Citation

Gonzalez EA, Loutzenheiser RC, Carter EC. Transp. Res. Rec. 1975; 533: 97-109.

Copyright

(Copyright © 1975, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

As part of a continuing research study directed to alleviate traffic congestion in the Baltimore Harbor Tunnel, the Howard policy-iteration technique was applied off-line to different traffic control alternatives. Three data- acquisition stations were used inside the tunnel for control purposes. One was at the tunnel's bottleneck, and 2 were upstream of this location. The value of the traffic concentration, which was used as the control variable, at these 3 stations defined 1 out of 18 possible states of tunnel traffic flow. State transitions in the system were assumed to occur every 30 sec. Extensive data collected in the Baltimore Harbor Tunnel was used to determine the state transition probabilities of the system under each alternative. The rewards associated with state transition were obtained by applying a model that considers actual flow at the bottleneck and average speed associated with the flow during each transition interval. Five different control alternatives were considered. One of the alternatives was no control; the remaining 4 were 2- to 4-min cycle lengths of a traffic signal located upstream of the tunnel entrance.