應用遺傳演算法於建築設計平面空間配置之研究

Applications of Genetic Algorithms to Architectural Plan Design

本文嘗試以「遺傳演算法（Genetic Algorithms，簡稱GA）」做為建構一套電腦輔助建築設計平面空間配置系統之方法，並藉以協助建築設計者在從事建築平面空間配置設計時之輔助參考工具。在本研究中，我們希望能從找出住宅設計時所須考量的因子，例如：空間量的大小安排、空間配置的規則等，同時應用遺傳演算法為運算基礎以使電腦能夠設計出理想的住宅平面方案，並進一步使電腦能夠輸出符合基地配置考量因素的平面設計方案。如此，本研究應可對在建築設計過程中所可能遭遇的設計問題提出較佳之解答方案，並進而達到真正之「電腦輔助設計」的目的。

The proposed research focuses on developing a computer aided architecture design （CAAD） system via genetic algorithm （GA）. Some important housing design factors such as the spatial massing, the individual space allocation etc. are utilized by the genetic algorithm for designing the better even 'optimal' architecture design plan. Thus, the CAAD system can really reach the objective of aiding designer while designing a specific architecture design plan. The main contents of this research include four aspects. The first one is to recollect the fine case studies in the world of their spatial allocation existing in architecture design plan. The second one is to study the feasibility of applying GA technique to CAAD. The third one is to build an appropriate system for housing design plan aid. And finally, the fourth aspect of this research is to adopt AutoLISP built in AutoCAD for program coding of the proposed CAAD system. Concluded from the four aspects mentioned above, the main contribution of this research is the development of a methodology （i.e., GA technique） for aiding architecture designer to design an architecture design plan more efficiently and effectively.