基于多项式混沌展开的人员疏散时间不确定性研究

发布时间：2019-07-09 09:25

【摘要】：目前,建筑防火设计常用的两种方法是“处方式”规范和性能化防火设计。“处方式”防火设计对建筑物的设计参数和各项指标都进行了较为详细的规定,但缺乏一定的灵活性。由于火灾发生和人员行为的不可预测性,建筑火灾下的人员疏散时间呈现显著的不确定性。性能化防火设计在处理火灾发生和人员疏散的随机性问题上,则主要采用安全系数或保守取值法。而安全系数的选取在较大程度上依赖于火灾专家,其取值的可靠性有待于验证。保守取值法通常根据最极端的情况对建筑物进行防火设计,这样往往会造成建筑资源的浪费。针对上述问题,本文提出了基于多项式混沌展开的人员疏散时间不确定性分析方法以及在不确定性参数影响下人员疏散时间最优出口参数取值的确定方法,主要研究工作与成果如下：根据人员疏散过程和参数可控的难易程度,将人员疏散时间的相关参数分为不确定性参数和出口参数。与人员疏散相关的不确定性参数较多,如人员密度、人群组成和预动时间等。出口参数是指与建筑物出口相关的参数,如出口宽度、出口间距、出口个数和出口位置等。为降低人员疏散时间不确定性分析的计算时间成本,本文耦合人员疏散模型和多项式混沌展开法建立了人员疏散时间的不确定性分析方法。该方法与Monte Carlo模拟方法在计算精度上较为接近且计算时间成本较少。此外,采用本文提出的不确定性分析方法比较了预动时间为确定值和随机变量情况下人员疏散时间的不确定性,结果表明所考虑的不确定性参数越多,安全系数并不一定越大。为减少人员疏散时间参数敏感性分析的计算成本,本文耦合多项式混沌展开和方差分解法提出了人员疏散时间参数敏感性分析方法。基于人员疏散时间计算模型的一阶多项式混沌展开给出了人员疏散时间参数的线性敏感度指标。由于人员疏散模型是高度非线性的,基于二阶多项式混沌展开给出了人员疏散时间参数的非线性敏感度指标。研究结果表明：通过本文提出的敏感性分析方法与基于抽样的敏感性分析方法所获得的各参数对人员疏散时间影响的重要度排序是一致的,且本文提出的敏感性分析方法的计算时间成本较少。当出口宽度较小时,人员疏散时间对不确定参数的敏感性受出口宽度影响显著。当人员密度较大时,人员疏散时间对出口参数的敏感性受人员密度影响较小。为降低计算成本并保证建筑物的安全性,经济性和功能适用性,耦合人员疏散模型、多项式混沌展开法和两阶段嵌套式的Monte Carlo模拟建立了不确定性参数作用下人员疏散时间最优出口参数的确定方法。以建筑内一单室防火分区为例,对人员疏散时间最优的出口宽度、出口间距和出口位置进行研究。研究结果表明本文提出的耦合多项式混沌展开和两阶段嵌套式的Monte Carlo模拟技术使得在不确定参数影响下疏散出口参数优化所需的计算成本显著减少。随着人员疏散时间取值可靠性的增加,人员疏散时间的最小值将显著增加。对于人员疏散时间不同的取值可靠性,最小的出口宽度是显著不同的。随着出口宽度的增加,人员疏散时间的均值和标准差显著下降。然而,当出口宽度足够大的情况下,人员疏散时间的均值和标准差受出口宽度影响较小,可将其视为常数。当百人有效出口宽度在0.1～0.5m内设计时,出口宽度对安全系数的取值可靠性影响较小。此外,通过最大疏散距离最小化所确定的出口间距和出口位置往往不是最优的。随着人员疏散时间取值可靠性的增大,最优的出口间距呈增大的趋势,且最大疏散距离最小化所确定的出口位置与人员疏散时间最优出口位置的偏离度逐渐增加。当出口间距较小时,安全系数的取值可靠性受出口间距影响较小；而当出口间距增大到某一值时,人员疏散时间一取值可靠性所对应的最小安全系数开始下降。另外,随着出口位置与最大疏散距离最小化所确定出口位置偏离度的增加,人员疏散时间一取值可靠性下的最小安全系数将逐渐减少。
[Abstract]:At present, the two methods commonly used in the design of fire-proof design are the "the way of the place" specification and the performance-based fire-proof design. The design parameters and various indexes of the fire-proof design of the "the way of the place" are specified in more detail, but some flexibility is lacking. As a result of the unpredictability of the occurrence of the fire and the behavior of the personnel, the evacuation time of the personnel in the building fire presents significant uncertainty. The safety factor or conservative value method is used to deal with the randomness of fire and personnel evacuation. The choice of safety factor depends on the fire expert to a large extent, and the reliability of the safety factor is to be verified. The conservative value method is usually used for fire-proof design of buildings according to the most extreme cases, which often leads to the waste of construction resources. In view of the above problems, this paper presents a method for determining the time uncertainty of the personnel evacuation time based on the polynomial hybrid system and the method for determining the optimal exit parameters of the personnel under the influence of the uncertainty parameters. The main research work and results are as follows: According to the difficulty of the personnel evacuation process and the controllable parameters, the relevant parameters of the personnel evacuation time are divided into the uncertainty parameters and the export reference parameters. Number. There are more uncertainty parameters associated with the evacuation of personnel, such as the personnel density, the composition of the population, and the pre-motion time And the like. The outlet parameter refers to the parameters related to the outlet of the building, such as the outlet width, the outlet space, the number of outlets and the outlet position. In order to reduce the calculation time cost of the uncertainty analysis of the evacuation time of the personnel, this paper establishes the uncertainty of the personnel evacuation time by means of the personnel evacuation model and the polynomial hybrid expansion method. The method and the Monte Carlo simulation method are very close to the calculation precision and the calculation time is In addition, using the uncertainty analysis method presented in this paper, the uncertainty of the pre-moving time as the determination value and the evacuation time of the personnel in the case of random variables is compared. The results show that the more the uncertainty parameters are considered, the safety factor is not one. In order to reduce the cost of the sensitivity analysis of the time-parameter sensitivity of personnel evacuation, the coupling polynomial hybrid system and the variance decomposition method are used to make the personnel evacuation time parameter sensitive. The first order polynomial hybrid system based on the personnel evacuation time calculation model gives the linearity of the personnel evacuation time parameter Sensitivity index. Because the personnel evacuation model is highly nonlinear, the non-linearity of the personnel evacuation time parameter is given based on the second order polynomial hybrid system. The results show that the sensitivity analysis method presented in this paper is consistent with the importance of the parameters obtained by the sampling-based sensitivity analysis method to the evacuation time of the personnel, and the sensitivity analysis method proposed in this paper The cost is less. When the outlet width is small, the sensitivity of personnel evacuation time to uncertain parameters is controlled by the outlet width The influence of personnel's evacuation time on the export parameters is affected by the personnel's density when the density of the personnel is large In order to reduce the computational cost and to ensure the safety, economic and functional applicability of the building, the coupled personnel evacuation model, the polynomial hybrid expansion method and the two-stage nested Monte Carlo simulation establish the optimal exit for personnel evacuation time under the influence of the uncertainty parameters. The method of the determination of the parameters is to take a single-room fire-proof partition in the building as an example, and the exit width, the outlet space and the exit distance of the personnel with the optimal evacuation time are given. The research results show that the coupling polynomial hybrid system and the two-stage nested Monte Carlo simulation technology in this paper make it possible to optimize the parameters of the outlet parameters under the influence of uncertain parameters. The calculation cost is significantly reduced. With the increase of the reliability of the evacuation time of the personnel, the personnel evacuation time is the most important. The small value will increase significantly. For personnel evacuation time, the value is reliable and the minimum outlet width The degree is significantly different. With the increase of the exit width, the mean value of the person's evacuation time and the standard deviation is significantly reduced. however, when the exit width is large enough, the mean and standard deviation of the person's evacuation time are less affected by the exit width, It can be regarded as a constant. When the effective exit width of a hundred people is designed in the range of 0.1 to 0.5m, the value of the outlet width on the safety factor The reliability impact is small. In addition, the determined exit spacing and outlet location are minimized by minimizing the maximum evacuation distance It is often not optimal. With the increase of the reliability of the evacuation time of the personnel, the optimal exit distance is increased, and the maximum evacuation distance is the optimal exit position of the exit position and the personnel evacuation time. The deviation degree is gradually increased. When the outlet interval is small, the reliability of the safety factor is less influenced by the outlet spacing; and when the outlet interval is increased to a certain value, the minimum safety corresponding to the value reliability of the personnel evacuation time The total coefficient starts to decrease. In addition, as the exit position and the maximum evacuation distance minimize the deviation of the exit position, the minimum safety under the reliability of the personnel evacuation time
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TU998.1

【参考文献】