Traffic signal systems in a developing country

The major cities of many developing countries are beginning to experience high levels of congestion due to an extremely rapid rise in car ownership levels. There is scope for relieving this congestion through the use of advanced fixed-time traffic control techniques like TRANSYT, a British program to design co-ordinated signal plans. This Dissertation investigated the value of such an application testing the performance of TRANSYT plans at a corridor in Santiago de Chile. Operational definitions and methods of measurements of different parameters, as well as the behaviour and simulation of priority intersections were also investigated. For the assessmment of TRANSYT benefits, in an arterial corridor of Great Santiago, four Scenarios were developed: Scenario 0 or before situation; Scenario 1 involving a direct application of TRANSYT plans to the site; Scenario 2 produced after a `fine-tuning' following the experience with Scenario 1; Secenario 3, similar to Scenario 2 but allowing some stages to be called when demand for them was detected; and finally Scenario 4 requiring improved modelling and calibration of bus traffic behaviour on the network. It was found that the direct implementation of TRANSYT plans (Scenario 1) produced poor benefits, but the other Scenarios generated improvements in journey time and stops similar to those encountered elsewhere in the developed world. However the Scenario 4 - that used a purpose designed model to calibrate TRANSYT including not only the typical parameters of platoon dispersion, but also the calibration of many other parameters traditionally not considered for calibration as well as operational redefinitions of free flow speed and a method of measuring it accurately, saturation flows as discharge rates on kerb-side lanes, cruise times in areas with high bus traffic friction, time spent on bus stops, proportions of buses that stop, and accurate recommendations of how to model networks with such typical developing countries conditions - produced considerable improvements in performance, more than doubling the benefits of plans prepared with more conventional TRANSYT values. The results of priority intersections showed that no significant difference is produced amongst parameters used in the British practice with the PICADY program and the Chilean values produced. Finally recommendations for future applications and for further research are given.