Computer-Aided Multiagent Simulation for Traffic Organization in Highway Construction Zones

Highway construction zones form critical bottlenecks that disrupt traffic flow and elevate safety risks. This study proposes a multi-agent simulation framework incorporating heterogeneous driver behavior models (aggressive, normal, conservative) to evaluate traffic organization strategies. A case study on a four-lane highway compared five strategies under varying traffic volumes. Results demonstrate that a combined strategy, integrating extended warning areas and variable speed limits, outperforms conventional methods, reducing travel time by up to 25.4% and safety incidents by 62% in high-volume scenarios. The framework effectively bridges theoretical models and practical management challenges by capturing key influencing factors such as lane change timing, driver response heterogeneity, and context-dependent lane closure configurations.