欢迎来到天天文库
浏览记录
ID:46611523
大小:944.13 KB
页数:4页
时间:2019-11-26
《基于神经网络的拖靶高度自抗扰控制研究》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、2014୍8ᄅ宇航计测技术Aug.,201434ज4௹JournalofAstronauticMetrologyandMeasurementVol.34,No.4文章编号:1000-7202(2014)04-0053-04中图分类号:V249.12文献标识码:A基于神经网络的拖靶高度自抗扰控制研究李双明(中国人民解放军92419部队,辽宁125106)摘要ູਔࢳथࡹ৫ϛඔ࿐ଆၛࠣۚϛຩࣜപႿࠎਔԛิ,ཙ႕ҵ༂ਈҩۚሱॆಠ॥ᇅྐۚބྐࡆཟԼၛ。ཙ႕Ӯᄯᇅ॥ྛ٦ؓಒሙ҂ଆਔࢆᇅ॥ಠॆሱ。مٚቔູപࣜຩൻೆྐۚਔ
2、ࢆ࠻,ඔ༢ሰၹӊҀఖᇅ॥ಠॆሱ,іҩਈ༂ҵ႕ཙ,Ⴛሱॆಠ॥ᇅҀӊၹሰൌൈ٦ۚޚਔղ,ಠۄࢸຓਔᇅၝིႵ,ಒࣚࡆ۷ିྟᇅ॥ϛمٚھ,ૼіᆇٟ。ᆞ۷ྛଢѓ。关键词ϛപࣜຩሱॆಠ॥ᇅٟᆇResearchaboutActiveDisturbancesRejectionControlofTowTargetHeightbasedonNeuralNetworkLIShuang-ming(92419UnitofPLA,Liaoning125106)AbstractInordertosolvethedifficultiestoes
3、tablishthemathematicalmodeloftowtargetandtheinfluenceofheightmeasurementerror,activedisturbancesrejectioncontrolmethodbasedonneuralnet-workisproposed.Theinfluenceofmathematicalmodelinaccuracytotheflightisreducedbydesignofactivedisturbancesrejectioncontrol.Withheig
4、htandverticalaccelerationastheinputsignalstoneuralnetwork,thecompensationfordisturbancerejectioncontrolfactorcoefficientisgot,whichdecreasestheinfluenceofheightmeasurementerror,andmakesthecompensationofcontrollorbecorrectedinrealtime.Simulationindicatsthatthismeth
5、odmakestheflightperformanceofthetowtargetmoreaccurate,effectivelyrestrainstheexternalinterference,andaccomplishesgoalofaconstanthighflying.KeywordsTowtargetNeuralnetworkActivedisturbancesrejectioncontrolSimulation1引言؟ູॢݚ٦ྛ,ສສ൳ݚশ、గੀߌ႕ཙ,ղ҂ޚۚ٦ྛଢ,മᇀᄯӮϛሖݚ。ଢభ,ϛ҂Ⴟ၂Ϯ
6、ۚϛؓང໓؟ྸႵၘौ၇ྶсദሱః,ఖྛ٦॥ᇅᅚषਔ࣮。໓ངࠏఠႄࡆݖ๙,ൔٚᇅ॥DPࠣଆ߄ྟཌႨҐ]1[ၛ෮,ѓଢྛ٦ູۚקޚקഡၛ,ྛ٦ྐۚϛѯੲҀྐۚཌބնۚྛ٦ႿႮ。ඌܱ༢ϛ൞ᇅ॥ۚҩਈ收稿日期:2013-11-28,修回日期:2014-03-01作者简介:චૼ(1986-),ଳ,ᇹ۽ӱഽ,ᇶေ࣮ٚཟ:ᇅა॥ᇅ。·54·宇航计测技术2014୍ᆴ,ཌྷбֆ၂ҩਈ൞ଆۚϛ༂ਈҩۚ༯ঃ౦ᇕᆃ,مٚࢨ༢,ఃሱॆಠҵ߶ࢆ,ႮႿඔ࿐ଆ҂ሙಒྟބ॥ᇅمෘఖᇅ॥مٚೂ༯ॆಠି৯,॥ᇅ
7、ᇉࠣ໗קࣚഈթfh=fhan(v1-v0,v2,r0,h)ᄝ҂ቀ。໓ང[2]ҐႨਔሱॆಠ॥ᇅ(ActiveDis-v1=v1+hv2turbanceRejectionControl)مٚ,ቋնႪ൞҂v2=v2+hfh၇ঠࡹ৫ଆ,ିܔሱҩ༢ଆބຓಠe=z1-yൌൈቔႨ۳ჍҀӊ,ಖطᆃᇕٚم҂ିࢳथۚҩਈ༂ҵၛࠣሱॆಠ॥ᇅఖҀӊҕඔ༂ҵ႕ཙ。Ֆfe=fal(e,0.5,h),fe1=fal(e,0.25,h)ၛഈਆԛ,ิԛਔࠎႿപࣜຩϛሱॆಠz1=z1+h(z2-β01e)(2)॥ᇅ۷ۚྛ٦,مٚࡆࣚሙ。z2
8、=z2+h(z3-β02fe+b0δz)z3=z3+h(-β03fe1)2拖靶的非线性模型e1=v1-z1,e2=v2-z2u0=-fhan(e1,ce2,r,h1)ϛᄝ٦ྛݖӱᇏ,࿊౼ሑэਈx1=H,x2=δz=(u0-z3)/b0Vyd,ఃᇏHູ٦ྛۚ,Vydູശࢆ,ᄵ٤ཌྟൈ[2]ൔᇏ:v0
此文档下载收益归作者所有