基于神经网络和遗传算法的注射成型优化设计

摘要

To minimize the fracture of lock springs during the use,a neural network model was developed to map the complex non-linear relationship between the processing conditions and quality indexes of the lock springs,and a genetic algorithm was used to optimize the molding process parameters on the basis of the neural network model mentioned above.The results indicated that the combination of neural network and genetic algorithm method was feasible to improve the product quality,and the residual stress decreased by 16.02 ％.Moreover,the sum of packing time and cooling time was shortened by 5.4 s.This suggested that the production time was reduced,and the production efficiency was improved when the lock springs were mass produced.%采用神经网络来描述工艺参数与品质指标之间的复杂非线性关系,并基于神经网络,利用遗传算法来优化成型工艺参数以减少锁簧在使用过程中易发生断裂这一现象.结果表明,基于神经网络和遗传算法的优化使塑件的残余应力减少了16.02％,提高了制品品质;同时保压时间与冷却时间总共减少了5.4s,从而缩短了锁簧的批量生产时间,提高了生产效率.