Abstract: The key technique of cellular automata operation is to establish transition rules suitable to corresponding problem. Traditionally, the indirect transition rules are established by experience or according to the results obtained through other arithmetic. To overcome the shortcomings of localization and long operation time caused by adopting indirect transition rules, genetic algo-rithms based direct transition rules are discussed. The direct local transition rules established by genetic algorithms are used in the thin-plate topological layout minimum weight optimization. Multi-objective optimization model is created and optimum transition rules are obtained by the evolution of genetic algorithms. From the simulation results, it is observed that the local direct rules obtained by evolution are effective in the self-organization process of the complex problems.

Key words: Cellular automata, Finite element method (FEM), Genetic algorithm, Multi-objective optimization, Structural topology optimization