SCIPRESS FORMA
Forma, Vol. 15 (No. 2), pp. 133-139, 2000
Original Paper

Multi-Objective Optimization of Spatial Truss Structures by Genetic Algorithm

Yasuhiro Kida, Hiroshi Kawamura, Akinori Tani and Atsushi Takizawa

Department of Architecture and Civil Engineering, Kobe University, Rokkodai, Nada 1-1, Kobe 657-8501, Japan
E-mail: kawamura@arch. kobe-u.ac.jp

(Received November 19, 1999; Accepted January 27, 2000)

Keywords: Genetic Algorithm, Artificial Life, Fuzzy Theory, Spatial Truss Structure, Optimization

Abstract. In this paper, architectural constructions composed of spatial truss units are considered by using a new technique of artificial life. Such spatial truss structure has flexibility and adaptability to environments. At first, equilateral tetrahedral (regular tetrahedral) units, which are basic space units having three dimensions, are proliferated to specified directions at random. At every proliferation step, various architectural evaluations and selections are performed according to obtained degrees of adaptation to environments. The purpose of this paper is to compose a system for a variety of automatical formation, which satisfies the given architectural and environmental conditions by using Genetic Algorithm (GA) originated by HOLLAND and GOLDBERG. The effectiveness of the proposed system is shown by digital and graphical simulations.


[Full text] (PDF 1.1 MB)