基于性能模拟与优化算法的既有办公建筑绿色改造敏感度预测模型——以寒冷地区为例

Facing a huge amount of existing office buildings with aging structure, poor comfort, obsolete equipment , high energy consumption and other issues. the current research still lacks the trade-off judgments of the retrofit measures and quantitative analysis of the comprehensive improvement of multiple performance targets in the early design stage. The existing office buildings in cold area are selected as the research object in this research program. By use of building performance simulation tools and building performance optimization algorithm as the main technical means, this research focuses on the Sensitivity Model of office building retrofit in the early design stage. Based on the field research and case study of the influence mechanism for building performance, the design strategies of green retrofit for existing office buildings are proposed. Then, this research presented a working procedure that is employed in the decision-making process of office building retrofits in China from an architect’s perspective, coupling multiple performance targets. Following a genetic optimization logic, the retrofit strategies are served as the variables, environmental comfort indicators and energy consumption indicators are served as objective functions, and the incremental costs are served as constrains. “Multi-variate exhaustiveness” and “Multi-objective optimization" sensitivity prediction model are established. The research results of the project will provide the scientific basis for the optimization of the strategic measures in the existing residential buildings, and provide guiding significance and design logic to both of the performance-oriented retrofit practice and architectural form innovation.

面对既有办公建筑量大面广、功能过时、舒适性差、设备陈旧、用能强度大等问题，当前的研究尚缺乏设计阶段改造措施的权衡判断和多性能目标综合提升的量化分析。本项目以寒冷地区既有办公建筑为研究对象，以性能模拟和优化算法为主要技术手段，对设计阶段的既有办公建筑绿色化改造设计方法和预测模型展开研究。通过建筑性能影响机制研究，提出绿色化改造设计策略，建立耦合多性能目标的绿色化改造设计方法，并进一步构建以改造措施为变量、以环境舒适度指标和运行能耗指标为目标函数、以增量成本为约束条件的遗传优化模型，探索既有办公建筑绿色化改造中的“多变量穷举”和“多目标优化”敏感度预测工具。项目研究成果能够为既有办公建筑绿色改造中的措施优选提供科学依据，对性能导向的改造实践有指导意义，同时为建筑形式创新提供设计逻辑。

通过对31个既有办公建筑典型案例的分析及调研，建立寒冷地区绿色化改造样本库探索建筑性能影响机制，提出相应的形式、构造、系统、设备改造优化策略，并结合典型案例剖析，形成4种以“热工/采光/通风”性能判断为原则的寒冷地区办公建筑分类，扩充寒冷地区改造策略框架下适用的具体措施。以此为基础，构建了针对设计阶段的、基于量化模拟技术的、耦合风、光、热环境的绿色化改造设计方法，并详细探讨了其在“性能诊断——改造设计——预测反馈”设计过程中的风、光、热性能指标、技术实现方法、参数设置。继而，基于性能表现的建筑分类筛选典型模型进行绿色化改造设计方法应用，研究得到每一类典型模型中对风、光、热优化效果明显的不同单项措施所带来节能敏感度、减排敏感度和成本效益敏感度，为绿色化改造中策略措施的权衡提供定量依据，同时为建筑形式创新提供设计逻辑。研究发现，经过合理设计的致力于风、光环境优化的被动形式策略，不会对能耗与碳排放产生明显负面影响，甚至能够达到节能减排效果；办公建筑典型模型以节能为目标应优选系统策略，以减排为目标应优选设备策略。进一步，运用非支配精英遗传算法，实现典型模型的多目标综合优化，探究在舒适度提升与碳排放降低目标下最优的策略组合，辅以成本增量因素的判断，确定性能导向的改造的最优解。最后，量化分析24种形式措施，20种构造措施，15种系统措施和7种设备措施对于不同方案的性能优化效果，用拉丁超立方抽样（LHS）方法生成各设计变量的随机组合，通过标准化回归系数（SRC）确定各设计变量的全局敏感度排序。为了方便建筑师使用，本研究基于JAVA编程开发了多目标优化敏感度预测模型。项目研究成果能够为既有办公建筑绿色改造中的措施优选提供科学依据，对性能导向的改造实践有指导意义，同时为建筑形式创新提供设计逻辑。