An Adaptive Traffic Light Control Algorithm Considering Driver Behavior and Wet Weather Effects: A SUMO-Based Study

Traffic congestion remains a persistent challenge in urban environments, often exacerbated by improper lane usage and adverse weather conditions. This study proposes an Efficient Dynamic Adaptive Traffic Light Control Algorithm (EDATLC) that dynamically adjusts green light durations by considering lane utilization, heterogeneous vehicle types, and lane-change-related driver behavior, which is modeled in this study through lane-change error scenarios, under both dry and wet surface conditions. The proposed algorithm is implemented and evaluated within the Simulation of Urban Mobility (SUMO) environment using a simulation-based framework with representative traffic demand patterns generated to reflect typical urban intersection conditions. EDATLC is compared against a baseline fixed-time control strategy and the classical Webster method under identical simulation settings. The results indicate that EDATLC improves intersection performance across all tested scenarios. Under wet surface conditions, the proposed method achieves up to 66.4% reduction in average vehicle delay compared to the baseline fixed-time control and 35.5% reduction compared to the Webster method. Additionally, the algorithm demonstrates improvements in environmental performance, including an 11.2% reduction in CO₂ emissions and fuel savings exceeding 23 liters during peak-hour operations. It should be noted that these improvements are scenario-dependent and represent relative performance gains within the simulation framework. Overall, the findings highlight that incorporating lane-change-related behavioral factors and environmental conditions into adaptive signal control strategies can contribute to enhanced operational efficiency and sustainability in urban traffic systems.