AI 中文总结
该研究采用CFD正向风险预测模型，通过准蒙特卡洛模拟发现乘客头部前后及侧向位置显著影响空气传播病原体感染风险，且现有统计矩方法无法充分捕捉该依赖关系。
AI 中文摘要
我们采用基于计算流体动力学（CFD）的正向感染风险预测模型，研究自然头部运动对空气传播病原体引发的感染风险的影响。通过表征每位乘客呼吸区位置与朝向的参数概率分布来呈现头部运动，采用准蒙特卡洛模拟获取感染风险分布。研究发现，头部前后及侧向位置对感染风险存在显著影响，当利用座位局部感染风险预测指导设计与政策决策时，需纳入该影响以提升预测稳健性。与基于采样的蒙特卡洛方法不同，采用一阶二阶矩及更高阶方法计算的风险分布统计矩与敏感性估计，未能充分捕捉头部运动分量与感染风险间的依赖关系。
英文摘要
We investigate the influence of natural head movement on the infection risk posed by airborne pathogens using a CFD-based forward risk prediction model. Quasi-Monte-Carlo simulations are used to obtain the resulting infection risk distributions by representing head movement via probability distributions of parameters describing the position and orientation of each passenger's breathing zone. A significant impact of fore/aft and lateral head position on infection risk was found and should be accounted for to increase robustness when predictions of local, seat-specific infection risks are used to guide design and policy decisions. Unlike the sampling-based Monte Carlo approach, estimates of the statistical moments of the risk distributions and sensitivities, calculated using first-order second-moment and higher-order methods, were found to inadequately capture the dependencies between head movement components and infection risk.
Comments21 pages, 15 figures
