SA/GA Hybrid Optimization Method Using a Procedure of Mesh Size Change

• Published in: TRANSACTIONS OF THE JAPAN SOCIETY OF MECHANICAL ENGINEERS Series C Vol. 77; no. 784; pp. 4367 - 4381
• Main Authors: SAKAMOTO, Yuichiro, BONKOBARA, Yasuhiro, KONDOU, Takahiro, ABE, Kenji, TANIGUCHI, Tomoyuki
• Format: Journal Article
• Language: Japanese; English
• Published: The Japan Society of Mechanical Engineers 2011
• Subjects: Framed Structure; Genetic Algorithm; Hybrid Optimization Method; Mesh Size Change; Multi-Objective Optimization; Optimal Design; Simulated Annealing
• ISSN: 0387-5024; 1884-8354
• DOI: 10.1299/kikaic.77.4367

SA/GA hybrid optimization method developed in a previous study is improved in order to achieve high computational performance. In the previous study, design variable vector elements are placed on the lattice point in order not to spawn too many similar solutions. Although this approach restrict the number of combination patterns related to design variable vector, if the interval of lattice points is narrow, members of search individuals are more likely to fall into local solutions because narrow interval makes huge combination pattern. In the proposed method using a procedure of mesh size change, in order to improve the computational performance, mesh size which means interval of the lattice points is set larger at the beginning of the optimization. Then, interval is narrowed as calculation is more progressed. The validity of the present method is confirmed by the calculation examples. In addition, assistance method which determine when mesh size should be changed using convergence ratio is applied to the proposed method in order to achieve more high computational performance. The calculation example shows the validity of this additional method.