摘要
针对二维连续体形状优化过程中解析灵敏度求解困难的情况,利用响应面方法将目标函数和约束函数近似显式化,建立序列二次规划模型。为了保证优化模型的可靠性,结合试验设计方法,以较少的结构分析代价构造约束响应面,从而能够快而稳定地收敛,本文重点研究并建立了二次近似评价函数用以计算自适应运动极限的策略。算例说明这种策略是有效而稳定的。
Based on RSM (Response Surface Methodology), the obiective function and constraint conditions are approximately explicit to overcome the difficulty of sensitivity analysis in two-dimensional continuum structures, sequential quadratic programming model is used. Combined with experiment design method, very few design points were used to obtain constraint conditions. In order to ensure the reliability of the optimization model, a second-order approximation function to each constraint function is researched and formed to compute move-limits for the current design point and adaptive move-limits evaluation is proposed. An example is given to show the efficiency and stability of this optimization strategy.
出处
《计算力学学报》
EI
CAS
CSCD
北大核心
2007年第4期447-452,共6页
Chinese Journal of Computational Mechanics
基金
国家自然科学基金(10472003)
北京市自然科学基金(3042002)资助项目
关键词
形状优化
响应面方法
运动极限
优化策略
shape optimization
response surface methodology
move-limits
optimization strategy