Study on Structural Optimization Technique Using Evolutionary Cellular Automata. Local Direct Rule Based on New Transition Function.

Abstract

In this paper, a local direct rule based on a new transition function of the Cellular Automata (CA) is presented. It is the structural optimization method by the Evolutionary Cellular Automata (ECA). The local direct rule uses the neighborhood states as the CA's input. The ECA is the CA optimized by the Genetic Algorithm (GA) for the certain problems. This new transition function makes to decrease drastically the degrees of local direct rule freedom depending on the number of the neighborhood cells. The design conditions of the structure are introduced as virtual cells to apply the same rule in all the design space. A local direct rule here is decided by the evolutionary calculation of GA. The method is applied to the minimum weight design problems and the adaptive solutions are obtained. From the results, it is observe that the local direct rule based on a new transition function is effective in the self-organization problem of the complex system.