Stacking sequence optimization of laminated composite structures using genetic algorithm with local improvement

Interest in the application of genetic algorithms (GA) to the stacking sequence of composite laminates has grown in recent years. However, the huge calculation time of GA is a major problem for designers. In this study, a local improvement is inserted into a standard GA, and the real calculation, by finite element analysis (FEA) for example, required in the local search is replaced by a regression model. Accordingly, the improved GA converges much sooner than a standard GA and the calculation time is greatly reduced. The regression analysis applies chosen trigonometric functions as base functions, and estimates objective function values accurately using only few sample points. The GA with local improvement is then applied to a sandwich plate and composite propeller, the amount of calculation is reduced by over half.