Satellite structure optimization based on high fidelity surrogate model

Here, to improve design quality and computational efficiency of satellite structure optimization, a global optimization method based on high fidelity dynamic surrogate model (HFDSM) was proposed by combining radial basis function (RBF) surrogate model and adaptive simulated annealing (ASA) algorithm. With this method, a search space adaptive updating strategy was constructed according to the global optimization results. After updating search space in optimization process, sample points were added and the near surrogate model was reconstructed. The prediction error of the optimal solution and the decline degree of the objective function were taken as the criteria for the convergence of optimization process to guarantee the global convergence of optimization and the model accuracy at the optimal solution. The optimization results of high-dimensional testing function and I-beam example showed that the proposed method can not only obtain high-precision optimization results, but also significantly improve the efficiency of optimization solving. Finally, the proposed method was used to solve the structural optimization problem of a certain high-dimensional satellite. The optimization results showed that the maximum prediction error of constraint functions, such as, structural fundamental frequency and dynamic response is only 0.65%; time cost is reduced by more than 50% compared with directly using ASA algorithm; very high accuracy and computational efficiency of the proposed method for solving satellite structural design optimization problem are verified. © 2021, Editorial Office of Journal of Vibration and Shock. All right reserved.