Using agent-based modelling for simulating the effects of space syntax values on pedestrian movement patterns

By Simon Veen

Abstract
Pedestrian movement has been studies for more than five decades. Space syntax is a network analysis method that investigates the global and local urban grid using axial lines. It also has techniques for observing movement within a network. While originally created for the analysis of spatial configurations, a correlation between the measured values and pedestrian movement patterns was found. Also included in this thesis are the angular segment analysis, which is an improvement on the space syntax and includes the angles between streets into the measurements, and PageRank measurements, which values are found to better correlate with movement flows than integration values. These values are calculated for a study area in the city centre of Amsterdam and used in an ABM. In the model, the agents display exploratory behaviour and base their wayfinding on the characteristics of the network. The purpose of the model is to gather movement patterns for the space syntax and PageRank values and compare them to each other. Three different movement pattern maps can be distinguished, and the choice values give the most logically human movement patterns. The other values have the same result as a non-preference walk. A reason for this could be the size of the study area, and a smaller study area could provide more meaningful results.

Keywords: Space Syntax; ABM; Human movement patterns; PageRank; Integration; Choice; Axial lines; Angular segment analysis