Title: An evolutionary-inspired optimisation algorithm suitable for solid mechanics engineering inverse problems
Authors: J.F. Caseiro; A. Andrade-Campos
Addresses: Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, Portugal. ' Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, Portugal
Abstract: In this work, a direct search optimisation algorithm based on differential and evolutionary methods, as well as in the behaviour of animals is developed. This algorithm is applied to inverse engineering problems. Firstly, the developed algorithm is validated and compared with existing optimisation algorithms using a set of composite functions specially design for that purpose. In the second phase, the algorithm is applied to engineering and computational mechanics problems. The 120-bar dome truss problem that is solved using the finite element method (FEM) is analysed. Subsequently, the compression of a cylindrical billet and the plate with a central cut-out problems are analysed. In these two problems, the function to minimise is given by a commercial FEM code. Finally, the algorithm is applied to a constitutive model parameter identification problem. The develop algorithm obtains good results and a good convergence rate.
Keywords: differential evolution; particle swarm optimisation; PSO; finite element method; FEM; structural optimisation; parameter identification; shape optimisation; metal forming; composite functions; Levenberg-Marquardt algorithm; simulated annealing; evolutionary algorithms; solid mechanics; inverse problems; compression; cylindrical billets.
International Journal of Mechatronics and Manufacturing Systems, 2011 Vol.4 No.5, pp.415 - 440
Published online: 07 Mar 2015 *Full-text access for editors Access for subscribers Purchase this article Comment on this article