ACSPRI Conferences, RC33 Eighth International Conference on Social Science Methodology

Font Size:  Small  Medium  Large

An agent based approach to building a synthetic population for transport planning for Sydney

Nam Huynh, Francois Lamy, Matthew Berryman, Pascal Perez, Mohammad Reza Namazi Rad

Building: Law Building
Room: Breakout 6 - Law Building, Room 022
Date: 2012-07-11 03:30 PM – 05:00 PM
Last modified: 2012-06-19


This paper presents an algorithm to generate the synthetic population from the census data provided by Australia Bureau of Statistics (ABS). The sets of ABS data used for this purpose include individual related tables (e.g. distribution of age by gender, relationship in household by age by gender, gross of individual weekly income) as well as household related tables (e.g. family composition, family composition by gender of persons in family, household weekly income). The core of the algorithm aims at merging individual and household level data sets, which centred on the correction of any discrepancies in common attributes among datasets (as random errors are introduced to values in ABS tables to avoid release of confidential data). An example of the synthetic population generated by this algorithm for a census district shows that the algorithm, while adjusting ABS tables to satisfy a set of assumptions and conditions, preserves quite well the key statistics/proportions of the population described in those ABS tables. The paper will also discuss how the data from Household Travel Survey could be incorporated to represent the transport attributes of the synthetic population.