二级多点近似方法在整星结构优化设计中的应用

以整星结构为研究对象,针对含有离散变量的混合变量优化问题,应用二级多点近似方法给出从卫星拓扑布局设计到尺寸的一体化优化求解过程。首先,对整星结构进行仿真和振动试验,设计满足总体指标要求且已得到在轨验证。其次,建立以极小化卫星质量为目标的多工况优化模型后利用分段多点逼近函数建立了显式近似问题。最后,采用遗传算法和变尺度法分别对离散/连续两类变量寻优,并以拓扑构型一致性和目标函数的下降程度作为收敛的判定依据。结果显示,在满足频率和静力的约束下,卫星质量降低了7.9%,且结构分析迭代次数仅为50~60次,从而验证了该方法在求解整星结构多变量多工况下的一体化设计问题时具有很高的计算效率和准确性。

Application of two-level multipoint approximation method in optimal design of whole satellite structure

Taking the whole satellite structure as the research object,an integrated optimization process from topological layout design to satellite size is presented by using the two-level multipoint approximation method for the mixed variable optimization problem with discrete variables.Firstly,simulation and vibration test of the whole satellite structure are carried out.The design is found to meet the requirements of the general specification and has been verified in orbit.Secondly,an explicit approximation problem is established by using a piecewise multipoint approximation function after describing a multi-conditions optimization model aiming at minimizing the satellite mass.Finally,a genetic algorithm and variable metric method are used to optimize the discrete and continuous variables respectively,and the convergence is judged by the consistency of topology configuration and the decreasing value of the objective function.The results are as follows: under the constraints of frequency and static force,the satellite mass is reduced by 7.9% and the number of structural analysis iterations is only 50~60,which verifies that this method has high computational efficiency and accuracy in solving the integrated optimization design problem of the whole satellite structure with multiple variables and under working conditions.