Print

Smith, Mark C.

Civil and Environmental Engineering

2007

MS

Microsimulation is a relatively new and vastly more capable way of modeling various aspects of highway systems compared to traditional deterministic models. It can be used to model large road networks while showing the ramifications of detailed intersection geometry and complex traffic control systems. Most importantly, it will predict changes in route choice due to recurring congestion, reflecting the dynamic nature of traffic behavior. In addition, the flexibility in accepting different geometric configurations, along with a wide array of user controlled parameters makes microsimulation well suited for studying unusual alignment or behavioral situations. This high degree of complexity, flexibility and user control comes at a cost. Matching the performance of a particular model to reality is not guaranteed, and thus great care must be taken when calibrating a model to reflect a desired condition.
This paper looks at two case studies at the extreme ends of the spectrum of traffic analysis. Case 1 involves calibrating a model of the very large and diverse road network of Chittenden County, Vermont. The model includes 18 municipalities, 6 interstate interchanges, 175 traffic signals, over a 500 square mile area, and almost 55,000 cars modeled in the afternoon peak hour. It is not only one of the largest models ever attempted, but one with the widest range of traffic conditions, from quiet villages to dense urban streets.
Case 2 involves calibrating a model of a modern roundabout, as defined in the recent FHWA Roundabout Informational Guide. Suitable roundabouts that match these guidelines in the New England area were cataloged and the most suitable one was chosen for study. Video data is collected and broken down into measured queues, average delay, and traffic volumes to seed a comparable simulation model. Various parameters are studied for their suitability and effect on controlling the output of simulation runs.