两种不同摄动算法的优化方法增强门板的抗弯曲刚度  

Two kinds of perturbation optimization algorithms for strengthening bending stiffness of door sheet

作　　者：付远[1,2] 陈志宝[1] 吴禄慎[2] 刘觐[1] 张友亮[1] 郭炜[1] 

机构地区：[1]江西省科学院,南昌330028 [2]南昌大学机电工程学院,南昌330028

出　　处：《应用力学学报》2018年第5期1065-1071,共7页Chinese Journal of Applied Mechanics

基　　金：国家自然科学青年基金(51505203);江西省科学院青年基金重点项目(2017-YZD2-10);江西省科学院科研开发专项基金博士项目(2014-YYB-17)

摘　　要：分别采用了两种壳体结构的形状优化算法来提高门板结构的抗弯曲刚度,它们分别是控制算法与敏度算法的壳体珠状优化方法。优化结果表明,对于壳状门板的刚度优化,这两种算法都是采用材料单元的摄动迭代来降低壳单元的最大应变能。两类算法之间的明显不同在于:控制算法在优化完成后会产生典型的珠状结构,而敏度算法产生的珠状结构则更平缓且不明显;同时,对于门板结构刚度的优化计算,控制算法的优化结果将最大单元应变能从1.83×10^(-3)J降低至1.35×10^(-3)J,而敏度算法则降低至1.49×10^(-3)J。因此对于该目标优化方法来说,控制算法更优于敏度算法。Bending stiffness of door sheet is increased by using two algorithms of shape optimization in shell structures separately. These two algorithms are called bead-controller and bead-sensitive of bead optimization. The results show that, both of these algorithms of door optimization reduce shell element maximum strain energy by matrix perturbation algorithm. The visible differences of these two algorithms are that: bead-controller generates typical bead structures after optimal iterative, and the bead structures generated by bead-sensitive are more smoothness. Meanwhile, for the optimizing calculation of door sheet structures, maximum strain energy is reduced from 1.83×10^(-3) J to 1.35×10^(-3) J by bead-controller algorithm and the optimal value via the bead-sensitive algorithm is 1.49×10^(-3) J. Therefore, the algorithm of bead-controller is better than bead-sensitive in this goal optimization.

关 键 词：珠优化 刚度 门板 壳体优化 摄动算法 

分 类 号：O325[理学—一般力学与力学基础]