Development of Simulation-Based Dynamic Signal Control System for Julgaha Junction, Galle

Abstract

Sri Lankan road networks have often been blamed for their infrastructure inadequacies and operational inefficiencies. Along these lines, unsignalized junctions face substantial challenges, including prolonged travel time, conflicts among road users, delays, and compromised road safety. Most traffic systems use fixed-time patterns, which are simple yet fail to accommodate fluctuating traffic volumes. To address these hurdles, studies developed adaptive control systems. However, such advanced systems have been underutilized due to high installation and operation costs. This study therefore remedies this limitation by proposing a cost-effective adaptive queue-length-based dynamic signal system (AQuLeDSS)at Julgaha Junction, Galle City, which is encompassed by the intersection of four trunk roads and experiences severe congestion during peak hours. The model was calibrated using field-collected traffic data in the SUMO to improve the fidelity. Key operational parameters were adjusted iteratively, thereby ensuring that the simulated environment closely replicated the observed traffic behaviour at the selected intersection form. The complete control algorithm, along with the You Only Look Once (YOLO)-based vehicle detection module, was implemented in Python in the SUMO environment. Signal phase allocation was determined based on predefined control rules and threshold values. The combined SUMO with Python package triggers AQuLeDSS, employing queue-length-amenable green, red, and vehicle clearance times. Thereunto, results indicate that the intersection capacity is not fully utilised when it is left unsignalized. AQuLeDSS implemented simulations achieved a 60–65% decrease in queue lengths, along with overall economic savings and a 48% reduction in emissions. This article recommends deploying an integrated system using edge-based GPU platforms and a Real-Time Operating System (RTOS)with carved-out zonal hub controls to enhance road safety in urban centres.