【摘要】 针对航天器舱门展收过程中由根铰间隙引起的振动问题,提出一种基于数字孪生和径向基模型的材料参数优化方法。首先建立根铰间隙动力学模型,并应用数字孪生技术构建舱门的数字孪生模型。以根铰材料的接触刚度系数、阻尼系数和动摩擦系数为设计变量,以最大角加速度为优化目标,利用优化拉丁方试验设计构建样本空间;应用径向基神经网络模型构建近似模型;采用结合的序列二次规划算法和多岛遗传算法求解近似模型的最优值。优化结果表明,与优化前相比,舱门的角加速度峰值和根铰间隙接触力明显减小,舱门的动态性能得到提升。更多还原

【Abstract】 To improve vibration caused by clearance of root-joints during the process of the spacecraft hatch deployment,a material parameters optimization method using digital twin and Radial Basis Functions(RBF)model was proposed.The dynamic model of clearance was established,and then the digital twin model of hatch was constructed by using digital twin technology.The contact stiffness coefficient,damping coefficient and dynamic friction coefficient of the root-joints’material were taken as the design variables,and the maximum angular acceleration as the optimization objective to build the sample space with the optimal Latin hypercube experiment design.The approximate model was constructed based on radial basis function neural network model.The combined Sequential Quadratic Programming Algorithm(SQP)and Multi-Island Genetic Algorithm(MIGA)was applied to calculate the optimal design.The optimization results indicated that the peak angular acceleration of hatch and the contact force of rootjoints were significantly reduced compared to original,and the dynamic performance of the hatch was improved.更多还原