Multi-Objective Optimization of Traffic Signal Timing at Typical Junctions Based on Genetic Algorithms
Zeyu Zhang,Han Zhu,Wei Zhang,Zhiming Cai,Linkai Zhu,Zefeng Li
DOI: https://doi.org/10.32604/csse.2023.039395
IF: 4.397
2023-01-01
Computer Systems Science and Engineering
Abstract:With the rapid development of urban road traffic and the increasing number of vehicles, how to alleviate traffic congestion is one of the hot issues that need to be urgently addressed in building smart cities. Therefore, in this paper, a nonlinear multi-objective optimization model of urban intersection signal timing based on a Genetic Algorithm was constructed. Specifically, a typical urban intersection was selected as the research object, and drivers’ acceleration habits were taken into account. What’s more, the shortest average delay time, the least average number of stops, and the maximum capacity of the intersection were regarded as the optimization objectives. The optimization results show that compared with the Webster method when the vehicle speed is 60 km/h and the acceleration is 2.5 m/s2, the signal intersection timing scheme based on the proposed Genetic Algorithm multi-objective optimization reduces the intersection signal cycle time by 14.6%, the average vehicle delay time by 12.9%, the capacity by 16.2%, and the average number of vehicles stop by 0.4%. To verify the simulation results, the authors imported the optimized timing scheme into the constructed Simulation of the Urban Mobility model. The experimental results show that the authors optimized timing scheme is superior to Webster’s in terms of vehicle average loss time reduction, carbon monoxide emission, particulate matter emission, and vehicle fuel consumption. The research in this paper provides a basis for Genetic algorithms in traffic signal control.
computer science, theory & methods, hardware & architecture