EVOLUTIONARY OPTIMIZATION OF MECHANICAL STRUCTURES

Abstract

In this paper an optimization method for mechanical structures, based on a combination of genetic algorithms and neural networks (NN), is discussed. The NN is used as an approximation of the finite element analysis. Three examples show that using these methods in order to optimize mechanical systems is an efficient way. Examples presented are optimization of the position of dampers on beams, optimization of the position of stiffeners on plates or shells, and optimization of the damping by viscoelastic material constraints.

Keywords:

• Evolutionary optimization
• genetic algorithms
• neural networks
• mechanical structures