Soft computing in structural dynamics

Abstract

The paper is devoted to new computational techniques in structural dynamics where one tries to study,model, analyse and optimise very complex phenomena, for which more precise scientific tools of the past were incapable of giving low cost and complete solution. Soft computing methods differ from conventional (hard) computing in that, unlike hard computing, they are tolerant of imprecision, uncertainty, partial truth and approximation. The paper deals with an application of the bio-inspired methods, like the evolutionary algorithms (EA), the artificial immune systems (AIS) and the particle swarm optimisers (PSO) to optimisation problems. Structures considered in this work are analysed by the finite element method (FEM) and the boundary element method (BEM). The bio-inspired methods are applied to optimise shape, topology and material properties of 3D structures modelled by the FEM and to optimise location of stiffeners in 2D reinforced plates modelled by the coupled BEM/FEM. The structures are optimised using the criteria dependent on frequency, displacements or stresses. Numerical examples demonstrate that the methods based on the soft computation are effective for solving computer aided optimal design problems.