基于径向基函数的卫星平台桁架结构优化设计

为了降低结构优化计算成本,同时保证搜索优化问题的全局优化解,本文将径向基函数代理模型应用于桁架式卫星平台结构设计优化中。针对桁架结构涉及离散的拓扑和尺寸变量,本文采用连续松弛变量的方式处理离散变量。另外,本文采用增广拉格朗日函数将涉及非线性约束的优化问题转换为一个只包含边界约束的优化问题,然后应用径向基函数代理模型对转换后的优化模型进行近似,并采用具有全局寻优能力的优化算法对所构造的代理模型进行优化。在桁架结构优化过程中,逐次更新朗格朗日罚系数和代理模型,直至搜索到可行优化点。最后,本文将所提出的优化策略应用于一个桁架式卫星平台设计优化实例中,通过优化结果验证了本文所提出的优化策略的高效性。

Satellite Bus Truss Structure Optimization Using Radial Basis Function

In orde r to reduce computational cost during satellite bus structure optimization and s atisfy the global convergence, the radial basis function is applied to dealling with structure optimization. As the truss structure optimization involves size, t opology and geometry variables, continuous relaxation is used to trea t discrete variables. Furthermore, the Lagrangian multiple method is applied to converting a constrained optimization problem to an equivalent unconstrained opt imi zation problem with bound constraint. Then the radial basis function is used to approximate the Lagrangian function, and the optimum is obtained through using g lobal optimization strategy for the approximated model. During the satellite bus structure optimization process, the parameter of Lagrange function and radial b asis are updated until the process searches a feasible optimum. Finally, the propo sed strategy is applied to a practical satellite bus truss structure, thus demon strating the high efficiency.