Multiscale Modeling of Virtual Urban Environments and Associated Populations

Abstract

This chapter introduces a novel approach to model Virtual Urban Environments (VUE) by taking into account: (1) the multiscale issue, (2) activities’ locations, multi-modal transport networks and their relationships, and (3) the associated synthetic population. This new modeling approach is based on a scale-free design pattern, which introduces the notion of Service as a way of representing locations’ accesses as well as modal shift opportunities using the transportation network. An example of creating such a VUE is given for Quebec-city’s case study. Simulating daily travel activities of an entire population can be performed with our TransNetSim software at the meso-scale. In order to forecast the travel demand we use a weighted bipartite matching algorithm that assigns activities’ locations to agents representing people, according to their predicted travel times. Using a precompiled routing matrix, TransNetSim is able to simulate 1 million trips on the Quebec-city’s network (around 32,000 nodes and 81,000 links) in less than a minute. TransNetSim is also coupled with the commercial traffic microsimulator Vissim, which provides an environment that is compatible with our micro-VUE model. This work is a first step toward a long-term objective of modeling VUEs as a framework in which many multiscale urban phenomena might be integrated.