The increasing number of traffic lights on main roads and the resulting delay for main road traffic, is the reason why traffic engineers are restudying the influence of traffic lights on traffic flow and the alternative: i.e. the priority rule operation. In the past, several methods have been put forward to calculate the capacity of priority junctions. It is however not clear which method gives the best results. In this paper a number of experiments will be discussed to evaluate the accuracy of six models used to calculate capacity and waiting times; the models are: - Siegloch, W. - Jessen, G.D. - Harders, - "Kantonales Tiefbauamt Solothurn", - "Forschungsgesellschaft fur Strassen- und Verkehrswesen", - Picady. The parameters used in these models (except Picady), can be measured in several ways, which in turn can affect the results of the calculations. Five methods have been used. This results in 26 sets used to calculate capacities, waiting times and queue length. The experimental data were gathered at 3 t-junctions and one four-legged cross-road. One of the main requirements for choosing the junction is that there must be a proper lane for every single movement, because the calculation methods are only valid under those circumstances. Conclusions have been reached by comparing the experimental data and the calculated capacities. Linear regression has been used to calculate the correlation coefficient. For the acquisition of data, extensive use has been made of video recordings. The following models give good results: - harders combined with different methods to measure the parameters; - "Forschungsgesellschaft fur Strassen- und Verkehrswesen"; - Siegloch and Jessen, but only for one method. Results obtained by the "Kantonales Tiefbauamt Solothurn"-model and the Picady-model are less reliable. For the covering abstract of the seminar see IRRD 273510.
Abstract