塑性工程学报
塑性工程學報
소성공정학보
JOURNAL OF PLASTICITY ENGINEERING
2009年
4期
79-84
,共6页
吴彦骏%赵震%梁艳迁%胡成亮%高崇晖
吳彥駿%趙震%樑豔遷%鬍成亮%高崇暉
오언준%조진%량염천%호성량%고숭휘
多工位高速锻造%拉丁超立方采样%Kriging%BP神经网络%优化
多工位高速鍛造%拉丁超立方採樣%Kriging%BP神經網絡%優化
다공위고속단조%랍정초립방채양%Kriging%BP신경망락%우화
high-speed multi-stage forging%latin hypercube sampling (LHS)%kriging%BP neural network%optimization
多工位高速锻造是一种节能减耗的先进成形技术,与传统的单工位多工序锻造有较大区别,不合适的预锻工位模具设计在生产中会造成载荷过大、锻件充填质量差、折叠等问题.文章提出了基于近似替代模型的多工位高速锻造预锻模具优化设计方法,采用拉丁超立方抽样方法构造初始样本点集,应用有限元软件获得响应值,并应用Kriging模型和BP神经网络模型建立真实问题的替代模型,来近似拟合预锻工位模具形状与成形载荷、成形质量之间的关系,并结合基于惩罚函数法的遗传算法求得最优解.以3工位高速锻造问题为例,经过优化取得了良好的效果,与初始设计相比,成形质量显著提高,成形载荷下降了40%左右.
多工位高速鍛造是一種節能減耗的先進成形技術,與傳統的單工位多工序鍛造有較大區彆,不閤適的預鍛工位模具設計在生產中會造成載荷過大、鍛件充填質量差、摺疊等問題.文章提齣瞭基于近似替代模型的多工位高速鍛造預鍛模具優化設計方法,採用拉丁超立方抽樣方法構造初始樣本點集,應用有限元軟件穫得響應值,併應用Kriging模型和BP神經網絡模型建立真實問題的替代模型,來近似擬閤預鍛工位模具形狀與成形載荷、成形質量之間的關繫,併結閤基于懲罰函數法的遺傳算法求得最優解.以3工位高速鍛造問題為例,經過優化取得瞭良好的效果,與初始設計相比,成形質量顯著提高,成形載荷下降瞭40%左右.
다공위고속단조시일충절능감모적선진성형기술,여전통적단공위다공서단조유교대구별,불합괄적예단공위모구설계재생산중회조성재하과대、단건충전질량차、절첩등문제.문장제출료기우근사체대모형적다공위고속단조예단모구우화설계방법,채용랍정초립방추양방법구조초시양본점집,응용유한원연건획득향응치,병응용Kriging모형화BP신경망락모형건립진실문제적체대모형,래근사의합예단공위모구형상여성형재하、성형질량지간적관계,병결합기우징벌함수법적유전산법구득최우해.이3공위고속단조문제위례,경과우화취득료량호적효과,여초시설계상비,성형질량현저제고,성형재하하강료40%좌우.
The high-speed multi-stage forging is an advance forming technique. It is different from the traditional die forging process. The improper preform design will lead to some problem in actual production, such as too large load, underfill, folding and so on. In order to obtain the optimal design, an new optimization method based on surrogate models is researched in this pa-per. This methodology applies the Latin Hypercube Sampling (LHS) to generate initial experiment samples. Then the values of response functions are obtained by numerical simulation. Afterwards, the kriging model and BP neural network are established the surrogate models to fit the relationships between variables and responses. These surrogate models are used in optimization cal-culation to obtain the optimal design by using Genetic Algorithm based on penalty function. A 3-station high-speed forging case was studied with the above-mentioned method. After optimization, the quality of the product was improved significantly and the forming load decreased more than 40% compared to the initial design.