Seminar

  • November 14, 2012

    PacTrans Seminar Series: Calibration of Micro-Simulation Models Using ITS Data: Lessons Learned

    When: Wednesday, November 14, 2012 from 2:30pm to 4:30pm
    Where: In UW Seattle Campus – HUB 332
    View the PDF here

    In order for a traffic micro-simulation model to be effective, it must be able to replicate both supply and demand characteristics, as well as their interaction. That is, the simulation model must be properly calibrated. The recent widespread deployment of ITS in North America has provided an opportunity to obtain significant amounts of traffic related data on a point, link, and trip basis. In addition, there has been a corresponding increase in the quality and quantity of micro-simulation models that can be used to model transportation systems. A methodology for calibrating traffic micro-simulation models so that the simulation results, such as travel time, represent the observed distributions obtained from the field will be presented. The focus of the talk will be on lessons learned from various calibration-related research projects over the past ten years. The seminar will conclude with a discussion of current and future research initiatives in this area.

  • October 24, 2012

    PacTrans Seminar Series: Discrete Choice Analysis for Travelers: A Semi-parametric Approach

    When: October 24, 2012 from 2:30pm to 3:30pm
    Where: More 234

    Professor Baibing Li from Loughborough University of UK will visit the University of Washington and deliver a seminar talk on campus. You are cordially invited to join his seminar.

    Discrete choice modelling is a widely used econometric approach to analyse the behaviour of travellers. The multinomial logit model is one of the most commonly used models in discrete choice analysis. This research develops a new model, semi-parametric multinomial logit model. The developed model links travellers’ attributes and alternatives to the choice probabilities via a sensitivity function. This sensitivity function reflects the degree of travellers’ sensitivity to the changes in the travelling costs. A Bayesian approach is investigated to draw statistical inference for the semi-parametric logit model. An empirical study on travellers’ value of time that involves stated preferences about two train-related alternatives and two bus-related alternatives is conducted to illustrate the developed model.
    Read More

    • 1 … 14 15 16