Comparison of Optimization Techniques for Delay Minimization in Signalized Intersections: PSO vs GA

Authors : Abdullah Karadağ, Murat Ergün

Pages : 162-172

Doi:10.31590/ejosat.1270905

View : 70 | Download : 92

Publication Date : 2023-08-31

Article Type : Research Paper

Abstract :Minimizing intersection delays is an important challenge in today’s smart cities. Even there are different approaches for delay minimization most of them uses the same nonlinear delay formula defined by Highway Capacity Manual (US). As a result choosing a fast and precise algorithm for finding the optimum inputs minimizing the delay output is a critical decision. In this paper we share our experience in selection of best optimization algorithm as a part of our work of developing an innovative system to minimize person delays in intersections. We compared two best known algorithms: Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). It is shown that using the same population size and number of iterations PSO is 7x faster and 17x more precise than GA.
Keywords : Adaptif trafik sinyal kontrolü, Genetik algoritmalar, Parçacık sürü optimizasyonu, Kavşak gecikme minimizasyonu, Gerçek zamanlı eniyileme