Buildings are the major energy consumers in cities. To promote building energy efficiency and urban sustainability, researchers developed comprehensive digital simulation tools, such as CitySim, CityBES, and Autozoner, efforts to investigate, understand, and predict energy dynamics of urban buildings. However, different from independent buildings in rural areas, urban buildings have complicated interactions with surrounding buildings and the environment. Owing to the computational complexity of large building numbers, these models chose top-down or regression methods with universal weather conditions. However, a reliable prediction of building energy dynamics needs a comprehensive understanding of the thermal process of building physical elements, which is often weld by the bottom-up methods. Therefore, this study proposed a novel local-weather extended distributed adjacency blocks (LW-DAB) model that incorporates bottom-up distributed adjacency blocks (DAB) simulation and allows customizable local micro-climate clusters. This study also deployed a city building digital model with 1175 buildings and compared the results with the conventional typical morphological year weather-based simulation in three different climate regions. By introducing the impacts of urban island effects (UHI) based on building the morphological relationship, the results suggest that the proposed LW-DAB model not only allows independent micro-climate assignment for each building but also half the simulation time with 5% accuracy loss compared with the conventional whole city simulation model.

建筑物是城市中的主要能源消耗者。为了提高建筑的能效和城市的可持续性，研究人员开发了综合的数字仿真工具，例如CitySim，CityBES和Autozoner，致力于研究，理解和预测城市建筑的能源动态。但是，与农村地区的独立建筑物不同，城市建筑物与周围建筑物和环境的相互作用复杂。由于大建筑物数量的计算复杂性，这些模型选择了具有通用天气条件的自顶向下或回归方法。但是，对建筑物能量动力学的可靠预测需要全面了解建筑物物理元素的热过程，而热过程通常是采用自下而上的方法进行焊接。因此，这项研究提出了一种新颖的本地天气扩展分布式邻接块（LW-DAB）模型，该模型结合了自下而上的分布式邻接块（DAB）模拟，并允许自定义本地小气候集群。这项研究还部署了一个具有1175座建筑物的城市建筑物数字模型，并将结果与​​在三个不同气候区域的常规典型的基于形态学年天气的模拟进行了比较。通过引入基于建筑物形态关系的城市岛屿效应（UHI）的影响，结果表明，所提出的LW-DAB模型不仅允许为每座建筑物进行独立的微气候分配，而且还可以将模拟时间减少一半，而精度损失为5％与常规的整个城市模拟模型相比。