拓扑优化结构再设计中的神经网络求解方法

Neural Network Solving Method in Redesign of Topology Optimization Structure

针对不同工况下的拓扑优化结构,为了提高该结构的可制造性,提出了拓扑优化结构再设计中的神经网络求解方法。该方法以拓扑结构为设计基础,通过有限元方法和神经网络求解方法对该结构进行二次设计。以结构综合性能改变最小为优化目标,用规则孔洞对拓扑结构中的不规则孔洞进行替换,从而实现易制造的目的。最后通过数值仿真验证该方法的可行性和有效性。结果表明:经过二次设计的拓扑结构与初始拓扑结构相比,其综合性能有略微下降,但由于结构变得规则从而降低了加工难度。

Aiming at the topology optimization structure under different working conditions,in order to improve the manufactur⁃ability of the structure,a neural network solution method of redesign of the topology optimization structure is proposed.This meth⁃od is based on a designed topological structure,and use the finite element method and the neural network solution method to rede⁃sign the topological structure.Taking minimum the change of structural overall performance as the optimization goal,the irregu⁃lar holes in the topological structure are replaced with regular holes,which achieves the purpose of facilitating manufacturing.Finally,the feasibility and effectiveness of the method are verified by numerical simulation.The results show that the overall perfor⁃mance of the topological structure after the secondary design is slightly worse than that of the initial topological structure,but the processing difficulty is reduced because the structure becomes regular.