为了在保持室内空气质量（IAQ）的同时降低能耗，基于CO2的需求控制通风（DCV）得到了广泛的研究，并成为未来通风系统的发展趋势。在各种先进的传感器中，实时占用估计（人员计数）传感器在基于二氧化碳的需求控制通风（DCV）系统中发挥了重要作用。然而，由于在瞬态条件下，室内颗粒物扩散速度慢于乘员存在的动态变化，因此实际CO2浓度通常会滞后于实际乘员数量，而动态渗透可能会缓解滞后效应。因此，本文进行了模拟，以深入了解CO2职业估算传感器的应用- 基于DCV和现实世界中的动态渗透。分别基于马尔可夫链模型和高方差高斯模型，研究了两种不同空间类型的两种不同的随机占用剖面。已经观察到，由于滞后效应，与马尔可夫链占用模型相比，高斯占用模型的实际CO2浓度更有可能超过设定值。幸运的是，模拟结果表明，由于可操作窗户等建筑围护结构的动态操作，动态渗透消除了滞后效应。因此，对于不同占用剖面的空间，建议通过将系统与建筑围护结构的动态操作相结合来运行DCV，以改善室内空气质量。 引用:2019年冬季会议，佐治亚州亚特兰大，会议论文

In order to reduce energy consumption while maintain high indoor air quality (IAQ), CO2-based Demand controlled ventilation (DCV) has beeninvestigated a lot and become trend of the future ventilation system. Among different advanced sensors, the real-time occupancy estimation (people counting)sensor has played an important role in CO2-based demand control ventilation (DCV) systems. However, since particle diffusion indoors is slower thandynamic changes of occupant presence in transient conditions, the actual CO2 concentration will generally lag behind actual number of occupants whiledynamic infiltration may mitigate the lag effect. Therefore, simulations were conducted in this paper to provide insight into the application of occupancyestimation sensors for CO2-based DCV and dynamic infiltration in the real world. Two different stochastic occupancy profiles for two different space typeshave been investigated, which were generated based on Markov-chain model and Gaussian model with high variance, respectively. It has been observed thatdue to the lag effect, the actual CO2 concentration with Gaussian occupancy model is more likely to exceed the set-point value than that with Markov-chainoccupancy model. Fortunately, the simulation results showed that dynamic infiltration owing to dynamic operation of building envelopes such as operablewindows eliminated the lag effect. Therefore, for the space with variant occupancy profile, it is suggested to operate DCV by integrating the system withdynamic operations of building envelopes in order to improve the indoor air quality.