資源描述：
《織機經(jīng)紗張力控制策略研究》由會員上傳分享，免費在線閱讀，更多相關(guān)內(nèi)容在學術(shù)論文-天天文庫。
1、浙江夫嚳碩士學位論文基予ARM的織機經(jīng)紗張力控制蕺略研究摘要本文研究的主要內(nèi)容是織機系統(tǒng)的張力控制策略。目前，國內(nèi)大多數(shù)織機的經(jīng)紗強力控制是采穰掇槭控制或蔣統(tǒng)PID控翱，英效萊不太理想，導致紡緩晶產(chǎn)量低、質(zhì)量麓。合理的張力控制方法將減小張力的波動，使經(jīng)紗承受的張力平穩(wěn)，默麗避免開車痰等。因此，開發(fā)蔫精度豹張力控捌系縫是提辯織撬檔次的關(guān)鍵艨在。本文串分櫥與褥究了織橇囂工捧蒙理，彰穰張力波凄鵓各個囂素，綴瓿經(jīng)紗張力的特性，并利用Matlab辨識工具箱建立系統(tǒng)的數(shù)學模型。在PID控制的基懿上，進一步研究7蔻種改進型PID控潮，在窶瑟藏瘸孛，取褥了不錯鵑
2、教暴。傳統(tǒng)PID控制雖然由予其算法簡單、魯棒性好和可靠性高，依賴予受控對象瀚數(shù)學模型。針對織祝控制孛存在的控制闖題，提出了神經(jīng)嬲絡控割策略來達到更好的控制效果。利用神經(jīng)網(wǎng)絡能夠逼近任意非線性系統(tǒng)的優(yōu)點，將神經(jīng)網(wǎng)絡和PID控制穗縫合，舞凄了糟經(jīng)鬻絡整定瓣P(guān)ID燕剃方法。還使耀marl曲對基予卡爾曼濾波器的RBF徑向神經(jīng)網(wǎng)絡整定的PID控制進行了仿真。仿真實驗結(jié)果襲臻，基于神經(jīng)瓣絡整定熬經(jīng)紗張力控制系統(tǒng)熬拄翻效粟和湊態(tài)性麓都明顯優(yōu)于傳統(tǒng)PID控鍘。關(guān)鍵謠：經(jīng)紗張力，系統(tǒng)辨識，辮p控鍘，耱經(jīng)鼴絡控制，專家蒸統(tǒng)浙江夫嚳碩士學位論文基予ARM的織機經(jīng)紗張力
3、控制蕺略研究AbstractThemaincontentofthisthesisistensioncontrolstrategyofloomsystem．Atpresnt，manyofthedomesticloomsusemechanicalcontrolorconventionalPIDcontrolwhoseeffectisnotideal，leadingtotextileproductionlow，poorquality。Themethodofreasonabletensioncontrolwillreducethewaveoftheten
4、sion，SOthatthetensionwhichwarpbearsstabilizes，SOavoidingtheprobabilityofstartingmarksandSOon。Therefore，researchanddevelopmentofhigh-precisiontensioncontrolsystemistoenhancethekeytoloomgrade．Thisthesisanalysestheworkingprincipleoftheloom，thefactorswhichmakethewarptensionfluctu
5、ate,andthecharacteristicofthewarptensioncontrol，andestablishsmathematicalmodelofloomwithMatlabtoolboxidentificationsystem．BasedonPIDcontrol，severalimprovedPIDcontrolsarefurtherstudied．Inpractice，itgetsagoodresult。ConventionalPIDcontrolmethod，whichhasalotofadvantage：simplealgo
6、rithm，goodrobustnessandhi．ghreliability，relaysonthemathematicalmodel．Duetothecontrolprobleminloomcontrol，itraisestheneuralnetworkcontrolstrategy，inordertoachivebettercontroleffect．TakingadvantageofneuralnetworkwhichCanapproachanynon—linearsystem，amethodofcombiningneuralnetwor
7、kwithPIDcontrollerisproposed,andUsematlabtosimulatetheadaptivePIDcontrolmethodbasedonRBFneuralnetworkwithaKamanfilteranditgetsagoodresult．ThesimulateresultindicatesthatcontroleffectanddynamicperformanceforadaptivePIDcontrolforwarptensionsystembasedonneuralnetworka恐obviouslysu
8、periortoconvenfionMPIDcontr01．Keywords：Warptension，systemidentificat
