algorithms; Optimization; Performance; Roads and streets; Iterative methods; Simulated annealing