資源描述:
《新一代深潛救生艇動(dòng)力定位系統(tǒng)設(shè)計(jì)與實(shí)現(xiàn)》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫(kù)。
1、哈爾濱I:程大學(xué)碩十學(xué)位論文摘要潛艇失事后,深潛救生艇是保證艇員逃生的重要手段,特別是近年來發(fā)生的俄羅斯“庫(kù)爾斯克號(hào)”沉沒事件后,各國(guó)海軍對(duì)潛艇救生工作更加重視,加大了對(duì)深潛救生艇的研發(fā)工作。本文針對(duì)深潛救生艇動(dòng)力定位系統(tǒng)的需求,設(shè)計(jì)了動(dòng)力定位系統(tǒng)的總體結(jié)構(gòu)。本文所設(shè)計(jì)的動(dòng)力定位系統(tǒng)基于3臺(tái)RTD公司的PCI04嵌入式計(jì)算機(jī),每臺(tái)計(jì)算機(jī)負(fù)責(zé)不同的任務(wù),并根掘任務(wù)需求安裝Windows或者VxWorks操作系統(tǒng)?;赪indows操作系統(tǒng)平臺(tái),利用VisualC++6.0實(shí)現(xiàn)了深潛救生艇信息顯示系統(tǒng),該系統(tǒng)基于模塊化思想設(shè)計(jì),方便系統(tǒng)的實(shí)現(xiàn)、移植和維護(hù)。針對(duì)PCI04計(jì)算機(jī)進(jìn)行了VxWork
2、s操作系統(tǒng)的移植和附加模塊驅(qū)動(dòng)程序的編寫,之后基于VxWorks操作系統(tǒng)平臺(tái),利用該操作系統(tǒng)的多任務(wù)機(jī)制,對(duì)動(dòng)力定位系統(tǒng)控制功能進(jìn)行了任務(wù)劃分并開發(fā)了實(shí)時(shí)控制系統(tǒng)。在深入分析深潛救生艇動(dòng)力定位系統(tǒng)需求及實(shí)際工作環(huán)境的情況下,實(shí)現(xiàn)了一種多模態(tài)RBF神經(jīng)網(wǎng)絡(luò)自校正控制算法,該算法具有多個(gè)工作模態(tài),并可以利用RBF神經(jīng)網(wǎng)絡(luò)獲得的Jacobian信息進(jìn)行參數(shù)優(yōu)化。最后將該控制算法在VxWorks操作系統(tǒng)平臺(tái)上實(shí)現(xiàn)。通過半實(shí)物仿真研究表明,本文所設(shè)計(jì)實(shí)現(xiàn)的動(dòng)力定位系統(tǒng)具有易操作性、人機(jī)界面友好性和控制實(shí)時(shí)性。同時(shí),仿真結(jié)果也表明本文所提出的多模態(tài)RBF神經(jīng)網(wǎng)絡(luò)自校正控制算法在海流干擾特別是變海流干擾
3、的情況下的有效性和實(shí)用性。本文的研究成果將用于深潛救生艇動(dòng)力定位系統(tǒng)的產(chǎn)品設(shè)計(jì)中。關(guān)鍵詞:深潛救生艇,動(dòng)力定位系統(tǒng),多模態(tài)控制,RBF神經(jīng)網(wǎng)絡(luò),VxWorks●Aftersubmarinewreck,deepsubmergencerescuevehicleistheimportantmethodforsubmarinersescape.Especially,aftertheincidentofRussia’S”Kursk”sankinrecentyears,statespaymoreattentiontotheworkofSubmarineRescue,andincreaseresearc
4、handdevelopmentofdeepsubmergencerescuevehicle.Inthispaper,wedesigntheoverallstructureofdynamicpositioningsystemaccordingtotherequirementofdeepsubmergencerescuevehicledynamicpositioningsystem.Thispaperachievedadeepsubmergencerescuevehicleinformationdisplaysystem,whichbasedWindowsoperatingsystemplat
5、form,usingVisualC++6.0IDE.BasedontheVxWorksoperatingsystemplatform,thispaperdivisedthetasksfordynamicpositioningsystemcontrolfunctionsanddevelopedareal—timecontrolsystemusingthemulti—taskingmechanism.Uponin—depthanalysisofrequirementofthedynamicpositioningsystemofdeepsubmergencerescuevehicle,thisp
6、aperdesignedamulti-modalRBFneuralnetworkself-tuningcontrolalgorithm,whichisachivedontheVxWorksoperatingsystemplatforms.Semi—physicalsimulationstudieshaveshownthatdynamicpositioningsystemdesignedbythispaperhaseasyoperability,friendlyman-machineinterfaceandreal-timecontr01.Atthesametime,thesimulatio
7、nresultsalsoshowthattheproposedmulti—modalRBFneuralnetworkself-tuningcontrolalgorithmiseffectiveandpracticalintheoceancurrents,especiallyinchangingoceancurrents.Thisresearchwillbeusedintheproductdesignofdeepsubme