Multi-objective optimization of the dimensions of a dwelling building based in energy and structure simulation

Abstract The development of computer technology in the past few decades led optimization associated with parameterization to develop designs that perform better, with or without the use of computer simulation. The purpose of this study was to apply this method to optimize the shape of a single family residential building with the goal of finding the design that presented the best thermal performance with the lowest cost of construction materials, considering structural constraints. The software Rhinoceros, Energy Plus and plug-ins for parametric modelling, information exchange between software and conduction of optimization were used in this study. Two simulations were performed, one considering the cost of construction materials and combined heating and cooling degree-hours, the second considering cost and heating degree-hour, which produced more relevant results. Two cases of each Pareto Front were selected for analysis, out of 19.8 thousand and 27.4 thousand cases obtained through evolutionary algorithms. Despite the limitations presented by some tools, the study demonstrated that there is great potential for implementation of this technology in the development of architectural projects.

摘要 近数十年来计算机技术的进步，使得结合参数化的优化方法得以催生性能更优的设计方案，无论是否借助计算机仿真（computer simulation）手段。本研究旨在将该方法应用于独栋住宅建筑的形状优化，目标是在兼顾结构约束的前提下，寻得热性能最优且建筑材料建造成本最低的设计方案。本研究采用了Rhinoceros、Energy Plus软件，以及用于参数化建模（parametric modelling）、软件间信息交互与优化实施的插件工具。本次研究开展了两组仿真实验：第一组同时考量建筑材料建造成本与采暖制冷度日数（heating and cooling degree-hours），第二组仅考量建造成本与采暖度日数（heating degree-hour），后者产出了更具参考价值的结果。研究通过进化算法（evolutionary algorithms）共得到1.98万组与2.74万组候选方案，并分别从两组实验的帕累托前沿（Pareto Front）中选取两组案例进行分析。尽管部分工具存在一定局限性，但本研究证实了该技术在建筑项目开发中具备极大的应用潜力。