本文回顾了多区域室内空气质量（IAQ）模型在住宅建筑中应用的经验验证研究。虽然模型也经过了分析验证和模型间比较，但本综述侧重于实证验证工作。在大多数报告中，实验数据仅与一个模型的预测进行比较——通常是CONTAM或COMIS模型。然而，模型间比较表明，这些模型与其他多区域模型之间是一致的，因此大多数比较可以推广到所有多区域模型。 很少有实证检验报告了测量和预测之间比较的统计分析。如果文献中有足够的数据，则进行了额外的统计分析并报告。此外，大多数发表的报告都没有涉及测量不确定度的问题。由于范围的限制、对实验和/或建模程序的描述不够详细、缺乏严格的统计分析、仅包含小范围的气流和浓度、验证数据集的独立性问题以及其他缺陷，任何单一报告的多区域室内空气品质模型验证工作都不能被视为完整。 然而，如果考虑到已发布的验证工作，可以得出结论，知识渊博的用户可以合理预测以烟囱驱动或带有惰性污染物的通风流为主的住宅规模建筑的换气率、分区间流量和污染物浓度。相比之下，对于高风速、活性污染物或特殊情况（如环境污染）的应用，显然需要进行更多的工作。此外，如果进行更多的统计分析，并报告更多关于测量和建模的细节，未来的模型验证工作将更加有用。 单位:SICitation:专题讨论会，ASHRAE交易，第107卷，pt。2.

This paper reviews empirical validation studies of the application of multizone indoor air quality (IAQ) models to residential-scale buildings. This review focuses on empirical verification efforts, although models have also been subjected to analytical verification and inter-model comparisons. In most reports, experimental data were compared to predictions of only one model - typically, either the CONTAM or COMIS models. However, inter-model comparisons have demonstrated consistency between these and other multizone models so most comparisons can be generalized to all multizone models. Few of the empirical verifications reported statistical analyses of the comparison between measurements and predictions. Where sufficient data were available in the literature, additional statistical analyses have been performed and reported. Also, most published reports did not address the issue of measurement uncertainty. No single reported multizone IAQ model validation effort can be considered to be complete due to limitations in scope, inadequate detail describing experimental and/or modeling procedures, lack of rigorous statistical analysis, inclusion of only small ranges of airflows and concentrations, questions on independence of validation datasets, and other shortcomings. However, if one considers the body of published validation work, it may be concluded that a knowledgeable user can expect to make reasonable predictions of air change rates, interzonal flows, and contaminant concentrations for residential-scale buildings dominated by stack-driven or ventilation flows with inert pollutants. In contrast, more work is clearly needed for applications with high wind speeds, reactive contaminants, or specialized situations such as ambient Additionally, future model validation efforts will be more useful if more statistical analyses are performed and if more detail on both the measurements and modeling are reported.Units: SI