MOGA和FEM相结合实现万能型钢轧机机架圆角的多目标优化  

Multi-objective optimization of the housing's fillet angle of the universal rolling mills using MOGA combined with FEM

作　　者：马劲红[1] 张文志[1] 宋剑锋[1] 董永刚[1] 

机构地区：[1]燕山大学轧机研究所,秦皇岛066004

出　　处：《塑性工程学报》2008年第1期146-149,177,共5页Journal of Plasticity Engineering

摘　　要：文章把多目标遗传算法(MOGA)和有限单元法(FEM)有机地结合起来,建立了一种新的结构优化方法。用MATLAB遗传算法工具箱函数编制了MOGA和FEM相结合的多目标优化程序,对万能型钢轧机机架圆角进行多目标优化,充分发挥了有限单元法数值计算准确及遗传算法求极值的高效性和全局性的优点。与ANSYS自带的零阶和一阶优化方法相比,本优化方法取得了较好的优化效果,为解决复杂的多目标工程优化问题提供了一种新的更有效的途径。In this paper, a new design optimization procedure is set up, based multi-objective genetic algorithm （MOGA） combined with the finite element method （FEM）. The procedure combined MOGA with FEM is developed by MATLAB GA toolbox functions, optimizing the housing＇s fillet angle of the universal rolling mill, Thus, taking full advantage of the unique features of genetic algorithms that the optimal solution is achieved by a rapid global search and taking advantage of the unique features of FEM that the computation is accuracy. Compared with the zero order and first order optimization method of ANSYS program, the result of this optimal method is better. This method can solve multi-objective problem with more complex constraint condition in the engineering, providing a new and effective means for solving large, complex problem in the engineering.

关 键 词：万能型钢轧机 机架圆角 多目标遗传算法 有限元 多目标优化设计 

分 类 号：TG333.13[金属学及工艺—金属压力加工] TB11[理学—数学]