An optimization method for the distance between exits of buildings considering uncertainties based on arbitrary polynomial chaos expansion. (October 2016)

• Record Type: Journal Article
• Authors: Xie, Qimiao; Wang, Jinhui; Lu, Shouxiang; Hensen, Jan L.M.
• Is Part Of: Reliability engineering & system safety. Volume 154(2016:Oct.)
• Page Start: 188
• Page End: 196
• Publication Date: 2016-10
• Subjects: Evacuation time -- Uncertainty analysis -- Performance-based fire protection design -- Optimization under uncertainties -- Distance between exits; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2016.04.018
• Languages: English
• ISSNs: 0951-8320

Abstract

Abstract: The distance between exits is an important design parameter in fire safety design of buildings. In order to find the optimal distance between exits under uncertainties with a low computational cost, the surrogate model (i.e. approximation model) of evacuation time is constructed by the arbitrary polynomial chaos expansion. Through a two-stage nested Monte Carlo simulation of this surrogate model, the optimal distance between exits under uncertainty is found efficiently. In order to demonstrate the proposed method, a single room with two exits is presented as a fire compartment and uncertainties of occupant density and child-occupant load ratio are also considered. In this case, the results showed that the optimal distance between exits changes with the level of probability of evacuation time, and there is a critical level of probability for the transition of the optimal value of the distance between exits. Furthermore, the traditional Monte Carlo simulation method is used to compare the accuracy of the surrogate model with the computer evacuation model FDS+Evac developed by the VTT Technical Research Center of Finland [1]. The results indicate that the proposed surrogate-based optimization method can achieve a similar accuracy with a much lower computational cost. Highlights: Arbitrary polynomial chaos expansion can be used to construct the surrogate model of evacuation time. The computational cost of this proposed optimization method is reduced significantly. The optimal distance betwwen exits changes with the level of probability. There is a critical level of probability for the transition of the optimal distance betwwen exits under small exit widths.