Mostafa Abbasi Kia - Department of Computer Science, University of Lorestan, Khorramabad, Lorestan, Iran
Mohammad Nadjafi - Aerospace Research Institute, Ministry of Science and Research, Tehran, Iran
Adel Nadjafi - Enerchimi Engineering Company, Tehran, Iran

As we know, people are primarily at risk of different incidents during their life, especially when they encounter unpredictable accidents. For example, fires in public places such as governmental or trade centers during their daily activities make them obliged to evacuate the building rapidly. This research deals with the fire safety of mentioned people by means of the probabilistic method. For this purpose, fire safety is addressed by modeling the egress of the people from the fire to a safe zone. A trade center building with a common layout has been chosen for safety analysis and a limit state function has been developed according to the timeline evacuation model and fire scenario. To define the safety of building visitors, the safety index method has been selected for computing the probability of trapping in fire (fatality) and safety index (beta index). The harmony search algorithm has been used to obtain Hasfoer and Lind reliability index. A sensitivity analysis of the model’s variables has been done to find the most important and effective parameters for fire safety. Results show response time to the fire, area of buildings and length of escape route are more critical in buildings. In other words, the safety of people will improve with decreasing response time and length of evacuation route and increasing dimensions of interior space of buildings.

Fire Safety, public buildings, the probabilistic method, optimization, Harmony Algorithm