Optimization of Traffic Network Signal Durations with Heuristic Algorithm and the Effect of Number of Individuals

Authors : Cihan Karakuzu, Emin Topal

Pages : 78-88

Doi:10.34088/kojose.1159113

View : 83 | Download : 50

Publication Date : 2023-05-31

Article Type : Research Article

Abstract :In the traffic network that we frequently use in our daily life, the primary demand of people has been to reduce the time they spend in traffic and to travel to the points they want to reach as quickly as possible. Developing countries want to meet this demand with the least cost in order to meet this demand. This study aims to manage the traffic network with the best times by optimizing the traffic signal durations in order to minimize the travel time for a road network chosen as a benchmark. For the optimization process, it is aimed to run a population-based heuristic algorithm with different numbers of individuals and obtain the best travel time. With the help of an open-source code traffic simulation program, which was run by modeling the benchmark road network, the received traffic data was also visually analyzed and compared. The effects of the heuristic algorithms applied with different numbers of individuals on the travel times according to the starting-destination points were examined before and after the optimization. As a result of the study, it has been observed that travel times and traffic signal times can be reduced with heuristic algorithms. Based on both numerical metrics and visual results, it has been determined that optimized traffic light durations give better results than non-optimized ones.
Keywords : Differential Evolution Algorithm, Heuristics Algorithms, Modeling and Simulation, Traffic Lights Optimization, Traffic Network Simulation, Soft Computing