Results in Physics 54 (2023) 107023

Contents lists available at ScienceDirect

Results in Physics
journal homepage: www.elsevier.com/locate/rinp

Complex dynamical analysis of fractional differences Willamowski–Röossler

chemical reaction model in time-scale analysis
Yu-Ming Chu a ,1 , Taher Alzahrani b ,1 , Saima Rashid c,d ,1 ,∗, Hisham Alhulayyil b ,1 ,
Waleed Rashidah b ,1 , Shafiq ur Rehman b ,1
a
School of Science, Hunan City University, Yiyang 413000, P.R. China
b Information Systems Department, College of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh 11432, Saudi
Arabia
c Department of Mathematics, Government College University, Faisalabad 38000, Pakistan
d
Department of Computer Science and Mathematics, Lebanese American University, Beirut, Lebanon

ARTICLE INFO ABSTRACT

MSC: Real mechanisms that are advancing in a complex network exhibit chaotic behaviour. This behaviour is crucial
26A51 in physical and complex systems involving numerical modelling frameworks because it essentially determines
26A33 the framework’s evolutionary process. In this context, notwithstanding its difficulty, the potential of intentional
26D07
oversight of the phenomenon has feasible effects; this is why theoretical approaches are advantageous in
26D10
such scenarios. This study investigates the functioning of a Willamowski–Rössler (W–R) mechanism, including
26D15
the synchronization of two minimal W–R structures depending on the responsive suggestion technique for
Keywords:
regulation, with the purpose of achieving chaos influence in chemical interactions. We investigate the reliability
Willamowski–Rössler system
of the steady state at various fractional order (FO) factors. Employing maximum Lyapunov exponents (MLEs),
Fractional difference equation
Chaotic attractors phase depictions, bifurcation schematics, the 0–1 evaluation and approximated entropy, it is demonstrated
Bifurcation that adjusting the FOs causes a system’s behavioural pattern to undergo a transition from steady to chaotic. In
Lyapunov exponent addition to demonstrating that the proposed scheme fits chaotically under certain circumstances, simulation
Complex systems outcomes demonstrate that mathematical modelling is used to illustrate theoretical debates. To verify that the
Community detection community detects chaos, the MLE and bifurcation illustrations, whose hallmark factors are plotted, display
erratic behaviour while effectively attempting to control the chaos.

Introduction while working on a three-dimensional climate simulation that sim-

ulated ambient-level transpiration [6–9]. In 1976, Rössler observed
Several of the unpredictable behavioural patterns identified in dif- a three-dimensional chaotic structure [10] that is logically far more
ferent exploratory response mechanisms currently have prototype ver- straightforward compared to the Lorenz mechanism. Following the
sions accessible. These are dynamic frameworks alongside only a hand- identification of the aforementioned conventional mechanisms, nu-
ful of processes and creatures that predominantly demonstrate evolving merous three-dimensional chaotic frameworks have been identified,
occurrences, including hyperventilation between multiple reliable and including the Arneodo mechanism [11], Sprott structure [12], Chen
steadfast states, dampened or sustained waves, and possibly chaos [1– framework [13], Lü–Chen mechanism [14], Cai structure [15], Tigan
3]. Massive response structures incorporating various structures accord- framework [16] and henceforth.
ing to numerous circumstances can be generated merely by combin- Chemically reactive processes have historically been an iconic topic
ing these factors restricted representations. Recently, the concept of
in nonlinear variability. The Belousov–Zhabotinsky response is the
chaos has been applied in a broad spectrum of domains, from physical
most well-known example of this dynamical model. The Belousov–
phenomena to quantum stability behaviour. To be chaotic, an ensem-
Zhabotinsky interaction is an assortment of chaotic chemical pathways
ble must possess the following characteristics: boundedness, unending
in which bromate, as the catalyst ions, undergo reduction in an alkaline
occurrence, and an acute reliance on initial conditions (ICs) [4,5].
environment by a substance that is organic, regardless of the presence
Lorenz observed the initially well-known chaotic phenomenon in 1963

∗ Corresponding author at: Department of Mathematics, Government College University, Faisalabad 38000, Pakistan.
E-mail addresses: [email protected] (Y.-M. Chu), [email protected] (T. Alzahrani), [email protected] (S. Rashid),
[email protected] (H. Alhulayyil), [email protected] (W. Rashidah), [email protected] (S.u. Rehman).
1
Contributed equally.

https://doi.org/10.1016/j.rinp.2023.107023
Received 15 June 2023; Received in revised form 12 September 2023; Accepted 24 September 2023
Available online 5 October 2023
2211-3797/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

of an intermediary. Rössler initially examined the chemical substances Mathematical model

processes and responses using a simplified chaos system [17]. Willam-
owski and Rössler seemed initially to propose that chemical interac- Here, our primary objective is to investigate certain elements of
tions can produce predictable chaos [18,19]. The examination of these the microscopic behaviour of a structure demonstrating unpredictable
chaotic algorithms of chemical interactions decreases the fascination behaviour at the macroscopic threshold. We must take into account
with regulating these erratic behaviours. Some procedures, such as hypothetical structures whose equilibrium formulas are accessible to
combustibles, benefit from erratic behaviour since turbulent gasoline an explicit procedure with an established microscopic counterpart.
distribution improves ignition speed [20–22]. In mechanisms including chemical interactions, this means all of the
measures taken into account are deemed ‘‘elementary’’ and thus adhere
Fractional calculus (FC) has proven to serve as an advantageous
to mass effect kinetics. An intriguing design presented by Geysermans
instrument for modelling mathematically an assortment of open-ended
and Nicolis [45] that meets these specifications is:
challenges in physical phenomena [23–25]. Atangana and Koca [26]
𝜅1 𝜅2
presented a model of thin viscous fluid sheet flow within the scope 1 + 𝐗1 ⇌ 2𝐗1 , 𝐗1 + 𝐗2 ⇌ 2𝐗2 ,
𝜅−1 𝜅−2
of fractional calculus. Moreover, Atangana [27] contemplated a new 𝜅3 𝜅4
fractional derivative and application to nonlinear Fisher’s reaction– 5 + 𝐗2 ⇌ 2 , 𝐗1 + 𝐗3 ⇌ 3 ,
𝜅−3 𝜅−4
diffusion equation. In recent decades, a significant number of re- 𝜅5
searchers have become interested in discrete FC [28]. Researchers 4 + 𝐗3 ⇌ 2𝐗3 . (1)
𝜅−5
are becoming increasingly worried regarding its potential functions
It consists of two autocatalytic phases, including components 𝐗1 and
in image encryption, artificial intelligence, image processing, the life
𝐗3 , followed by three additional processes (one of which serves as au-
sciences and numerous other domains. Numerous multifaceted inter-
tocatalytic) featuring 𝐗1 , 𝐗3 and another component 𝐗2 . It is apparent
actions, featuring chaos, hyperchaos and interacting attractors, were
that the beginning (1 , 4 , 5 ) and ultimate (2 , 3 ) product accumu-
just successfully identified in FO reaffirmed visualizations [29]. For lations remain constant; thus, their corresponding values influence the
instance, authors [30] examined the hyper-chaotic excitement of the maintenance of the equilibrium temperature. The rate constants 𝜅𝜄 , 𝜄 =
discrete Hénon map, while Peng et al. [31] contemplated the chaos 1, 2, … , 5 are further information that must be stipulated in order to
in the discrete memristor-based system with a FO difference, Cao establish an adequately structured state. To minimize the quantity of
et al. [32] presented the distributed coordination of networked FO optional parameters (15 in the initially proposed framework) as sig-
systems, Mehmet [33] proposed the FO sliding mode control with nificantly as feasible while maintaining support for frantic behaviour,
reaching law approach, whereas Alsharidi et al. [34] demonstrated we will be looking at a streamlined form of responses (1) where 𝜅2 =
an absence of disturbances in the discrete fractional semiconductor 𝜅3 = 𝜅4 = 0 in the following section. The final two relationships
network. In the meantime, the researchers of [35] found the fascinating are additionally analogous to assuming that 2 and 3 continue to be
chaotic behaviours of an innovative three-dimensional biophysical neu- eliminated inside the reactor. In this region, the flow formulas assume
ron model with an indestructible accordance of a stable state. Due to a perfect combination and a well-stirred reactor (see [45]).
such distinguishing features, FO iterated maps have received extensive ⎧𝐱̇𝟏 (𝐭̄ ) = 𝜗1 𝐱1 − 𝜗2 𝐱2 − 𝐱1 𝐱2 − 𝐱1 𝐱3 ,
attention in scientific communities (see [36,37]). ⎪ 1
⎨𝐱̇𝟐 (𝐭̄ ) = 𝐱1 𝐱2 − 𝜗3 𝐱2 , (2)
Interestingly, when an unpredictable dynamic mechanism crosses ⎪ ̇ ̄ 2
⎩𝐱𝟑 (𝐭 ) = 𝜗4 𝐱3 − 𝐱1 𝐱3 − 𝜌𝐱 , 3
the bifurcation in this regard, it undergoes unexpected descriptive
[ ] [ ] [ ]
transformations [38]. To conduct a more comprehensive examination where 𝜗1 = 𝜅1 1 , 𝜗2 = 𝜅−1 , 𝜗3 = 𝜅5 5 , 𝜗4 = 𝜅4 4 and 𝜌 =
of the procedure feedback that occurs in these nonlinear dynamical 𝜅−5 . Furthermore, it is predicated that 𝜅2 = 𝜅4 = 1. Every aspect of
structures, it is necessary to determine both component plane anoma- parametric information connected with framework (2) is non-negative
lies (bifurcation indications, chaos) as well as phase plane supernovae and the ICs for mechanism (2) are outlined below:
(steady states, periods, unaltered closed curves) [39–41]. At this point, ⎧𝐱1 (0) = 𝐱10 > 0,
as far as we comprehend, nobody else has attempted the dynamic ⎪
⎨𝐱2 (0) = 𝐱20 > 0, (3)
inspection of a discrete fractional W–R model with the help of a discrete ⎪
Caputo-type formulation [42,43]. This piqued our interest, prompting ⎩𝐱3 (0) = 𝐱30 > 0.
us to investigate the occurrence and behaviour of a discrete fractional The researchers suggest that the integer-order system (2) displays
W–R framework with commensurate and incommensurate FOs. chaos by contrasting the outcomes of phase profiles as well as time
Adopting the above propensity, we developed the discrete fractional series with information from the real world (see; [45]). In this in-
representation of the W–R framework, whose non-fractional system has vestigation, the Caputo fractional derivative will be introduced to
been examined in the scientific literature [44]. Taking into account the previous integer-order model in [46], yielding the FO derivative
framework as follows:
both commensurate and incommensurate FO factors, we aim to analyse
the stability of the steady states of the proposed model with varying FO ⎧𝑐 𝐃𝛶1 𝐱1 (𝐭̄ ) = 𝜗1 𝐱1 − 𝜗2 𝐱2 − 𝐱1 𝐱2 − 𝐱1 𝐱3 ,
⎪𝑐 𝛶 1
components. The FO of structures that are similarly situated to their ⎨ 𝐃 2 𝐱2 (𝐭̄ ) = 𝐱1 𝐱2 − 𝜗3 𝐱2 , (4)
classical-order algorithms has MLEs. ⎪𝑐 𝛶 3 ̄ 2
⎩ 𝐃 𝐱3 (𝐭 ) = 𝜗4 𝐱3 − 𝐱1 𝐱3 − 𝜌𝐱 , 3
The following constitutes the manuscript’s layout: Section ‘‘Math-
ematical model’’ presents a mathematical framework for the W–R where 𝛶𝜄 , 𝜄 = 1, … , 3 are the FOs such that 𝛶𝜄 ∈ (0, 1). 𝑐 𝐃𝛶𝜄 is the Caputo
formulation which is calculated using the subsequent procedure:
chemical reaction model. In Section ‘‘Preliminaries on discrete FC’’, we
𝜓
identify certain key preliminaries to discrete FC while demonstrating 1
𝑐
𝐃𝛶𝜄  (𝜓) = (𝜓 − 𝜒)−𝛶𝜄  ′ (𝜒)𝑑𝜒. (5)
the framework’s fractional discrete version. Section ‘‘Dynamic analysis 𝛤 (1 − 𝛶𝜄 ) ∫𝜇0
of W–R model’’ investigates the changing behaviour of the theoretical It is worth mentioning that a commensurate FO framework is estab-
framework and the reliability of its steady states in an environment of lished when 𝛶1 = 𝛶2 = 𝛶3 ; alternatively, an incommensurate FO
commensurate, incommensurate and synchronization analysis. Section framework develops.
‘‘0–1 test and complexity’’ employs the 0–1 analysis view and ApEn In accordance to the ICs (𝐱10 , 𝐱20 , 𝐱30 ) = (0.21, 0.01, 0.12) and the
techniques with graphical illustrations. settings specified as (𝐱10 , 𝐱20 , 𝐱30 ) = (0.22, 0.02, 0.25) and (𝐱10 , 𝐱20 , 𝐱30 ) =

2
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 1. Phase illustrations of (4) for commensurate and incommensurate FO.

∑𝐪 ( )
(50, 60, 70) the commensurate continuous FO W–R system (4) has chaotic 𝐪
= (−1)𝐪−𝜏  (𝜓 + 𝜏), 𝜓 ∈ N𝐚 . (9)
attractors when 𝛶1 = 𝛶2 = 𝛶3 = 0.976, 𝛶1 = 𝛶2 = 𝛶3 = 0.981 and the 𝜏=0
𝜏
incommensurate continuous FO W–R system (4) has chaotic attractors
Currently, we require the resulting hypothesis [47] to figure out the
when (𝛶1 , 𝛶2 , 𝛶3 ) = (1, 0.98, 1) and (𝛶1 , 𝛶2 , 𝛶3 ) = (1, 0.98, 0.95). The
stability manipulations to stay on the critical highlights of a fractional
multifaceted chaotic attractor has been demonstrated by the dynamic
portrait of the W–R framework (4) displayed in Fig. 1. discrete mechanism with commensurate FO characteristics:
( )𝐓
Preliminaries on discrete FC Theorem 1 ([47]). Assume that there is a FO ℏ(𝜓) = ℏ1 (𝜓), … , ℏ𝐪 (𝜓) ,
𝛽 ∈ (0, 1) and  ∈ R𝐪×𝐪 . the zero steady states of the commensurate
Prior to defining our fractional discrete framework, we will go over discrete FO framework
certain crucial terms and mathematical principles of discrete FC. 𝑐 𝛽
𝛥𝐚 ℏ(𝜓) = ℏ(𝜓 − 1 + 𝛽) (10)
{ {
Definition
} 1 ([42]). Assume that there is a time scale N𝐚 = 𝐚, 𝐚 + 1, 𝐚 + | |
for all 𝜓 ∈ N𝐚+1−𝛽 is asymptotically stable if ℘𝜆 ∈ 𝜔 ∈ ̃ ∶ |𝜔| ≤
2, … , 𝐚 ∈ R. The fractional sum of order 𝛽 for a mapping  can be | |
( | |
defined as | arg 𝜔|−𝜋 )𝛽 | |
}
2 2 cos | 2−𝛽| 𝛽𝜋
𝑎𝑛𝑑 | arg 𝜔| ≥ 2 , where ℘𝜆 indicates the eigenvalues
| |
1 ∑
𝜓−𝐚
𝛥−𝛽 (𝜓 − 1 − 𝜒)(𝛽−1)  (𝜒), ∀𝐚 > 0, 𝜓 ∈ N𝐚+𝛽 . of the matrix .
𝐚  (𝜓) = (6)
𝛤 (𝛽) 𝜒=0
However, the reliability and robustness concept of the dynamic FO
Definition 2 ([43]). The 𝛽-Caputo fractional difference operator is incommensurate framework is formulated as follows:
described as follows
Theorem 2 ([48]). Assume the system
𝑐 𝛽 −(𝐪−𝛽)𝛥𝐪
𝛥𝐚  (𝜓) = 𝛥𝐚  (𝜓) ( )
⎧𝑐 𝛥𝛽𝐚1 𝐱1 (𝜓) = ℏ1 𝐱(𝜓 − 1 + 𝛽1 ) ,
1 ∑
𝜓−(𝐪−𝛽)
⎪ 𝛽 ( )
= (𝜓 − 1 − 𝜒)(𝐪−𝛽−1) 𝛥𝐪  (𝜒), (7) ⎪𝑐 𝛥𝐚2 𝑥2 (𝜓) = ℏ2 𝐱(𝜓 − 1 + 𝛽2 ) , 𝜓 = 0, 1, … ,
𝛤 (𝐪 − 𝛽) 𝜒=0 ⎨ (11)
⎪⋮
where 𝜓 ∈ N𝐚+𝐪−𝛽 , 𝛽 ∉ N and 𝐪 = ⌈𝛽⌉ + 1. 𝛥𝐪  (𝜒) and (𝜓 − 1 − 𝜒)(𝐪−𝛽−1) ⎪ 𝑐 𝛽𝐧 ( )
⎩ 𝛥𝐚 𝑥𝐧 (𝜓) = ℏ𝐧 𝐱(𝜓 − 1 + 𝛽𝐧 ) ,
represents the 𝑚th integer difference operator and falling factorial
mapping, respectively, indicated as ( )𝐓
where ℏ = (ℏ1 , … , ℏ𝐧 ) ∶ R𝐧 ↦ R𝐧 and 𝐱1 (𝜓) = 𝐱1 (𝜓), … , 𝐱𝐧 (𝜓) ∈ R𝐧 .
𝛤 (𝜓 − 𝜒) Suppose that 𝛽𝜄 ∈ (0, 1), 𝜄 = 1, ̃ 𝐧 and  is the least common multiple of
(𝜓 − 1 − 𝜒)(𝐪−𝛽−1) = (8)
𝛤 (𝜓 − 𝜒 − 𝐪 + 𝛽 + 1) the denominators 𝜇𝜄 of 𝛽𝜄′ 𝑠 having 𝛽𝜄 = 𝜔𝜄 ∕𝜇𝜄 , (𝜇𝜄 , 𝜔𝜄 ) = 1, 𝜔𝜄 , 𝜇𝜄 ∈ Z+ , 𝜄 =
and 1, 2, … , 𝐧. If every root of the equation
( ) (  )
𝛥𝐪  (𝜓) = 𝛥 𝛥𝐪−1  (𝜓) 𝑑𝑒𝑡 𝑑𝑖𝑎𝑔(℘ 𝛽1 , … ., ℘𝛽𝐧 ) − (1 − ℘ )J = 0, (12)

3
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

̃ 𝜉 , then the straightforward

If 𝐱10 = 𝐱1 (0) occurs inside the collection ∕ Dynamic analysis of W-R model
solution of the structure (10) is locally asymptotically stable (LAS), where
𝜉 = 1∕, J is the Jacobian matrix of (10) and In this section, we will investigate when the formerly suggested
{ ( discrete fractional W–R model (14) is steady or in chaos in both
| arg 𝜔| )𝜉 𝜉𝜋 }
𝜉 = 𝜔 ∈ ̃ ∶ |𝜔| ≤ 2 cos 𝑎𝑛𝑑 | arg 𝜔| ≤ . (13) instances: commensurate and non-commensurate FOs, respectively. The
𝜉 2 present research will make use of an assortment of computational tech-
niques, among them MLEs computations, bifurcation illustrations and
Analysis of discrete FO W-R system the visual representation of phase profiles in multifaceted estimations.
In addition, we use the 0–1 evaluation to determine whether or not
chaos is present.
The discrete version of the FO W–R system can be constructed by
𝛶
substituting the Caputo FO formulation 𝑐 𝐃𝐚1𝜄 in framework (4) with the Commensurate FO
𝛽
FO difference scheme defined in d 𝑐 𝛥𝐚𝜄1 , which is demonstrated below:
In what follows, the robustness of the stable state points within
⎧𝑐 𝛥𝛽1 𝐱 (𝜓) = 𝜗 𝐱 (𝜓 − 1 + 𝛽 ) − 𝜗 (𝐱 (𝜓 − 1 + 𝛽 ))2 the discrete fractional W–R model (14) with commensurate FO is
⎪ 𝐚1 1 1 1 1 2 1 1
investigated. For the purpose of determining the system’s steady state,
⎪ −𝐱1 (𝜓 − 1 + 𝛽1 )𝐱2 (𝜓 − 1 + 𝛽1 )
⎪ we address the subsequent collection of formulae:
⎪ −𝐱1 (𝜓 − 1 + 𝛽1 )𝐱3 (𝜓 − 1 + 𝛽1 ),
⎨ 𝑐 𝛽2 ⎧𝜗1 𝐱1 − 𝜗2 𝐱2 − 𝐱1 𝐱2 − 𝐱1 𝐱3 = 0,
𝛥
⎪ 𝐚1 2 𝐱 (𝜓) = 𝐱1 (𝜓 − 1 + 𝛽2 )𝐱2 (𝜓 − 1 + 𝛽2 ) − 𝜗3 𝐱2 (𝜓 − 1 + 𝛽2 ), ⎪ 1
⎪𝑐 𝛽3 ⎨𝐱1 𝐱2 − 𝜗3 𝐱2 = 0, (19)
⎪ 𝛥𝐚1 𝐱3 (𝜓) = 𝜗4 𝐱3 (𝜓 − 1 + 𝛽3 ) − 𝐱1 (𝜓 − 1 + 𝛽3 ) ⎪ 2
⎪ ⎩𝜗4 𝐱3 − 𝐱1 𝐱3 − 𝜗1 𝐱3 = 0.
⎩ −𝐱1 (𝜓 − 1 + 𝛽3 )𝐱3 (𝜓 − 1 + 𝛽3 ) − 𝜌(𝐱3 (𝜓 − 1 + 𝛽3 ))2 ,
Using the set of specifications identified as 𝜌 = 5.5 and 𝜗4 = 8, we
(14) are able to determine( two types ) of steady ( states: 𝐸0) = (0, 0, 0)(a trivial
equilibrium,
) 𝐸1( = 0, 0, 𝜗4 ∕𝜌 , 𝐸2 = 𝜗1 ∕𝜗2 , 0, 0 , )𝐸3 = 𝜗3 , 𝜗1 −
for 𝜓 ∈ N𝐚+1−𝛽 . 𝛽𝜄 , 𝜄 = 1, … , 3 are the FO values such that 𝛽𝜄 ∈ (0, 1], 𝜄 =
1, 2, 3. ( 4 = 𝜗4 − 𝜗1 𝜌∕1 − 𝜌𝜗2 , 0, 𝜗1 − 𝜗2 𝜗4 ∕1)− 𝜌𝜗2 are semi-trivial
𝜗2 𝜗3 , 0 , 𝐸
and 𝐸5 = 𝜗3 , 𝜌(𝜗1 − 𝜗2 𝜗3 ) + 𝜗3 − 𝜗4 ∕𝜌, 𝜗4 − 𝜗3 ∕𝜌 be the interior steady
The numerical procedure for the fractional discrete system can be state. The preceding hypothesis offers the steady-state requirements of
constructed using the subsequent hypothesis: the equilibrium location 𝐸5 :

Theorem 3 ([49]). Suppose there is a fractional difference equation Theorem 4. Assume that the steady states of the system (14) is LAS if
{ 𝛽 ( ) 𝛽 ≤ 0.7683.
𝑐 1
𝛥𝐚 𝐟(𝜓) = ϝ 𝜓 + 𝛽𝜄 − 1, 𝐟(𝜓 + 𝛽𝜄 − 1) ,
(15)
𝛥𝜅 𝐟(𝜓) = 𝐟𝜅 , 𝐪 = ⌈𝛽𝜄 ⌉ + 1, Proof. The Jacobian matrices about the equilibria 𝐸0 , 𝐸1 , 𝐸3 , 𝐸4 and
𝐸5 are presented as follows:
the unique solution of this IVP (15) is presented as
𝜌𝜗 −𝜗
⎛𝜗1 0 0⎞ ⎛ 1𝜌 4 0 0⎞
𝜓−𝛽𝜄
∑ ( ) ⎜ ⎟ ⎜ ⎟
1 (𝛽𝜄 −1) J𝐸0 = ⎜0 −𝜗2 0 ⎟ , J𝐸1 = ⎜ 0 −𝜗3 0 ⎟,
𝐟(𝜓) = 𝐟0 (𝜓) + (𝜓 − 𝜒 + 1) ϝ 𝜓 + 𝛽𝜄 − 1, 𝐟(𝜓 + 𝛽𝜄 − 1) ,
𝛤 (𝛽𝜄 ) ⎜0 0 𝜗4 ⎟⎠ ⎜ 𝜗4 ⎟
𝜒=𝐚+𝐪−𝛽𝜄 ⎝ ⎝ −𝜌 0 𝜗4 ⎠
𝜓 ∈ N𝐚+𝐪 , ⎛𝜗1 𝜗 𝜗
⎞
− 𝜗1 − 𝜗1
(16) ⎜ 2 2 ⎟
= ⎜0 0 ⎟,
𝜗1 −𝜗2 𝜗3
J𝐸2
⎜ 𝜗2 ⎟
where ⎜0 0
𝜗4 𝜗2 −𝜗1 ⎟
⎝ 𝜗2 ⎠
∑
𝐪−1
(𝜓 − 𝐚)𝜅 𝜅 ⎛ −𝜗2 𝜗3 −𝜗3 −𝜗3 ⎞
ϝ0 (𝜓) = 𝛥 ϝ(𝐚). (17) ⎜ ⎟
𝜅=0
𝛤 (𝜅 + 1) J𝐸3 = ⎜𝜗1 − 𝜗2 𝜗3 0 0 ⎟,
⎜ 0 0 𝜗4 − 𝜗3 ⎟⎠
Therefore, the mathematical representation of the discrete fractional ⎝
W–R framework (14) is developed using the aforesaid approach: ⎛ 𝜗2 (𝜗4 −𝜗1 𝜌) 𝜗4 −𝜗1 𝜌
−
𝜗4 −𝜗1 𝜌
⎞
⎜ 𝜌𝜗2 −1 𝜌𝜗2 −1 𝜗2 𝜌−1
⎟
= ⎜ ⎟.
𝜗4 −𝜗1 𝜌
⎧ ∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽1 )
𝐧−1 J𝐸4 0 0 (20)
⎪𝐱 (𝐧) = 𝐱 (0) + 1 ⎜ 1−𝜗2 𝜌 ⎟
⎪ 1 1 𝛤 (𝛽1 ) 𝛤 (𝐧 + 1 − 𝜆 − 1) ⎜ 𝜗1 −𝜗4 𝜗2 0
𝜌(𝜗1 −𝜗2 𝜗4 ) ⎟
⎪ ( 𝜆=0 ) ⎝ 𝜗2 𝜌−1 𝜌𝜗2 −1 ⎠
⎪ × 𝜗1 𝐱1 (𝜆) − 𝜗2 (𝐱1 (𝜆))2 − 𝐱1 (𝜆)𝐱2 (𝜆) − 𝐱1 (𝜆)𝐱3 (𝜆) , Furthermore, it is obvious that 𝜗1 , − 𝜗2 and 𝜗4 are eigenvalues of
⎪
⎪ 1
∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽2 ) (
𝐧−1 ) J(𝐸0 ), implying that insignificant equilibrium is unsteady. Analogously,
⎨𝐱2 (𝐧) = 𝐱2 (0) + 𝛤 (𝛽 ) 𝐱1 (𝜆)𝐱2 (𝜆) − 𝜗3 𝐱2 (𝜆) , 𝜗1 𝜌−𝜗4
, − 𝜗3 and −𝜗4 are eigenvalues of J(𝐸2 ). This demonstrates that
⎪ 2
𝜆=0
𝛤 (𝐧 + 1 − 𝜆 − 1) 𝜌
⎪ 𝐸2 is an accumulation if 𝜗1 𝜌 < 𝜗4 and unsteady if 𝜌𝜗1 > 𝜗4 . However,
∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽3 )
𝐧−1
𝜗 −𝜗 𝜗 𝜗 𝜗 −𝜗
⎪𝐱 (𝐧) = 𝐱 (0) + 1 −𝜗1 , 1 𝜗 2 3 , − 2 𝜗4 1 are eigenvalues Jacobian J(𝐸3 ) and as a result,
⎪ 3 3 𝛤 (𝛽3 ) 𝛤 (𝐧 + 1 − 𝜆 − 1) 2 2
𝐸2 is a collapse if 𝜗2 𝜗4 < 𝜗1 < 𝜗2 𝜗3 . Furthermore, 𝐸3 is unsteady if
⎪ ( 𝜆=0
⎪ × 𝜗4 𝐱3 (𝜆) − 𝐱1 (𝜆) − 𝐱1 (𝜆)𝐱3 (𝜆) − 𝜌(𝐱3 (𝜆))2 , 𝜗1 > 𝜗2 or 𝜗2 𝜗4 < 𝜗1 . In addition, the characteristic polynomial for
⎩
J(𝐸4 ) is as outlined below:
(18)
𝐷(℘) = (℘ − 𝜗4 + 𝜗3 )(℘2 + 𝜗2 𝜗3 ℘ + 𝜗1 𝜗3 − 𝜗2 𝜗23 ), (21)
where 𝐱1 (0), 𝐱2 (0) and 𝐱3 (0) indicate ICs. This is an entirely novel kind
demonstrating that 𝐸4 is an accumulation if 𝜗4 < 𝜗3 and unsteady if
of chemical reaction system that has ‘‘memory impacts’’. As shown in 3, 𝜗4 > 𝜗3 . Likewise, the characteristic polynomial for J(𝐸5 ) is as outlined
the configurations of 𝐱1 (𝐧), 𝐱2 (𝐧) and 𝐱3 (𝐧) are interdependent concern- below:
ing every previous factor 𝐱1 (0), 𝐱1 (1), … , 𝐱1 (𝐧−1), 𝐱2 (0), 𝐱2 (1), … , 𝐱2 (𝐧− ( (℘ + 𝜗 )(1 − 𝜌𝜗 ) + 𝜗 𝜌 − 𝜗 )
3 2 1 4
1) and 𝐱3 (0), 𝐱3 (1), … , 𝐱3 (𝐧 − 1). 𝐷1 (℘) =
1 − 𝜌𝜗2

4
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 2. Time-fractional illustrations of the W–R model configurations (14) for 𝛽 = 0.76.

Fig. 3. Time-fractional illustrations of the W–R model configurations (14) for 𝛽 = 0.76.

5
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 4. Bifurcation plots of the commensurate FO W–R system (14) via (a-c) 𝜌 having 𝛽 = 0.76 and (d) MLE exponent.

( ( 𝜗 𝜌 + 𝜗 𝜗 − 𝜗 𝜌(𝜗 + 𝜗 ) )
1 2 4 2 1 4
× ℘2 − eigenvalues ℘2 and ℘3 into the requirement mentioned in Theorem 1,
( )𝛽
𝜌𝜗2 − 1 | arg ℘|−𝜋
we have: |℘| ≤ 2 cos 2−𝛽
(𝜗 − 𝜗2 𝜗4 )(𝜗4 − 𝜗1 𝜌) )
. Thus, we attain 𝛽 ≤ 0.7683. So that
× ℘+ 1 . (22) the steady state 𝐸5 is asymptotically stable if 𝛽 ≤ 0.7683.
𝜗2 𝜌 − 1
Take into account 𝛽 = 0.76 in order to confirm ( this significant
)
Thus, 𝐸5 is a sink if 𝜗4 − 𝜗1 𝜌 < 𝜗3 1 − 𝜗2 𝜌 and it is unsteady if 𝜗4 − 𝜗1 𝜌 > | arg ℘𝜆 |−𝜋
result. We may infer concerning this that |℘𝜆 | ≤ 2 cos 2−𝛽
and
𝜗3 − 𝜗2 𝜌.
| arg ℘𝜆 | ≥ 𝛽𝜋 , ∀ 𝜆 = 1, 2, 3. According to Theorem 1, the steady state
Then, we make the assumption that 𝜗4 > 𝜗3 and 𝜗4 +𝜌𝜗2 𝜗3 < 𝜌𝜗1 +𝜗3 2
are accurate. The Jacobian matrix of the structure (14) concerning 𝐸5 of the discrete W–R system having commensurate order (14) is
unique non-negative steady states 𝐸5 is presented by: asymptotically stable. Fig. 2 illustrates the steady state situation for the
framework under the ICs (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = (40, 50, 60).Furthermore,
⎛ −𝜗2 𝜗3 −𝜗3 −𝜗3 ⎞ suppose that 𝛽 ) = 0.845. Therefore, we can draw the conclusion |℘𝜆 | ≥
⎜ 𝜌(𝜗 −𝜗 𝜗 )+(𝜗 −𝜗 ) ⎟ (
| arg ℘𝜆 |−𝜋
=⎜ 0 ⎟. 2 cos , ∀ 𝜆 = 2, 3. In view of Theorem 1, the steady states 𝐸5
1 2 3 3 4
J𝐸5 0 (23) 2−𝛽
⎜ 𝜌
⎟ of the discrete fractional W–R system having commensurate order (14)
𝜗3 −𝜗4
⎜ 0 𝜗3 − 𝜗4 ⎟
⎝ 𝜌 ⎠ is unsteady; thus, this framework could represent a frantic attractor.
A straightforward calculation produces the subsequent characteristic Fig. 3 demonstrates the time progression of the configurations liable to
polynomial for the above system 𝐷1 (℘) ∶ the ICs (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = (40, 50, 60). The framework is undoubtedly
( ) unsteady, with frantic attractors. □
𝐷2 (℘) = ℘3 + 𝜗3 (𝜗2 − 1) + 𝜗4
( 𝜗 (2𝜗 − 2𝜗 + 𝜌𝜗 + 𝜌𝜗 (𝜗 − 2𝜗 )) )
× ℘2 + ℘ 3 3 4 1 2 4 3 Bifurcation and Lyapunov exponents (LEs)
𝜌
𝜗 (𝜗 − 𝜗4 )(𝜗4 − 𝜗1 𝜌 + 𝜗3 (𝜗2 𝜌 − 1)) To investigate the fluctuations of the commensurate FO W–R system
+ 3 3 . (24)
𝜌 described in (14) in terms of FO, commence with the ICs (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) =
In view of the Routh–Hurwitz principal [50], non-negative steady-states (40, 50, 60) and framework settings as shown in ‘‘Mathematical model’’.
of (14) is asymptotically stable if ℘1 = −3.2000, ℘2 = 0.086132 + Fig. 4(a-b) depicts the bifurcation illustration of the discrete FO frame-
0.543714𝑖 and ℘3 = 0.086132−0.543714𝑖. As a result of incorporating the work (14) where 𝜌 fluctuates in the range [2, 3] along with multiple FO

6
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 5. Phase illustrations of (14) for commensurate FOs.

values 𝛽. When 𝛽 = 0.76, the framework is steady at 2 < 𝜌 < 2.53182, where
however, as 𝜌 increases, the configuration 𝐱1 (𝐧) of the mathematical ⎧
⎪ ∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽1 )
𝐧−1
framework (14) displays erratic motion when coupled with recurring 𝜍 (𝐧) = 𝜍 (0) + 1
⎪ 𝜄 𝜄 𝛤 (𝛽1 ) 𝛤 (𝐧 − 𝜆)
views in the range 𝜌 ∈ [2.53182, 2.9301] and when 𝜌 ∈ [2.9301, 2.9786], ⎪ 𝜆=0
( )
the structure (14) becomes erratic. As 𝛽 declines, the structure becomes ⎪ × 𝛿1 𝐱1 (𝜆) − 2𝛿2 𝐱1 (𝜆)𝜍𝜄 (𝜆) − 𝐱1 (𝜆)𝛶𝜄 (𝜆) − 𝐱1 (𝜆)𝜑𝜄 (𝜆) ,
⎪
less unstable and the erratic zone shifts to the their freedom. For ⎪ 1
∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽2 ) (
𝐧−1
)
instance, for 𝛽 = 0.73, the structure is steady between 𝜌 ∈ [2, 2.53182] ⎨𝛾𝜄 (𝐧) = 𝛾𝜄 (0) + 𝛤 (𝛽 ) 𝜍𝜄 (𝜆)𝐱2 (𝜆) − 𝛿3 𝐱2 (𝜆) , (26)
⎪ 2
𝜆=0
𝛤 (𝐧 − 𝜆)
as well as exhibits a chaotic domain between 𝜌 ∈ [2.008, 2.12143]. ⎪ ∑ 𝛤 (𝐧 − 𝜆 − 1 + 𝛽3 )
𝐧−1
Fig. 4(c) illustrates the bifurcation illustration of the commensurate ⎪𝜑 (𝐧) = 𝜑 (0) + 1
⎪ 𝜄 𝜄 𝛤 (𝛽3 ) 𝛤 (𝐧 − 𝜆)
discrete fractional chemical model (14) for ICs (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = ⎪ ( 𝜆=0
)
(40, 50, 60) while maintaining the identical parameter settings as the ⎪ × 𝛿 𝐱
4 3 (𝜆) − 𝐱3 (𝜆)𝜍 1 (𝜆) − 2𝛿5 𝐱3 (𝜆)𝜑𝜄 (𝜆) , 𝜄 = 1, … , 3.
⎩
model described in ‘‘Mathematical model’’ where 𝛽 = 2.010. Precisely
Therefore, the 𝐿𝐸𝑠 be calculated as follows:
will be viewed, the framework demonstrates steady behaviour initially,
however as the FO expands, the mechanism forfeits strength, exhibiting ln |℘(𝜄)
𝜅 |
℘𝜅 (𝑢10 ) = lim , 𝑓 𝑜𝑟 𝜅 = 1, … , 3, (27)
regular movement prior to transforming into erratic in each period 𝜄↦∞ 𝜄
𝛽 ∈ [2, 2.9999]. When we continue to boost the FO, the mathematical ℘𝜅 indicates the eigenvalues of (26). We employed an application in
structure will eventually reach endlessly. MATLAB for determining the maximum 𝐿𝐸𝑠 of the discrete fractional
The 𝐿𝐸𝑠 are a crucial mechanism for demonstrating chaos in dis- W–R system (14), and the outcomes are displayed in Fig. 4c via the ICs
(𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = (40, 50, 60) and the identical framework settings as
crete fractional systems when combined with a bifurcation schematic.
listed in ‘‘Mathematical model’’. The framework has negative 𝐿𝐸𝑠 at
The Jacobian matrix approach is implemented to figure out the 𝑀𝐿𝐸𝑠
the lowest levels of 𝛽 and non-negative 𝐿𝐸𝑠 as expands, indicating that
(see; [51]), that are determined in an identical behaviour as the con-
the W–R system (14) adjustments regarding an equilibrium to a chaotic
figurations in the discrete mechanism (14). The tangent map 𝐉𝜄 can be state, resulting in affirms the findings acquired from the bifurcation
described as follows: representation depicted in Fig. 4(d). To help clarify the aforementioned
⎛ 𝜍1 𝜍2 𝜍3 ⎞ results, Fig. 5(a-d) depicts phase schematics of the discrete fractional
𝐉𝜄 = ⎜ 𝛾1 𝛾2 𝛾3 ⎟ , (25) W–R designs (14) in the three-dimensional planes corresponding to
⎜ ⎟ different FOs.
⎝𝜑1 𝜑2 𝜑3 ⎠

7
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 6. ((a)-(c)) Bifurcation illustration of the incommensurate FO W–R framework (14) via 𝛽1 for 𝛽1 = 0.7 and 𝛽3 = 0.8 (b)-(d) MLEs.

Incommensurate FO with incommensurate orders, we examine the consistency of steady

states 𝐸6 using the requirement for stability mentioned in Theorem 2.
The current study investigates the behaviour of the incommensurate Assume (𝛽1 , 𝛽2 , 𝛽3 ) = (0.6, 0.7, 0.8) and implement the specifications 𝜌.
discrete fractional W–R framework (14) in an approach identical to that Take note that the LCM  = 10 and thereafter
of commensurate orders. Specifically, we analyse the implications of ( )
det 𝑑𝑖𝑎𝑔(℘6 , ℘6 , ℘7 ) − (1 − ℘10 )J(𝐸5 ) = 0,
incommensurate fractional factors on the fluctuating practises of the
discrete W–R model (14). The bifurcation representation and 𝑀𝐿𝐸 of ⇔
the incommensurate discrete fractional W–R system (14) are shown 0.7712 + 0.5678℘6 − 0.78543℘7 − 1.2345℘10 + 0.1234℘12
in Fig. 6(a-b), where 𝛽1 is the bifurcation element and 𝛽2 = 0.7 and
+ 0.5443℘13 + 2.5643℘16 + 2.3109℘17
𝛽3 = 0.8. The ICs (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = (40, 50, 60) and the setting values
specified in ‘‘Mathematical model’’ where 𝜌 = 2.010 are used to generate + 0.00002℘19 + 2.4590℘20 − 1.8976℘22 − 1.7883℘23
this schematic. We are determine to observe that the discrete model + 0.9845℘26 − 0.8965℘27 − 0.7712℘30 = 0. (28)
is steady when 𝛽2 = 0.6564. When 𝛽2 expands, the maximum 𝐿𝐸 ′ 𝑠
After simplification, it is worth mentioning that for 𝜆 = 26, 27, 30,
measures shift from non-negative to negative, implying chaos emerges ( ( ))1∕10
alongside the manifestation of particular regular orbital motions. Ad- we attain |℘𝜆 | > 2 cos 10| arg ℘𝜆 | , whereas the entire system’s
𝜋
ditionally, when 𝛽2 ∈ [0.7641, 0.8023], the 𝑀𝐿𝐸 is non-negative, the additional ℘𝜆 techniques (28) fulfil | arg ℘𝜆 | ≥ 20 . This means that
structure ends up in chaos. Fig. 6(c-d) also demonstrates the bifurcation ℘𝜆 ∈ ∕ ̃ 1∕10 , (𝜆 = 1, ̄30). As a result, according to Theorem 2,
illustration and its associated 𝑀𝐿𝐸 for 𝛽1 = 0.7 and 𝛽2 = 0.8 for the discrete W–R system having incommensurate FO (14) has a lo-
analysing the fluctuating behaviours of the discrete fractional W–R sys- cally asymptotically stable steady point 𝐸5 . Fig. 8 demonstrates the
tem (14) when the FO 𝛽3 is varied. In the fractional discrete W–R system stability scenario of the incommensurate mechanism under the ICs
(14), boosting 𝛽3 effectively leads to a transition from steady to erratic (𝐱1 (0), 𝐱2 (0), 𝐱3 (0)) = (40, 50, 60). Take (𝛽1 , 𝛽2 , 𝛽3 ) = (0.7, 0.8, 0.8) verify
behaviour. More specifically, the framework is steady when 𝛽3 < 0.632, that the incommensurate discrete fractional W–R system (14) is capable
where the 𝑀𝐿𝐸 is negative, and chaotic when 𝛽3 ∈ ]0.743, 0.865], of creating erratic attractors. Evidently,  = 𝐿𝐶𝑀(10, 50, 50) = 50 and
where the 𝑀𝐿𝐸 is non-negative. The results presented here show that thus:
incommensurate orders have an effect on the fluctuating behavioural ( )
det 𝑑𝑖𝑎𝑔(℘30 , ℘37 , ℘35 ) − (1 − ℘50 )J(𝐸5 ) = 0,
patterns of the W–R system (14). The respective phase attractants of
the discrete fractional W–R system techniques using incommensurate which produces
orders (14) are depicted in Fig. 7 for comprehensiveness.
For the purpose of to gain additional insight into the various multi- 0.7712 + 1.4557℘35 − 0.78543℘7 − 1.2345℘10 + 0.1234℘12
faceted evolving practises of the discrete fractional W–R system (14) + 0.5443℘13 + 2.5643℘16 + 2.3109℘17

8
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 7. Phase illustrations of (14) for incommensurate FOs.

+ 0.00002℘19 + 2.4590℘20 − 1.8976℘22 − 1.7883℘23 The elements of the sliding technique to stay within the erroneous
26 27 30
+ 0.9845℘ − 0.8965℘ − 0.7712℘ = 0. (29) parameters are determined by the reaching laws as follows:

Consequently, by addressing the structure (29), it is possible to deter- ( ) ̄ ̄

mine that the system has outcomes ℘𝜆 = (−0.0373934 ± 1.1547228041𝑖) ⎧𝐫1 = 𝑐 𝐃𝛶1 + 𝜁1 ∫0𝐭 𝜖1 (𝜒)𝑑𝜒 = 𝜖1 + 𝜁1 ∫0𝐭 𝜖1 (𝜒)𝑑𝜒,
( ( ))1∕50 ⎪ (𝑐 𝛶 ) 𝐭̄ 𝐭̄
which indicates | arg ℘𝜆 | ≤ 2 cos 50| arg ℘𝜆 | 𝜋
and | arg ℘𝜆 | ≤ 200 . ⎨𝐫2 = 𝐃 2 + 𝜁2 ∫0 𝜖2 (𝜒)𝑑𝜒 = 𝜖2 + 𝜁2 ∫0 𝜖2 (𝜒)𝑑𝜒, (32)
⎪ (𝑐 ) 𝐭̄ 𝐭̄
As a result, based on Theorem 2, the steady state 𝐸5 = (1.3, 1.1971, ⎩𝐫3 = 𝐃𝛶3 + 𝜁3 ∫0 𝜖3 (𝜒)𝑑𝜒 = 𝜖3 + 𝜁3 ∫0 𝜖3 (𝜒)𝑑𝜒.
0.6129) is unstable. Consequently, the discrete fractional W–R system
(14) may exhibit chaos attractor, as evidenced by the trajectory of time The fractional version of (32) can be presented as:
formation displayed in Fig. 9.
⎧𝑐 𝐃𝛶1 𝐫1 = 𝑐 𝐃𝛶1 𝜖1 + 𝜁1 𝜖1 ,
Synchronization ⎪𝑐 𝛶 𝑐
⎨ 𝐃 2 𝐫2 = 𝐃𝛶2 𝜖2 + 𝜁2 𝜖2 , (33)
⎪𝑐 𝛶3 𝑐 𝛶
⎩ 𝐃 𝐫3 = 𝐃 3 𝜖3 + 𝜁3 𝜖3 .
To synchronize the FO W–R systems, we suggest a FO sliding
technique in this subsection. Let us consider the FO system as follows:
In view of the concept defined by [52]:
⎧𝑐 𝐃𝛶1 𝐱1 (𝐭̄ ) = 𝜗1 𝐱1 − 𝜗2 𝐱2 − 𝐱1 𝐱2 − 𝐱1 𝐱3 ,
⎪𝑐 𝛶 1
⎨ 𝐃 2 𝐱2 (𝐭̄ ) = 𝐱1 𝐱2 − 𝜗3 𝐱2 , (30) ⎧𝑐 𝐃𝛶1 𝜖1 = −𝜚1 𝑠𝑔𝑛(𝐫1 ) − 𝛯1 𝐫1 ,
⎪𝑐 𝛶3 ̄ 2
⎪𝑐 𝛶
⎩ 𝐃 𝐱3 (𝐭 ) = 𝜗4 𝐱3 − 𝐱1 𝐱3 − 𝜗1 𝐱 . ⎨ 𝐃 2 𝜖2 = −𝜚2 𝑠𝑔𝑛(𝐫2 ) − 𝛯2 𝐫2 , (34)
3
⎪𝑐 𝛶3
The FO framework using the sliding technique regulator is described as ⎩ 𝐃 𝜖3 = −𝜚3 𝑠𝑔𝑛(𝐫3 ) − 𝛯3 𝐫3 .
follows:
From (33) and (34), we have
⎧𝑐 𝐃𝛶1 𝑣1 (𝐭̄ ) = 𝜗1 𝑣1 − 𝜗2 𝑣2 − 𝑣1 𝑣2 − 𝑣1 𝑣3 + 𝑢1 ,
⎪𝑐 𝛶 1
⎨ 𝐃 2 𝑣2 (𝐭̄ ) = 𝑣1 𝑣2 − 𝜗3 𝑦2 + 𝑢2 , (31) ⎧𝑐 𝐃𝛶1 𝜖1 + 𝜁1 𝜖1 = −𝜚1 𝑠𝑔𝑛(𝐫1 ) − 𝛯1 𝐫1 ,
⎪𝑐 𝛶3 ̄ 2 ⎪𝑐 𝛶
⎩ 𝐃 𝑣3 (𝐭 ) = 𝜗4 𝑣3 − 𝑣1 𝑣3 − 𝜗1 𝑣3 + 𝑢3 , (35)
⎨ 𝐃 2 𝜖2 + 𝜁2 𝜖2 = −𝜚2 𝑠𝑔𝑛(𝐫2 ) − 𝛯2 𝐫2 ,
where 𝑢1 , 𝑢2 and 𝑢3 are the synchronization regulators structure. ⎪𝑐 𝛶3
⎩ 𝐃 𝜖3 + 𝜁3 𝜖3 = −𝜚3 𝑠𝑔𝑛(𝐫3 ) − 𝛯3 𝐫3 .

9
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 8. Time-fractional illustration of the W–R system (14) for (𝛽1 , 𝛽2 , 𝛽3 ) = (0.6, 0.7, 0.8).

As determined from the error behaviour as 𝜖𝜄 = 𝐯𝜄 − 𝐱𝜄 , 𝜄 = 1, 2, 3, the The sloping technique regulator’s Lyapunov candidacy is described as
fractional error interactions are as follows: follows:
⎧𝑐 𝐃𝛶1 𝜖 (𝐭̄ ) = 𝜗 (𝑣 − 𝐱 ) − 𝜗 (𝑣2 − 𝐱2 ) 1 2
⎪ 1 1 1 1 2 1 1  (𝐫1 , 𝐫2 , 𝐫3 ) = (𝐫 + 𝐫22 + 𝐫32 ). (39)
2 1
⎪ − (𝑣1 𝑣2 − 𝐱1 𝐱2 ) − (𝑣1 𝑣3 − 𝐱1 𝐱3 ) + 𝑢1 ,
⎨𝑐 𝛶2 ̄ (36) After performing the first order differentiation, then considering the FC
⎪ 𝐃 𝜖2 (𝐭 ) = (𝑣1 𝑣2 − 𝐱1 𝐱2 ) − 𝜗3 (𝑣2 − 𝐱2 ) + 𝑢2 ,
⎪𝑐 𝐃𝜐3 (𝐭̄ ) = 𝜗 (𝑣 − 𝐱 ) − (𝑣 𝑣 − 𝐱 𝐱 ) − 𝜗 (𝑣2 − 𝐱2 ) + 𝑢 . derivative scheme, we have
⎩ 4 3 3 1 3 1 3 1 3 3 3
̇ = 𝐫1 𝐃1−𝛶1 .𝐃𝛶1 𝐫1 + 𝐫2 𝐃1−𝛶2 .𝐃𝛶2 𝐫2 + 𝐫3 𝐃1−𝛶3 .𝐃𝛶3 𝐫3 . (40)
Comparing (35) and (36), we have
It can be challenging to figure out the sign of the Lyapunov first
⎧𝜗 (𝑣 − 𝐱 ) − 𝜗 (𝑣2 − 𝐱2 ) − (𝑣 𝑣 − 𝐱 𝐱 ) − (𝑣 𝑣 − 𝐱 𝐱 ) + 𝑢 + 𝜁 𝜖
⎪ 1 1 1 2 1 1 1 2 1 2 1 3 1 3 1 1 1 derivative from (40). As a result, we suggest a novel approach to
⎪ = −𝜚1 𝑠𝑔𝑛(𝐫1 ) − 𝛯1 𝐫1 , address this challenge. To overcome a problem, use the Lyapunov
⎪
⎨(𝑣1 𝑣2 − 𝐱1 𝐱2 ) − 𝜗3 (𝑣2 − 𝐱2 ) + 𝑢2 + 𝜁2 𝜖2 = −𝜚2 𝑠𝑔𝑛(𝐫2 ) − 𝛯2 𝐫2 , stability approach, we have
⎪ 2 2
⎪𝜗4 (𝑣3 − 𝐱3 ) − (𝑣1 𝑣3 − 𝐱1 𝐱3 ) − 𝜗1 (𝑣3 − 𝐱3 ) + 𝑢3 + 𝜁3 𝜖3 ( ) ( )
⎪ = −𝜚 𝑠𝑔𝑛(𝐫 ) − 𝛯 𝐫 . ̇ ≤ 𝐫1 −𝜚1 𝑠𝑔𝑛(𝐫1 ) − 𝛯1 𝐫1 + 𝐫2 −𝜚2 𝑠𝑔𝑛(𝐫2 ) − 𝛯2 𝐫2
⎩ 3 3 3 3 ( )
+ 𝐫3 −𝜚3 𝑠𝑔𝑛(𝐫3 ) − 𝛯3 𝐫3
(37)
≤ −(𝜁1 |𝐫1 | + 𝜁2 |𝐫2 | + 𝜁3 |𝐫3 |) − (𝛯1 𝐫12 + 𝛯2 𝐫22 + 𝛯3 𝐫32 ). (41)
To identify the regulators to stay in the sliding technique regulation
As a result, (41) is a negative function that deduces that the framework
scheme to synchronize the FO systems,
is steady and exists for all bounded ICs where 𝛯1 , 𝛯2 and 𝛯3 are
⎧𝑢 = −(𝜗 + 𝜁 )𝜖 + 𝜗 (𝑣2 − 𝐱2 ) + (𝑣 𝑣 − 𝐱 𝐱 ) non-negative fixed numbers.
⎪ 1 1 1 1 2 1 1 1 2 1 2
We characterize two separate ICs for FO W–R systems (4) in the
⎪ + (𝑣1 𝑣3 − 𝐱1 𝐱3 ) − 𝜚1 𝑠𝑔𝑛(𝐫1 ) − 𝛯1 𝐫1 ,
⎨ synchronization section. The ICs of the configurations of the FO system
⎪𝑢2 = −(𝑣1 𝑣2 − 𝐱1 𝐱2 ) + (𝜗3 − 𝜁2 )𝜖2 − 𝜁2 𝜖2 − 𝜚2 𝑠𝑔𝑛(𝐫2 ) − 𝛯2 𝐫2 ,
⎪𝑢 = −(𝜗 + 𝜁 )𝜖 + (𝑣 𝑣 − 𝐱 𝐱 ) + 𝜗 (𝑣2 − 𝐱2 ) − 𝜚 𝑠𝑔𝑛(𝐫 ) − 𝛯 𝐫 . are (𝐱1 , 𝐱2 , 𝐱3 ) = (40, 50, 60) and (50, 60, 70), whereas the ICs of the FO
⎩ 3 4 3 3 1 3 1 3 1 3 3 3 3 3 3
W–R systems (4) are 𝑣1 = 1, 𝑣2 = 3 and 𝑣3 = 1. The synchronization
(38) oversights with 𝛯1 = 𝛯2 = 𝛯3 = 9 are shown in Fig. 10.

10
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 9. Time-fractional illustrations of W–R model (14) for (𝛽1 , 𝛽2 , 𝛽3 ) = (0.75, 0.7, 0.85).

0–1 test and complexity the mechanism interactions are frantic. The mean square displacement
is defined by the following procedure:
In this section, we will address the significant impact of the 0–1 test  {
1 ∑ ( )2 ( )2 }
and the complexity of the behavioural pattern of the W–R model (14). 𝜉 (𝐪) = lim 𝜆𝜉 (𝚥+𝐪)−𝝀𝜉 (𝚥) + 𝐰𝜉 (𝚥+𝐪)−𝐰𝜉 (𝚥) ,  ≥ 10𝐪.
 ↦∞  𝚥=1

0–1 text (43)

In view of the asymptotic rate of progression 𝜉 , we have

The 0−1 examination technique [53] is capable of helping distin-
log 𝜉 (𝐪)
guish between erratic and regular behavioural patterns in evolving 𝜉 = lim . (44)
𝐪↦∞ log(𝐪)
structures. The evaluation takes an assortment of information as input
parameters, and its outcome addresses 0 or 1, depending on the extent When the processes of the fractional W–R system (14) are either chaotic
to which the patterns of variation are erratic or not. or not, the asymptotic rate of advancement enables an assessment. 
approaching 0 indicates that the entire system is non-frantic, while 
To express the procedure, we identify the transformation compo-
̃ for neighbouring to 1 indicates that the mechanism is chaotic.
nents in the following way, employing the temporal sequence (𝚥)
The 0–1 examination of the W–R system (14) was executed instan-
𝚥 = 1, 2, … ,  : ̃
taneously on the collection of statistics (𝜅) in this case. Fig. 11 shows
∑
𝐪 the 𝜆 − 𝐰 displays for periodic and chaotic behaviour of the discrete
𝜆𝜉 (𝐪) = ̃ cos(𝚥𝜉),
(𝚥) fractional W–R (14) for different FO values 𝛽1 , 𝛽2 and 𝛽3 .
𝚥=1
Specifically, one may observe that whenever 𝛽 = 0.73, 0.85, 0.93,
∑ 𝐪
(𝛽1 , 𝛽2 , 𝛽3 ) = (0.7, 0.7, 0.8), (𝛽1 , 𝛽2 , 𝛽3 ) = (0.72, 0.82, 0.85) and (𝛽1 , 𝛽2 , 𝛽3 ) =
𝐰𝜉 (𝐪) = ̃ sin(𝚥𝜉),
(𝚥) (42)
𝚥=1
(0.7, 0.8, 0.0.77), Fig. 11(a-f) demonstrates bounded type pathways,
that reveals the fact that the structure is oscillatory, while when
where (𝜉 ∈ (0, 𝜋)) denotes an insignificant quantity. The representation 𝛽 = 0.89, (𝛽1 , 𝛽2 , 𝛽3 ) = (0.7, 0.8, 0.84) and (𝛽1 , 𝛽2 , 𝛽3 ) = (0.7, 0.8, 0.78),
of 𝜆𝜉 and 𝐰𝜉 in the (𝜆𝜉 − 𝐪𝜉 ) plane can be utilized to determine whether Fig. 11(g-i) demonstrate Brownian-type illustrations, which affirms the
or not chaos develops. frantic pattern of the discrete FO W–R system (14). The outcome of the
If the performance of 𝜆𝜉 and 𝐰𝜉 is restricted, the framework’s 0–1 validate are broadly in accordance with the earlier outcomes of the
interactions are periodic; however, if the behaviour is Brownian-like, 𝑀𝐿𝐸 and bifurcation schematics.

11
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 10. Time-fractional illustrations of W–R system (14) for (a) 𝛽1 = 𝛽2 = 𝛽3 = 0.62, (b) (𝛽1 , 𝛽2 , 𝛽3 ) = (0.6, 0.7, 0.8), (c) (𝛽1 , 𝛽2 , 𝛽3 ) = (0.75, 0.7, 0.85).

Approximate entropy (ApEn) Table 1

The ApEn estimation of the FO W–R system (14).

The ApEn technique is employed throughout this part to measure 𝛽1 𝛽2 𝛽3 ApEn 𝛽1 𝛽2 𝛽3 ApEn
the degree of complexities of the discrete fractional W–R system (14). 0.73 0.73 0.73 0.2276 0.7 0.7 0.77 0.2389
In broad terms, progressively additional complicated time periods pos- 0.85 0.85 0.85 0.6432 0.89 0.89 0.89 0.2178
0.93 0.93 0.93 0.2181 0.7 0.8 0.84 0.4331
sess greater ApEn outcomes. The ApEn is determined employing a
0.7 0.7 0.8 0.4361 0.7 0.8 0.78 0.3021
procedure [54]: 0.72 0.82 0.85 0.2053 – – – –

𝐴𝑝𝐸𝑛 = 𝛩𝐪 (𝐬) − 𝛩𝐪+1 (𝐬),

where 𝛩𝐪 (𝐬) is presented as
discrete fractional difference Caputo-like approach and compared it
1 ∑
𝐧−𝐪+1
𝐪 with the existing results. It is demonstrated that the structure has
𝛩𝐪 (𝐬) = log ̃𝜆 (𝐬)
𝐧 − 𝐪 + 1 𝜆=1 six scientifically achievable fixed points, one of which is a particular
̃ where 𝑆𝑇 𝐷() ̃ stands for non-negative stable state. The framework’s local evolving behaviour in
and 𝐬 can be expressed as 𝐬 = 0.2 𝑆𝑇 𝐷()
relation to the aforementioned states of equilibrium is addressed. The
the positive square-root of variance. Table 1 shows the ApEn outcomes
Routh–Hurwitz specifications are applied specifically to determine the
for the fractional W–R system (14) in the two scenarios in which
the FO information are commensurate and incommensurate. Since parametric approach that manipulates the framework’s asynchronous
may be noticed, the ApEn culminates exhibit greater measures when steadiness concerning its non-negative stable state. In view of the
(𝛽1 , 𝛽2 , 𝛽3 ) = (0.85, 0.85, 0.85), (0.7, 0.8, 0.84), (0.7, 0.8, 0.77) suggesting commensurate and incommensurate FOs, MLEs and bifurcated plots are
that the technique’s sequence becomes increasingly intricate in those used to investigate the technique’s chaotic behaviour. As previously
situations. As a consequence, the aforementioned outcome agrees via stated in [45], chemical processes primarily occur in thermodynam-
the prior 𝑀𝐿𝐸 leads to and affirms the presence of unpredictability in ics, where erratic interactions are capable of being noticed subject
the W–R model (14). to the predicted volume dimensions of the mechanism. In addition,
we presented an innovative approach for achieving the MLEs for FO
Conclusion sliding mechanism processors. Mathematical modelling is employed for
confirming the chaotic patterns using the 0–1 test and ApEn techniques.
To achieve chaos oversight of chemical processes, we initially in- This subject definitely merits greater consideration in the future, partic-
vestigated the behaviour of the W–R framework incorporating the ularly when addressing spatiotemporal chaotic behaviour as well as the

12
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

Fig. 11. 0–1 depictions of FO W–R system (14).

development of turbulent motion [45]. With the enactment of a certain Acknowledgements

suitable variable-order fractional discretization of the framework, addi-
tional rich evolving behaviour and chaos influence can be mentioned. The work was supported and funded by the Deanship of Scientific
Maintaining our sights on certain current evolving studies of discrete Research at Imam Mohammad Ibn Saud Islamic University (IMSIU),
time frameworks (see; [55]), our forthcoming work will concentrate on Saudi Arabia (grant number IMSIU-RG23061). All authors read and
the expanding analysis of a specific discrete equivalent of this chemical
approved the final manuscript.
response mechanism.

Availability of supporting data References

No data were used to support this study. [1] Atangana A, Owolabi KM. New numerical approach for fractional differential
equations. Math Model Nat Phenom 2018;13. http://dx.doi.org/10.1051/mmnp/
2018010.
Funding
[2] Owolabi KM, Atangana A. On the formulation of Adams–Bashforth scheme
with Atangana- Baleanu-Caputo fractional derivative to model chaotic problems.
This work was supported and funded by the Deanship of Scientific Chaos 2019;29:023111. http://dx.doi.org/10.1063/1.5085490.
Research at Imam Mohammad Ibn Saud Islamic University (IMSIU), [3] Owolabi KM, Atangana A. Modelling and simulation of a dynamical system
Riyadh 11432, Saudi Arabia, (IMSIU) (grant number IMSIU-RG23061). with the Atangana-Baleanu fractional derivative. Eur Phys J Plus 2018;133:15.
http://dx.doi.org/10.1140/epjp/i2018-11863-9.
Declaration of competing interest [4] Azar AT, Vaidyanathan S. Chaos modeling and control systems design. Studies
in computational intelligence, vol. 581, New York, USA: Springer; 2015.
We declare that all authors contributed equally and there are no [5] Azar AT, Vaidyanathan S. Computational intelligence applications in modeling
and control. Studies in computational intelligence, vol. 575, New York, USA:
conflicts of interest. Springer; 2015.
[6] Lorenz EN. Deterministic non periodic flow. J Atmos Sci 1963;20:130–41.
Data availability [7] Toufik M, Atangana A. New numerical approximation of fractional derivative
with non-local and non-singular kernel: application to chaotic models. Eur Phys
No data was used for the research described in the article. J Plus 2017;132:444. http://dx.doi.org/10.1140/epjp/i2017-11717-0.

13
Y.-M. Chu et al. Results in Physics 54 (2023) 107023

[8] Atangana A, Gomez-Aguilar JF. Hyperchaotic behaviour obtained via a nonlocal [33] Mehmet OEFE. Fractional order sliding mode control with reaching law
operator with exponential decay and Mittag-Leffler laws. Chaos Solit Fract approach. Turk J Elec Eng Comp Sci 2010;18(5).
2017;102:285–94. http://dx.doi.org/10.1016/j.chaos.2017.03.022. [34] Alsharid AH, Rashid S, Elgan SK. Short-memory discrete fractional difference
[9] Atangana A. Fractal-fractional differentiation and integration: connecting fractal equations wind turbine model and their inferential regulate of a chaotic stable
calculus and fractional calculus to predict complex system. Chaos Solit Fract magnet synchronous transformer in time-scale analysis. AIMS Math 2023. in
2017;102:396–406. press.
[10] Rössler OE. An equation for continuous chaos. Phys Lett A 1976;57:397–8. [35] Qurashi MA, Rashid S, Jarad F, Ali E, Egam RH. Dynamic prediction modelling
[11] Arneodo A, Coullet P, Tresser C. Possible new strange attractors with spiral and equilibrium stability of a fractional discrete biophysical neuron model. Res
structure. Comm Math Phys 1981;79:573–9. Phy 2023;48:106405.
[12] Sprott JC. Some simple chaotic flows. Phys Rev E 1994;50:647–50. [36] Atangana A, Alkahtani BST. Modeling the spread of rubella disease using the
[13] Chen G, Ueta T. Yet another chaotic attractor. Int J Bifur Chaos 1999;9:1465–6. concept of with local derivative with fractional parameter: Beta-derivative.
[14] Lü J, Chen G. A new chaotic attractor coined. Int J Bifur Chaos 2002;12:659–61. Complex 2016;21:442–51.
[15] Cai G, Tan Z. Chaos synchronization of a new chaotic system via nonlinear [37] Atangana A, Baleanu D. New fractional derivatives with nonlocal and non-
control. J Uncertain Syss 2007;1:235–40. singular kernel: theory and application to heat transfer model. Therm Sci
[16] Tigan G, Opris D. Analysis of a 3D chaotic system. Chaos Solitons Fractals 2016;20:1–16.
2008;36:1315–9. [38] Jabli N, Khammari H, Mimouni MF, Dhifuoui R. Bifurcation and chaos phenom-
[17] Gaspard P. Rössler systems. In: Alwyn Scott, editor. Encyclopedia of nonlinear ena appearing in induction motor under variation of PI controller parameters.
science. New York: Routledge; 2005, p. 808–11. WSEAS Tran Syst 2010;9(7):784–93.
[18] Stucki J, Urbanczik R. Entropy production of the Willamowski-Rössler oscillator. [39] Mira C, Carcasses JP, Simo C, Tatjer JC. Crossed areaspring area transition-(II)
Z Naturforsch 2005;60:599–607. foliated parametric representation. Int J Bifur Chaos 1991;1(2):339–48.
[19] Willamowski KD, Rössler OE. Irregular oscillations in a realistic abstract [40] Nizar J, Khermneri H, Mimouni MF. Chaos-low periodic orbits transition in a
quadratic mass action system. Z Naturforsch 1980;35a:317–8. synchronous switched circuit. WSEAS Trans Circuits Syst 2008;7(12).
[20] Ali R, Asjad MI, Aldalbahi A, Rahimi-Gorji M, Rahaman M. Convective flow of [41] Cafagna D, Grassi G. Fractional order systems without equilibria: The firs-
a max well hybrid nano fluid due to pressure gradient in a channel. J Therm texample of hyperchaos and its application to synchronization. Chin Phys B
Anal Calorim 2021;143:1319–29. 2015;24(8):080502.
[21] Ali B, Ali L, Abdal S, Asjad MI. Significance of Brownian motion and thermo [42] Atici F, Eloe P. Discrete fractional calculus with the nabla operator. Electron J
phoresis influence on dynamics of Reiner–Rivlin fluid over a disk with non- Qual Theory Differ Equ 2009;1–12. http://dx.doi.org/10.14232/ejqtde.2009.4.3.
Fourier heat flux theory and gyrotactic micro organisms: A numerical approach. [43] Abdeljawad T. On Riemann and Caputo fractional differences. Comput Math Appl
Phys Scr 2021;96:094001. http://dx.doi.org/10.1088/1402-4896/ac02f0. 2011;62:1602–11.
[22] Chu YM, Ali R, Asjad MI, Ahmadian M, Senu N. Heat transfer flow of maxwell [44] Niu H, Zhang Q, Zhang AY. The chaos synchronization of a singular chemical
hybrid nano fluids due to pressure gradient in to rectangular region. Sci Rep model and a Willamowski–Rössler model. Int J Inf Syst Sci 2010;6(4):355–64.
2020;10:16643. [45] Geysermans P, Nicolis G. Thermodynamic fluctuations and chemical chaos in
[23] Li J, Jahanshahi H, Kacar S, Chu Y, Gómez-Aguilar J, Alotaibi N, et al. On a well-stirred reactor: A master equation analysis. J Chem Phys 1993;99:8964.
the variable-order fractional memristor oscillator: data security applications http://dx.doi.org/10.1063/1.465566.
and synchronization using a type-2 fuzzy disturbance observer-based robust [46] Din Q. Dynamics and hopf bifurcation of a chaotic chemical reaction model.
control. Chaos Solitons Fractals 2021;145:110681. http://dx.doi.org/10.1016/j. MATCH Commun Math Comput Chem 2022;88:351–69. http://dx.doi.org/10.
chaos.2021.110681. 46793/match.88-2.351D.
[24] Wang Y, Jahanshahi H, Bekiros S, Bezzina F, Chu Y, Aly A. Deep recurrent [47] Cermak J, Gyori I, Nechvátal L. On explicit stability conditions for a linear
neural networks with finitetime terminal sliding mode control for a chaotic fractional difference system. Fract Calc Appl Anal 2015;18:651–72. http://dx.
fractionalorder financial system with market confidence. Chaos Solitons Fractals doi.org/10.1515/fca-2015-0040.
2021;146:110881. http://dx.doi.org/10.1016/j.chaos.2021.110881. [48] Shatnawi M, Djenina N, Ouannas A, Batiha I, Grassi G. Novel convenient con-
[25] Wang B, Jahanshahi H, Bekiros S, Chu Y, Gómez-Aguilar J, Alsaadi F, et al. ditions for the stability of nonlinear incommensurate fractional-order difference
Tracking control and stabilization of a fractional financial risk system using systems. Alex Eng J 2022;61:1655–63. http://dx.doi.org/10.1016/j.aej.2021.06.
novel active finite-time fault-tolerant controls. Fractals 2021;29(06):2150155. 073.
http://dx.doi.org/10.1142/S0218348X21501553. [49] Wu G, Baleanu D. Discrete fractional logistic map and its chaos. Nonlinear
[26] Atangana A, Koca I. Model of thin viscous fluid sheet flow within the scope Dynam 2013;75:283–7. http://dx.doi.org/10.1007/s11071-013-1065-7.
of fractional calculus: fractional derivative with and no singular kernel. Fund [50] Kelley WG, Peterson AC. The theory of differential equations: Classical and
Inform 2017;151(1–4):145–59. qualitative. New York: Springer; 2010.
[27] Atangana A. On the new fractional derivative and application to nonlinear [51] Wu G, Baleanu D. Jacobian matrix algorithm for Lyapunov exponents of the
Fisher’s reaction–diffusion equation. Appl Math Comput 2016;273:948–56. discrete fractional maps. Commun Nonlinear Sci Numer Simul 2015;22:95–100.
[28] Edelman M, Macau EE, Sanjuan MA. Chaotic, fractional, and complex dy- http://dx.doi.org/10.1016/j.cnsns.2014.06.042.
namics: New insights and perspectives. Berlin, Germany: Springer International [52] Vaidyanathan S, Sampath S. Sliding mode controller design for the global chaos
Publishing; 2018. synchronization of Coullet systems. Lecture notes of the institute for computer
[29] Nizar J, Hedi K, Mimouni MF. Chaos-low periodic orbits transition in a sciences, social-informatics and telecommunications engineering, vol. 84, 2012,
synchronous switched circuit. WSEAS Trans Circuits Syst 2008;7(12):2008. p. 103–10.
[30] Shukla M, Sharma B. Investigation of chaos in fractional order generalized [53] Gottwald GA, Melbourne I. The 0-1 test for chaos: A review. In: Skokos C,
hyperchaotic henon map. AEU - Int J Electron Commun 2017;78:265–73. http: Gottwald G, Laskar J, editors. Chaos detection and predictability. Lecture notes
//dx.doi.org/10.1016/j.aeue.2017.05.009. in physics, Berlin, Heidelberg: Springer; 2016.
[31] Peng Y, He S, Sun K. Chaos in the discrete memristorbased systemwith fractional- [54] Pincus S. Approximate entropy as a measure of system complexity. Proc Natl
order difference. Res Phys 2021;24:104106. http://dx.doi.org/10.1016/j.rinp. Acad Sci USA 1991;88:2297–301. http://dx.doi.org/10.1073/pnas.88.6.2297.
2021.104106. [55] Huang L-L, Park JH, Wu GC, Mo ZH. Variable-order fractional discrete-time
[32] Cao YG, Li Y, Ren W, Chen YQ. Distributed coordination of networked fractional recurrent neural networks. J Comput Appl Math 2020;370:112633.
order systems. IEEE, Trans Syst, Man Cybern 2010;40(2):362–70.

14