Multi-objective Optimization and Decision-making Study of Ring-stiffened Cylindrical Shell Using NSGA-II and PCA

The NSGA-II is employed to solve the Multi-objective optimization problem presented in this paper. The objectives seek the maximum two stresses, three critical loads of stability and minimum weight subjected to the behavior constraint and frequency constraint. The Pareto based on all generations is put forward and compared with the Pareto got from the last generation. The effects of frequency constraint on the optimum results are analyzed and the results show some noteworthy characteristics of the samples that have lower frequency. To obtain the final compromise solution from the Pareto set, the PCA and cluster analysis are applied in the decision-making. To reduce the workload and reflect the subjective intention of designer, the Pareto set is classified into six categories by hierarchical cluster. One of the six categories is selected to be evaluated by PCA to obtain the final solution.