VISWALK Microsimulation-Based Evacuation Modeling of a Critical Infrastructure: A Social Force Model Approach

Given the continuous increase in occupancy in critical urban infrastructures, the adoption of evacuation strategies has become a widely recognized solution to tackle the challenges posed by natural or manmade disasters. Unfortunately, even in this modern day and age, Bangladesh has witnessed many disastrous accidents resulting in thousands of casualties due to the absence of effective emergency evacuation plans in high-occupancy infrastructures. Effective development of evacuation plans necessitates the collection of diverse data, encompassing details concerning evacuation duration along with several other factors. In light of this, the primary objective of this study is to establish a methodology to determine the total evacuation time of a 5-storeyed readymade garments factory (RMG), representing a densely populated critical infrastructure. A PTV VISWALK microsimulation model of the study infrastructure was developed to simulate evacuation scenarios to reflect the real-life occupancy data. The Social Force Model (SFM) of pedestrian dynamics has been used to represent the panic amidst people during disaster situations by tuning the walking behavior parameters. Integration of VISWALK and SFM parameters ensures the understanding of pedestrian characteristics and movement, which helps to emulate different types of evacuation scenarios. Machine learning techniques have been incorporated with the Latin Hypercube Sampling (LHS) method in our study to calibrate the parameters for creating emergency conditions to obtain the total evacuation time. Results of the microsimulations show that the minimum total evacuation time (TET) or required safe egress time (RSET) is 8 minutes 7 seconds, exceeding the available safe egress time (ASET) 5 minutes, as found in several fire drill surveys of structure with similar geometries. The findings of this study will enhance the understanding of evacuation dynamics and provide insights into the social force model parameters which can be further utilized in the optimization of emergency response strategies within similar infrastructure. The results showcased in this research will inform stakeholders regarding occupants’ safety and provide insights into the potential risks associated with the layout of analogous structures, thereby ensuring more resilient urban environments. Journal of Engineering Science 15(2), 2024, 1-12