The Third Transportation Student Research Symposium
Abstract Detail
| Name | School | Title of presentation |
| Varun Ramanujam, Vikrant Vaze, Yang Wen | MIT | Application of multi-class vehicles counts in the calibration of micro-simulation models |
|
Different classes of vehicles have different impacts on the real-world traffic. For example, the
traffic dynamics of trucks is distinct from common passenger cars, and trucks may be restricted from
entering the parkways. To replicate and study the network performance in a realistic way, a
micro-simulation model should capture the characteristics of multi-class vehicles in its modeling and
calibration approaches.
We develop a methodology to use multi-class vehicles counts collected at toll stations to calibrate a micro-simulation model. We discuss assumptions that are necessary for our modeling approach. Simultaneous Perturbation Stochastic Approximation (SPSA) is used as the solution algorithm. Practical issues occurred during the implementation are also discussed. We demonstrate the methodology in two micro-simulation models: MITSIMLab, a microscopic simulator, and DynaMIT, a mesoscopic simulator. The approach is tested on the network of lower Westchester County, New York, using real data. Computed goodness-of-fit measures indicate a clear improvement as the result of the calibration with multi-class vehicles counts. |
Events Page
Visit the New England Section of ITE events page for news about other upcoming events around New England, including a Mass ITE Chapter meeting .