Building performance is focused on maintaining and increasing sustainability while enhancing occupants' comfort levels. Occupants' comfort depends on uncertain, interrelated personal, social and building factors. However, the relationships between these factors are not covered by performance assessment tools based on linear analysis approaches. This study proposes a probabilistic model based on Bayesian networks (BNs) to assess the comfort performance of a building. An extensive review and evaluation of the causal factors of building performance, supported by the results of a satisfaction survey, are the basis for the BN model. Sensitivity analysis is used to verify the BN model, and the proposed approach is tested on an existing building. Findings from this research can guide decision-makers in the facility management industry to assess and understand causal factors of occupants’ comfort to properly establish sustainable and cleaner production strategies.

建筑性能的重点是维持和增加可持续性，同时提高居住者的舒适度。乘员的舒适度取决于不确定，相互关联的个人，社会和建筑因素。但是，基于线性分析方法的性能评估工具未涵盖这些因素之间的关系。这项研究提出了一种基于贝叶斯网络（BNs）的概率模型来评估建筑物的舒适性。BN模型的基础是对建筑性能因果关系的广泛审查和评估，并得到满意度调查结果的支持。灵敏度分析用于验证BN模型，并在现有建筑物上对所提出的方法进行了测试。