Simulation-Based Optimization of Dynamic Merging Strategies in Work Zones

The establishment of a robust transportation network has led to significant traffic safety challenges in highway maintenance zones characterized by lane compression and traffic flow interweaving. The current static confluence control approach struggles to accommodate traffic flow dynamics, resulting in suboptimal peak resource utilization and increased average traffic delays. In response, this study introduces a dynamic convergence control framework termed “perception-decisionoptimization.” Through traffic simulation and a Poisson regression model, the impact of lane numbers and other variables on merging conflicts is assessed. Leveraging the SUMO platform and orthogonal testing, a Poisson regression model is constructed to quantify data, revealing the interplay between traffic volume and speed limits, as well as junction positioning and speed limits. An adaptive merging optimization algorithm is devised to real-time adjust merging distances, striking a balance between conflict risk and traffic efficiency. Simulation outcomes demonstrate the superiority of the dynamic strategy over uncontrolled and fixed merging approaches, showcasing reduced peak lane change areas and extended merging distances during peak traffic periods. This research offers quantitative insights for dynamic safety management in construction zones and holds promise for future extensions into multi-source information integration and tailored scene adaptation.