Earticle

초록

영어

This study focused on the problems of generative-based optimization design, recently become increasingly popular in the field of architectural design, such as overlooking form due to performance-oriented evaluation and designers' difficulties in analyzing huge number of solutions. In response to this, the study aims to incorporate a form-based clustering technique into the existing optimization process. To achieve this, we use Wallacei, an optimization tool, to perform optimization on the set example. Afterwards, the cluster composition and process characteristics of the derived optimal solution group were analyzed using various clustering techniques. In detail, K-means clustering was applied to the optimal solution group derived for the example set as the first step to confirm the adequacy and characteristics of clustering according to the form criteria. Second, we perform comparison analysis on the changes and differences in various clustering results derived from K-means clustering and three hierarchical clustering to derive rationality and efficient utilization of its application. Through this, the application of the clustering technique to generative design simplifies the huge solutions through form-based evaluation and clustering to increase user interpretation of the solutions and its usability, thereby complementing the completeness of optimization process required in architectural design.

목차

Abstract
1. 서론
1.1 연구배경 및 목적
1.2 연구범위 및 방법
2. 선행연구 및 관련 이론 고찰
2.1 선행연구 고찰 및 시사점
2.2 다중목적 최적화와 월라시
2.3 다양한 군집화 기법
3. 최적화 연산 및 결과 도출
3.1 분석 과정 및 방법
3.2 최적화 예제 및 데피니션 구성
3.3 최적화 연산 및 솔루션 도출
3.4 성능 기준에 따른 형태적 유형 도출
4. 군집화 연산 및 결
4.1 K-평균 군집화를 통한 파레토 프론트군의 분석
4.2 다양한 군집화 기법을 통한 군집화 구성의 분석
4.3 분석결과 해석 및 군집화 기법의 활용방안
5. 결론
REFERENCES

저자

• 이우형 [ Lee, Woo-Hyoung | 남서울대학교 건축학과, 부교수 ] Corresponding Author

참고문헌