Calibration methodology for microscopic simulation of pedestrian traffic

Simulation of pedestrian traffic has great potential as decision support tool when designing pedestrian infrastructure, such as public transport stations. Such simulations are already used to visualize prognosticated pedestrian traffic at proposed stations, but quantitative predictions of traffic characteristics, such as delay or level of service, are still unavailable due to the lack of calibrated models.

Calibration of microscopic models of pedestrian traffic is a challenging task, and even though many calibration methods have been proposed, it is still unclear which, if any, of the proposed methods that are suitable for various situations and applications. This project takes a first step towards answering this question by reviewing and quantitatively comparing proposed calibration methods. The project also aims to calibrate, using the most promising method found in the literature, one of the most used model types using publicly available data, and estimate the uncertainty of the resulting set of parameter values.

The project is carried out by VTI and is funded by Trafikverket via Centre for Traffic Research (CTR).

Project leader: Fredrik Johansson, VTI