基于地理信息系统的城市建筑能耗模型建立方法

Energy Models of Urban Building Stock Based on Geographic Information System

为降低城市快速膨胀导致的能源消耗及温室气体总量,城市环境建筑能耗分析已经成为研究热点,但是城市环境中区域建筑能耗模型尚处于初步研究阶段。随着我国城市地理信息系统及相关数据库的完善,为建立准确且预测功能完善的建筑能耗模型提供了便利条件。重点探讨根据城市地理信息系统及相关统计数据,建立三维城市环境建筑动态能耗模型。重点分析了可能遇到的主要问题及解决方法,包括如何简化地理信息系统的数据、考虑城市建筑复杂性、模型建立的自动化完成、全局敏感性分析节能措施等。并且用2个实例说明处理空间规模不同的建筑群可采用的不同建模方法。研究不仅为城市节能提供指导,也为建筑信息模型的发展和促进城市规划领域中建筑能耗模型的广泛应用提供了坚实基础。

Assessment of urban building energy has become an active research field in order to reduce energy consumption and associated carbon emission due to fast urban expansion. However,building energy models at urban scale are still in their infancy. G ISs（ G eographic Information Systems） at urban environments are becoming more mature and increasingly available,which provides a new opportunity to develop the G IS-based urban energy models. This paper focuses on the main problems when creating building dynamic energy models at urban scale based on the G IS information and data from other sources.The procedure of constructing energy models at urban scale can be divided into seven steps： collecting original data related to building energy models in a city; pre-processing these data for the specific purpose of building energy analysis; arranging these data in a readable way suitable for the specific energy program（ such as Energy Plus or De ST）; creating energy models using the dynamic building energy program and computer language for parametric analysis; running these energy models using multi-core workstations or other high-performance computers to expedite the calculation for a number of building energy models; validating these energy models in comparison with measured data or typical energy data;finally applying the global sensitivity analysis to identify effective energy saving measures at urban scale.The issues are discussed, including simplifying the G IS data, considering the complexity of urban buildings,automation of creating building energy models,global sensitivity analysis for energy saving analysis. The G IS data usually have unnecessary data for the purpose of building energy analysis and,therefore,it is necessary to delete the information unrelated to building energy analysis from the G IS data.The automation of creating energy models should be implemented using the computer languages,such as R or MATLAB,because a large number of buildings are involved for urban-scale energy analysis. Two examples are presented to demonstrate how different energy methods for building stocks can be created for various spatial scales. One is to create the dynamic building energy models for the buildings（ around 100,000 models） located at the Westminster in London,using the Energy Plus program. The other is to construct the 3D energy models（ approximate 100 models） for the campus buildings of Tianjin University of Science and Technology,China. For the Westminster model,the G IS data are more readily available,while the G IS data need to be created from the conventional urban map for the Tianjin campus model. The other difference between these two projects is that the Westminster energy analysis implements the one-zone-per-floor energy models,whereas the five zones（ one interior and four perimeter zones） are created for per floor level in the Tianjin campus model. As a result,the Tianjin campus model can better take into account the characteristics of heating/cooling loads in different directions of a building. This research would be helpful to provide guidance for creating the building stock energy models at urban scale.Moreover,it is also useful for the development of building information modeling and urban planning by using the methods proposed from this paper.