Skip to Main Content
The back-analysis of mechanics parameters needs iterative forward calculating, resulting in low efficiency; meanwhile the Particle Swarm Optimization (PSO) algorithm and other optimization algorithms are exposed to local optimum possibilities. In this paper, DEPSO-ParallelFEM, a system integrated of an algorithm of hybrid particle swarm with Differential Evolution (DE) operator, termed DEPSO, and parallel Finite Element Method (FEM), is proposed to solve these problems. DEPSO guarantees the particle to escape from local minima by enhancing particlepsilas diversity through the combination of PSO operator and DE operator; and parallel FEM is applied to improve the computing speed and precision by adopting the techniques of Cluster of Workstation (COW), MPI, Domain Decomposition Method (DDM), and Object-Oriented Programming (OOP) and so on. A computational example proves that this system is of excellent parameter exploration capability and high speed; thus it is of great academic value and significant applicable value.
Computational Intelligence and Natural Computing, 2009. CINC '09. International Conference on (Volume:1 )
Date of Conference: 6-7 June 2009