论文部分内容阅读
针对直接蒙特卡洛模拟效率较低的不足,利用基于拉丁超立方抽样(Latin Hypercube Sampling,LHS)的蒙特卡洛模拟分析了有喷淋作用时可用安全疏散时问(ASET)的不确定性;以秩相关系数和偏秩相关系数为指标,对影响ASET的参数进行了敏感性分析。分析表明,同样的精度条件下,拉丁超立方抽样所需要的抽样次数要远小于简单随机抽样的次数。算例分析结果表明,对于一个防火分区面积大小的无排烟单层商场,在有喷淋作用时,可用安全疏散时间具有对数正态分布的特征;敏感性分析表明影响可用安全疏散时间的主要参数为喷头距火源中心线距离、火灾增长系数和总热损失系数。
In view of the low efficiency of direct Monte Carlo simulation, Monte Carlo simulation based on Latin Hypercube Sampling (LHS) was used to analyze the uncertainty of available evacuation time (ASET) with sprinkler function. The rank correlation coefficient and partial rank correlation coefficient were used as indices to analyze the sensitivity of ASET parameters. The analysis shows that under the same precision conditions, the number of samples required for Latin hypercube sampling is much smaller than the number of simple random samples. The results of the example analysis show that for a smoke-free single store with a fireproof area size, the available evacuation time has the characteristics of lognormal distribution when there is sprinkling effect. The sensitivity analysis shows that the impact of the available evacuation time The main parameters for the nozzle from the center of the fire source distance, fire growth factor and total heat loss coefficient.