綠建築改善策略之成本效益研究－應用遺傳算法於最佳化及多目標分析

Abstract

目前市面上的既有建築之節能改善策略在選擇上非常多樣化，而相異之既有建築物因類型和設計的不同，往往也會有其最適用並且最有效率的改善策略組合。以往在策略的挑選上只能依靠建築師的工作經驗做為判斷的依據，卻缺乏一套有系統的挑選標準。所以本論文針對既有建築在改善策略的選擇問題上，透過遺傳演算法對眾多策略做適當的挑選，再結合建築能源模擬軟體EnergyPlus計算建築物的全年耗能，用以進行改善策略組合的最佳化運算分析，最後得到在固定的改善成本下之最佳改善策略搭配組合。 此外，在既有建築改善的過程中，除了考慮改建後的節能量之外，改善策略所需花費的成本往往也會影響最後選擇的決策。所以本論文將多目標規劃中的限制式法與權重法運用於挑選改善策略之系統上，對施工成本與節能效益之關係進行討論分析，並找出此兩個衝突的目標之間的權衡曲線(Trade-off line)，以提供決策者在挑選改善策略組合時之參考依據。 模擬結果顯示出，將遺傳演算法結合EnergyPlus應用於既有建築的改善策略組合挑選上，可以有效的降低最佳改善策略的挑選時間，而挑選到的改善策略組合雖然不保證為最佳組合，但會是非常接近最佳解的近似策略組合。最後由多目標規劃所得到的權衡曲線，可以有效預估使用成本與降低耗能量之間的關係，以提供決策者做策略選擇的判斷依據。

Nowadays, there are various choices of reformation strategies for saving energy for existing buildings. The different building types and designs also have the most appropriate and efficient combination of renovation strategies. However, in the past, the selection of strategies can only rely on architects’ experiences without any systemic principle. Thus, this study attempts to obtain the best combination of reformation strategies in fixed cost through the simulation model, EnergyPlus coupled with genetic algorithm. The former is utilized to analyze the annual energy consumption for improving the optimization of reformation strategies, and the latter is used to make a proper choice among alternatives. In addition, the final decision will be affected by both the energy saving and the cost of reformation strategies during the building renovation process. Hence, this study also makes an effort to find a trade-off line for a collision between two objectives to provide a basis of selection, by means of the constraint method and weight method in multi-objective programming for selections of reformation strategies. The results represent that the use of Energyplus coupled with genetic algorithm for the application of selections of building renovation strategies can effectively shorten the searching time. Besides, albeit the selected combination of reformation strategies is not guaranteed to be the best, it will be close to the optimal solution at least. Finally, the trade-off line derived from multi-objective programming can effectively estimate the correlation between the use cost and the decrease of energy consumption to offer a basis of selection of strategies to decision-makers.