Following previous research in Chile that showed the influence of physical and operational characteristics of traffic signal junctions on capacity, in-depth research was done to capture the influence of turning vehicles as opposed to other vehicles sharing a lateral lane. Currently, equivalency factors and then saturation flows are estimated as a function of lane type and width, period of the day, and the proportion of public transport vehicles in the flow. A database containing more than 400 discharges and about 4,000 headways was collected at 12 signalized junctions in Santiago, Chile, and using the synchronous regression method, new formulations were obtained to estimate equivalency factors. These formulations were defined, besides the above-mentioned factors, as a function of turning-vehicle proportion, turn radius, and pedestrian level that interfere with turning vehicles. Validation of the results was successful. The difference between observed and estimated saturation flows was about 1%, which is an excellent accuracy level. In fact, it is far better than current estimations. As a result, a better and more realistic estimation for saturation flow in turning lanes in developing countries was obtained, and therefore, in order to improve traffic signal junction performance, the new formulations can be introduced into computational programs for simulation and optimization of different urban traffic projects.
Abstract