Performance Evaluation of a Dynamic Signal Control System for Mixed Traffic Conditions Using Sparse Data

Dynamic signal control systems combine traditional traffic signals with an array of sensors to intelligently control vehicle and pedestrian traffic. Under mixed traffic conditions similar to that prevailing in India, where there are a variety of vehicle classes moving without lane discipline, implementation of such dynamic signal control systems is challenging. This is because automated sensors such as loop detectors used for data collection under homogeneous traffic conditions do not work well under mixed traffic conditions. Under such conditions, the possibility of utilizing sparse data obtained with respect to sampled travel times from different sensors is currently being explored. The present study focuses on developing a dynamic signal control system using sparse data collected using radio frequency identification (RFID) sensors under mixed traffic conditions. An existing delay equation was reformulated as an optimization problem for the dynamic signal control. The evaluation of the dynamic signal control system was carried out using a calibrated VISSIM microsimulation model. Python programming was used to integrate the dynamic signal control algorithm through the COM interface in VISSIM. The results indicated a good degree of improvement in overall average intersection delay of 12.6% compared to fixed time signal control. Thus, the proposed optimization method using delay from RFID sensors is suitable for implementing dynamic signal control under mixed traffic conditions.