The capacity of unsignalised intersections is generally based on a gap acceptance relationship in which the entering drivers are assumed to accept a gap greater than the critical gap and reject all those less than the critical gap. The entering drivers are assumed to enter the intersection at headways of the follow-on time if the gap in the conflicting streams is long. The analysis of unsignalised intersections has been developed from the early work of Tanner and extended by Tanner, Cowan, Troutbeck and others. Others have used a M/G/1 process to represent the effect of this gap acceptance behaviour. Others, notably the Swedish Capacity Manual, have used a M/G2/1 queueing process to represent behaviour at unsignalised represent behaviour at unsignalised intersections. The M/G2/1 process has two generalised service time distributions and is thought to give better results. This paper describes the qualities of the service time distribution and compares the distribution with results considered to be appropriate in the literature. The usefulness of the M/G2/1 distribution over the M/G/1 distribution will be discussed. (A)
Samenvatting