Kong DP, Lu SX, Feng L, Xie QM (2011) Uncertainty and sensitivity analyses of heat fire detector model based on Monte Carlo simulation. Journal of Fire Sciences 29(4), 317-337. [In English]

Web link: http://dx.doi.org/10.1177/0734904110396314

Keywords:

heat fire detector, Latin hypercube sampling, Monte Carlo simulation, uncertainty analysis, sensitivity analysis, predictor smoothing methods, sampling-based methods, complex-systems, input variables, assessments, example, design

Abstract: Estimating activation time for heat fire detectors is essential in fire safety engineering. Due to the complexity of the heat detector model, many uncertainties are associated with the input parameters. It is necessary to consider the influence of the uncertainties associated with the input parameters on predicted activation time. For this purpose, uncertainty and sensitivity analyses method for the heat fire detection model based on Latin hypercube sampling are introduced. The uncertainty analysis results indicate that the predicted activation time would be a variable approximating to a unimodal, positively skewed distribution rather than the strictly normal distribution, with the consideration of uncertainties associated with the input parameters. Various statistical coefficients and the result of stepwise regression, both confirm that location of the heat detectors and fire growth rate have the most significant influence on the predicted activation time.