论文研究 - 分数阶元素的电磁解释.pdf-资料库

Journal of Modern Physics, 2017, 8, 2209-2218 http://www.scirp.org/journal/jmp ISSN Online: 2153-120X ISSN Print: 2153-1196 Electromagnetic Interpretation of Fractional-Order Elements Guishu Liang, Jiawei Hao*, Dongqing Shan Department of Electric Engineering, North China Electric Power University, Baoding, China How to cite this paper: Liang, G.S., Hao, J.W. and Shan, D.Q. (2017) Electromagnetic Interpretation of Fractional-Order Elements. Journal of Modern Physics, 8, 2209-2218. https://doi.org/10.4236/jmp.2017.814136 Received: November 8, 2017 Accepted: December 18, 2017 Published: December 21, 2017 Copyright © 2017 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access Abstract Fractional circuits have attracted extensive attention of scholars and research- ers for their superior performance and potential applications. Recently, the fundamentals of the conventional circuit theory were extended to include the new generalized elements and fractional-order elements. As is known to all, circuit theory is a limiting special case of electromagnetic field theory and the characterization of classical circuit elements can be given an elegant electro- magnetic interpretation. In this paper, considering fractional-order time de- rivatives, an electromagnetic field interpretation of fractional-order elements: fractional-order inductor, fractional-order capacitor and fractional-order mu- tual inductor is presented, in terms of a quasi-static expansion of the fraction- al Maxwell’s equations. It shows that fractional-order elements can also be in- terpreted as a fractional electromagnetic system. As the element order equals to 1, the interpretation of fractional-order elements matches that of the clas- sical circuit elements: L, C, and mutual inductor, respectively. Keywords Fractional Circuits, Fractional-Order Elements, Maxwell’s Equations, Electromagnetic Field 1. Introduction Recently, the realization and generalization of fractional calculus has become of great significance in many fields [1] such as determining voltage-current rela- tionship in a non-ideal capacitor, fractal behavior of a metal insulator solution interface, electromagnetic waves, and recently in electrical circuits such as filters [2]-[7], oscillators [8] [9] [10], passive realization [11] [12] and energy-related issues in supercapacitors [13]. Furthermore, applications of fractional calculus have been reported in many areas such as physics [14], nonlinear oscillation of earthquakes [15], and mathematical biology [16]. DOI: 10.4236/jmp.2017.814136 Dec. 21, 2017 2209 Journal of Modern Physics

G. S. Liang et al. DOI: 10.4236/jmp.2017.814136 Fractional calculus has higher capability of providing accurate description than integer dynamical systems. In the circuit theory, the fractional-order ele- ment with impedance proportional to sα is considered as a generalized ele- ment that covers the conventional circuit elements when the order of elements equals to integer [17]. And the fractional-order mutual inductance is the genera- lized concept of the mutual inductance in the fractional-order domain [18]. Due to the extra fractional-order parameters, more flexibility is added. During the last 10 years, several promising trials have been introduced for the realizations of the fractional-order element and based on different techniques such as chemical reactions [19], fractal shapes [20], and graphene material [21]. Moreover, many finite circuit approximations were suggested to model fractional-order elements [22] [23] [24] [25]. In the framework of fractional calculus, there are lots of literatures about Maxwell’s equations. Tarasov [26] formulated the fractional Green’s, Stokes’ and Gauss’s theorems and realized the proofs of these theorems for simple regions. He considered fractional nonlocal Maxwell’s equations and the corresponding fractional wave equations. Tarasov [27] used the fractional integrals to describe fractal distribution. Tarasov [28] suggested a fractional generalization of the integral Maxwell equations. Lazo [29] proposed a spatially symmetric and causal gauge invariant fractional electromagnetic field from a Lagrangian formulation. Electromagnetic field propagation has also been modeled using the fractional calculus [30]. It is well known that circuit theory is a limiting special case of electromagnetic field theory. In particular, the characterization of the four classical circuit ele- ments can be given an elegant electromagnetic interpretation by the quasi-static expansion of Max-well’s equations [31] [32] [33]. The electromagnetic interpre- tation of classical circuit elements in Laplace domain is introduced in [34]. The fractional-order elements appeared, but there was no electromagnetic field in- terpretation of fractional-order elements yet. So, by considering fractional-order time derivatives, this paper is to give an analogous interpretation for the charac- terization of fractional-order elements. This paper is organized as follows: in Section 2, the notions of fractional cal- culus and fractional-order elements and fractional Maxwell’s equations are pre- sented. After that, the quasi-static expansion of fractional Maxwell’s equations and the electromagnetic field interpretation of fractional-order elements are shown in Section 3. Finally, conclusions are found in Section 4. In this section, some preliminary definitions are presented. 2. Preliminaries This section is devoted to introducing some preliminaries, which help us to state the results of the paper. 2.1. Fractional Calculus The first definition is about fractional derivative operator as a generalization for 2210 Journal of Modern Physics

G. S. Liang et al. classical derivative operator. There are three kinds of definitions about the frac- tional derivative, the Riemann-Liouville definition, the Grunwald-Letnikov defi- nition and the Caputo definition. According to the Caputo definition for frac- tional derivative, the derivative of order α for function is given by ( ) t f α tD f a ( ) t = 1 ( 1 Γ − ) α t ∫ 0 ( t ) τ − − fα ′ ( ) τ τ d , for )0,1α∈ ( (1) where α tD a α d α= t d , ( )Γ  is Euler’s Gamma function. 2.2. Fractional-Order Elements The template is used to format your paper and style the text. All margins, column widths, line spaces, and text fonts are prescribed; please do not alter them. You may note peculiarities. For example, the head margin in this template measures proportionately more than is customary. This measurement and others are deli- berate, using specifications that anticipate your paper as one part of the entire journals, and not as an independent document. Please do not revise any of the current designations. ( ) i t and ( )u t If ( are respectively the voltage and current of a fraction- )0,1α∈ and the pseudo-capacitance Cα, then ( ) α u t d α t d (2) ( ) i t C α = al-order capacitor with order The impedance of a fractional-order capacitor with order α and the pseu- Similarly, for ( ) z s do-capacitance Cα equals ( ) and i t )0,1β∈ ( ( )u t inductor with order relation holds, j α ( ) C α ω 1 in frequency domain. = as the voltage and current of a fractional-order and the pseudo-inductance Lβ, the following ( ) u t = L β d ( ) β i t β t d (3) The impedance of a fractional-order inductor with order β and pseu- do-inductance Lα equals ( ) z s ) β ω= L ( j β in frequency domain. For fractional coupling mutual inductance, we assume the primary and sec- ,αβ respectively. Conse- ondary inductors are of different fractional orders quently, in this case the mutual inductance is not symmetric, and the mutual inductance could be considered of the fractional-order γ, + M 12 = L 11  u  1   u M  2 = α i d 1 α t d γ d 21 d t i 1 γ 21 21 + L 22 i 2 γ 12 γ d 12 d t d d β i 2 β t (4) In frequency domain, we have, ) ( α j ω ) γ j ω U   U     = ( 21 1 2     γ 12 ( L 11 M 21 ) j ω ( ) j ω β 12 M L 22        I 1 I 2    (5) DOI: 10.4236/jmp.2017.814136 2211 Journal of Modern Physics

G. S. Liang et al. , , , ( ) 0,1 21 12 αβγ γ ∈ where L L M and tors and mutual inductance, 11 mutual inductance. For the traditional case represents the matrix equation of the integer-order mutual inductor. is the fractional order of primary, secondary induc- 21M are primary, secondary, αβγ γ = , Equation (5) , 1 22 12 21 12 , , , , 2.3. Fractional Maxwell’s Equations According to Faraday’s law for induction and with a general fractional-order de- rivative α, we assume that tσ [35] (auxiliary parameter, introduced to be consistent with the time dimensio- nality.) which represents the fractional time components in the system, we have the first Maxwell’s curl equation [36], in this paper, and the parameter )0,1α∈ ( ∇ × E = − µ 1 1 σ− α t α H ∂ α t ∂ (6) Similarly, for some general fractional-order derivative α, the 2nd Maxwell curl equation can be written as ∇ × H J = + ε 1 1 σ− t α E ∂ α α t ∂ (7) The other two equations are independent of time and stay the same, ∇ ⋅ =B =D ∇ ⋅ 0 ρ (8) (9) 3. Electromagnetic Interpretation of Fractional-Order Elements In terms of a quasi-static expansion of fractional Maxwell’s equations, the elec- tromagnetic interpretation of the fractional-order elements is presented in this section. 3.1. Quasi-Static Expansion of Fractional Maxwell’s Equations Quasi-static approximation is used to simplify Maxwell’s equations involving a family of dynamic fields, produced by accelerating charges and time varying currents. For wave velocities less than that of light, these fields can be decom- posed into “static” fields for a finite time duration called family time, to avoid atτ= complications due to retarded potentials. Family time is defined as, where a is the time-rate parameter. Then we have, = α a α E , ∂ ατ ∂ α H ∂ α t ∂ = α a α H . Applying the equations to (6) and (7), we ∂ ατ ∂ α E ∂ α t ∂ get, ∇ × E = − µ aα α H ∂ 1 α σ τ− ∂ t α (10) DOI: 10.4236/jmp.2017.814136 2212 Journal of Modern Physics ∇ × H J = + ε aα α E ∂ 1 α σ τ− ∂ t α (11)

Since the electric field and magnetic field are functions of the parameter a as x y z , and τ, the fields can be expressed as a power series , , well as functions of in aα, G. S. Liang et al. E E = 0 + α E a 1 + ( α a ( α a + )2 E  (12) 2 )2 H  (13) (  , called 0,1,2, k = α a , + ) 2 k    α a = 0 + α H a 1 + x y z , , , , τ α a ) ∂ H H = ( E x y z , , where 1    k ! the kth-order electric field. ) , τ = k ∂ k E 0 ( k = 0,1,2, kH , And , are defined in a similar way, called the kth-order magnetic field, all the other electromagnetic quantities can be expressed, simi- larly, as power series in aα. Equating coefficients of ( )kaα , we get the kth-order fractional Maxwell’s eq- uations For the zero-order set ∇ × ∇ × 0 0 =E (14) =H J (15) 0 0 For the first-order set ∇ × E 1 = − µ α H ∂ 0 α σ τ− ∂ 1 1 t α (16) ∇ × H J = 1 + 1 And, in general, for the kth-order set, ε α E ∂ 0 σ τ− α ∂ 1 1 t α (17) ∇ × E k = − µ 1 1 α − σ t α H ∂ k 1 − α ∂ τ (18) ∇ × H J = k + k ε 1 1 α − σ t α E ∂ k 1 − α τ ∂ (19) Assuming transverse electromagnetic waves travelling in an isotropic me- dium, i.e. electric field along x-axis, magnetic field along y-axis and wave propa- gation along z-axis, the fractional Maxwell's equations in the transform domain are given as follows, ρ ε 0 = k E j z ⋅ x = (20) j k H z ⋅ y (21) 1 j αω 1 σ− t H (22) µ ) α y j k E z × x = − ( j k z × H y = ) σ ω E + ( j α 1 1 σ− α t E (23) ε x DOI: 10.4236/jmp.2017.814136 where ω is angular frequency, k is propagation vector, λ is wavelength, 2213 Journal of Modern Physics

G. S. Liang et al. σ is electrical conductivity and ρ is charge density. Then we get the zero- and first-order fractional Maxwell’s equations, − 0 0 = kE = (24) 0 j Eσ jkH (25) 0 1 j αω 1 σ− t (26) H µ ( ) α 0 − j kE 1 = − − j kH 1 = E σ 1 + ( α ) j ω 1 1 α σ− t E ε 0 (27) 3.2. An Electromagnetic Interpretation of Fractional-Order Elements Consider the characteristic impedance of a fractional electromagnetic system, given by the ratio of the electric and magnetic fields with a factor of “ j” since the fields are orthogonal, Z = E j H = j ( ( E E 1 0 H H + + 0 1 ) ) (28) Then, a resistor is identified to be a fractional electromagnetic system whose first-order fields are negligible compared to its zero-order fields. i.e. E H = . 0 1 , 1 Z = k E j 2π 0 H σ σλ = = 0 (29) The characteristic impedance is a resistor. A fractional-order inductor is identified to be a fractional electromagnetic sys- tem which only considers the first order electric field and the zero-order magnetic field i.e. E H = . 0 0 , 1 Z = E j 1 H 0 = E j 1 H 0 = ( α ) j ω µλ 1 σ− α 2π t (30) The characteristic impedance is a fractional-order inductor. As 1α= , the sys- tem reduces to an inductor. A fractional-order capacitor is identified to be a fractional electromagnetic sys- tem where only consider the first order magnetic field and the zero-order electric field i.e. E H = . 0 1 , 0 Z = E j 0 H 1 = E j 0 H 1 = 1 − α 2π σ t ( α j ω ελ ) (31) The characteristic impedance is a fractional-order capacitor. As 1α= , the system reduces to a capacitor. The fractional-order mutual inductor can also be seen as a fractional electro- ,αβ is the order of prima- magnetic system. We assume that, for convenience, ry and secondary inductor. What we care about is the effect of the magnetic field on the electric field in fractional mutual coupled circuits, from (10), we obtain, In primary inductor, 2214 Journal of Modern Physics DOI: 10.4236/jmp.2017.814136

In secondary inductor, G. S. Liang et al. ∇ × E t 1 = − µ 1 αα H a ∂ t 1 α σ τ− ∂ t α 1 (32) ∇ × E t 2 = − µ 2 H a t 1 β β τ− σ t ββ ∂ ∂ 2 (33) where E H E H is the total fields in primary and secondary inductor. t , , , 1 1 2 2 t t t As the order of primary inductor is different from that of the secondary induc- H H can be regarded as the superposition of the magnetic fields in li- ,t 1 tor, near system, we assume that, 2 t H t 1 = l 11 H c 1 + l 12 1 − α γ σ 12 t − γ α ∂ 12 τ ∂ γ α 12 − H t 2 = l 21 1 − β γ σ t 21 − γ β ∂ 21 ∂ τ γ β 21 − H c 1 + l 12 H (34) c 2 H (35) c 2 c 2 , 21 l 12 H H is the fields produced by primary and secondary inductor, where ,c 1 l l l , functions of the space coordinates alone, represent the contribution , , 22 11 to the total magnetic field from ,γ γ is the order of mu- 12 tual inductance in primary and secondary inductor, respectively. And when α β= 2cH , and 1cH or 21 = . γ γ α 12 = 21 , we have Substituting (34) (35) into (32) (33) respectively, we get, γ ∂ 12 γ σ τ ∂ 12 ∇ × = − µ 1 µ 1 l 11 H E − c 1 1 t α a 1 γ − 12 t l 12 l 22 H (36) c 2 H (37) c 2 α α a ∂ 1 α α − σ τ ∂ t a 1 t γ ∂ 21 σ τ− γ ∂ β γ 21 21 ∇ × E t 2 = − µ 2 l 12 H c 1 − µ 2 β β a ∂ 1 β − β σ τ ∂ t E H H , as power series in By expressing the fields in primary inductor, t aα, and equating coefficients of ( )kaα , the corresponding first-order fractional Maxwell’s equation of the primary inductor can be deduced as we get (26). In the transform domain, we have c 1 , , 1 2 c − j k E t 1 11 = − ( α ) 1 j αω 1 σ− t µ 1 H t 01 (38) where E t 11 = E ∂ t 1 α a ∂ α a = 0 is the total first order electric field in primary inductor, t 01 = H and l H 11 1 α γ ω − σ t field in primary inductor. α a = + ( c 1 j 12 0 γ α − ) 12 l H 12 c 2 α a = 0 , the total zero-order magnetic In the similar way, we can get the corresponding equation in secondary induc- tor, − j k E t 2 12 = − ( β ) 1 j βω 1 σ− t µ 2 H t 02 (39) DOI: 10.4236/jmp.2017.814136 where E t 12 = E ∂ t 2 β a ∂ β = a 0 is the first order electric field in secondary inductor, and 2215 Journal of Modern Physics

G. S. Liang et al. H t 02 = l H 21 c 1 β = a 0 + l H 22 c 2 β = a 0 , the total zero-order magnetic field 1 β γ ω − σ t ( j 21 γ β − ) 21 in secondary inductor. As the first term of the power series expansion of magnetic field in aα is the same as the first term of the power series expansion of magnetic field in aβ, let H H H H = = , and . c 02 c 1 β = a 0 c 2 α a = 0 c 2 β = a 0 = H H c 01 Then we obtain, α a = = c 1 0 − j k E t 1 11 = − µ 1 − j k E t 2 12 = − µ 2 1 1 α − σ t 1 1 σ− γ 21 t ( α ) j ω l H 11 c 01 − µ 1 1 1 − γ σ 12 t ( j ω γ ) 12 l H 12 c 02 (40) ( γ 21 ) j ω l H 21 c 01 − µ 2 1 1 − β σ t ( β ) j ω l H 22 c 02 (41) when cH = , 0 02 z 11 = E j t H c 11 01 = k ( ( 1 α − 2π α ω µ ω µλ σ 1 t 1 − σ t l 11 1 1 l 11 1 = ) ( ) α α j j ) (42) z 21 = E j t H c 12 01 = ( when cH = , 0 01 γ 21 j ω µ 2 21 l ) k 2 1 − γ σ 21 t = ( γ 21 ( 1 − γ j ω µλ σ 21 t l 21 2 2 ) 2π ) (43) z 22 = β j ω µ 2 l 22 ) = ( 12 02 z 12 = = ( γ 12 j ω µ 1 l 12 ) c E j t H E j t H 11 c 02 z k 2 1 β − σ t k 1 1 γ − σ 12 t = ( = ( β ( 1 β − j ω µλ σ t l 22 2 2 ) 2π ) (44) γ 12 ( 1 γ − j ω µλ σ 12 t l 12 1 1 ) 2π ) (45) , where z 11 z 12 represent the element of characteristic impedance matrix. Then we have the characteristic impedance matrix of the fractional electro- z 22 21 , , magnetic system, Z =    z 11 z 21 z 12 z 22    =     ( α ) l 11 1 1 ( ( 1 α − j 2π ω µλ σ t ( ) 1 γ − j ω µλ σ 21 t ) 2π 2 2 γ 21 21 l ) γ 12 l 12 ( ( ) 1 γ − j ω µλ σ 12 t ( ( ) 1 β − j ω µλ σ t l 22 2 2 1 1 β ) 2π ) 2π     The characteristic impedance matrix is a fractional-order mutual inductor. As = , the fractional-order mu- α β γ γ = the element order equal to 1, i.e. 12 tual inductor reduce to the mutual inductor. = = 21 1 4. Concluding Remarks In terms of quasi-static expansion of the fractional Maxwell’s equations consider- ing fractional-order time derivatives, we get the electromagnetic interpretation of several fractional-order elements: fractional-order inductor, fractional-order capacitor, and fractional-order mutual inductance. It shows that fractional-order elements can also be interpreted as a fractional electromagnetic system. As the element order equal to 1, the interpretation of fractional-order elements matches that of the classical circuit elements: L, C, and the mutual inductance, respectively. Future research will aim to get the electromagnetic interpretation of fraction- al-order memristor fractional high-order elements. 2216 Journal of Modern Physics DOI: 10.4236/jmp.2017.814136

资料库

论文研究 - 分数阶元素的电磁解释.pdf

相关推荐

开发技术

热门标签

最新资料