依赖EBF神经网络的注塑成型熔接痕长度优化

Weld Mark Length Optimization of Injection Molding Based on EBF Neural Network

分析了注塑成型中熔接痕形成原因,分析了浇口位置对形成熔接痕的影响,提出了依赖椭球基函数(EBF)神经网络的注塑成型熔接痕长度优化。通过将EBF神经网络作为注塑成型熔接痕条件,根据注塑成型熔接痕依次构建得到注塑成型熔接痕长度函数与注塑成型熔接痕适应度函数,并以这二个函数作为优化目标。在实现注塑成型熔接痕的基础上,选择李雅普诺夫第二方法来验证EBF神经网络具有稳定解。最后通过仿真结果表明:该算法能够均衡注塑成型熔体长度,提高注塑成型熔接痕长度优化实用性。

The causes of weld line formation in injection moulding were analyzed. The influence of gate position on the formation of weld line was analyzed. The optimization of weld line length in injection moulding based on ellipsoidal basis functional( EBF) neural network was proposed. By using EBF neural network as the condition of weld mark in injection moulding,the length function of weld mark in injection moulding and the fitness function of weld mark in injection moulding were constructed according to the weld mark in injection moulding,and these two functions were taken as optimization objectives. On the basis of realizing the weld line in injection moulding,Lyapunov’s second method was selected to verify the stable solution of EBF neural network. Finally,the simulation results show that the algorithm could balance the melt length of injection moulding and improve the practicability of optimizing the weld line length of injection moulding.