Generation of a synthetic population

The basis of any ABM is a synthetic population. Visum does not have the functionality to generate one. We recommend using PopulationSim. To get you started, this topic describes roughly how to generate a synthetic population using PopulationSim. Since the software is not developed at PTV, we do not offer any guarantees or support for it. If you have any questions about PopulationSim, please contact the manufacturer.

Note: In the directory C:\Users\Public\Documents\PTV Vision\PTV Visum 2024/Examples, you can find an example of use on this topic. Among other things, the Demand ActivitySim example describes in detail how to create and import a synthetic population and provides instructions for installing PopulationSim.

A synthetic population is based on household and personal datasets as well as statistics on the frequency of certain characteristics (so-called marginal distributions). Household and personal data often come from a census or mobility/household survey. The data need not necessarily come from the planning area. It is sufficient if they were collected in a comparable area, for example, data from rural areas for a model of a rural area. The statistics, on the other hand, must relate to the planning area. They mostly include population figures differentiated by age groups, employment and unemployment figures, school and university places, and other available data. For the above example, freely invented data sets were generated for a section of the city of Karlsruhe (see folder "PopulationSim" in the examples directory). The generation of a synthetic population is similar to a weighting process. Clones of the original households together with the persons living in them are distributed in the planning area in such a way that the given marginal distributions are achieved with the best possible accuracy.