Study on a Pedestrian Simulation Model of Natural Movement

Abstract

The purpose of this study is to construct a pedestrian simulation model of natural movement by applying rules that influence pedestrian route selection from previous experimental studies. Based on wall vector field, the multi agent-based simulation model which evolves through the relationship between vision-driven agents and configuration of environment was programmed by NetLogo. It considered visual field angle, visual distance, visual noise (looking around), moving speed, attractive factor, and other agents. The results of the simulation in several environments including a real architectural space such as the Tate Gallery produced better movement patterns than a previous model (Turner's EVA model) both locally and globally. Since the simulation model included variables related to visual factors which affect natural movement, it could help towards understanding the relationship between visibility and movement, its effects on spatial usage, and the affordance of configuration in a built environment for natural movement.