Form for Free Surface Shell Structure Using Swarm Intelligence With Manipulation for Decent Solution Search

Abstract

In this paper, the particle swarm optimization (PSO) and the artificial bee colony (ABC) are applied to the structural morphogenesis for free surface shell structure. In general, the structural morphogenesis is treated as the optimization problem. However, searching of the decent solutions to maintain diversity on the design variable space and the objective function space is important in the structural morphogenesis. The decent solutions have comparatively high evaluation containing a global optimal solution and local optimal solutions, and diversity of the design variable space. These solutions are a lot of free surface shell forms satisfying the design constraint, and a designer's idea will receive stimulation from those forms. Therefore, PSO and ABC implementing a manipulation for decent solution search are proposed here. First, the proposed procedures and these computational properties are described. Next, our schemes are applied to the structural morphogenesis for free surface shell structure. Here, the single-objective optimization problem is treated. The objective functions are the total and the bending strain energy. From numerical results, a lot of free surface shell forms with diversity and structural stability are shown. Finally, we show evaluation scheme for robustness concerning the structural form, and the relation between the decent solution forms and the structural robustness forms is clarified.