Synthetic Population Techniques in Activity-Based Research

Synthetic Population Techniques in Activity-Based Research

Sungjin Cho (Hasselt University, Belgium), Tom Bellemans (Hasselt University, Belgium), Lieve Creemers (Hasselt University, Belgium), Luk Knapen (Hasselt University, Belgium), Davy Janssens (Hasselt University, Belgium) and Geert Wets (Hasselt University, Belgium)
Copyright: © 2014 |Pages: 23
DOI: 10.4018/978-1-4666-4920-0.ch003
OnDemand PDF Download:
No Current Special Offers


Activity-based approach, which aims to estimate an individual induced traffic demand derived from activities, has been applied for traffic demand forecast research. The activity-based approach normally uses two types of input data: daily activity-trip schedule and population data, as well as environment information. In general, it seems hard to use those data because of privacy protection and expense. Therefore, it is indispensable to find an alternative source to population data. A synthetic population technique provides a solution to this problem. Previous research has already developed a few techniques for generating a synthetic population (e.g. IPF [Iterative Proportional Fitting] and CO [Combinatorial Optimization]), and the synthetic population techniques have been applied for the activity-based research in transportation. However, using those techniques is not easy for non-expert researchers not only due to the fact that there are no explicit terminologies and concrete solutions to existing issues, but also every synthetic population technique uses different types of data. In this sense, this chapter provides a potential reader with a guideline for using the synthetic population techniques by introducing terminologies, related research, and giving an account for the working process to create a synthetic population for Flanders in Belgium, problematic issues, and solutions.
Chapter Preview

Synthetic population techniques can be largely divided into two groups: IPF and CO. Most techniques in these two groups have a similar concept of fitting seed data to a target marginal distribution, but they generate the required synthetic population in totally different way. This section covers the different ways by introducing related research in each group.

Complete Chapter List

Search this Book: