Characterization and Optimization of Traffic Flow Using Large Scale Simulation and User Data

• Activity Number: 466 - Contemporary Statistical Graphics: Methods and Applications
• Type: Contributed
• Date/Time: Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Arindam Fadikar* and Stefan Wild
• Companies: Argonne National Laboratory and Argonne National Laboratory
• Keywords: compute model calibration; traffic forecasting; agent based model

Abstract

In order to forecast traffic flows during adversaries, one needs to identify and characterize the distributions of vehicles on road at different locations and time of the day. In this work, we aim to establish an unified approach to characterize and forecast the traffic pattern by using a simulation model combined with user-based and road sensor information. The approach requires two-fold consideration: (1) combining observations from heterogeneous sources, which later serves as training data for the simulation model, and (2) calibration of that computer model. Inferences are drawn by combining actual observations with the calibrated computer model. We demonstrate our approach on the real traffic situation around O’Hare International Airport in Chicago using anonymous user-generated real trajectories, Illinois road statistics and an agent-based traffic simulation model.