Novel system for road traffic optimisation in large cities

• Published in: IET smart cities Vol. 4; no. 2; pp. 143 - 155
• Main Authors: Alkhatib, Ahmad A. A., Abu Maria, Khulood, Alzu'bi, Shadi, Abu Maria, Eman
• Format: Journal Article
• Language: English
• Published: Melbourn John Wiley & Sons, Inc 01.06.2022; Wiley
• Subjects: Artificial intelligence; Automation; Cities; Control systems; Driving conditions; Genetic algorithms; Indicators; Neural networks; road traffic control; road traffic management; Roads & highways; Scheduling; Simulation; smart cities; smart traffic light; Traffic congestion; Traffic control; Traffic flow; Traffic intersections; Traffic jams; Traffic management; Traffic signals; Transportation planning; urban traffic control; UTC; Vehicles; United States > US; Michigan
• DOI: 10.1049/smc2.12032
• ISSN: 2631-7680

Traffic congestion and road intersection management have become a significant issue, mainly with the highly increasing number of vehicles in cities. There is a common belief from vehicle drivers that installing traffic lights with some consideration of traffic flows will be dominant in traffic movements. This article proposes a novel system for Urban Traffic Control (UTC) with a continuous dynamic environment adaptation to improve traffic flow on large cities' network roads. The proposed system introduces vehicle counting method, lane evaluation of the current status and controlling method considering the effect on the whole traffic network—not just the intersection itself—to provide an efficient traffic scheduling. The main objective of the authors’ system is to reduce traffic jam, by reducing waiting time and trip time for vehicles at intersections. Some indicators and models are introduced in this work to assign traffic flow schedules with minimum traffic congestion and vehicle waiting time. These indicators and models include a traffic jam indicator, vehicle priority and lane weight. A multi‐agent urban traffic control system is proposed as the simulation environment using NetLogo simulator. (A total of 150) Vehicles are generated with random behaviour distributed over 25 intersections for 9 h duration to compare the normal fixed cycle traffic light and the authors’ smart traffic control. Results show a reduction in the total average waiting time of all vehicles for the simulation period by more than (29.98%). Hence, it is more suitable for the complexity of the current traffic condition with minimum changing infrastructure.