Abstract:
The dispersion of vehicle platoons as they progress along a street from a traffic signal is very important in determining the optimum fixed-time signal offset settings. A traffic model developed by Robertson for predicting the dispersion of platoons of vehicles released from a traffic signal, to be used in the signal offset setting model TRANSYT, was tested to see whether it was applicable for use under Bandung traffic conditions. The most significant characteristic of traffic in Bandung was the substantial number of motorcycles.
Platoons of cars and motorcycles were first treated as two different platoons, each dispersing separately as they moved along two separate but parallel links. Parameters observed were the flows in successive time slices at the entry to the link and at stations 100, 200, and 300 metres downstream. The measured average car and motorcycle flow histograms of the study site links were compared with the flow histograms predicted by Robertsons formula. The formula provided a good prediction of the profiles of the dispersed platoon of cars as well as motorcycles.
The observed average cruise times of cars and motor-cycles showed that they tended to have a common mean journey time on short links, but deviations might occur. Treating cars and motorcycles as one platoon at the sites where they had different average cruise times showed that Robertsons formula was sensitive to differences between short average cruise times for the two vehicle categories, but not to any differences between long average cruise times. Tests showed that treating cars and motorcycles together as one platoon in terms of p.c.u./hour was more convenient than treating them as vehicles because time slice flow frequencies in terms of p.c.u./hour were insensitive to any differences in average cruise times as between cars and motorcycles. Also the total delay could be calculated directly from the flow profile expressed in p.c.u./hour.
It was concluded that the traffic model which is used in TRANSYT to calculate the performance index of a network was applicable to Bandung traffic, and that TRANSYT might be expected to co-ordinate traffic signals in Bandung satisfactorily.