建立高效准确的自然通风换气量的仿真计算方法是开展住宅自然通风研究的关键。然而, 住宅的自然通风呈现强随机性、高动态性和多区域性的特点, 增加了获取住宅自然通风换气状况的难度。该文建立了一种住宅多区域动态自然通风换气量的仿真计算方法, 以人体释放CO₂作为示踪气体, 基于各态遍历的思路和空气质量守恒原则, 建立多区域空气流动模型, 对每个区域的CO₂瞬态质量守恒方程实施Kalman滤波法, 进而判断获取各时刻的空气流动模型和自然通风换气量。通过在可控风量和气流方向的实验环境中的实测分析, 验证了该方法对住宅多区域动态自然通风换气量预测的可靠性; 针对实际住宅计算了多区域动态自然通风换气量, 评价了该方法在工程实际中的适用性。该研究将有助于住宅室内环境的准确预测和室内环境健康状况的合理评价。

Objective: Natural ventilation in real residential environments is characterized by dynamic multizone airflow. Ignoring either of the two features can lead to inaccurate assessments of natural ventilation in residences. The multiple tracer gas method has traditionally been used to investigate this multizone airflow within residential settings. However, it is impractical for large-scale applications owing to the high costs associated with the measurement process, potential disruption to occupants' daily lives, and the need for stable ventilation conditions. As a result, accurately measuring air exchange rates remains a significant challenge. A more in-depth study of the measurement method for dynamic multizone airflows is urgently required. This study proposes a simulation method for identifying dynamic multizone airflows in naturally ventilated residences. Methods: This method utilizes CO₂ emitted by occupants as a tracer gas to study multizone airflow in residential buildings. It considers all feasible multizone airflow patterns using a traversal approach and the air volume conservation principle. Furthermore, to strike an optimal balance between effectively tracking the dynamic characteristics of natural ventilation and minimizing noise sensitivity, a transient indoor CO₂ mass balance equation associated with the Kalman filter is applied to each zone. The resulting time series of air exchange rates can be presented for each airflow pattern. These rates are then evaluated to identify the pattern that most closely aligns with the actual airflow pattern and the corresponding outdoor-indoor air exchange rates and interzonal airflow rates. Furthermore, two validation experiments were conducted in an unoccupied two-bedroom apartment with controllable ventilation patterns to validate the method. Subsequently, the method was employed in an occupied apartment, utilizing measured indoor CO₂ concentrations and occupancy data for each zone to produce the time series of air exchange rates. Results: The comparison among the calculated air exchange rates using the proposed method and experimental data indicates that 85% of the calculated values have absolute errors within RHHZ_177;0.2 h^-1, and 95% fall within RHHZ_177;0.4 h^-1. Furthermore, 75% of the calculated values have relative errors within RHHZ_177;10%, and 95% are within RHHZ_177;20%. The calculated air exchange rates and airflow directions closely match the experimental conditions, indicating that the method proposed in this study effectively represents the multizone aspects of natural ventilation in residential environments. Moreover, the applicability of the method to real residences is demonstrated through its application in an occupied apartment. The calculated air exchange rates for each zone during the measurement period, after filtering out anomalous results, fall within a reasonable range. These results present the airflow patterns that characterize the multizone nature of dynamic natural ventilation. Conclusions: Natural ventilation is complex owing to its multizone nature and time dependence, leading to data scarcity. This method effectively addresses this gap by quantifying the multizone representation of dynamic airflows in residences on a large scale. Understanding indoor–outdoor air exchange rates and interzonal airflow rates is pivotal, as these parameters significantly influence indoor thermal conditions and air quality. In this regard, this study offers a valuable and practical approach to comprehensively understanding natural ventilation and its effects on occupants' health conditions in real residential environments.