荊永明,梁貴書,趙海成
(1.華北電力大學(xué) 電氣與電子工程學(xué)院, 河北 保定 071003; 2.山西耀光煤電有限責(zé)任公司,山西 晉中 031100)
分?jǐn)?shù)階雙二次型導(dǎo)抗函數(shù)的無(wú)源綜合
荊永明1,梁貴書1,趙海成2
(1.華北電力大學(xué) 電氣與電子工程學(xué)院, 河北 保定 071003; 2.山西耀光煤電有限責(zé)任公司,山西 晉中 031100)
無(wú)源網(wǎng)絡(luò)綜合一直是備受關(guān)注的研究課題,近幾年,隨著分?jǐn)?shù)階微積分的廣泛應(yīng)用,人們開始研究含有分?jǐn)?shù)階階次的策動(dòng)點(diǎn)函數(shù)綜合?;谌诩冸娮杈W(wǎng)絡(luò)與兩個(gè)分?jǐn)?shù)階儲(chǔ)能元件的連接,提出了分?jǐn)?shù)階雙二次型導(dǎo)抗函數(shù)最少儲(chǔ)能元件無(wú)源綜合方法。推導(dǎo)了其實(shí)現(xiàn)的充分必要條件。此外,對(duì)于無(wú)法用電阻矩陣表示三口電阻網(wǎng)絡(luò)的情況,通過(guò)相應(yīng)的串并聯(lián)電路得到其實(shí)現(xiàn)條件及對(duì)應(yīng)元件值。當(dāng)兩個(gè)分?jǐn)?shù)階階次均為1時(shí),所提方法可退化為傳統(tǒng)雙二次型導(dǎo)抗函數(shù)綜合,因而具有更廣的應(yīng)用范圍,最后,通過(guò)數(shù)值例子驗(yàn)證了所提方法的準(zhǔn)確性。
分?jǐn)?shù)階雙二次導(dǎo)抗函數(shù);三口電阻網(wǎng)絡(luò);網(wǎng)絡(luò)綜合
近些年,分?jǐn)?shù)階電路和系統(tǒng)已廣泛應(yīng)用于不同實(shí)際工程中。分?jǐn)?shù)階電子電路的研究包括其穩(wěn)定性分析[1]、時(shí)域響應(yīng)分析[2]、頻域響應(yīng)分析[3]、能源效率分析[4]、共振條件分析[5]和靈敏度分析[6]。對(duì)于分?jǐn)?shù)階電路的綜合,傳統(tǒng)思路是通過(guò)整數(shù)階方法逼近分?jǐn)?shù)階系統(tǒng),再對(duì)其進(jìn)行物理實(shí)現(xiàn)[7,8]。然而,因整數(shù)階元件不能直接體現(xiàn)分?jǐn)?shù)階階特性,逼近時(shí)所需元件數(shù)較多,結(jié)構(gòu)復(fù)雜。目前,已有不同方法搭建分?jǐn)?shù)階電容器和電感器[9,10],基于這些電子元件,分?jǐn)?shù)階系統(tǒng)可以簡(jiǎn)單精確地實(shí)現(xiàn)[11]。因此,研究基于分?jǐn)?shù)階元件的無(wú)源網(wǎng)絡(luò)綜合具有至關(guān)重要的意義。然而,分?jǐn)?shù)階網(wǎng)絡(luò)函數(shù)很難有一般的實(shí)現(xiàn)方法,文獻(xiàn)[12]通過(guò)RLC雙口網(wǎng)絡(luò)與分?jǐn)?shù)階電容連接,在一定條件下可以綜合分?jǐn)?shù)階策動(dòng)點(diǎn)阻抗函數(shù)。
(1)
1.1 電路結(jié)構(gòu)
欲通過(guò)三口電阻網(wǎng)絡(luò)與分?jǐn)?shù)階儲(chǔ)能元件的連
接實(shí)現(xiàn)如式(1)分?jǐn)?shù)階雙二次導(dǎo)納函數(shù)。因不同類型儲(chǔ)能元件的連接可實(shí)現(xiàn)更廣的網(wǎng)絡(luò)函數(shù),所以采用一個(gè)分?jǐn)?shù)階電容和一個(gè)分?jǐn)?shù)階電感與三口電阻網(wǎng)絡(luò)的連接方式進(jìn)行綜合,如圖1所示。
圖1 三口電阻網(wǎng)絡(luò)與分?jǐn)?shù)階儲(chǔ)能元件連接Fig.1 Three port resistive network connected with fractional energy storage elements
純電阻網(wǎng)絡(luò)為基本網(wǎng)絡(luò),端口2接α階電感,端口3接β階電容。三口純電阻網(wǎng)絡(luò)電阻矩陣為
(2)
(3)
(4)
其中,
將R矩陣轉(zhuǎn)換可得X矩陣,轉(zhuǎn)變過(guò)程如式(5)
(5)
其中,X為半正定矩陣,且
(6)
則式(3)可以轉(zhuǎn)化為式(4)。
1.2 分?jǐn)?shù)階策動(dòng)點(diǎn)導(dǎo)納函數(shù)實(shí)現(xiàn)
給定如式(1)策動(dòng)點(diǎn)導(dǎo)納函數(shù),首先需轉(zhuǎn)換為式(4)含有X系數(shù)的導(dǎo)納函數(shù),得到半正定矩陣X,再轉(zhuǎn)為式(3)含有Rij(i,j=1,2,3)系數(shù)的導(dǎo)納函數(shù),得到Paramount矩陣R。最后,根據(jù)Tellegen電路實(shí)現(xiàn)矩陣,根據(jù)k,b確定分?jǐn)?shù)階電容和分?jǐn)?shù)階電感值,即可得到圖1電路結(jié)構(gòu)。
1.2.1 導(dǎo)納函數(shù)系數(shù)應(yīng)滿足的條件
定理1:根據(jù)文獻(xiàn)[13]定理4的思路可得,給定一般的分?jǐn)?shù)階策動(dòng)點(diǎn)導(dǎo)納函數(shù)如式(1)所示,其可以轉(zhuǎn)換為式(4)的條件為
(7)
(8)
(9)
X3=a0X2
(10)
(11)
(12)
d1≥a1
(13)
a0d1-d0a1≥0
(14)
a0≥a1a2
(15)
(16)
(17)
1.2.2 導(dǎo)納函數(shù)的實(shí)現(xiàn)
式(4)對(duì)應(yīng)的系數(shù)為X矩陣,當(dāng)且僅當(dāng)其可以轉(zhuǎn)化為式(3),且對(duì)應(yīng)R為Paramount矩陣[13],k>0,b>0,才能通過(guò)圖1的方式實(shí)現(xiàn)。下面討論式(4)應(yīng)滿足的條件。
文獻(xiàn)[13]中的引理1和引理2介紹并證明了半正定矩陣X轉(zhuǎn)換為Paramount矩陣R應(yīng)滿足的條件,本文不再贅述。
1.2.3 綜合示例
考慮式(18)分?jǐn)?shù)階雙二次型導(dǎo)納函數(shù)的實(shí)現(xiàn)。
(18)
(19)
(20)
圖2 式 (18)電路實(shí)現(xiàn)結(jié)構(gòu)Fig.2 Circuit for equation (18)
上文給出了采用兩個(gè)分?jǐn)?shù)階元件和電阻組成網(wǎng)絡(luò)實(shí)現(xiàn)策動(dòng)點(diǎn)導(dǎo)納函數(shù)的一般實(shí)現(xiàn)方法,對(duì)于符合相應(yīng)條件的導(dǎo)納函數(shù),可按照上述流程進(jìn)行綜合。然而,當(dāng)三口Tellegen電路(如圖3所示)的某一元件值為無(wú)窮大時(shí),式(2)電阻矩陣R無(wú)法表示,上述方法失效。盡管可用其對(duì)偶矩陣G求解,本文針對(duì)特殊情況提出了更簡(jiǎn)單綜合方法。
圖3 電阻R矩陣對(duì)應(yīng)Tellegen電路Fig.3 Tellegen circuit corresponding to matrix R
2.1 特殊網(wǎng)絡(luò)結(jié)構(gòu)分類
圖3中,為了不短接端口2的分?jǐn)?shù)階電感,G1和G2不能同時(shí)為無(wú)窮大;為了不短接端口3的分?jǐn)?shù)階電容,G4不能為無(wú)窮大。因此,特殊情況的串并聯(lián)電路可分為兩大類,8種對(duì)應(yīng)電路,如表1所示。
表1 不能用R矩陣表示三口電阻網(wǎng)絡(luò)
Tab.1 Three port resistive circuits which cannot be expressed by matrixR
特殊三口電路結(jié)構(gòu)分類G1=∞,G2=0①G3=G4=0②G3=0,0 2.2 特殊結(jié)構(gòu)實(shí)現(xiàn)的充分必要條件 針對(duì)每種電路,推導(dǎo)了其實(shí)現(xiàn)的充分必要條件及相應(yīng)元件值,以②號(hào)電路為例,過(guò)程見定理5。 (21) 圖4 G1=∞、G2=0、G3=0、0 其他7種策動(dòng)點(diǎn)導(dǎo)納函數(shù)所對(duì)應(yīng)的電路結(jié)構(gòu)及其可以實(shí)現(xiàn)的充分必要條件匯總?cè)绫?(見附錄)。 2.3 綜合示例 考慮式(22)分?jǐn)?shù)階雙二次型導(dǎo)納函數(shù)的實(shí)現(xiàn)。 (22) 圖5 式(22)的電路實(shí)現(xiàn)結(jié)構(gòu)Fig.5 Circuit for equation (22) 本文提出了分?jǐn)?shù)階雙二次型導(dǎo)抗函數(shù)的最少儲(chǔ)能元件綜合方法,即通過(guò)三口電阻網(wǎng)絡(luò)與分?jǐn)?shù)階電容和分?jǐn)?shù)階電感連接進(jìn)行實(shí)現(xiàn),推導(dǎo)了其實(shí)現(xiàn)的充分必要條件。此外,對(duì)于無(wú)法用電阻矩陣表示三口電阻網(wǎng)絡(luò)的情況,通過(guò)相應(yīng)的串并聯(lián)電路得到其實(shí)現(xiàn)條件及對(duì)應(yīng)元件值。最后通過(guò)示例驗(yàn)證了所提方法的準(zhǔn)確性。當(dāng)階次α=β=1時(shí),此方法可退化為傳統(tǒng)雙二次導(dǎo)抗函數(shù)的綜合,因而具有更寬的應(yīng)用范圍,為分?jǐn)?shù)階無(wú)源網(wǎng)絡(luò)綜合提供了新方法。 [1]RADWANAG,Stabilityanalysisofthefractional-ordercircuit[J].FractionalCalculusAppl,2012,3(1): 1-13. [2]JACYNTHOLA,TEIXEIRAMCM,ASSUNCAOE,etal.Identificationoffractional-ordertransferfunctionsusingastepexcitation[J].CircuitsandSystemsII:ExpressBriefs,IEEETransactionson, 2013, 62(9): 896-900. [3]GALVORKH,HADJILOUCASS,KIENITZKH,etal.Fractionalordermodelingoflargethree-dimensionalRCnetworks[J].CircuitsandSystemsI:RegularPapers,IEEETransactionson, 2013, 60(3): 624-637. [4]HARTLEYTT,VEILLETTERJ,ADAMSJL,etal.Energystorageandlossinfractional-ordercircuitelements[J].Circuits,Devices&Systems,IET, 2013, 9(3): 227-235. [5]RADWANAG.Resonanceandqualityfactorofthefractionalcircuit[J].EmergingandSelectedTopicsinCircuitsandSystems,IEEEJournalon, 2013, 3(3): 377-385. [6]DIAOLJ,ZHANGXF,CHENDY.Fractional-ordermultiplecircuit[J].ActaPhysicaSinica, 2014,63(3): 038401. [7]MAIONEG.Closed-formrationalapproximationsoffractional,analoganddigitaldifferentiatorsintegrators[J].EmergingandSelectedTopicsinCircuitsandSystems,IEEEJournalon, 2013, 3(3): 322-329. [8]MAIONEG.Laguerreapproximationoffractionalsystems[J].ElectronicsLetters, 2002, 38(20): 1234-1236. [9]KRISHNAMS,DASS,BISWASK,etal.Fabricationofafractionalordercapacitorwithdesiredspecifications:Astudyonprocessidentificationandcharacterization[J].ElectronDevices,IEEETransactionson, 2011, 58(11): 4067-4073. [10]MONDALD,BISWASK.PackagingofSingleComponentFractionalOrderElement[J].DeviceandMaterialsReliability,IEEETransactionson, 2013, 13(1): 73-80. [11]TAVAZOEIMS,TAVAKOLI-KAKHKIM.Minimalrealizationsforsomeclassesoffractionalordertransferfunctions[J].EmergingandSelectedTopicsinCircuitsandSystems,IEEEJournalon, 2013, 3(3): 313-321. [12]SARAFRAZMS,TAVAZOEIMS.RealizabilityofFractional-OrderImpedancesbyPassiveElectricalNetworksComposedofaFractionalCapacitorandRLCComponents[J].CircuitsandSystemsI:RegularPapers,IEEETransactionson, 2013, 62(12): 2829-2835. [13]CHENMZQ,SMITHMC.Electricalandmechanicalpassivenetworksynthesis[M].RecentAdvancesinLearningandControl.SpringerLondon, 2008: 35-50. [14] 王鍇, 陳志強(qiáng). 雙二次阻抗函數(shù)的低復(fù)雜度實(shí)現(xiàn)問(wèn)題 [C].第 24 屆中國(guó)控制與決策會(huì)議論文集. 2012. 附錄 Passive Synthesis of Fractional Biquadratic Immittance Function JING Yongming1, LIANG Guishu1,ZHAO Haicheng2 (1.School of Electrical and Electronic Engineering, North China Electric Power University, Baoding 071003,China;2.Shanxi Yaoguang Coal & Power Limited Liability Company, Jinzhong 031100,China) Passive network synthesis is a research topic which has been a focus one. In recent years, with the wide application of fractional calculus, researchers began to study and analyze the driving-point function synthesis containing fractional order. The passive synthesis method which contains minimal energy storage element for fractional biquadratic immittance function is proposed based on the connection of three-port pure resistive network and two fractional order storage elements. The necessary and sufficient conditions for the realization of the proposed method are deduced. In addition, if the three-port pure resistive network cannot be represented by resistance matrix, the realization condition and corresponding component values can be obtained by relevant series-parallel circuit. When the both orders are equal to 1, the proposed method degenerates into traditional biquadratic immittance function synthesis, which indicates it has wider application. At last, the proposed method is verified by numerical examples. fractional biquadratic immittance function; three-port resistive network; network synthesis 表2 特殊分?jǐn)?shù)階雙二次策動(dòng)點(diǎn)導(dǎo)納函數(shù)的實(shí)現(xiàn)結(jié)構(gòu)、實(shí)現(xiàn)條件及對(duì)應(yīng)元件值 Tab.2 Synthesis structure, realization condition and component values for special fractional biquadratic driving-point admittance function 編號(hào)對(duì)應(yīng)網(wǎng)絡(luò)函數(shù)實(shí)現(xiàn)結(jié)構(gòu)元件值與系數(shù)間關(guān)系滿足條件①Ys()=CsβEsα+β+Gsβ+HLα=EC,Cβ=CH,R6=GCC、E、H>0,G≥0,當(dāng)G=0時(shí),R6=0②Ys()=Csβ+DEsα+β+Fsα+Gsβ+HLα=EC,Cβ=CH-DGH2,G4=DH,R6=GHCH-DGC、D、E、F、H>0,G≥0DE=CF,CH-DG>0,當(dāng)G=0時(shí),R6=0③Ys()=Asα+β+Csβ+DEsα+β+Gsβ+HLα=EHCH-DG,Cβ=CH-DGH2,G3=DH,R6=FGAH-BGA、C、D、E、H>0,G≥0,AH=DE,CH-DG>0AH-BG>0,當(dāng)G=0時(shí),R6=0④Ys()=Asα+β+Bsα+Csβ+DEsα+β+Fsα+Gsβ+HLα=F2DF-BH,Cβ=DF2G-BEGH-BCFGCF2H,G3=BFG4=DF-BHHF,R6=CF2DF2-BEH-BCFA、C、D、E、H>0,G≥0ABFH2+BDEFH+BDF2G=ADF2H+B2EH2+2B2FGHDEFH+DF2G=CF2H+BEH2+DF2GDF2G-BEGH-BCFG>0,DF-BH>0DF2-BEH-BCF>0,當(dāng)G=0時(shí),R6=0⑤Ys()=Asα+β+Bsα+Csβ+DEsα+β+FsαLα=FD,Cβ=AD-BCDF,G4=BF,R6=EFAF-BEA、B、D、F>0,C、E≥0DE=CF,AF-BE>0,AD-BC>0當(dāng)C=E=0時(shí),R6=0⑥Ys()=Asα+β+Csβ+DEsα+β+FsαLα=FD,Cβ=AF,R6=EAA、D、F>0,C、E≥0DE=CF,當(dāng)C=E=0時(shí),R6=0⑦Ys()=Asα+β+Csβ+DEsα+β+Fsα+Gsβ+HLα=FD,Cβ=AF,G3=DH,R6=FGAHA、C、D、F、H>0,E、G≥0AEH2+DEFG=CF2G,FG=HE當(dāng)E=G=0時(shí),R6=0⑧Ys()=Asα+β+Bsα+Csβ+DEsα+β+Fsα+Gsβ+HLα=F2DF-BH,Cβ=AH-BGHFG3=DF-BHHF,G4=BF,R6=FGAH-BGA、C、D、F、H>0,E、G≥0AEH2+DEFG=BEGH+CF2G,FG=HEAH-BG>0,DF-BH>0當(dāng)E=G=0時(shí),R6=0 10.3969/j.ISSN.1007-2691.2017.01.07 2016-05-16. 國(guó)家自然科學(xué)基金資助項(xiàng)目(51177048);河北省自然科學(xué)基金資助項(xiàng)目(E2012502009). TN711.1 A 1007-2691(2017)01-0046-06 荊永明(1987 -),男,博士研究生,研究方向?yàn)殡娋W(wǎng)絡(luò)理論及其在電力系統(tǒng)中的應(yīng)用;梁貴書(1961 -),男,教授,主要研究方向?yàn)殡娋W(wǎng)絡(luò)理論及其應(yīng)用,電力系統(tǒng)電磁兼容和電力信息分析與處理等。3 結(jié) 論