Agent-based modelling framework for evacuation planning

• Main Author: Ahmad Mustaqim Jumat
• Other Authors: Cai Wentong
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/144507
• Institution: Nanyang Technological University

As urbanization continues, public places in cities such as transportation hubs and shopping centres often attract high volume of pedestrians. In the case of life-threatening accidents such as an earthquake or fire, efficient evacuation strategies can evacuate large crowds in a short period of time while minimizing the risk of injuries and deaths. It has been widely accepted that crowd simulation is a useful and cost-effective tool to test various evacuation strategies without performing resource and time-consuming experiments. The objective of this research is to develop an agent-based simulation framework and identify the most efficient evacuation strategy for the City Hall MRT Station in Singapore as a basis of realism. The simulation framework is built with AnyLogic Software, specifically the AnyLogic Pedestrian Library (APL) is used to simulate the movements and behaviours of pedestrians in a physical environment on a day-to-day basis. In addition, this study also evaluated the impact of social distancing on pedestrian traffic flow in the City Hall MRT Station, given that the social distancing measures are implemented due to COVID-19. Through computer simulations, the total evacuation time and pedestrian flow rate were estimated for a set of candidate evacuation strategies: Random evacuation, Shortest-Distance First evacuation and Sub-Region evacuation. It was observed that the Shortest-Distance First evacuation strategy was the most efficient yielding the least evacuation time and highest pedestrian flow rate amongst the evacuation strategies.