Abstract:To reduce the low-order elastic vibration risk of a slender missile structure under coupled aerodynamic, static, dynamic and parametric uncertainty effects, a robust multidisciplinary design optimization method integrating Kriging surrogate modeling, Sobol global sensitivity analysis and collaborative optimization is proposed. A coupled aerodynamic-static-structural dynamic framework is established, in which one-way fluid-structure interaction is used to transfer aerodynamic pressure loads to the structural model. A dynamic equivalent model based on mass elements and rotational inertia is then developed to improve modal-analysis efficiency. Kriging surrogate models are constructed using optimal Latin hypercube sampling, and Sobol total-effect indices are used to screen dominant variables and reduce input dimensions. Deterministic optimization results obtained using multidisciplinary feasible and collaborative optimization architectures are compared, and a robust objective combining the mean and standard deviation of the first natural frequency is further introduced into the collaborative optimization framework. The equivalent modal model keeps the relative errors of the first five elastic modes within 5%, while reducing a single modal analysis from about 6 h to about 10 min. Collaborative deterministic optimization increases the first natural frequency from 132.70 Hz to 173.01 Hz. The robust optimal design achieves a nominal first natural frequency of 148.05 Hz, with a drag coefficient of 0.4738, a maximum stress of 29.11 MPa and a mass of 178.04 kg. Monte Carlo verification with 10000 samples shows that the robust design has a frequency mean of 142.18 Hz and a standard deviation of 0.808 Hz, and all 6σ intervals satisfy the prescribed constraints. The proposed method enables simultaneous frequency improvement and fluctuation suppression for missile overall structures using limited high-fidelity samples.