Bat Optimization For Crowd Evacuation Simulation in Classroom
Loading...
Files
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Malaysia Sarawak (UNIMAS)
Abstract
Description
The goal of this project is to improve crowd evacuation methods in classroom settings by developing a simulation model that employs the Bat Algorithm (BA), a nature-inspired optimization method. In emergencies such as fires or natural disasters, the ability to quickly and efficiently evacuate individuals from confined spaces is crucial for reducing injuries and deaths. Conventional evacuation models often lack the flexibility to adapt to changing factors such as variations in crowd behaviour, immediate obstacles, and modifications in exit routes. This research employs the Bat Algorithm, inspired by bats' echolocation skills to optimize evacuation routes dynamically and in real time, addressing these constraints. The simulation considers factors such as crowd density, exit accessibility and obstacles, utilizing agent-based modelling to replicate human movement during evacuations. The model was testing in three different scenarios. In Scenario 1, featuring a single available exit and a moderate crowd density (21–50 agents), the shortest average evacuation time was 1.01 minutes. Scenario 2, featuring two available exits and higher density with clustered agent placement (65–105 agents), recorded the longest average time of 1.25 minutes due to flow congestion. In Scenario 3, with all four exits open and high-density with clustered & random placement (80–140 agents), an average evacuation time of 1.02 minutes was recorded, indicating that several exits can efficiently accommodate larger crowds. The results indicate that the Bat Algorithm successfully adapts to varying conditions, reducing evacuation times and improving route effectiveness. This research presents a more flexible, data-driven approach for classroom evacuations that responds to real-time situations, ensuring safer and more effective evacuation procedures. The results are anticipated to enhance emergency response strategies and have wider applications for public safety management.