Developing an ABM framework to identify pedestrian movements of visitors of Amsterdam based on behavioural processes.

Abstract

The main objective of this thesis is to develop an agent-based modelling framework that identifies pedestrian movements of visitors of the city of Amsterdam, based on behavioural processes. To do this it is necessary to study pedestrian movements and dynamics by first developing a substantiative theoretical background. The developed pedestrian visitor model simulates how visitors with different characteristics move towards their destinations. The visitors are derived from four different types and are given a set of characteristics and intrinsic preferences. The movement of visitors is defined by a destination-choice aspect and a route-choice aspect. The destination-choice aspect gives meaning to the characteristics, preferences and itineraries of the agents in the model, whereas the route-choice aspect depicts the route to follow based on dynamic values derived from the interaction of the agents on the pedestrian street network. To determine crowdedness and the action taken when confronted with crowdedness, a crowdedness threshold value has been created based on the visitor type and planned activity in the model. The outcomes have been analysed with the use of two scenarios, a scenario without crowdedness thresholds and a scenario with crowdedness thresholds to show the movement patterns generated by the pedestrian visitor model. The movement patterns from the first scenario concentrate more in the city centre, whereas the movement patterns from the second scenario show a more dispersed output of movement patterns.