摘要
提出了一种具有区间不确定性的机翼颤振优化方法。采用拉丁超立方方法建立仿真试验表,基于M SC.N as-tran平台进行颤振仿真分析。获得仿真数据之后,应用K rig ing方法构造了包含区间不确定性参数的机翼颤振分析代理模型,并进行有效性检验。基于建立的代理模型并按照区间序数关系,将不确定性优化目标和约束条件转化为确定性表达形式,从而形成区间不确定性的结构优化设计方法。该方法将区间法优化和代理模型相结合,同时综合有限元仿真和遗传算法的优点,计算效率较高且应用范围较广。以某复杂机翼结构为例进行了含区间不确定性的颤振优化计算。分析结果表明了所提方法的正确性和可行性。
A novel method for wing structures flutter optimization containing uncertainties based on interval theory is pro- posed. The simulation experiment table is arranged by Latin Super Cubic Method, and executed by MSC. Nastran. The sur- rogate model is established by using Kriging method based on the simulation results and its validity is checked. The objective function and constraint conditions of wing flutter optimization which contain uncertainties are changed to traditional determin- istic form using interval theory and surrogate model firstly, and then been solved efficiently. This method can combine the ad- vantages of simulation analysis, surrogate model and genetic algorithm together perfectly. Finally the correctness and feasibil- ity is shown by numerical examples.
出处
《振动工程学报》
EI
CSCD
北大核心
2011年第5期461-467,共7页
Journal of Vibration Engineering
基金
国家自然科学基金资助项目(10872089)
关键词
气动弹性
颤振优化
不确定性
区间法
代理模型
aeroelasticity
flutter optimization
uncertainty
interval method
surrogate model
