现有自适应成长技术根据数学规划理论中的库恩塔克条件推导迭代公式,对以应变能为目标,体积为约束的板壳结构加强筋分布优化问题,设计效率高且结果好。但对不同类型的优化问题需要重新推导迭代公式,通用性差,且难以推广到多约束问题。针对现有技术的不足,采用移动渐近线算法(The method of moving asymptotes,MMA)迭代更新设计变量,将自适应成长技术推广应用于桁架结构拓扑优化设计中。讨论了单约束和多约束的典型桁架结构设计问题,并对多载荷工况进行了研究。算例结果表明,所提方法可以得到清晰的杆件分布和具体尺寸信息,设计效率高,适用性好,便于实际工程加工,具有较好的应用前景。
Abstract
The existing adaptive growth technique deducted the optimization iteration formula according to the condition named Karush-Kuhn-Tucker condition. It is high efficiency and good, which is aimed to strain energy and bound to the volume against optimization problem of stiffener distribution for plate and shell structures. However, different optimization iteration formula, which deducted by different kinds of problem, is poor and difficult to generalize, especially for the optimization problems with more than one constraint. Against with the lack of adaptive growth technique, a new method that is moving asymptotes(MMA) is suggested by this article and the adaptive growth technique is applied to truss structure topology optimization design. The design of typical truss structures with single constraint and multiple constraints is discussed, and the multi load cases are studied. The results of numerical examples show that the proposed method can obtain clear information of rod distribution and dimension, and has high design efficiency and good applicability. It is convenient for practical machining and has a good application prospect.
关键词
自适应成长技术移动渐近线算法桁架结构拓扑优化
Keywords
Adaptive growth techniqueMethod of moving asymptotesTruss structureTopology optimization