基于最优拉丁超立方对抑爆球注塑工艺参数在变量范围内抽样并仿真分析,选用预测结果与仿真结果基本一致的RBF代理模型作为多 目标优化设计的代理模型.通过主效应分析发现,抑爆球体积收缩率和末端填充压力变化存在矛盾性.基于MIGA算法,以最小体积收缩率、末端填充压力和生产周期作为优化目标,完成多 目标优化设计.提出一种解决多 目标优化综合权重分配的改进FAHP方法,并且,得到注塑工艺参数优化权重分配,使目标优化效果显著提升.利用帕累托最优解得到最佳工艺参数组合并仿真求出最小体积收缩率为6.616%,最小末端填充压力为74.18 MPa,最短生产周期为17.04 s,分别减少了7.8%、7%和13.2%,并且,通过试验得到验证.
Based on the optimal Latin hypercube,the process parameters of explosion suppression ball injection molding were sampled and simulated within the range of variables,and the RBF agent model with the prediction results basically consistent with the simulation results was selected as the agent model for multi-objective optimization design.Through the main effect analysis,it was found that there was a contradiction between the volume shrinkage of the explosion suppression ball and the change of the end filling pressure.Based on the MIGA algorithm,the multi-objective optimization design was completed with the minimum volume shrinkage,end filling pressure and production cycle as the optimization objectives.An improved FAHP method was proposed to solve the multi-objective optimization comprehensive weight distribution,and the weight distribution of injection molding process parameters was obtained,which significantly improved the objective optimization effect.Through Pareto optimal solution,the best combination of process parameters was obtained,and the minimum volume shrinkage was 6.616%,the minimum end filling pressure was 74.18 MPa,and the shortest production cycle was 17.04 s,which were decreased by 7.8%,7%and 13.2%respectively,and verified by experiments.