Comparative evaluation of dynamic TRANSYT and SCATS- Based signal control systems using paramics simulation

Abstract

With the development of Advanced Traveler Information Systems (ATIS), it has become more important to compare the performances of different signal control systems, such as the widely be used TRANSYT and SCATS, for assessing the performance of traffic management strategies, and for provision of route-guidance information.The objective of this research is to compare the performances of traffic operations under TRANSYT and SCATS control, in two different networks under time-varying traffic demands by means of microscopic traffic simulation. TRANSYT is used to compute and update the signal timing plans at short intervals to simulate its operations in adaptive mode. For the simulation and update of TRANSYT signal control plan, TRANSYT has been interfaced with the PARAMICS microscopic traffic simulation model. At regular time intervals, TRANSYT collects input data from the simulation model, and calculates new signal timing plans which will to be implemented in the simulation. For the simulation of SCATS signal control using the algorithms that decides on the cycle length, split plan and offsets between adjacent intersections in a sub-system have been coded in an Application Programming Interface plug-in that can be used with PARAMICS. One real network with four intersections and one hypothetical network with two signalized intersections were selected for the simulation experiments. The real network is modelled after major arterials in the CBD in Singapore, using real traffic demand covering the morning peak period, including the transition before and after the morning peak. The hypothetical network, consists of two intersections, is designed to test the TRANSYT and SCATS signal control logic under relatively high fluctuation of traffic flow, including the change in peak traffic demand in different directions, in a relatively short period.The three output indices of the network namely (1) mean delay; (2) number of vehicles; and (3) mean speed were gathered and evaluated. From the three indices, TRANSYT appears slightly more effective than SCATS in the real network. However, SCATS is more effective in handling the relatively high fluctuation in traffic demand in the hypothetical network.