Self-organization of traffic lights for minimizing vehicle delay

Abstract

Problems of delay and resultant driver frustration continue to get worse as congestion increases in large cities. The delays can be minimized by optimization of traffic signals. Traffic signal optimization is a complex problem making improvements difficult. Recent advances in sensors and communication have enabled the use of novel algorithms for improving traffic signal efficiency. This paper investigates the use of a distributed traffic signal control algorithm based on the concepts self-organization. The algorithm is benchmarked against traditional traffic signal algorithms. The self-organization based algorithms perform significantly better compared to the existing algorithms. A simulation model was created based on an abstraction of city road with multiple intersections using data on actual traffic counts. The results are presented in this paper along with the details of the algorithm.