Citation

Heydecker B. Transp. Traffic Theory 1984; 9: 113-132.

Copyright

(Copyright © 1984, Publisher varies)

DOI

unavailable

PMID

unavailable

Abstract

Buses can experience considerable savings in delay at junctions where priority is provided for them by selective vehicle detection. However, these priority methods necessarily disrupt the normal operation of the traffic signals. This paper investigates the consequences of this for non-priority traffic. The methods developed here can be used to investigate proposed priority schemes in advance of their implementation. Formulae are identified which are appropriate to estimate the mean overflow and delay in a stream where the durations of the effective red and green periods are random variables. The streams at a junction can be divided into 7 classes, each of which can be analysed individually to estimate the statistics of service which are required to use the delay formula. A direct analysis of 3 of these classes provides sufficient information to deduce the required statistics for any stream. An embedded Markov chain is identified in the control process. This enables the analysis to proceed with the estimation of certain conditional expectations. One of the classes of stream is analysed in detail to estimate these quantities. This provides sufficient information to investigate a range of junctions. Example calculations are included for a junction based upon a real priority experiment. (Author/TRRL)