A computer simulation which models the behaviour of traffic at a signal-controlled intersection is described. The approach road may include a bus priority lane, the end of which can be set back from the intersection. Calibration of the model is explained, and the model is successfully validated against two real situations. The model is used to determine the community benefit provided by the bus lane in a large scale experimental situation, and also in a working bus priority scheme. The benefit is evaluated for saturated traffic flows, and for flows below saturation, and recommendations are made for the design of bus lanes in real situations. Guidelines for selection of the optimum setback are suggested for situations where the flow is constantly undersaturated, constantly saturated, and where the flow rises to a peak, so that the intersection runs at saturation for a limited time. The response of the system to changes in the proportion of turning traffic, different signal phasing times and bus flows, is investigated. The bus lane is found to have no effect on the bunching of buses at the intersection. (Author/publisher)
Abstract