本文發展與建構回應表面式取代設計限制條件函數以求解含允許模糊值的結構最佳化設計問題。重點方向在於發展適於求解大型結構的模糊最佳化設計解題方法與程序,可得近似最佳解並減少結構計算分析。得知縮減設計空間,建立相繼縮小回應表面式的搜尋解策略,及佈點策略對解題效果與效率造成影響。在含允許模糊值最佳化設計問題中,以相同回應表面式計算設計目標之最大值及最小值的解題策略,可得到近似唯一解,有效處理無顯函數的困擾及相對減少需用有限元素法的結構分析次數。本文以三維72桿,216桿及360桿桁架多變數及限制的設計為例研究,解說含允許模糊值及無顯函數的大型結構最佳化設計問題。 This paper developed a design process for dealing with large-scale fuzzy design optimization problems with complex explicit or implicit constrained functions under allowable fuzzy values. This paper provides a strategy for determining the intermediate maximum and minimum objective value that are sensitive to the search of optimum. Using successive reducing response surface method is important for searching the global optimum. A unique level-cut design solution technique combined with the presenting algorithm so that a definite final optimum design can be obtained. Large-scale three dimensional truss optimization problems including 72-bar, 216-bar and 360-bar truss are presented in the paper to demonstrate the method and process. Using RSM combined with suitable solution technique shown in this paper can effectively reduce the number of finite element analysis. The presenting iteration algorithm is stable and can obtain the efficient and reasonably accurate design.
Relation:
2007 CSME CONF中國機械工程學會第24屆全國學術研討會論文集=Proceedings of the 24th National Conference on Mechanical Engineering the Chinese Society of Mechanical Engineers,6頁