BP神经网络与GA算法相结合的双色成型保压曲线优化

提出采用前馈神经网络（BP神经网络）与遗传算法(GA算法)相结合优化产品保压曲线,通过改善2种材料的顶出时体积收缩率，进而改善双色产品的翘曲问题。得到优化的工艺参数组合为：聚丙烯（PP）保压压力55 MPa，保压时间12.5 s；丙烯腈苯乙烯丁二烯共聚物（ABS）保压压力75 MPa，保压时间3.5 s; 模拟验证得到优化保压曲线下优化目标为 4.411，小于各实验方案; 双色产品的翘曲由原来的1.696 mm降为0.7427 mm。

Optimization of Pressure in Bi-color Molding Through a Hybrid of Back Propagation Neural Network and Genetic Algorithm

A method of combining back propagation neural network (BP neural network) and genetic algorithm was proposed to optimize the pressure curve in order to improve the volume shrinkage at ejection and the warping of bi-color products. Using the method, the optimal pressure curve could be obtained: PP holding pressure 55 MPa, holding time 12.5 s, ABS holding pressure 75 MPa，holding time 3.5 s. Finally, the CAE simulation was simulated under the optimized parameters. The target variable was 4.411，less than the experimental scheme. And the warpage of bi-color products was reduced from 1.696 to 0.7427 mm.