论文部分内容阅读
建立了一种遗传算法耦合反向传播(BP)神经网络的非线性计算模型,并将该计算模型用于以塑料收纳盒盖板为对象的注塑成型工艺参数的分析与控制。首先通过Taguchi正交试验法设计了一系列工艺参数数据组合,然后利用Moldflow软件得到相关工艺参数对应的翘曲变形结果,再将数据分别导入BP网络模型及遗传算法优化过的BP网络模型。结果表明:与普通BP神经网络计算模型相比,优化后的计算模型具有更好的稳定性和更高的计算精度,能够更好地应用于注塑成型工艺参数分析与控制。
A nonlinear computational model of BP neural network coupled with genetic algorithm (GA) was established. The computational model was used to analyze and control the injection molding process parameters of plastic storage box cover. Firstly, a series of process parameters data combination was designed by Taguchi orthogonal test. Then, the results of warpage of corresponding process parameters were obtained by using Moldflow software. Then the data were imported into BP network model and BP network model optimized by genetic algorithm respectively. The results show that compared with the ordinary BP neural network model, the optimized calculation model has better stability and higher calculation precision, and can be better applied to the analysis and control of injection molding process parameters.