基于多种群并行遗传算法的板材拉深过程压边力优化  

Blank Holder Force Optimum of Drawing Based on Multi-Population Parallel Genetic Algorithm

作　　者：戴护民[1] 李赞[2] 夏巨谌[2] 王新云[2] 胡国安[2] 

机构地区：[1]武汉船舶职业技术学院机械工程系,湖北武汉430050 [2]华中科技大学模具技术国家重点实验室,湖北武汉430074

出　　处：《武汉船舶职业技术学院学报》2006年第4期25-27,共3页Journal of Wuhan Institute of Shipbuilding Technology

摘　　要：针对板材拉深成形过程,在增量法有限元数值模拟的基础上,以板料成形厚度变量作为目标函数,采用神经网络、多种群并行遗传算法对压边力进行了优化,对优化后结果进行了有限元模拟,取得很好效果。Aimed at blank drawing process and based on addendum method of finite element simulation, blank holder force optimum was realized by using neural network and multi-- population parallel genetic algorithm which employed the depth variable as objective func- tion. Finite element simulation was performed to verify the optimized results and the better effect was obtained.

关 键 词：有限元 遗传算法 压边力 优化 

分 类 号：TG35[金属学及工艺—金属压力加工]