The paper is devoted to an application of the evolutionary algorithms and the finite element methodto genetic generation of 3-D structures. The shape, topology and material of the structure aregenerated for various optimization criteria: the minimum of volume, minimum of the stressfunctional, the minimum displacement functional, the maximum of the first eigenfrequency and themaximum of difference between first and second eigenfrequency.Evolutionary methods have found various applications in mechanics, especially in structuraloptimization [1]. The main aim of these methods is the simulation of biological processes based onheredity (genetics) and on the natural selection (the theory of evolution) to create the optimalindividuals (solutions) presented by single chromosomes. Evolutionary models of computation canbe performed by using genetic algorithms.During the evolutionary process the domain, its boundary, and the field of Young’s modulus canchange for each generation. The distribution of values of Young’s modulus and as a consequence theshape and topology of the structure is described by hyper surface which is controlled by genes. Genescreate a chromosome which is a member of a population of solutions.Numerical example: Optimization of 3-D structure resembling stairs – minimum of the volume.Some results are shown in Fig. 1.
展开▼
