The aim of this report is to study the effectiveness of different stategies for setting out vehicle counting detectors at signalised intersections. Detectors are used for generating road data for a traffic forecast module. Using traffic forecasts as a basis, an optimisation of the time schedules for systems of signalised intersections was carried out with the TRANSYT programme. In order to rate how links in a traffic network feed adjacent links with traffic, a network model has been developed in this study. This model permits of flow balance at all intersections and an automatic dispersion of traffic in the various outward directions. Comparisons between different strategies concerning the use of existing detectors on two roads in the city of Uppsala produced the following result: (1) an extensive provision of detectors at intersections means sound national economics; (2) it is most important to put detectors in a position where they record the largest traffic flows; (3) it is favourable to use detectors at exits connected to different signal groups for the detection of flows from different directions if all vehicles passing the stop line are to be detected; and (4) experience shows that the work and costs of maintaining data on non-detected flows in the forecast module should be weighed carefully against the additional cost of installing more detectors, thus minimising the overall cost. (A)
Abstract