Reanalysis assisted metaheuristic optimization for free vibration problems of composite laminates

The metaheuristic optimization methods have been widely used in complex composite material design. However, it is still difficult to apply these methods to complicated problems due to expensive cost of evaluations. Therefore, to reduce the expensive cost of finite element (FE) analysis in the composite laminate optimization problems, the surrogate assisted optimization (SAO) method has been well developed. However, considering the low accuracy of surrogate model for complicated cases, the SAO is always difficult to converge to the global optimum. In this study, a fast and accurate solver, the reanalysis algorithm is extended to vibration problem and integrated to improve the efficiency of metaheuristic optimization of composite laminates instead of surrogate model. In the proposed method, the shifting method is used in each iterative step. To validate the performance of suggested method, several numerical examples are presented. The results demonstrate that reanalysis assisted particle swarm optimization (RPSO) is much more efficient for free vibration problems of composite laminates than the PSO. Moreover, such strategy can be extended to other metaheuristic algorithms easily.