欢迎来到天天文库
浏览记录
ID:46606606
大小:482.12 KB
页数:7页
时间:2019-11-26
《基于神经网络响应面的机翼气动稳健性优化设计》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、第31卷2010钜第6期6月航空学报ACTAAERONAUTICAETASTRONAUTICASINICAVoI.31No.6June2010文章编号:1000—6893(2010)06—1134—07基于神经网络响应面的机翼气动稳健性优化设计蒙文巩,马东立,楚亮(北京航空航天大学航空科学与工程学院,北京100191)WingAerodynamicRobustnessOptimizationBasedonNeuralNetworkResponseSurfaceMengWengong,MaDongli,ChuL
2、iang(schoolofAeronauticScienceandEngineering,BeijingUniversityofAeronauticsandAstronautics,Beijing100191,China)摘要:针对不确定性因素引起飞机性能波动的现象,探讨了机翼气动优化设计过程的稳健性问题;建立了面向速度和扭转角两个不确定性因素的气动性能稳健性约束模型;在利用MATI,AB构造基于均匀设计法的BP(BackPropagation)神经网络响应面基础上,应用遗传算法对机翼分别进行考虑稳健性约束和
3、不考虑稳健性约束的气动优化设计,得到两种不同的优化方案。计算结果表明:两种优化方案的最大升阻比都比初始方案的大;在巡航马赫数下,与不考虑稳健性约束的优化方案相比,考虑稳健性约束的优化方案的最大升阻比小0.0279,但在马赫数、扭转角对应范围内其最大升阻比的变化量分别小0.0340和0.0016,其他气动性能参数也更加稳定,波动更小,气动性能具有更好的稳健性,从而证明本文方法进行机翼气动稳健性优化设计是可行、有效的。关键词:稳健性;不确定性因素;BP神经网络响应面;遗传算法;机翼;气动优化中图分类号:V211.
4、4l文献标识码:AAbtract:TherobustnessproblemintheaerodynamicoptimizationofanaircraftwingisdiscussedinreferencetOtheundulationofaircraftperformancederivedfromuncertaintyfactors.Aerodynamicperformancerobust—nessconstrainedmodelsarebuiltwhicharesubjecttOtheuncertaint
5、yfactorsofvelocityandtwistangle.BydintoftheBP(BackPropagation)neuralnetworkresponsesurfacebasedontheuniformdesignwhichiscon—struttedthroughMATLAB,tWOschemes。whosedifferenceliesinwhetherornotrobustnessistakenintoac—count,arerespectivelyobtainedbasedongenetic
6、algorithm.Theresultssuggestthatthemaximumlift-dragratiosatcruisingspeedsforbothschemesarehigherthanthoseoftheinitialscheme.Thoughtheschemewiththeconsiderationofrobustnessis0.0279lowerthanthatoftheschemewithoutit。thevariationofmaximumlift—dragratiooftheforme
7、rschemeisrespectively0.0340and0.0016lessthanthelatterwithintherangeofthecruiseMathnumberandthetwistangle.Otheraerodynamicperformancesofthedesignwhichtakesrobust—nessintoconsiderationarealsomuchmorestablethanthosewhichdoesnot.Thereforetheaerodynamicro—bustne
8、ssoptimizationmethodinthisarticleisshowntObeusefulandefficient.Keywords:robustness;uncertaintyfactor;BPneuralnetworkresponseSUrface;geneticalgorithm;wing;aer—odynamicoptimization稳健性(Robustness)指产品抵抗本身或
此文档下载收益归作者所有