In order to improve the design efficiency of the multi-openings cylindrical pressure structure, the sensitivity analysis and dimension reduction of the design variables were carried out. The optimal Latin hypercube method was used to selecting the sample points, and the sample points were calculated and analyzed. The sample points were fitted, and the approximate models with the accuracy meeting the engineering needs were obtained. The deterministic multi-objective and multi-objective robust optimization models of the multi-openings cylindrical pressure structure were established. The deterministic multi-objective optimization, 3σ robustness optimization and 6σ robustness optimization were carried out respectively, and the corresponding optimization scheme was determined by the minimum distance method. The results show that compared with 3σ optimization, as the weight increases, the ultimate load increases. Compared with deterministic multi-objective optimization, the weight and ultimate load increase greatly, indicates that the design parameters must be increased to meet the requirements of structural robustness, so that the weight is increased, and the quality level of all constraint functions and objective functions of 6σ optimization is improved. The quality level of many design variables has been raised to more than 6σ. It can be seen that the robustness optimization of 6σ can greatly improve the robustness of multi-openings cylindrical pressure structure. © 2021, Editorial Board of Journal of Huazhong University of Science and Technology. All right reserved.
