Whole model empirical validation on a full-scale building

This paper describes an empirical validation study undertaken on two identical full-size buildings within the scope of the IEA ECB Annex 58 project. Details of the experimental configuration and monitoring are included, together with results from measurements and from predictions made by 21 modelling teams using commercial and research simulation programmes. The two-month, side-by-side experiment was undertaken on buildings with high levels of thermal mass and in a period with high solar gains. The detailed specification and associated measurement data provide a useful empirical validation dataset for programme testing. Results from the modelling demonstrate good agreement between measured data and predictions for a number of programmes, in both absolute predictions of temperatures and heat inputs as well as dynamic response. On the other hand, a significant number of user input errors resulted in poor agreement for other programmes, especially in the blind validation phase of the modelling methodology.本论文介绍了一项在国际能源署（International Energy Agency, IEA）建筑与社区计划（Energy in Buildings and Communities Programme, ECB）附件58项目框架下，针对两栋完全一致的全尺寸建筑开展的实证验证研究。研究涵盖了实验配置与监测方案的细节，同时包含了21个建模团队利用商用及科研仿真程序得到的测量结果与预测结果。该项为期两个月的并排对照实验，针对高热质量建筑开展，实验时段内太阳辐射得热量较高。详尽的参数规范与配套测量数据，可为仿真程序的测试提供极具价值的实证验证数据集。建模结果显示，多款仿真程序的预测结果与测量数据吻合度良好，在温度、热输入的绝对数值预测以及动态响应方面均表现优异。但另一方面，大量用户输入错误导致其余部分程序的预测结果与实测数据吻合度较差，尤其在建模方法的盲验证阶段这一问题更为突出。