fractional active disturbance rejection...

11
Research Article Fractional active disturbance rejection control Dazi Li a,n , Pan Ding a , Zhiqiang Gao b a Institute of Automation, Beijing University of Chemical Technology, Beijing 100029, PR China b Center for Advanced Control Technologies, Cleveland State University, Cleveland, OH 44115, USA article info Article history: Received 3 June 2015 Received in revised form 5 January 2016 Accepted 29 January 2016 Available online 28 February 2016 Keywords: Fractional active disturbance rejection con- trol (FADRC) Fractional extended state observer (FESO) Fractional proportional-derivative con- troller Linear fractional order system (FOS) abstract A fractional active disturbance rejection control (FADRC) scheme is proposed to improve the performance of commensurate linear fractional order systems (FOS) and the robust analysis shows that the controller is also applicable to incommensurate linear FOS control. In FADRC, the traditional extended states observer (ESO) is generalized to a fractional order extended states observer (FESO) by using the fractional calculus, and the tracking differentiator plus nonlinear state error feedback are replaced by a fractional proportional-derivative controller. To simplify controller tuning, the linear bandwidth-parameterization method has been adopted. The impacts of the observer bandwidth ω o and controller bandwidth ω c on system performance are then analyzed. Finally, the FADRC stability and frequency-domain characteristics for linear single-input single-output FOS are analyzed. Simulation results by FADRC and ADRC on typical FOS are compared to demonstrate the superiority and effectiveness of the proposed scheme. & 2016 ISA. Published by Elsevier Ltd. All rights reserved. 1. Introduction Fractional calculus is the generalization of ordinary integer order calculus. Systems described by fractional order calculus are known as fractional order systems (FOS). Fractional calculus pro- vides a preferable method to describe complicated natural objects and dynamical processes such as electrical noises, chaotic system, and organic dielectric materials [16]. As a consequence, scientists show more and more interests in identication of FOS [7,8]. Commensurate linear FOS is a special kind of FOS, with a simple model and proportional orders [9,10]. Controllers with fractional order operator are naturally suitable for the FOS [11,12]. There are mainly four kinds of fractional order controllers, which are CRONE (Contrôle Robuste dOrder Non Entier) controller, TID (Tilt Integral Derivative) controller, frac- tional order PID controller, and fractional order lead-lag compen- sator [1319]. Considering the industrial universal controller design requirements, such as compact structure, repeatability, model independence, easy parameter turning and strong robust- ness, active disturbance rejection control (ADRC) provides an alternative paradigm for FOS control [2022]. The central objective of ADRC is to treat the internal and external uncertainties as the total disturbance and to reject them actively. Compact frame, effortless turning and sufciently good performance make ADRC popular in the world of industrial control [2325]. ADRC was rstly used to control FOS in [26], where fractional order is regarded as a part of the total disturbances, and an extended state observer (ESO) is used to estimate and reject it. Because the known or available model information is neglected and under- used, it would require higher observer bandwidth for accurate state estimation. In this paper, a distinct fractional active dis- turbance rejection control (FADRC) is proposed as a generalized and enhanced ADRC solution for the FOS. ESO is redesigned as a fractional one according to the highest fractional order of FOS. The modied fractional extended states observer (FESO) not only accurately estimates the total disturbance but also the fractional order dynamic states, leading to a reduced observer bandwidth. In addition, a fractional order PD controller is used to replace the tracking differentiator and the nonlinear state error feedback. Although FADRC is designed for commensurate linear FOS ori- ginally, the robustness analyses demonstrate that FADRC is also appropriate for incommensurate linear FOS. Simulation results show that FADRC has more inherent superiority and potential for FOS control. Due to the difculty brought by the nonlinearity and uncer- tainty, theoretical studies of ADRC are still lagging behind its industrial applications. Recent research focuses on time domain convergence, frequency response and describing function in ana- lyzing nonlinearity [2729]. Stability analysis has been sub- stantially studied for FOS in [30]. An extended root locus method by Patil [31] provides a simple way to construct root locus of general FOS and is employed for FADRC analysis and design. Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/isatrans ISA Transactions http://dx.doi.org/10.1016/j.isatra.2016.01.022 0019-0578/& 2016 ISA. Published by Elsevier Ltd. All rights reserved. n Correspondence to: School of Information Science and Technology, Beijing University of Chemical Technology, P.O. box No. 81, Beijing 100029, PR China. Tel.: þ86 10 64434930; fax: þ86 10 64437805. E-mail address: [email protected] (D. Li). ISA Transactions 62 (2016) 109119

Upload: others

Post on 06-Aug-2020

5 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

ISA Transactions 62 (2016) 109–119

Contents lists available at ScienceDirect

ISA Transactions

http://d0019-05

n CorrUniversTel.: þ8

E-m

journal homepage: www.elsevier.com/locate/isatrans

Research Article

Fractional active disturbance rejection control

Dazi Li a,n, Pan Ding a, Zhiqiang Gao b

a Institute of Automation, Beijing University of Chemical Technology, Beijing 100029, PR Chinab Center for Advanced Control Technologies, Cleveland State University, Cleveland, OH 44115, USA

a r t i c l e i n f o

Article history:Received 3 June 2015Received in revised form5 January 2016Accepted 29 January 2016Available online 28 February 2016

Keywords:Fractional active disturbance rejection con-trol (FADRC)Fractional extended state observer (FESO)Fractional proportional-derivative con-trollerLinear fractional order system (FOS)

x.doi.org/10.1016/j.isatra.2016.01.02278/& 2016 ISA. Published by Elsevier Ltd. All

espondence to: School of Information Scienity of Chemical Technology, P.O. box No. 816 10 64434930; fax: þ86 10 64437805.ail address: [email protected] (D. Li).

a b s t r a c t

A fractional active disturbance rejection control (FADRC) scheme is proposed to improve the performanceof commensurate linear fractional order systems (FOS) and the robust analysis shows that the controlleris also applicable to incommensurate linear FOS control. In FADRC, the traditional extended statesobserver (ESO) is generalized to a fractional order extended states observer (FESO) by using the fractionalcalculus, and the tracking differentiator plus nonlinear state error feedback are replaced by a fractionalproportional-derivative controller. To simplify controller tuning, the linear bandwidth-parameterizationmethod has been adopted. The impacts of the observer bandwidth ωo and controller bandwidth ωc onsystem performance are then analyzed. Finally, the FADRC stability and frequency-domain characteristicsfor linear single-input single-output FOS are analyzed. Simulation results by FADRC and ADRC on typicalFOS are compared to demonstrate the superiority and effectiveness of the proposed scheme.

& 2016 ISA. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Fractional calculus is the generalization of ordinary integerorder calculus. Systems described by fractional order calculus areknown as fractional order systems (FOS). Fractional calculus pro-vides a preferable method to describe complicated natural objectsand dynamical processes such as electrical noises, chaotic system,and organic dielectric materials [1–6]. As a consequence, scientistsshow more and more interests in identification of FOS [7,8].Commensurate linear FOS is a special kind of FOS, with a simplemodel and proportional orders [9,10].

Controllers with fractional order operator are naturally suitablefor the FOS [11,12]. There are mainly four kinds of fractional ordercontrollers, which are CRONE (Contrôle Robuste d’Order NonEntier) controller, TID (Tilt Integral Derivative) controller, frac-tional order PID controller, and fractional order lead-lag compen-sator [13–19]. Considering the industrial universal controllerdesign requirements, such as compact structure, repeatability,model independence, easy parameter turning and strong robust-ness, active disturbance rejection control (ADRC) provides analternative paradigm for FOS control [20–22]. The central objectiveof ADRC is to treat the internal and external uncertainties as thetotal disturbance and to reject them actively. Compact frame,effortless turning and sufficiently good performance make ADRC

rights reserved.

ce and Technology, Beijing, Beijing 100029, PR China.

popular in the world of industrial control [23–25]. ADRC wasfirstly used to control FOS in [26], where fractional order isregarded as a part of the total disturbances, and an extended stateobserver (ESO) is used to estimate and reject it. Because theknown or available model information is neglected and under-used, it would require higher observer bandwidth for accuratestate estimation. In this paper, a distinct fractional active dis-turbance rejection control (FADRC) is proposed as a generalizedand enhanced ADRC solution for the FOS. ESO is redesigned as afractional one according to the highest fractional order of FOS. Themodified fractional extended states observer (FESO) not onlyaccurately estimates the total disturbance but also the fractionalorder dynamic states, leading to a reduced observer bandwidth. Inaddition, a fractional order PD controller is used to replace thetracking differentiator and the nonlinear state error feedback.Although FADRC is designed for commensurate linear FOS ori-ginally, the robustness analyses demonstrate that FADRC is alsoappropriate for incommensurate linear FOS. Simulation resultsshow that FADRC has more inherent superiority and potential forFOS control.

Due to the difficulty brought by the nonlinearity and uncer-tainty, theoretical studies of ADRC are still lagging behind itsindustrial applications. Recent research focuses on time domainconvergence, frequency response and describing function in ana-lyzing nonlinearity [27–29]. Stability analysis has been sub-stantially studied for FOS in [30]. An extended root locus methodby Patil [31] provides a simple way to construct root locus ofgeneral FOS and is employed for FADRC analysis and design.

Page 2: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

D. Li et al. / ISA Transactions 62 (2016) 109–119110

The FOS is translated into its integer order counterpart and thenanalysis method of general integer system can be directly adopted.

The rest of this paper is organized as follows. In Section 2, anintroduction of commensurate linear FOS and fractional orderstate observer with full-dimensionality are presented. In Section 3,the framework of FESO and FADRC and the corresponding algo-rithm are introduced. Section 4 presents the stability andfrequency-domain characteristics of FADRC. Simulation results ofFADRC and ADRC are then compared in Section 5. Finally, con-clusions are given in Section 6.

2. Fractional order systems and fractional order state observerwith full-dimensionality

The conventional integer order single input single output(SISO) transfer function can be extended to the case of the FOS.There exist various alternative definitions of the fractional deri-vative, and the Grünwald–Letnikov (GL), Riemann–Liouville (RL)and Caputo definitions are mainly used. In all of the three defini-tions, the fractional operator acts as a non-local operator, and thatis to say fractional derivatives have a memory of the past values.

It is difficult to directly implement the fractional order operatorin time-domain for the complicated FOS by using the standarddefinitions. To solve this problem, the normative integer-orderoperators are applied to approximate fractional order operators. Alot of works and researches have been done in this area. Pichégives the discrete-time approximations of fractional order opera-tors based on numerical quadrature [32]. Freeborn proposed amethod to reduce the second-order approximation ripple error ofthe fractional order differential operator [33]. The approximationlisted in [34] is adopted in this paper, which is based on networktheory approximations, and the approximation can give desiredaccuracy over any frequency band.

2.1. Transfer function representation

A SISO linear FOS can be described as [35]:

y tð ÞþXni ¼ 0

ait0Dtϕi y tð Þ ¼

Xmj ¼ 0

bjt0Dtφj u tð Þ; ð1Þ

where u tð Þ is the input, y tð Þ is the output, ϕi 1r irnð Þ and φj0r jrmð Þ are real positive numbers, and ϕ1oϕ2⋯ oϕn;φ1oφ2⋯oφm φmoϕn

� �. Model coefficients ai 1r irnð Þ and bj 0r jð

rmÞ are constants. t0Dαt α¼φi or α¼φj

� �is a fractional order

differential operator, t0 and t denote the upper and lower limit ofthe integral interval, respectively. The Caputo's fractional deriva-tive of order α with variable t and starting point t0¼0 is defined asfollows:

0Dαt y tð Þ ¼ 1

Γ 1�γ� � Z t

0

y mþ1ð Þ τð Þt�τð Þγ dτ; ð2Þ

where Γ Zð Þ is Euler's gammafunction, and α¼mþγ; mANþ ;0oγr1. In the fractional differential Eq. (1), if the order

1/sα+

+

+

+

+

X

X

U YB

A

C

1/sα

L

B

A-LC

Fig. 1. Block diagram of fractional order state observer.

differentiations are integer multiples of a single based order: i.e.ϕi ¼ iα;φj ¼ jα, the system will be termed as commensurate orderand takes the following form:

y tð ÞþXni ¼ 0

ai0Dtiαy tð Þ ¼

Xmj ¼ 0

bj0Dtjαu tð Þ: ð3Þ

With zero initial conditions, the Laplace transform of Eq. (3)becomes

G sð Þ ¼ Y sð ÞU sð Þ ¼

Pmj ¼ 0

bjsjα

1þ Pni ¼ 1

aisiα: ð4Þ

2.2. Fractional order state observer with full-dimensionality

As is well known, a commensurate FOS admits the followingstate-space representation:

0Dαt xðtÞ ¼ AxðtÞþBuðtÞ

yðtÞ ¼ CxðtÞ ;

�ð5Þ

where matrix A;B; and C are constants. A new fractional orderstate observer with full-dimensionality for the commensuratelinear FOS is obtained by generalizing the classical Luenbergerstate observer [36], and the structure of the state observer isshown in Fig. 1. where L is the undetermined coefficient matrix,and 1=sα represents the fractional order integer operator. Theobserver error can be expressed as E¼ X� X , where X is the actualstate and X is the estimated state.

0Dαt E¼ A�LCð ÞE: ð6ÞWhen the eigenvalues of matrix (A–LC) stay in the stable

region, Eq. (6) will be asymptotically stable (refer to Section 4 fordetails).

3. The structure of the fractional active disturbance rejectioncontroller

The traditional ADRC consists of three main parts: the trackingdifferentiator (TD), the ESO, and the nonlinear state error feedback(NLSEF). TD is used to provide the transient process and its deriva-tive of the input signal. ESO is used to estimate the states plus thetotal disturbance. After the aforementioned state variables havebeen obtained, NLSEF is applied to combine them and obtain thecontrol signal. By the efforts of Gao, the bandwidth-parameterization method was proposed to linearize the ESO andPD controller without losing high precision and efficiency [37,38]. Inthis section, the structure of FADRC is presented. Compared withtraditional ADRC, ESO is replaced by a FESO while NLSEF is replacedby a linear fractional PD controller. A second-order FADRC is shownin Fig. 2, where v0 tð Þ and y tð Þ represent the setpoint and output,respectively, while u0 tð Þ denotes the output of linear fractional PDcontroller, u tð Þ is the control signal, and w tð Þ is the external dis-turbance. Particularly, z1 tð Þ, z2 tð Þ and z3 tð Þ are the outputs of FESO,and

_b is a system-dependent coefficient.

3.1. Design of fractional extended states observer

A second order linear FOS with commensurate order α isassumed as follows

Y sð ÞU sð Þ ¼

bs2αþa2sαþa1

: ð7Þ

Page 3: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

D. Li et al. / ISA Transactions 62 (2016) 109–119 111

where a2; a1; b and α 0oαo1ð Þ are constants. Its differentialequation form is

y 2αð Þ ¼ �a2y αð Þ �a1y� �þbu¼ f y αð Þ; y; t

� �þbu: ð8Þ

Particularly, let x1 ¼ y; x2 ¼ y αð Þand x3 ¼ f y αð Þ; y; t� �

, amongwhich x1; x2 represent the system states and x3 is the external

w(t)

Linear fractionalPD controller

Fractional ExtendedState Observer

Plant1b-

0 (t)u (t)u (t)y( )0v t

( )1z t( )2z t

( )3z t

Fig. 2. Configuration of a second-order FADRC.

ωo

0 1 2 3 4 50

0.20.40.60.81.01.2

t

outp

ut

outputFESO z1

ωo

ωo=

ωo=

0 1 2 3 4 50

0.20.40.60.81.01.2

t

outp

ut

outputFESO z1

0 1 2 3 4 50

0.20.40.60.81.01.2

t

outp

ut

outputFESO z1

0 1 2 3 4 50

0.20.40.60.81.01.2

t

outp

ut

outputFESO z1

Fig. 3. Tracking responses with different ωo. (a) ωo¼

state. Then, the augmented state space form of Eq. (7) can berepresented as:

x αð Þ ¼ AxþBuþEh

y¼ Cx;

(ð9Þ

where x αð Þ ¼x1 αð Þ

x2 αð Þ

x3 αð Þ

264

375; x¼

x1x2x3

264

375;A¼

0 1 00 0 10 0 0

264

375;B¼

0b0

264

375;

C ¼ 1 0 0� �

; E¼001

264

375; and h¼ f αð Þ ∙ð Þ

A linear FESO is designed to estimate the states x1, x2, and x3 in

the following.

z αð Þ ¼ Azþ_BuþL y� y� �

y¼ Cz;

(ð10Þ

=5

0 1 2 3 4 5-15

-10

-5

0

t

outp

utdisturbanceFESO z3

=50

500

1000

0 1 2 3 4 5-15

-10

-5

0

t

outp

ut

disturbanceFESO z3

0 1 2 3 4 5-15

-10

-5

0

t

outp

ut

disturbanceFESO z3

0 1 2 3 4 5-20

-10

0

10

20

t

outp

ut

disturbanceFESO z3

5, (b) ωo¼50, (c) ωo¼500, and (d) ωo¼1000.

Page 4: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

0 1 2 30

0.2

0.4

0.6

0.8

1.0

1.2

t

outp

ut

Fig. 4. Desired response and system outputs with different ωo.

0 1 2 30

0.2

0.4

0.6

0.8

1

1.2

t

outp

ut

Fig. 5. System outputs with different ωc.

Re w

Im w

sheet 1

sheet 2

sheet 3

sheet v

sheet v-1

απ

Fig. 6. The correspondence between the w-plane and s Riemann sheets.

D. Li et al. / ISA Transactions 62 (2016) 109–119112

where

z αð Þ ¼ z1 αð Þ z2 αð Þ z3 αð Þh iT

; z¼ z1 z2 z3� �T

;_B ¼ 0

_b 0

h i, and

L¼ β1 β2 β3

h iTare observer gains. z1, z2 and z3 are the outputs

of FESO: z1 is the estimation of the state x1, z2 is the estimation ofthe state x2, and z3 is used to estimate the total disturbance x3. Inaddition,

_b is the estimated value of b, and

_b � b. To simplify the

tuning process, the bandwidth-parameterization method [39] isemployed. According to w-plane mapping (mentioned in Section4), the FOS is translated into an integer order system in w-plane.By placing the poles of the translated characteristic equation λ wð Þin one location, the following is obtained

λ wð Þ ¼w3þβ1w2þβ2wþβ3 ¼ wþωoð Þ3 ð11Þ

where, the observer gains can be linearized as

β1 ¼ 3ωo

β2 ¼ 3ωo2

β3 ¼ωo3;

8><>: ð12Þ

For the integer order system, variable ωo is referred to as thebandwidth of ESO. When it comes to FOS, ωo possesses thebandwidth characteristics. In order to facilitate distinction, thevariable ωo in (11) is considered to be the w-plane bandwidth ofFESO. The main objective of FESO is to estimate the total dis-turbance in real time, and wider w-plane bandwidth will result infaster response. In practice, however, upper limitation of thebandwidth is related with the sampling ratio, and exceeding thelimit will magnify sensor noises and dynamic uncertainties. Awell-tuned ωo must therefore make a balance between rapidityand stability [26,39,40]. The following Eq. (13) is used to test theFESO single parameter ωo. Both of the two poles �0:312571:0735i stay in the stable region (refer to Section 4 for details), thatmeans the system of Eq. (13) is open-loop stable. Parametersωc¼20,

_b ¼ b¼ 1:25 and ωo is set as {5, 50, 500, 1000}. The

tracking responses are shown in Fig. 3

Y sð ÞU sð Þ ¼

1:25s1:8þ0:625s0:9þ1:25

: ð13Þ

Fig. 3 shows that estimating ability of FESO is strengthenedwith the increase of the observer w-plane bandwidth. However,outputs of FESO become unstable when the w-plane bandwidth isbeyond the toplimit. The upper limit is related to the samplingrate, and a higher sampling rate leads to a higher upper limit.

3.2. Design of fractional PD controller

Referring to Fig. 3, a well-tuned observer can track the extendedstate f y; y αð Þ;ω

� �accurately. The control law can be designed as

u¼ �z3þu0_b

; ð14Þ

to obtain a desired response, where u0 is a common linear frac-tional PD control:

u0 ¼ kpðv0�z1Þþkd v0 αð Þ �z2� �

; ð15Þ

where kp and kd are controller gains. The parameters tuning isfurther simplified using the method in [39]

kd ¼ 2ωc

kp ¼ω2c

(ð16Þ

where ωc is the w-plane bandwidth of the controller. Timederivative of the setpoint is omitted to avoid the pulse in [39].However, because the fractional order differentiation of the stepsignal is a gradual process rather than the pulse signal, v0 αð Þ cannotbe omitted in (15). Then Eq. (7) becomes a commensurate cascadefractional order integrator form as

y 2αð Þ � u0: ð17Þ

Considering Eqs. (15) and (17), the following closed-looptransfer function can be obtained, namely, the desired responseof FADRC is

Y sð ÞR sð Þ �

kdsαþkps2αþkdsαþkp

: ð18Þ

Page 5: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

D. Li et al. / ISA Transactions 62 (2016) 109–119 113

Eq. (13) is taken as an example again, with fixed parameter_b ¼ 1:25. Fig. 4 shows the desired response and the output resultswith different ωo (ωc¼20 is fixed). Fig. 5 shows the outputresponses with different controllerw-plane bandwidthωc (ωo¼200is fixed).

As Fig. 4 shows, the output results tracks closely to the desiredresponse with the increase of ωo. This is because the largerobserver w-plane bandwidth can ensures precise disturbanceestimation and then compensator can be applied to reject the totaldisturbance. Fig. 5 shows that smallωc makes the system responseslower, while large ωc makes it faster.

In a practical application, the design procedure consists of twostages. In the first stage, a linear FESO is designed and a well-tunedωo is selected to ensure accurate estimations. In the second stage,a fractional order PD controller is designed under the assumptionthat the total disturbance is well estimated, and all the existingmethods could be used in this stage for designing linear or non-linear controllers.

4. Root locus and frequency domain analysis for fractionalactive disturbance rejection control

A FOS can be generally expressed in the following form:

H sð Þ ¼ bmsmv þ⋯þb1s

1vþb0

ansnvþ⋯þa1s

1vþa0

: ð19Þ

where ak k¼ 0;⋯;nð Þ; bk k¼ 0;⋯;mð Þ are constants, and v41. TheRiemann sheets are determined by using

s¼ sj jejϕ: ð20Þ

Φr sð Þ ¼kds4αþ kpþkdβ1

� �s3αþ kpβ1þkdβ2

� �s2αþ kdβ3þkpβ2

� �sαþkpβ3

_bGp

s3αþ β1þkd� �

s2αþ β1kdþβ2þkp� �

sα� � þ kpβ1þkdβ2þβ3

� �s2αþ kpβ2þkdβ3

� �sαþkpβ3

ð24Þ

where 2kþ1ð Þπoθo 2kþ3ð Þπ; and k¼ �1;0;⋯; v�2. The Rie-mann sheet is named as the Principal Riemann Sheet whenk¼ �1: Note that only roots lying on the Principal Riemann Sheetcan determine the time-domain behavior and stability perfor-mance. These sheets are mapped to the w-plane as defined by

w¼ wj jejθ ð21Þwhere w¼ sα and α¼ 1=v. The sheets can be projected to the w-plane by

α 2kþ1ð Þπoθoα 2kþ3ð Þπ ð22ÞThe correspondence between the w-plane and s Riemann

sheets is shown in Fig. 6. With the transformation of the w-plane,the stability of FOS can be predicted by the trend of root locus. Theregion of instability �απ=2rarg wð Þrαπ=2

� �in the w plane

corresponds to the right half plane �π=2rarg sð Þrπ=2� �

in the splane. The root locus branches never enter the unstable region,

Fig. 7. Block diagram of the FA

which implies the system remains stable. Otherwise, if the rootlocus branches never enter the stable region, the system remainsunstable. If the branches move from the stable region to theunstable region (or move from the unstable region to the stableregion), then the range of grain can be determined [30,31,41,42].

4.1. Root locus of linear fractional order systems

The steps to plot the root loci of FOS are listed as follow:a) Attain the open loop transfer function of FOS;b) Transform FOS into an integer order system on the w-plane;c) Obtain the root locus of the transformed system;d) Identify the Principal Riemann Sheets and the unstable

region on the w-plane;e) Perform stability analysis from the root locus on the w-plane.In order to analyze the stability of FADRC, the FADRC time

domain configuration shown in Fig. 2 is changed into a frequencydomain block diagram form [42]. As Fig. 7 shows, V0(s) and Y(s)represent the setpoint and output, respectively; Ω denotes theexternal disturbance; Gp(s) is the transfer function of a commen-surate linear FOS.

The other transfer functions in the blocks are listed in Appendix A.By the equivalent transformation of the block diagram, the close-looptransfer function can be obtained as follows when the effect of theexternal disturbance Ω is omitted.

Φr sð Þ ¼Y sð ÞVO sð Þ ¼

Gr1GcGpþGr2Gp_bþGpGf1þGcGpGf2

: ð23Þ

and

Taking Eq. (7) as an example, the parameters of FADRC aredesigned using Eqs. (12) and (16). Then, the open loop transferfunction of Φr sð Þ can be written as

Φo sð Þ ¼ Φr sð Þ1�Φr sð Þ

: ð25Þ

and

Φo sð Þ ¼ kds4αþ kpþkdβ1

� �s3αþ kpβ1þkdβ2

� �s2αþ kdβ3þkpβ2

� �sαþkpβ3

_bGp

s3αþ β1þkd� �

s2αþ β1kdþβ2þkp� �

sα� ��kds4α� kpþkdβ1

� �s3αþβ3s2α

ð26ÞFor different kinds of objects, two categories are studied to

discuss their root loci:

4.1.1. For the commensurate fractional order systemWhen Gp sð Þ ¼ b

s2α þa2sα þa1and

_bb � 1, the open loop transfer

function can be written as

DRC in frequency domain.

Page 6: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

Φo sð Þ ¼ kds4αþ kpþkdβ1

� �s3αþ kpβ1þkdβ2

� �s2αþ kdβ3þkpβ2

� �sαþkpβ3

s5αþ β1þa2� �

s4αþ β2þβ1a2þkda2þa1� �

s3αþ β1kda2þβ2a2þkpa2þβ1a1þkda1þβ3

� �s2αþ β1kda1þβ2a1þkpa1

� �sα

ð27Þ

D. Li et al. / ISA Transactions 62 (2016) 109–119114

Fig. 8 shows the root locus when α¼0.9, a1¼1.25, a2¼0.625,_b¼b¼1.25, ωo¼100 and ωc¼20, the open loop poles and zerosare given in Table 1.

4.1.2. For the incommensurate fractional order systemWhen Gp sð Þ ¼ b

s2α þa2sαþ δ þa1and

_bb � 1, the open loop transfer

function can be written as

Φo sð Þ ¼ kds4αþ kpþkdβ1

� �s3αþ kpβ1þkdβ2

� �s2αþ kdβ3þkpβ2

� �sαþkpβ3

_bb s2αþa2sαþδþa1� �

s3αþ β1þkd� �

s2αþ β1kdþβ2þkp� �

sα� ��kds4α� kpþkdβ1

� �s3αþβ3s

2αð28Þ

Fig. 9 shows the root locus when α¼0.9, δ¼�0.3 a1¼1.25,a2¼0.625,

_b¼b¼1.25, ωo¼100 and ωc¼20, the open loop poles

and zeros are given in Table 1. Fig. 10 shows the root locuswhen α¼0.9, δ¼0.3 a1¼1.25, a2¼0.625,

_b¼b¼1.25, ωo¼100

-250 -200 -150 -100 -50 0 50-40

-30

-20

-10

0

10

20

30

40

Root Locus

Real Axis (seconds-1)

Imag

inar

y A

xis

(sec

onds

-1)

Stability Boundary

Fig. 8. Root locus when α¼0.9, a1¼1.25, a2¼0.625,_b¼b¼1.25, ωo¼100 and

ωc¼20, where the dotted line is the stability boundary.

Table 1Open loop zeros and poles for different objects.

Objects Instability region

α¼ 0:9a1 ¼ 1:25a2 ¼ 0:625

δ¼ 0 (Fig. 8) � 920πrarg wð Þr 9

20π

b¼ 1:25_b ¼ 1:25

δ¼ �0:3 (Fig. 9) � 320πrarg wð Þr 3

20π

ωo ¼ 100ωc ¼ 20

δ¼ 0:3 (Fig. 10) � 320πrarg wð Þr 3

20π

and ωc¼20, and the open loop poles and zeros are given inTable 1.

From the root loci in Figs. 8–10, it is clear that all the curves arein the stable region and have no intersections with the stabilityboundary. In the other words, the system remains stable for allgain values. Although the objects are not the standard commen-surate FOS (δa0), the proposed method is also applicable.

4.2. Frequency-domain characteristics analysis

Frequency domain characteristics are widely used to analyzeautomatic control systems in classical control theory. Bode

Zeros Poles

w1;2;3 ¼ �100w4 ¼ �10

w1 ¼ 0w2 ¼ �121:4w3;4 ¼ �89:6720:5iw5 ¼ �0:05

w1;2;3 ¼ �4:64w4;5;6 ¼ 2:32þ4:02iw7;8;9 ¼ 2:32�4:02iw10 ¼ �2:15w11;12 ¼ 1:0871:87i

w1;2;3 ¼ 0w4;5 ¼ �4:7670:17iw6 ¼ �4:40w7;8 ¼ 2:4774:14iw9;10 ¼ 2:1474:11iw11;12 ¼ 2:3773:80iw13 ¼ �0:38w14;15 ¼ 0:1870:33i

w1;2;3 ¼ �4:64w4;5;6 ¼ 2:32þ4:02iw7;8;9 ¼ 2:32�4:02iw10 ¼ �2:15w11;12 ¼ 1:0871:87i

w1;2;3 ¼ 0w4;5 ¼ 2:4374:51iw6;7 ¼ 2:8373:73iw8;9 ¼ 1:6673:88iw10;11 ¼ �4:9670:42iw12 ¼ �3:92w13 ¼ �0:38w14;15 ¼ 0:1970:32i

-12 -10 -8 -6 -4 -2 0 2 4 6-10

-5

0

5

10Root Locus

Real Axis (seconds-1)

Imag

inar

y A

xis

(sec

onds

-1)

Stability Boundary

Fig. 9. Root locus when α¼0.9, δ¼�0.3 a1¼1.25, a2¼0.625,_b¼b¼1.25, ωo¼100

and ωc¼20, where the dotted line is the stability boundary.

Page 7: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

D. Li et al. / ISA Transactions 62 (2016) 109–119 115

diagram is an important component for the frequency-domainanalysis. The closed-loop stability, rapidity and accuracy can beanalyzed based on the bode diagram. A basic feedback frame canbe obtained from the frequency-domain block diagram of Fig. 7,and the equivalent open-loop transfer function is described in Eq.

-15 -10 -5 0 5 10-15

-10

-5

0

5

10

15Root Locus

Real Axis (seconds-1)

Imag

inar

y A

xis

(sec

onds

-1)

Stability Boundary

Fig. 10. Root locus when α¼0.9, δ¼0.3 a1¼1.25, a2¼0.625, b¼1.25, ωo¼100 andωc¼20. The dotted line is the stability boundary.

Fig. 11. Bode diagrams with different parameters. (a) Bode diagram with differen

(26). Taking Eq. (13) as an example, the Bode diagram with dif-ferent ωc (ωo¼200 is fixed), is shown in Fig. 11(a) and the bodediagram with different ωo (ωc¼50 is fixed) is shown in Fig. 11(b).

Fig. 11(a) shows that the crossover frequency becomes biggerwith the increase of ωc. This means that a bigger ωc makes thesystem response more quickly. In addition, the phase stabilitymargin keeps almost unchanged, which ensures the stability of thesystem. It can be seen from Fig. 11(b) that the system almost hasalmost the same bode diagram in the low frequency interval fordifferent ωo, which means the observer w-plane bandwidth ωo

has little effect on the stability.

5. Simulation and discussion

In this section, simulation results for four different FOS areimplemented to verify the superiority and effectiveness of theproposed method. The methods in [43] and [26] are adopted todesign ADRC for FOS control (see Table 2). In Table 2, comparativeresults with method in [43] are listed in No. 1 and comparativeresults with method in [26] are listed in No. 2–5. For the sake offairness, the transition process of the setpoint generated by TD isadded in FADRC and ADRC, and the second order TD can bedesigned as [21]

t ωc (ωo¼200 is fixed). (b) Bode diagram with different ωo (ωc¼50 is fixed).

Page 8: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

Table 2Controller parameters and performance indices.

No. Objects Parameters {ωo, ωc,_b, T} ISE (10�5)

FADRC Comparative method FADRC Comparative method

1. Commensurate nonlinear FOS {30, 5, 1, 0.001} ωo¼30, kp¼10, ki¼2, kd¼1 λ¼0.8,_b¼1, T¼0.001 3000 45,000

2. Gas-turbine model (at 90% rated speed) {100, 10, 14165, 0.001} 670 10003. Gas-turbine model (at 93% rated speed) {100, 10, 8533, 0.001} 340 6604. Heat-solid model {300, 10, 0.0252, 0.0001} 35.3 37.5

{30, 10, 0.0252, 0.0001} 35.3 1205. Solid-core active magnetic bearing {4000, 200, 5600, 0.0001} 1.2 68

0 2 4 6 8 100

2

4

6

t

outp

ut

transition processFADRCGao's Method

The output y

Control signal

0 2 4 6 8 10-15

-10

-5

0

5

10

t

cont

rol s

igna

l

FADRCGao's Method

Fig. 12. Outputs comparison between FADRC and Gao's method. (a) The output y.(b) Control signal.

D. Li et al. / ISA Transactions 62 (2016) 109–119116

_v1 ¼ v2_v2 ¼ f han v1�v0; v2; r;hð Þ ;

(ð29Þ

where v0 is the setpoint, v1 is the transition process of v0, v2 is thedifferential trajectory of v1, r and h are adjustable parameters, andthe function f han ∙ð Þ is defined as:

f han x1; x2; r;hð Þ ¼ �rsign að Þ; aj j4d

rad; aj jrd ;

(ð30Þ

where a and d are given as follows:

d¼ rh;

d0 ¼ hd;

y¼ x1þhx2;

a0 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffid2þ8r y

q;

a¼x2þ a0 �dð Þ

2 sign yð Þ; y 4d0

x2þ yh; y rd0

:

8<:

8>>>>>>>>>>><>>>>>>>>>>>:

ð31Þ

All the parameters and assessment indices are listed in Table 2,where T is the sampling time; ISE is the short for integral squareerror: ISE tð Þ ¼ R t

0 y tð Þ�v1 tð Þð Þ2dt.

5.1. A commensurate nonlinear fractional order system

The following commensurate nonlinear FOS is used to comparethe performance between FADRC and the Gao's method [43].

D0:8x1ðtÞ ¼ x2ðtÞD0:8x1ðtÞ ¼ sin x2ðtÞð Þþx2ðtÞþuðtÞyðtÞ ¼ x1ðtÞ

:

8><>: ð32Þ

The reference input is v(t)¼5, and the parameters of FADRCand Gao's method are listed in Table 2. Fig. 12(a) shows the tran-sition process and closed loop output of Eq. (32), and Fig. 12(b) shows the control signals.

As Fig. 12 shows, FADRC can track the transition process withless oscillation, and FADRC's control signal is also reasonable. It canbe seen that the commensurate nonlinear FOS is well controlled byFADRC. In addition, the total variation(TV) of the control effort isused to compare the performance of different controllers. Thedefinition of TV index is given as follows:

TV ¼X1i ¼ 1

juiþ1�uij: ð33Þ

TV obtained for FADRC is 26.1 and that for Gao's method is 44.1,which shows that FADRC owns smaller control effort. This isbeneficial for practical application.

5.2. Gas turbine plant

The model of the fractional order gas turbine plant, whichconverts the fuel energy into an useful form, is given by Nataraj

[44]. The input and output of the gas turbine are fuel rate andturbine speed, respectively. For the operating regime at 90% ratedspeed, the fractional order model is

G90% sð Þ ¼ 103:97050:00734s1:6807þ0:1356s0:8421þ1

: ð34Þ

For another operating regime at 93% rated speed, the fractionalorder model is

G93% sð Þ ¼ 110:92380:0130s1:6062þ0:1818s0:7089þ1

: ð35Þ

Fig. 13 shows the output responses for the operating regime at90% and 93% of rated speed demand. The well-tuned parametersof FADRC and ADRC are listed in Table 2.

Fig. 13(a) and (c) shows FADRC has faster adjustment capability.Fig. 13(b) and (d) shows the corresponding control signals.

5.3. Heat-solid model

A heat solid model is described by the following fractionalorder transfer function [45]

G sð Þ ¼ 139:69s1:26þ0:598

: ð36Þ

Page 9: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

Output y (at 90% rated speed)

Control signal (at 90% rated speed)

Output y (at 93% rated speed)

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1

1.2

t

outp

ut

transition processFADRCADRC

0 0.5 1 1.5 20

0.004

0.008

0.012

t

cont

rol s

igna

l

FADRCADRC

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1.0

1.2

t

outp

ut

transition processFADRCADRC

Control signal (at 93% rated speed)

0 0.5 1 1.5 20

0.004

0.008

0.012

t

cont

rol s

igna

l

FADRCcontrol signal ADRC

Fig. 13. Response comparisons between FADRC and ADRC on gas-turbine model.(a) Output y (at 90% rated speed). (b) Control signal (at 90% rated speed). (c) Outputy (at 93% rated speed). (d) Control signal (at 93% rated speed).

Output y (ωo =300)

Output y (ωo =30)

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1.0

1.2

t

outp

ut

transition processFADRCADRC

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1.0

1.2

t

outp

ut

transition processFADRCADRC

Fig. 14. Response comparison between FADRC and ADRC on heat-solid model.(a) Output y (ωo¼300). (b) Output y (ωo¼30).

D. Li et al. / ISA Transactions 62 (2016) 109–119 117

The input and output are voltages, and least squares method isused to identify the unknown parameters. A FADRC is designedwith the commensurate order α¼0.63, and the total disturbancesis f FADRC ¼ �0:598=39:69y. A first-order ADRC/second-order ESO isused for comparison. The total disturbance is f ADRC ¼ y 1ð Þ �y 1:26ð Þ �0:598=39:69y. Parameters of FADRC and ADRC are listed inTable 2. Simulation results with ωo¼300 and ωo¼30 are shown inFig. 14(a) and (b), respectively.

It can be seen from Fig. 14 and Table 2 that FADRC has a betterresponse. It is also important to note that FADRC is able to respondquickly even with a relatively small bandwidth (see Fig. 14(b)),which makes the solution more practical.

5.4. Solid-core active magnetic bearing

Due to the eddy current effect, a solid-core active magneticbearing (AMB) shows some fractional order characteristics and thetraditional magnetic equivalent circuit model has significantamounts of errors when predicting the actual system. A weightedleast-squares method was derived for the general fractional modelbased on the widely studied commensurate order fractional modelin [46]. The final results show that the identified fractional modelstructure is closer to the actual system, and the result is:

G sð Þ ¼ 5594:32s2:75þ259:08s1:83�85950:3s0:79�14240336:8

: ð37Þ

Typically, the AMB model is close to the third order linear FOSwith the commensurate order α¼0.9, and a third order FADRC istherefore adopted for the control of the plant. Parameters ofFADRC and ADRC are listed in Table 2.

Fig. 15 shows that third-order FADRC has faster adjustmentcapability. Furthermore, FESO can get precise estimations whenthe observer w-plane bandwidth is increased to 1000. For ESO,however, the precise estimations are not available until theobserver bandwidth is increased to 4000. This means that FADRCdoes not require higher bandwidth or sampling rate and is moresuitable for practical engineering.

5.5. Robust analysis

A robustness assessment is carried out about the parameters ofEq. (13). The following plant (38) with uncertain parameters δ andε are considered:

y 1:8þεð Þ ¼ �58y 0:9þδð Þ�10

8y

� �þ10

8u: ð38Þ

The normalized integral square error (ISE) trajectory trackingperformance indices are used as

ISEv tð Þ ¼ R t0 y tð Þ�v1 tð Þð Þ2dt

ISEf tð Þ ¼R t0 f tð Þ�z3 tð Þð Þ2dt :

8<: ð39Þ

Page 10: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

Fig. 16. Behavior of the output ISE trajectories for different values of the uncer-tainty order δ, when ε¼0 is fixed (Red full line indicates FADRC, blue dotted lineindicates ADRC and the step is 2ζ. For interpretation of the references to color inthis figure legend, the reader is referred to the web version of this article.)

Fig. 17. Behavior of the disturbance ISE trajectories for different values of uncer-tainty order δ, when ε¼0 is fixed (Red full line indicates FADRC, blue dotted lineindicates ADRC and the step is 2ζ. For interpretation of the references to color inthis figure legend, the reader is referred to the web version of this article.)

Output y

Control signal

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1

1.2

t

outp

ut

transition processFADRCADRC

0 0.5 1 1.5 2

-3000

-2000

-1000

0

1000

t

cont

rol s

igna

l

FADRCADRC

Fig. 15. Response comparison between FADRC and ADRC on solid-core activemagnetic bearing. (a) Output y. (b) Control signal.

Fig. 19. Behavior of the disturbance ISE trajectories for different values of uncer-tainty order ε, when δ¼0 is fixed (Red full line indicates FADRC, blue dotted lineindicates ADRC and the step is ζ. For interpretation of the references to color in thisfigure legend, the reader is referred to the web version of this article.)

Fig. 18. Behavior of the output ISE trajectories for different values of the uncer-tainty order ε, when δ¼0 is fixed (Red full line indicates FADRC, blue dotted lineindicates ADRC and the step is ζ. For interpretation of the references to color in thisfigure legend, the reader is referred to the web version of this article.)

D. Li et al. / ISA Transactions 62 (2016) 109–119118

where v1 tð Þ is the transition process of the setpoint and f tð Þ is thetotal disturbance. Considering the dynamic characteristic of thesystem and the accuracy of the identification method, the varia-tion of uncertain parameter ε is set between �0.25 and 0.25. Asthe second item produces small effects on the system, and para-meter δ is set between �0.5 and 0.5. Figs. 16 and 17 show thetrajectories of ISEv;δ tð Þ and ISEf ;δ tð Þ for different uncertain para-

meters of δ in step of 2ζ, when ε¼0 is fixed and ζ ¼ffiffiffi3

p�

ffiffiffi2

p� �=6.

Fig. 17 and Fig. 18 show the trajectories ISEv;ε tð Þ and ISEf ;ε tð Þ fordifferent uncertain parameters of ε in step of ζ, when δ¼0 is fixed.

From Fig. 16 to Fig. 19, the following conclusions can beobtained. 1). For different uncertain parameters δ, Figs. 16 and 17show that FADRC owns smaller ISE value in both the output anddisturbance trajectories. This means that FADRC could track tra-jectory (setpoint or total disturbance) more quickly and precisely.The constant steady states of these loci indicate that the corre-sponding outputs are ultimately close to the desired referencetrajectories. 2). When the uncertain parameter ε changes, the

advantages of FADRC are distinct. Figs. 18 and 19 show that FADRCobtains better performance when the first item 1.8þε is changed,while FESO is affected by the modification of ε. When 1.8þε iscloser to 2, the ESO achieve a better performance. This is a goodexample to prove that FADRC is necessary for a system with aninteger-distancing fractional order. 3). The highest order of FOSplays an important role in FADRC designing for different systembecause the no-highest items are treated as the total disturbanceand rejected by the controller. 4). More importantly, FADRC couldtrack the trajectory quickly and precisely even if there exist a largechange of uncertain parameter δ and ε, which means the FADRC isalso appropriate for incommensurate linear FOS.

6. Conclusion

The traditional ADRC solution for FOS is improved in this paper,where the ESO and the NLSEF are replaced by FESO and PDα

Page 11: Fractional active disturbance rejection controlmechatronics.ucmerced.edu/sites/mechatronics.uc... · ness, active disturbance rejection control (ADRC) provides an alternative paradigm

D. Li et al. / ISA Transactions 62 (2016) 109–119 119

controller respectively. Linear bandwidth-parameterization methodis applied to simplify the parameters tuning. In addition, stabilityand frequency-domain characteristics of FADRC for FOS are alsoanalyzed. Numerical simulations show the superiority and effec-tiveness of the proposed scheme over the existing ADRC solution.Moreover, robustness analysis shows that FADRC is also appropriatefor incommensurate FOS control. Furthermore, it is believed thatfurther improvements can be obtained by employing the nonlineargains in FADRC but it would require a thorough theoretical study ofnecessity.

Acknowledgments

This research has been supported by the National Natural Sci-ence Foundation of P.R. China (Grant No. 61573052).

Appendix: A

The transfer functions of the blocks in Fig. 7 are

Gr1 sð Þ ¼ kdsαþkps2αþ β1þkd

� �sαþ β1kdþβ2þkp

� � A�1

Gr2 sð Þ ¼ kds3αþ kpþkdβ1

� �s2αþ kpβ1þkdβ2

� �sαþkpβ2

s2αþ β1þkd� �

sαþ β1kdþβ2þkp� � A�2

Gf1 sð Þ ¼ kpβ1þkdβ2

� �sαþkpβ2

s2αþ β1þkd� �

sαþ β1kdþβ2þkp� � A�3

Gf2 sð Þ ¼ s2αþkdsαþkps2αþ β1þkd

� �sαþ β1kdþβ2þkp

� � A�4

Gc ¼β3

sαA�5

References

[1] Skaar SB, Michel AN, Miller RK. Stability of viscoelastic control system. IEEETrans Autom Control 1988;33(4):348–57.

[2] Chen Y, Moore KL. Analytical stability bound for a class of delayed fractional-order dynamic systems. In: Proceedings of the 40th IEEE conference ondecision and control; 2001. p. 1421�6.

[3] Mandelbrot B, Member S. Some noise with 1/f spectrum, a bridge betweendirect current and white noise. IEEE Trans Inform Theory 1967;2(13):289–98.

[4] Tsai JS, Chien T, Guo S, Chang Y, Shieh L. State-space self-tuning control forstochastic fractional-order chaotic systems. IEEE Trans Circuits Syst I. Regularpaper. 54(3); 2007. p. 632–42.

[5] Reyes-Melo ME, Martinez-Vega JJ, Guerrero-Salazar CA, Ortiz-Mendez U.Application of fractional calculus to modelling of relaxation phenomena oforganic dielectric materials. In: Proceedings of the 2004 International con-ference on solid dielectrics; 2004. p. 530–3.

[6] Friedrich C. Relaxation and retardation functions of the Maxwell model withfractional derivatives. Rheol Acta 1991;30(2):151–8.

[7] Liu D, Laleg-Kirati T, Gibaru O, Perruquetti W. Identification of fractional ordersystems using modulating functions method. In: Proceedings of the 2013American control conference (ACC); 2013. p. 1679–85.

[8] Idiou D, Charef A, Djouambi A. Linear fractional order system identificationusing adjustable fractional order differentiator. IET Signal Process 2013;4(8):398–409.

[9] Li D, Fan W, Gao Y, Jin Q. Iterative least square identification algorithm forfractional order system. J Jiangnan Univ: Nat Sci Ed 2010;9(4):404–8.

[10] Charef A, Idiou D, Djouambi A, Voda A. Identification of linear fractional sys-tems of commensurate order. In: Proceedings of the 3rd international con-ference on systems and control; 2013. p. 259�64.

[11] Chen Y. Ubiquitous fractional order controls. In: Proceedings of the SecondIFAC symposium on fractional derivatives applications; 2006. p. 481–92.

[12] Chen Y, Petras I, Xue D. Fractional order control-A tutorial. In: Proceedings ofthe 2009 American control conference; 2009. p. 1397–411.

[13] Oustalout A, Bansard M. First generation CRONE control. In: Proceedings ofconference on system, man and cybernetic, 1993' systems engineering in theservice of humans'; 1993. p. 130-5.

[14] Oustaloup A, Bluteau B, Lanusse P. Second generation CRONE control. In:Proceedings of confernce on system, man and cybernetic, 1993' systemsengineering in the service of humans'; 1993. p. 136�42.

[15] Oustaloup A, Melchior P, Lanusse P, Cois O, Dancla F. The CRONE toolbox forMatlab. CACSD. In: Proceedings of Conference on IEEE International sympo-sium on computer-aided control system design; 2000. p. 190–5.

[16] Lurie BJ. Three-parameter tunable tilt-integral-derivative (TID) controller.Patent, US; 1994.

[17] Podlubny I. Fractional-order systems and pid controllers. IEEE Trans AutomControl 1999;1(44):208–14.

[18] Li D, Liu L, Jin Q, Hirasawa K. Maximum sensitivity based fractional IMC–PIDcontroller design for non-integer order system with time delay. J ProcessControl 2015;31:17–29.

[19] Raynaud HF, Zergainoh A. State-space representation for fractional ordercontrollers. Automatica 2000;36(7):1017–21.

[20] Han J. Auto-disturbances-rejection controller and it's application. ControlDecis 1998;13(01):19–23.

[21] Han J. Active disturbance rejection control technique. Beijing: NationalDefense Industry Press; 2008.

[22] Han J. From PID to active disturbance rejection control. IEEE Trans Ind Electron2009;56(3):900–6.

[23] Tian G, Gao Z. Benchmark tests of active disturbance rejection control on anindustrial motion control platform. In: Proceedings of the 2009 conference onAmerican control conference; 2009. p. 5552–7.

[24] Sun L, Dong J, Li D, Lee KY. A practical multivariable control approach based oninverted decoupling and decentralized active disturbance rejection controller.Ind Eng Chem Res 2016. http://dx.doi.org/10.1021/acs.iecr.5b03738.

[25] Xia Y, Dai L, Fu M, Li C, Wang C. Application of active disturbance rejectioncontrol in tank gun control system. J Frankl Inst 2014;351(4):2299–314.

[26] Li M, Li D, Wang J, Zhao C. Active disturbance rejection control for fractional-order system. ISA Trans 2013;52(3):365.

[27] Yang X, Huang Y. Capabilities of extended state observer for estimatinguncertainties. In: Proceedings of the 2009 American control conference; 2009.p. 3700–5.

[28] Tian G, Gao Z. Frequency response analysis of active disturbance rejectionbased control system. In: Proceedings of the 16th IEEE international con-ference on control applications part of IEEE multi-conference on systems andcontrol; 2007. p. 1595–9.

[29] Wu D, Ken C. Frequency-domain analysis of nonlinear active disturbancerejection control via the describing function method. IEEE Trans Ind Electron2013;60(9):3906–14.

[30] Tan N, Faruk Özgüven Ö, Mine Özyetkin M. Robust stability analysis of frac-tional order interval polynomials. ISA Trans 2009;48(2):166–72.

[31] Patil MD, Vyawahare VA, Bhole MK. A new and simple method to constructroot locus of general fractional-order systems. ISA Trans 2014;53(2):380–90.

[32] Robert P. Noise-filtering properties of numerical methods for the inverse Abeltransform. IEEE Trans Instrum Meas 1992;41(4):517–22.

[33] Freeborn TJ, Maundy B, Elwakil A. Second order approximation of the frac-tional laplacian operator for equal-ripple response. In: Proceedings of the 53rdIEEE international midwest symposium on circuits and systems (MWSCAS);2010. p. 1173–6.

[34] Hartley TT, Lorenzo CF, Killory Qammer. H. Chaos in a fractional order Chua'ssystem. IEEE Trans Circuits Syst I: Fundam Theory Appl 1995;42(8):485–90.

[35] Podlubny I. Fractional differential equations. San Diego, CA: Academic Press;1999.

[36] Wang J, Xiao H. Design of fractional order state observer with full-dimensionality. J Chongqing Univ Posts Telecommun: Nat Sci Ed 2009;21(6):795–814.

[37] Gao Z. Active disturbance rejection control: a paradigm shift in feedbackcontrol system design. In: Proceedings of the 2006 American control con-ference; 2006. p. 2399�405.

[38] Ramírez-Neria M, Sira-Ramírez H, Garrido-Moctezuma R, Luviano-Juárez A.Linear active disturbance rejection control of underactuated systems: the caseof the Furuta pendulum. ISA Trans 2014;53(4):920–8.

[39] Gao Z. Scaling and bandwidth-parameterization based controller tuning. In:Proceedings of the 2003 American control conference; 2003. p. 4989�96.

[40] Wang G, He Z. Control system design. Beijing: Tsinghua University Press;2008.

[41] Monje CA, Chen Y, Vinagre BM, Xue D, Feliu V. Fractional-order systems andcontrols. Adv Ind Control 2010.

[42] Das. S. Functional fractional calculus for system identification and controls.Berlin: Spring; 2008.

[43] Gao Z. Acitve disturbance rejection control for nonlinear fractional-ordersystems. Int J Robust Nonlinear Control 2015.

[44] Nataraj PSV, Kalla R. Computation of spectral sets for uncertain linearfractional-order systems. Commun Nonlinear Sci Numer Simul 2010;15:946–55.

[45] Petráš I, Dorcák L, Koštial I. A comparison of the integer and the fractionalorder controllers on the laboratory object. In: Proceedings of ICAMC98/ASRTP'98; 1998. p. 451–4.

[46] Zhong J, Li L. Fractional-order system identification of a solid-core activemagnetic bearing. Proc CSEE 2013;33(18):170–7.