本文的目的是提供一种优化建筑立面的方法，以达到供暖、制冷和照明负荷的三重目标，从而获得最低的年能源成本。需要优化的变量是窗框的尺寸和遮阳系统的深度。使用建筑性能模拟程序（TRNSYS）计算能量负荷。使用模拟程序Daysim计算日光标准。该标准定义为照度高于500勒克斯阈值时的积分时间。当阈值低于500勒克斯时，则需要人造光。 这些变量对目标有相反的影响:窗户网格尺寸和遮阳深度可能对年度能源成本产生相反的影响，在冬季增加室内太阳热增益和日照，在夏季导致过热问题。因此，提出了一种方法来寻找总能源成本的最优解。为了寻找导致最低年能源成本的最优立面设计变量，采用了一种优化方法——遗传算法。这种方法应用于一个单独的办公室。结果表明，遗传算法在寻找具有对立目标的最优解时可以节省时间，有助于建筑师尽早做出决策- 设计阶段的决策。引用:ASHRAE论文CD:2014 ASHRAE冬季会议，纽约

The objective of this article is to provide a methodology for optimizing building facade with respect to the triple objective of heating, cooling and lighting load, therefore to get the lowest annual energy cost. The variables to optimize are the dimension of window grid and the depth of shading system. Energy load is computed using building performance simulation program (TRNSYS). A criterion of daylight was calculated using the simulation program Daysim. The criterion is defined as the integrated time when the illuminance is above a threshold of 500 lux. When the threshold is below 500 lux, then artificial light is required. The variables have antagonistic effects on the objectives: window grid dimension and shading depth may have opposite effects on annual energy cost, by increasing indoor solar heat gain and daylight during winter time and leading to overheating problems during summer time. Therefore, a methodology is proposed to find the optimal solutions for the total energy cost. An optimization method - genetic algorithm, was performed in order to find the optimal façade design variables leading to the lowest annual energy cost. This method was applied to a single office room. The result shows that genetic algorithm could save time when looking for the optimal solutions with antagonistic objectives and would help architects to make early-design-stage decisions.