中文摘要：

针对注塑过程当中影响塑件质量的多个工艺参数配置问题,提出改进粒子群算法、BP神经网络、灰色关联度相融合的成型工艺参数优化模型。首先,针对BP易陷入局部最优、收敛效率低的不足,改进粒子群算法中粒子速度与位置更新策略并优化BP算法的权值和阈值,从而构建起工艺参数预测模型。在此基础上,以正交实验数据为训练样本,Moldflow软件分析结果为输出样本,灰色关联度为粒子群适应度函数,进而由粒子群算法寻得最佳的工艺参数。实验结果表明,该方法能够更快、更好的获得注塑成型中的工艺参数,且以此工艺参数进行实验,塑件的翘曲变形量、收缩率均较小。

英文摘要：

As a complicate batch process,the final product qualities of injection molding process are easily influenced by improper given process parameters. With the modified PSO-BP algorithm as optimization model,process parameters are optimized to make part warpage and shrinkage arriving at their optimal values. Firstly,an improved particle position updating strategy is proposed to overcome the low convergence rate and easily falls into local optimization of BP neural network.Subsequently,a prediction model about process parameters and the optimized objectives is established. Then,taking the orthogonal experiment samples as input data and the output data generated from Moldflow software,the better process parameters are got through PSO with the grey correlation degree fitness function. The experimental analysis shows that the warpage and the shrinkage of the plastic parts both become smaller with the proposed method,which demonstrates this algorithm can provide a faster and better tool to optimize injection molding process parameters.