计算机辅助设计与图形学学报
計算機輔助設計與圖形學學報
계산궤보조설계여도형학학보
Journal of Computer-Aided Design & Computer Graphics
2015年
10期
2001-2007
,共7页
二分法%灰度过滤%导重法%RAMP模型%拓扑优化
二分法%灰度過濾%導重法%RAMP模型%拓撲優化
이분법%회도과려%도중법%RAMP모형%탁복우화
bi-sectioning algorithm%gray-scale filter%guide-weight method%RAMP scheme%topology optimization
为了提高导重法求解拓扑优化问题的计算效果,提出一种改进的导重法,并引入了灰度过滤技术抑制优化过程中灰度单元的产生。首先基于RAMP(the rational approximation of material properties)模型结合导重法求解最小柔度拓扑优化问题的迭代表达式,利用二分法对表达式中的拉格朗日乘子求法进行了改进;为减少优化后结构图像中的灰度单元数量,在迭代表达式中引入灰度过滤函数;最后将上述理论拓展到多工况拓扑优化问题中,采用归一化组合处理方法建立目标函数。对多工况拓扑优化问题的2个典型算例进行计算的结果表明,应用文中理论求解拓扑优化问题具有收敛稳定、求解快速、图像清晰的特点。
為瞭提高導重法求解拓撲優化問題的計算效果,提齣一種改進的導重法,併引入瞭灰度過濾技術抑製優化過程中灰度單元的產生。首先基于RAMP(the rational approximation of material properties)模型結閤導重法求解最小柔度拓撲優化問題的迭代錶達式,利用二分法對錶達式中的拉格朗日乘子求法進行瞭改進;為減少優化後結構圖像中的灰度單元數量,在迭代錶達式中引入灰度過濾函數;最後將上述理論拓展到多工況拓撲優化問題中,採用歸一化組閤處理方法建立目標函數。對多工況拓撲優化問題的2箇典型算例進行計算的結果錶明,應用文中理論求解拓撲優化問題具有收斂穩定、求解快速、圖像清晰的特點。
위료제고도중법구해탁복우화문제적계산효과,제출일충개진적도중법,병인입료회도과려기술억제우화과정중회도단원적산생。수선기우RAMP(the rational approximation of material properties)모형결합도중법구해최소유도탁복우화문제적질대표체식,이용이분법대표체식중적랍격랑일승자구법진행료개진;위감소우화후결구도상중적회도단원수량,재질대표체식중인입회도과려함수;최후장상술이론탁전도다공황탁복우화문제중,채용귀일화조합처리방법건립목표함수。대다공황탁복우화문제적2개전형산례진행계산적결과표명,응용문중이론구해탁복우화문제구유수렴은정、구해쾌속、도상청석적특점。
An improved guide-weight method is presented to improve the computational efficiency of the guide-weight method on solving topology optimization problems, and the gray-filter method is introduced to restrain the gray-scale element which is generated in the optimization process. Since the optimization model was established based on RAMP(rational approximation of material properties, RAMP) scheme, and the guide-weight method was used as an algorithm combining for proposing the iteration formulas of the least compliance in multi objective topology optimization problems. First, the Lagrange multipliers solution me-thod of the formulas were improved by the bi-sectioning algorithm. Then, with the purpose to suppress the generating of gray-scale element, the gray-scale filtering function was introduced into iterative formulas. Moreover, the proposed theory were extended to multi objective optimization problem, and ill-conditioning were avoided by normalizing single objective. Finally, two examples of multi objective optimizations prob-lems were calculated respectively. The results calculated by examples demonstrate that using the proposed theory above can solve topology optimization problems with steady convergence, fast calculate speed and image sharpening.