Agent-Based Fire Evacuation Model as a Tool for Fire Safety Design Assessment in Malaysia

Abstract

Evacuation model is an important tool in formulating performance-based building design. It is a tool aims at helping designers assess the robustness of their design against factors such as human behaviour in fire. Currently, performance-based code for fire safety design has been the hot trend all over the world. Nevertheless, even though Malaysia is applying prescriptive-based code, performance-based approach for special buildings and mega projects are allowed. This shows that it is just a matter of time before performance-based code be fully adopted in place of prescriptive-based code in Malaysia. This paper aims to highlight how evacuation modelling may help designers assess the performance of their design. In this paper, an example of evacuation scenario was modelled and simulated using an agent-based fire evacuation model -FDS+Evac. The output of the simulation provide graphical representation such as the fire curve, the evacuation time, and the time to untenable condition that can be used to analyse the life safety objective based on concepts of ASET and RSET. The FED values of all occupants throughout the evacuation process are less than tenability criteria (0.3 FED). Therefore, it is concluded that all occupants have evacuated safely in the event of fire.

Crowd Modeling and Simulation for Safer Building Design

“…Firstly, the user needs to define the value of control parameters involved in the ABC algorithm, which are exit door locations, the limit and the number of iterations. Exit door locations are randomly initialized by using equation (11) in a given search space area. (11) where x n,i represents the value for the i th dimension, where i = 1,....,D (D = the number of optimization parameters), and the n th solution, where n = 1,....,SN (SN = the number of solutions).…”

“…Exit door locations are randomly initialized by using equation (11) in a given search space area. (11) where x n,i represents the value for the i th dimension, where i = 1,....,D (D = the number of optimization parameters), and the n th solution, where n = 1,....,SN (SN = the number of solutions). x ni is bounded between a specified lower value l i = 0 and an upper value u i = Room Lenght.…”

“…14Finally, a scout bee will determine the abandoned solution by evaluating the limit parameter. If it exists, the scout bee will find a new random solution by using equation (11).…”

“…Firstly, define the values of control parameters involved in the algorithm, which are the number of population (exit door locations), the limit, the number of iterations and the fixed archive size. The exit door locations are randomly initialized by using equation (11).…”

“…It is straightforward but can be impractical as it can be dangerous if it involves large crowds, it is time consuming, and it only allows limited solutions to be proposed since any alterations to building layouts would be costly. The second approach to evaluating evacuation efficiency of a building is by checking the compliance of the building with the prescriptive building codes, which are documented standards and codes for building safety purposes that must be adhered to in building design [11]. This approach facilitates designer's work, but does not necessarily result in safe evacuations during emergency, and limits design freedom and flexibility.…”