Abstract
Traffic streams in most of the urban areas of developing countries are highly heterogeneous with poor lane discipline. Traffic simulation and optimization packages based on microscopic traffic flow models are popular for modeling traffic flow and optimizing traffic controls all around the globe. However, applying microsimulation packages to model undisciplined and heterogeneous traffic streams may not yield perfect results. But, macroscopic models can be easily calibrated for such heterogeneous and undisciplined traffic streams. This research applies the calibrated Cell Transmission Model (CTM)-based optimization tool to determine optimal signal plans and to evaluate the potential of reduction in control delay at signalized intersections. Most of the traffic signals in Karachi are pre-timed with a constant signal plan throughout the day, despite significant changes in traffic dynamics within a day. This study evaluates the performance of existing signal timing plans at two different locations in Karachi. The selected locations include a segment of urban arterial with three consecutive signals in close proximity. The second location was a standalone signalized intersection. Twelve-hour traffic and delay data on typical working days were collected through video recording technique. The data collected from the field were simulated using CTM-based signal optimization software, DISCO. The optimization results show an improvement of 45% in the existing delay at the selected segment of Khayaban-e-Ittehad. The delay at KDA intersection can be reduced by 29% by optimizing signal plans. The study shows that optimum signal timings based on accurate data can reduce delays and improve traffic network performance.