An aggregate model for transient and multi-directional pedestrian flows in public walking areas

A dynamic pedestrian propagation model applicable to congested and multi-directional flows is proposed. Based on the cell transmission model, several important extensions are made to describe the simultaneous and potentially conflicting propagation of multiple pedestrian groups. Their movement across space and time can be individually tracked, such that group-specific travel time distributions can be predicted. Cell potentials allow for en-route path choice that takes into account prevailing traffic conditions. The model is formulated at the aggregate level and thus computationally cheap, which is advantageous for studying large-scale problems. A detailed analysis of several basic flow patterns including counter- and cross flows, as well as two generic scenarios involving a corner- and a bottleneck flow is carried out. Depending on the choice of model parameters, various behavioral patterns ranging from disciplined queueing to impatient jostling can be realistically reproduced. Following a systematic model calibration using real data, an extensive case study involving a Swiss railway station is presented. A comparison with the social force model and pedestrian tracking data shows a good performance of the proposed model with respect to predictions of travel times and pedestrian density fields.


    • EPFL-REPORT-196055

    Record created on 2014-01-20, modified on 2016-12-16

Related material