資源描述:
《智能作動(dòng)器Preisach遲滯參數(shù)辨識(shí)及補(bǔ)償研究.pdf》由會(huì)員上傳分享,免費(fèi)在線(xiàn)閱讀,更多相關(guān)內(nèi)容在工程資料-天天文庫(kù)。
1、振動(dòng)與沖擊第33卷第2O期JOURNALOFVIBRATIONANDSHOCK智能作動(dòng)器Preisach遲滯參數(shù)辨識(shí)及補(bǔ)償研究王萍萍,劉磊,欒曉娜(1.哈爾濱工業(yè)大學(xué)航天學(xué)院,哈爾濱150001;2.大連理工大學(xué)航空航天學(xué)院,遼寧大連116024)摘要:針對(duì)實(shí)現(xiàn)智能作動(dòng)器精確運(yùn)動(dòng)需對(duì)遲滯非線(xiàn)性進(jìn)行辨識(shí)、補(bǔ)償問(wèn)題,采用奇異值分解方法辨識(shí)Preisach模型參數(shù),構(gòu)造變幅值諧波輸入信號(hào)滿(mǎn)足持續(xù)激勵(lì)條件;基于辨識(shí)模型設(shè)計(jì)逆模型前饋補(bǔ)償Preisach遲滯非線(xiàn)性;通過(guò)壓電平臺(tái)檢驗(yàn)該辨識(shí)、補(bǔ)償方法。實(shí)驗(yàn)結(jié)果表明,奇異值分解法可有效辨識(shí)遲
2、滯效應(yīng)非線(xiàn)性模型,基于辨識(shí)結(jié)果的逆模型補(bǔ)償可減小跟蹤誤差89.5%。關(guān)鍵詞:智能作動(dòng)器;Preisach遲滯效應(yīng);奇異值分解;逆模型中圖分類(lèi)號(hào):TP24;TP29文獻(xiàn)標(biāo)志碼:AParameteridentificationandcompensationofPreisachhysteresisinsmartactuatorsWANGPing-ping,Lei,LUANXiao—na(1.SchoolofAstronautics,HarbinInstituteofTechnology,Harbin150001,China;2.Da
3、lianUniversityofTechnology,SchoolofAeronauticsandAstronautics,Dalian116024,China)Abstract:Theworkingperformanceofsmartactuatorsisaffectedbythehystereticeffectnonlinearity.Thus,itisnecessarytoidentifyandcompensatethehysteresisnonlinearitytoachieveprecisemotion.Byvirt
4、ueofthesingularvaluedecomposition(SVD)method,theparametersofthePreisachmodelwereidentified,takingdesignedharmonicsignalswithvaryingamplitudesasinputstosatisfyPEcondition.Then,basedontheidentifiedPreisachmodel,themodel—inversionfeedforwardwasdesignedtocompensatethehy
5、steresisnonlinearity.Apiezoelectricstagewasusedtovalidatetheproposedidentificationandcompensationmethods.TheexperimentalresultsindicatethatthehysteresisnonlinearitycanbeidentifiedusingtheSVDmethod.a(chǎn)ndthetrackingerrorisreducedby89.5%usingthemodel-inversioncompensator
6、.Keywords:smartactuator;Preisachhysteresiseffect;singularvaluedecomposition;model—inversion智能作動(dòng)器廣泛用于微納米技術(shù)領(lǐng)域,但其遲滯也會(huì)產(chǎn)生小的奇異值,需進(jìn)行截?cái)嘁詼p小誤差。但若非線(xiàn)性效應(yīng)影響作動(dòng)器精度,而遲滯非線(xiàn)性效應(yīng)造成輸人信號(hào)不滿(mǎn)足持續(xù)激勵(lì)(PE)條件J,也易造成較的誤差可達(dá)位移的10%一15%_1J。為提高作動(dòng)器精大辨識(shí)誤差。因此需采用具有變幅值的輸入信號(hào)進(jìn)行度,需研究智能作動(dòng)器遲滯非線(xiàn)性辨識(shí)及補(bǔ)償J。參數(shù)辨識(shí),以滿(mǎn)足持續(xù)激勵(lì)條
7、件,提高辨識(shí)精度。為驗(yàn)遲滯效應(yīng)物理特性較復(fù)雜,可通過(guò)基于現(xiàn)象的數(shù)證辨識(shí)結(jié)果,本文設(shè)計(jì)Preisach逆模型前饋補(bǔ)償遲滯非學(xué)模型描述,尤其Preisach模型。經(jīng)典Preisach遲線(xiàn)性,利用壓電平臺(tái)實(shí)驗(yàn)驗(yàn)證遲滯辨識(shí)、補(bǔ)償方法的有滯非線(xiàn)性只依賴(lài)輸入信號(hào)路徑,與信號(hào)頻率無(wú)關(guān),而效性。Preisach具有全局記憶性_5J,即過(guò)去、當(dāng)前時(shí)刻輸入共1Preisach遲滯非線(xiàn)性模型同決定將來(lái)某時(shí)刻輸出。為辨識(shí)Preisach模型,Henze等通過(guò)假設(shè)Preisach模型分布函數(shù)利用數(shù)據(jù)點(diǎn)進(jìn)1.1Preisach遲滯非線(xiàn)性模型行參數(shù)辨識(shí),但
8、通常較難提前獲得某壓電作動(dòng)器的經(jīng)典Preisach遲滯非線(xiàn)性模型為遲滯算子與相應(yīng)分布函數(shù);Song等對(duì)輸入信號(hào)進(jìn)行微分求解Preisach密度函數(shù)的二重積分j,即密度函數(shù),但易受測(cè)量噪聲影響。本文用最小二乘與奇異值分解方法辨識(shí)Preisach模型參數(shù)。噪聲及擾動(dòng),(t)=『主