摘要
直接利用表面网格的几何信息计算最佳浇口位置,可显著提高浇口位置优化算法的计算速度与有效性。基于流动平衡的定义和权值流动长度,建立了优化目标函数。根据一定假设条件,将表面网格转化为带权值的无向图,结合遗传算法与爬山算法的优点,建立了Generic-Hillclimbing(GH)浇口位置搜索方法,并利用最短路径算法求解目标函数值。与Moldflow浇口优化功能相比较的实例表明,该方法能够高效准确地得到优化的浇口位置。
The efficiency and accuracy of the gate optimal algorithm of injection molding could be improved by directly using the geometrical information of the fusion mesh.The objective function is defined by introducing the weighted flow length to achieve the balanced flow of the melt.To solve the objective function,the fusion mesh is translated to weighted graphic firstly under some hypotheses.And then the weighted flow length of very node could be got by using the shortest path algorithm.A new optimal methodology called as Generic-Hillclimbing（GH）was developed by combining Generic algorithm and Hillclimbing algorithm.Generic algorithm is employed to find the suboptimum gate location.Hillclimbing program used the suboptimum as the beginning point to search the optimum result.The examples which compare optimization results of the presented GH and Moldflow indicate that the presented methodology is highly efficient and accurate.
出处
《高分子材料科学与工程》
EI
CAS
CSCD
北大核心
2011年第9期171-174,共4页
Polymer Materials Science & Engineering
基金
国家自然科学基金资助项目(50675080
0875095)
关键词
浇口优化
表面网格
遗传算法
爬山算法
gate location optimization
fusion
Generic algorithm
Hillclimbing algorithm
