Examination of Traffic Incident Management Strategies via Multi-Resolution Modeling with Dynamic Traffic Assignment

摘要

Transportation professionals have long faced the challenge of reliably analyzing howdifferently the transportation system will operate if certain facility improvements are made, anunexpected incident occurs or when certain operational strategies are deployed. This isparticularly difficult when these scenarios have system wide impacts. One of the main reasons isresulted from the many possible alternative routes in a transportation network and the differentpaths the drivers can be taking. Without appropriate methods to address this critical vehicle pathfinding issue and the resulting time-dependent volumes on the alternative routes, the reliability ofthe traditional transportation modeling tools are often questioned.Due to the limitations of both travel demand models (TDM) and micro-simulation modelsin analyzing complex traffic incident management scenarios, a new multi-resolution modelingapproach with mesoscopic simulation based Dynamic Traffic Assignment (DTA) was followed.The paper illustrates the approach through a case study. The case study investigates trafficmanagement strategies during a major freeway crash that occurred in the Phoenix metropolitanregion. These steps were followed: (1) obtain the initial vehicle origins, destinations and volumesfrom TDM; (2) apply Dynus-T, a mesoscopic simulation-based DTA model, to conduct scenarioanalysis and obtain the space and time dependent volumes, vehicle paths and resulting traveltimes; (3) import the output from the Dynus-T into VISSIM, a micro-simulation model, for moredetailed analysis at certain hot spots when necessary.The model calibration, scenario analysis and output of the DTA model will be discussedin detail. This multi-resolution modeling approach appears to be very useful to generate reliable,efficient and cost-effective solutions for complex Traffic Incident Management (TIM) tasks.