浅析多属性决策的气动隐身多目标优化

《浅析多属性决策的气动隐身多目标优化》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、学兔兔www.xuetutu.com第48卷第l3期机械工程学报V_o1．48NO．132012年7月JOURNAL0FMECHANICALENGINEERINGJu1．2012DoI：1O．3901／JME．2012．13．132基于多属性决策的气动隐身多目标优化木廖炎平刘莉龙腾(北京理工大学宇航学院北京100081)摘要：针对多目标优化结果排序与选择的多属性决策(Multi。attributedecisionmaking，MADM)I"n~题，将多目标优化与MADM相结合，提出基于MADM的多目标优化方法，并将该方法应用于跨声速前掠翼(Forward-s

2、weptwing，FSW)气动隐身多目标优化中，优化结果提高了跨声速FSW的气动和隐身性能。采用类别形状函数变换法(Class．shapefunctiontransformation，CST)方法对翼型几何外形进行描述，实现FSW气动和隐身多学科优化设计模型的参数化描述。建立基于N—s方程的计算流体力学方法的FSW气动分析模型和基于矩量法的计算电磁学方法的FSW隐身分析模型。将Pareto多目标遗传算法得到的Pareto非劣解集构成MADM矩阵，采用基于模糊熵权的改进的逼近理想解的排序法(Modifiedtechniquefororderpreference

3、bysimilari够toidealsolution，M．TOPSIS)方案评价方法进行Pareto非劣解排序，最终确定最佳的Pareto非劣解。研究结果验证了所提出方法的有效性，为多目标优化问题提供了一种新的解决途径。关键词：多属性决策多目标优化模糊熵权前掠翼Pareto遗传算法中图分类号：V22lMulti-objectiveAerodynamicandStealthyPerformanceOptimizationBasedonMulti-attributeDecisionMakingLIAOYanpingLIULiLONGTeng(SchoolofAe

4、rospaceEngineering，BeijingInstituteofTechnology,Beijing100081)Abstract：Inviewofmulti-attributedecisionmaking(MADM)problemsforrankingandselectingofmulti--objectiveoptimizationresults，themulti—objectiveoptimizationmethodbasedonMADMisproposedbycombiningthemulti-objectiveoptimizationwi

5、thMADM．Themulti—objectiveaerodynamicandstealthyperformanceoftransonicforward-sweptwing(FSW)issolvedbytheproposedmethod，whichCanimprovetheaerod3namicandstealthyperformanceoftransonicFSWefectively．Theclass-shapefunctiontransformation(CST)methodisusedtodescribetheparameterizedairfoilg

6、eometry．Theparameterizedmodelsforaerod)namicandstealthyperformanceofFSWareconstructed．TheaerodynamicanalysismodelofFSWisconstructedbycomputationalfluiddynamicsmethodbasedonN—Sequations．ThestealthyperformanceanalysismodelofFSWisconstructedbycomputationalelectromagneticsmethodbasedme

7、thodofmoments．TheMADMdecisionmarxisconstructedbytheParetooptimalsolutionssetsolvedfromParetomulti-objectivegeneticalgorithm,themodifiedtechniquefororderpreferencebysimilaritytoidealsolution(M-TOPSIS)approachbasedonfuzzyentropyweightisemployedtoranktheParetooptimalsolutionsandultima

8、telytoidentifythebestParet