Nonlinear Dyn (2024) 112:6205–6214

https://doi.org/10.1007/s11071-024-09370-9

ORIGINAL PAPER

Transient dynamics in a quasiperiodically forced nonsmooth

dynamical system
Gaolei Li · Jicheng Duan · Denghui Li ·
Chen Wang · Shuning Deng

Received: 26 September 2023 / Accepted: 29 January 2024 / Published online: 24 February 2024
© The Author(s), under exclusive licence to Springer Nature B.V. 2024

Abstract Transient chaos is very common in period- Keywords Transient dynamics · Strange nonchaotic
ically forced dynamical systems, but transient strange attractors · Lyapunov exponent · Phase sensitivity
nonchaotic dynamics have not been studied by resear-
chers. Therefore, it is meaningful to explore the dynam-
ical phenomena of transient chaos and transient strange 1 Introduction
nonchaos. In this paper, we investigate a quasiperi-
odically forced cantilever beam system with impacts Transient chaos means that the state of a nonlinear
and discover abundant transient dynamic phenomena. dynamical system starts from the initial state as chaos
Firstly, we determine the transient interval by the bifur- and eventually stabilizes in another relatively ordered
cation diagram and the maximum Lyapunov exponent state [1]. In other words, transient chaos occurs only
and investigate the correspondence between them. Sec- during a certain period of the dynamical systems, rather
ondly, we identify transient chaos and transient strange than throughout the entire motion process. Transient
nonchaos and distinguish transient attractors and per- chaos can be regarded as a kind of metastable state [2].
manent attractors using the maximum Lyapunov expo- The concept of transient chaos is ideally suited to the
nent. Furthermore, we verify the strange property of description of nonequilibrium processes [3–5].
attractors through the phase sensitivity and the power In the numerical study of attractors, the usual
spectrum. approach is to first discard a long sequence of the tra-
jectory and then focus on the asymptotic properties.
G. Li · C. Wang · S. Deng However, if only the asymptotic behavior of dynam-
School of Civil Engineering and Mechanics, Yanshan University, ics is studied, it means losing the interesting chaotic
Qinhuangdao 066004, China parts contained in the transient. [6] Transient chaos
G. Li is a dynamical phenomenon caused by nonattractive
Hebei Key Laboratory of Mechanical Reliability for Heavy chaotic saddles in phase space [7,8]. The chaotic sad-
Equipments and Large Structures, Yanshan University, Qin-
dle is a bounded set and has fractal structures in both
huangdao 066004, China
stable and unstable directions, while chaotic attractors
J. Duan (B) have fractal structures only in stable direction. Due to
School of Mathematics and Physics, China University of Geo-
sciences, Wuhan 430074, China the fractal structure in the unstable direction, there are
e-mail: [email protected] countless gaps of varying sizes on the unstable mani-
D. Li
fold of chaotic saddle. The initial condition is usually
School of Mathematics and Statistics, Hexi University, Zhangye attracted by the chaotic saddle along the stable direc-
734000, China tion, staying near it for a finite time, and then leaving

123
6206 G. Li et al.

the chaotic saddle through one of the gaps in the unsta-

ble direction [9]. In addition, the scattering process in
open Hamiltonian problems [10] and particle transport
in open hydrodynamical flows [11] both involve the
emergence of chaotic saddle and transient chaos.
SNA was first discovered by Grebogi et al. [12] in
1984. From a geometric perspective, SNAs are similar
to chaotic attractors, meaning that SNAs have com-
plex fractal structures. However, in terms of dynamics,
SNAs are nonchaotic, meaning that the maximum Lya-
punov exponent is nonpositive. At present, the study
of SNAs has attracted extensive attention from schol- Fig. 1 The model of the cantilever beam system with impacts
ars, and they are studies in various dynamical sys-
tems such as economic systems [13], circuit systems
stiffness K 1 and K 2 , respectively. e1 and e2 represent
[14], vibro-impact systems [15], neural systems [16],
the gaps between the mass M and the two constraint
and so on [17–20]. The research depth has gradu-
springs. Taking the upper edge of M as the origin of
ally increased, ranging from smooth systems [21] to
the coordinate system when the cantilever structure is
nonsmooth systems [22], from maps [23] to ordinary
stationary, we establish a Cartesian coordinate system
differential equations [24], from strange nonchaotic
with the positive direction oriented vertically down-
dynamical phenomena [25] to generation mechanisms
ward. When the displacement of the mass M reaches
[26], from numerical demonstrations [27] to theoretical
e1 or −e2 , M collides with the two constraint springs.
proofs [28], and the existence of SNAs [29]. SNAs and
The equation of motion of the cantilever beam sys-
chaotic attractors have many similar properties [30].
tem with impacts is as follows
Since the dynamic phenomenon of transient chaos is
relatively common, the dynamic phenomenon of tran- M Ẍ + c Ẋ + f (X ) + E(X ) = A1 cos(ωt) + A2
sient strange nonchaos should also exist, but no one has × cos(ϕt), (1)
studied it. SNAs have practical applications in multiple
fields, and studying the dynamical phenomena of tran- where f (X ) is the elastic restoring force of the system,
sient strange nonchaos can help uncover the underlying which can be described in a nonlinear form [31]
mechanisms of natural laws and engineering problems, 12E I 432E I 3
providing new solutions and design insights. f (X ) = X+ X , (2)
L3 35L 5
The remainder of this paper is organized as follows.
E(X ) is a piecewise function with gaps
In Sec. 2, we introduce a quasiperiodically driven can- ⎧
tilever beam system with impacts. In Sec. 3, we ana- ⎨ K 1 (X − e1 ) , X > e1 ,
lyze different types of transient dynamics. In Sec. 4, E(X ) = 0, −e2 ≤ X ≤ e1 , (3)
⎩
we focus on determining the strange property of SNAs. K 2 (X + e2 ) , X < −e2 .
Finally, in Sec. 5, we conclude our results.
The dimensionless form of Eq. (1) is obtained as fol-
lows. Let x = XL , ζ = Mc , α = 12E
M L3
I
, k1 = KM1 , k2 =
K2 A1 A2 e1 e2
2 The mechanical model M , β = M L , μ = M L , ē1 = L , ē2 = L . Then Eq.
(1) can be written as
 
We consider a cantilever beam system with impacts, 36 3
ẍ + ζ ẋ + α x + x + e(x) = β cos(ωt)
as depicted in Fig. 1. The system consists of a mass M 35
and two leaf springs with the length L and the bend- +μ cos(ϕt), (4)
ing stiffness EI. The boundary conditions restrict the
rotation of the mass, and the mass is excited by a force where
⎧
F, which can be decomposed into two components, ⎨ k1 (x − ē1 ) , x > ē1 ,
F1 = A1 cos(ωt) and F2 = A2 cos(ϕt). The upper and e(x) = 0, −ē2 ≤ x ≤ ē1 , (5)
⎩
lower constraints are formed by two linear springs with k2 (x + ē2 ) , x < −ē2 .

123
Transient dynamics in a quasiperiodically forced nonsmooth dynamical system 6207

Let θ = ωt, φ = ϕt, Eq. (4) can be transformed into on the basin boundary, converting the attractor into a
the following state equations nonattracting chaotic set and generating transient chaos
[1,34]. At an interior crisis, a chaotic attractor, typically
ẋ = y, of small size, collides with a coexisting nonattracting
 
36 3 chaotic set (or its stable manifold), resulting in a larger
ẏ = −ζ ẋ − α x + x + β cos θ
35 chaotic attractor. While boundary crisis is a mechanism
(6)
+ μ cos φ − e(x), for chaotic transients, interior crisis provides a mech-
θ̇ = ω, anism for sudden enlargement of a chaotic attractor in
φ̇ = ϕ. the phase space and intermittency.
The orbit is usually attracted by the chaotic attractors
Taking φ as the angle variable, we obtain a three- before the crisis, lasts for some time, and then suddenly
dimensional Poincaré map leaves the chaotic region (no return) and is attracted to
: → : other regions. We know that the definition of transient
  (7) chaos is the form of chaos due to nonattracting chaotic
≡ (x, v, φ) ∈ R × R × S 1 | φ mod 2π/ϕ = 0 . sets in the phase space. Transient SNAs are first discov-
This map has the form ered in the system, and their dynamical properties are
similar to transient chaos. To show the transient behav-
xn+1 = f 1 (xn , yn , φn ) , ior, we locally magnify Fig. 2b. It can be observed that
yn+1 = f 2 (xn , yn , φn ) , the attractors of different types appear alternately in
(8)
2π · ω the interval α ∈ [0.594, 0.609]. The transient attractors
φn+1 = φn + (mod2π ). exist in a relatively short time interval and eventually
ϕ
evolve into permanent attractors, as shown in Fig. 3.
In the parameter interval [0.594, 0.609], the tran-
3 Abundant transient dynamics sient dynamics is very complex, so we choose differ-
ent parameters for analysis. The attractors of the tran-
In general, we can determine the region of transient sient state are marked by blue color, and the permanent
dynamics with the help of the bifurcation diagrams attractors are marked by red color. First, we choose
and the maximum Lyapunov exponent. In this section, α = 0.60827 and observe that the transient attractors
we choose the system parameters β = 0.03, kc = appear for a long time in the time series diagram. When
0.12, μ = 0.02, k1 = 25, √ k2 = 25, ϕ = 1, e1 = t ≈ 50000, the transient attractors disappear and the
0.15, e2 = 0.15, and ω = 5−1 2 . Taking α as the con- motion of the system evolves into permanent attrac-
trol parameter, the bifurcation diagram and the max- tors, as shown in Fig. 4a. At this point, the maximum
imum Lyapunov exponent of the system with respect Lyapunov exponent of the system ( L E = −0.00903,
to α are plotted in the interval [0.55,0.63], as shown see Fig. 4b) is less than 0, indicating that the transient
in Fig. 2a, b. At the upper end of Fig. 2, we mark the attractors and the permanent attractors are not chaotic,
different types of attractors. The red region of Fig. 2 but the transient attractors have a strange geometric
corresponds to the chaotic attractors marked by CA, structure, as shown in the blue attractors in the (θn , xn )-
the green region corresponds to the transient region plane and (xn , yn )-plane in Fig. 4c, d. The strange prop-
marked by T, and the yellow region corresponds to erty can be verified by the phase sensitivity and singular
the quasiperiodic attractors marked by QA. It is shown continuous spectrum. For t ∈ [0, 50000], the value of
that for α ∈ [0.594, 0.609], transient dynamics appear the derivative of state variable with respect to phase
in the system. We find that the value of maximum Lya- increases, which shows that the attractors are strange.
punov exponent oscillates near 0 in this parameter inter- For t ∈ [50000, 100000], the value tends to a fixed
val, see Fig. 2b. value, indicating that the attractor does not have strange
Transient chaos is closely related to crises. There property. Then the permanent attractors are quasiperi-
are two major types of crises in dynamical systems, odic, and it can be observed that the attractor is smooth
including boundary crises and interior crises [32,33]. A in the local magnification, as shown in the red attractor
boundary crisis occurs when a chaotic attractor collides in Fig. 4c, d. The power spectrum of SNA has discrete
with an unstable periodic orbit (or its stable manifold) and continuous elements, so it is called the singular

123
6208 G. Li et al.

Fig. 2 The bifurcation and

maximum Lyapunov
exponent of the system with
respect to α

negative at this time, as shown in Fig. 6b. It can be

observed from the maximum Lyapunov exponent that
the attractors have chaotic property during the transient
state, but the permanent attractors are nonchaotic. We
can observe that the local magnification of the attractor
is nonsmooth, as shown in Fig. 6c, d. The quasiperiodic
attractors evolve into SNAs through fractal routes [35],
and the wrinkled geometry can be used as a prelude to
the appearance of SNAs.
Finally, the value of α is further reduced to 0.5974,
and the existence interval of transient dynamics is fur-
Fig. 3 The transient region T in Fig. 2a is selected for local ther reduced from the time series diagram, see Fig. 7a.
magnification
The dynamical properties of transient chaos do not
change, but the permanent attractors evolve from SNAs
continuous spectrum. The detailed procedure for veri- to chaotic
fying strange property in this set of parameters will be attractors, and both attractors are chaotic. Because the
given in Sec. 4. corresponding maximal Lyapunov exponent is always
Next, we take α = 0.5996, the transient dynam- positive, as shown in Fig. 7b. Since the time interval
ics occur over a relatively long time interval, and the t over which the transient dynamics appears is rela-
system becomes a permanent state at t ≈ 58000, as tively short, the transient chaotic attractors (blue attrac-
shown in Fig. 5a. The maximum Lyapunov exponent tors) are clearly less obvious in the phase diagram than
of the system also changes from positive to negative the permanent attractors (red attractors), as shown in
at t ≈ 58000, as shown in Fig. 5b. The blue chaotic Fig. 7c, d.
attractors evolve into quasiperiodic attractors, indicat-
ing that the transient dynamics under this parameter
is transient chaos, as shown in Fig. 5c, d. However, it 4 Verify the strange property of SNAs
can be observed that a part of the trajectory is smooth
and the other is nonsmooth in the local magnification According to the definition of SNAs, SNAs have
of Fig. 5d, so the red permanent attractors have local strange property, but it is difficult to judge the strangeness
unsmoothness. of SNAs based on its dynamics. There are a few more
We further decrease the value of α, taking α = effective tools to judge its strange property, such as
0.599, and transient dynamics were found to emerge rational approximation [36], and power spectrum [37].
over relatively short time interval [0, 35000], see In this section, we verify the strange property of
Fig. 6a. The transient phenomenon enters the perma- SNAs using the phase sensitive, and this approach is
nent state at t ≈ 35000, and the maximum Lyapunov based on the phase-sensitive dependence of SNAs [38].
exponent of the system also changes from positive to Because SNAs are piecewise nondifferentiable, there

123
Transient dynamics in a quasiperiodically forced nonsmooth dynamical system 6209

Fig. 4 For α = 0.60827,

the transient attractors and
the permanent attractors

Fig. 5 For α = 0.5996, the

transient attractors and the
permanent attractors

123
6210 G. Li et al.

Fig. 6 For α = 0.599, the

transient attractors and the
permanent attractors

Fig. 7 For α = 0.5974, the

transient attractors and the
permanent attractors

123
Transient dynamics in a quasiperiodically forced nonsmooth dynamical system 6211

Fig. 8 The phase

sensitivity rate τ N versus N

Fig. 9 The phase diagram

in the (xn , yn )-plane

are some special bifurcation points whose derivative ations and tends to infinity, the value of the derivative
value with respect to phase tends to infinity, and by of the attractor with respect to the phase is also infinite,
this property, the phase sensitivity indicates that the indicating that the attractor is nonsmooth, that is, the
attractors have nonsmoothness. According to Equation attractor is an SNA.
(8), the derivative of state variable with respect to phase For a small σ , an n 0 satisfying the phase difference
can be expressed as follows ε0 = θn 0 − θ0 < σ can be found, and the derivative
∂ fi with respect to the phase can be approximated by
Si (N ) = (i = 1, 2), (9)
∂θ
where N is the number of iterations. If N → +∞, N −n 0
∂ fi | f i (k + n 0 ) − f i (k)|
the value of SiN does not increase with the number of Si (N ) = ≈ (i = 1, 2),
iterations, but tends to a stable value, then the attractor ∂θ |θ (k + n 0 ) − θ (k)|
k=1
is not strange. If SiN increases with the number of iter- (10)

123
6212 G. Li et al.

slope of the increase of the two types of attractors is

slightly different, as shown in Fig. 8c. Finally, we ver-
ify the case that both transient attractors and permanent
attractors are chaotic attractors (α = 0.5974). Because
chaotic attractors have strangeness, the value is con-
tinuously increasing with the number of iterations, as
shown in Fig. 8d. The results show that the phase sen-
sitivity can explain whether the attractor has strange
property.
In order to further verify the feasibility of this
method, we select parameters without transient dynam-
Fig. 10 The phase sensitivity exponent τ N versus N , the ics to verify the strangeness of √ the attractors. Fixe
quasiperiodic attractor for α = 0.615 and SNA for α = 0.6005 parameter values ϕ = 1, ω = ( 5 − 1)/2, ē1 = ē2 =
0.25, β = 0.04, ζ = 0.1, k1 = k2 = 40 and μ = 0.04,
take α = 0.615, the attractors are quasiperiodic, as
where k, n 0 satisfy k + n 0 ≤ N , and f i (k) denotes the
shown in Fig. 9a. When α = 0.6005, the attractors have
k-th iteration of f i . The maximum value of SiN after N
the nonsmooth property and the corresponding Lya-
times iterations is denoted as
punov exponent is negative, then the attractor evolves
τ N = max {Si (N )} . (11) into an SNA, as shown in Fig. 9b.
If the derivative of the state variable with respect to We use the same calculation method for the strange
phase goes to infinity, the attractor has strange proper- properties under these two sets of parameters to illus-
ties. trate the properties of the attractor by the phase sensi-
Here we verify the strange properties of the tran- tivity rate. Because the quasiperiodic attractors have no
sient attractors and permanent attractors in Sec. 4 and strange properties, the value of τ N should be stationary
compute the phase-sensitive rate of these attractors. with the increasing number of iterations and the slope
Firstly, we verify the strangeness of attractors in Fig. 4 of τ N should be approximately equal to zero. However,
( α = 0.60827). At the moment, the transient attractors SNA has the strangeness that the value of τ N always
and permanent attractors are SNAs and 2T quasiperi- increases with the number of iterations to an infinite
odic attractors, respectively. According to the definition value, as shown in Fig. 10.
of phase sensitivity rate, SNAs have strange property To further verify the strange property of attractors,
and the value of τ N tends to infinity. However, the 2T we choose another important method, namely, singu-
quasiperiodic attractors have no strange properties and lar continuous spectrum [37]. Discrete and continuous
τ N should tend to a fixed value. Therefore, the calcu- power spectra are typically present in the power spec-
lated phase sensitivity rate should first increase and then trum, and the discrete spectrum corresponds to peri-
tend to a fixed value. In Fig. 8a, we can observe that the odic or quasiperiodic motions, while the continuous
value of τ N has been increasing in the transient state, power spectrum is associated with chaotic and random
but after the transient state ends, the attractors have motions. However, the power spectrum of SNAs is sin-
no strange properties, and τ N tends to a stable value. gular continuous spectrum, which includes both dis-
Secondly, the transient attractors and permanent attrac- crete and continuous components, and it is the special
tors in Fig. 5 (α = 0.5996) are chaotic attractors and spectrum that exists between the discrete and continu-
2T quasiperiodic attractors with wrinkles, respectively. ous spectra. The Fourier transform is defined as follows
The calculated phase sensitivity rate results should be N
the same as in Fig. 8a, which first increases and then S(ω, N ) = xn ei2π mω , (12)
tends to a fixed value, as shown in Fig. 8b. Then, the n=1
case that transient attractors are chaotic attractors and the power spectrum of the attractor is defined as [39]
permanent attractors are SNAs (α = 0.599) is verified.
Pω = lim |S(ω, N )/N |2 . (13)
Because both types of attractors have strange property, N →∞
the values of phase sensitivity rate are calculated to The power spectrum of quasiperiodic attractors is
increase with the increase of iteration times N , but the discontinuous and has countable δ-peaks, as shown in

123
Transient dynamics in a quasiperiodically forced nonsmooth dynamical system 6213

Fig. 11 The power

spectrum

Fig. 11a. However, the power spectrum of SNAs has National Natural Science Foundation of China (Nos. 12362002,
multiple δ-peaks and exhibits a fractal geometric struc- and 12172306).
ture; it is called singular continuous spectrum, as shown Data availability The datasets generated during and/or analyzed
in Fig. 11b. The singular continuous spectrum is a mix- during the current study are available from the corresponding
ture of discontinuous and continuous spectrum, which author on reasonable request.
can reflect the strange property of the attractor. There-
fore, we can distinguish SNAs and quasiperiodic attrac- Declarations
tors by the power spectrum. Conflict of interest All co-authors have no conflict of interest
to declare.

5 Conclusion
References
In this paper, a class of nonsmooth dynamical sys-
tems with quasiperiodic forcing is studied, and abun- 1. Lai, Y.C. , Tél, T.: Transient Chaos: Complex Dynamics on
dant transient dynamics phenomena are found, includ- Finite Time Scales. Springer. New York, Part I, pp. 17–160
ing transient chaos and transient strange nonchaos. The 2. de Moura, A.P.S., Lai, Y.C., Bird, J.P., et al.: Tunneling
and nonhyperbolicity in quantum dots. Phys. Rev. Lett. 88,
parameters of transient dynamics were determined by 236804 (2002)
the bifurcation diagram and the maximum Lyapunov 3. Dorfmann, J.R.: An Introduction to Chaos in Nonequilib-
exponent. The time intervals in which different types of rium Statistical Mechanics. Cambridge University Press.
transient dynamics exist are analyzed. In addition, the Cambridge, first edition (1999)
4. Ngamsa Tegnitsap, J.V., Fotsin, H.B.: Multistability, tran-
strange property of the attractors is verified by the phase sient chaos and hyperchaos, synchronization, and chimera
sensitivity in the transient and nontransient case. The states in wireless magnetically coupled VDPCL oscillators.
results show that the phase sensitivity is an effective Chaos Solitons Fractals 158, 112056 (2022)
tool to verify the strangeness of attractors in transient 5. Kong, L.W., Fan, H., Grebogi, C., et al.: Emergence of tran-
sient chaos and intermittency in machine learning. J. Phys.
dynamics. Complexity 2, 035014 (2021)
6. Tél, T.: The joy of transient chaos. Chaos 25, 097619 (2015)
Acknowledgements We sincerely thank the people who give 7. Chithra, A., Fozin, T.F., Srinivasan, K., et al.: Complex
valuable comments. The paper is supported by the National Nat- dynamics in a memristive diode bridge-based MLC cir-
ural Science Foundation of China (Nos. 11832009, 12072291, cuit: coexisting attractors and double-transient chaos. Int.
12362002, and 12172306). J. Bifurc. Chaos 31, 2150049 (2021)
8. Boccaletti, S., Grebogi, C., Lai, Y.C., et al.: The control of
Author contributions Gaolei Li helped in methodology, for- chaos: theory and applications. Phys. Rep. 329, 103–197
mal analysis, writing—original draft. Jicheng Duan contributed (2000)
to formal analysis, writing—review & editing. Denghui Li and 9. Dhamala, M., Lai, Y.C.: Controlling transient chaos in deter-
Shuning Deng validated the study. Chen Wang was involved in ministic flows with applications to electrical power systems
methodology. and ecology. Phys. Rev. E 59, 1646–1655 (1999)
10. Seoane, J., Sanjuán, M.: Exponential decay and scaling laws
Funding The paper is supported by the Natural Science Founda- in noisy chaotic scattering. Phys. Lett. A 372, 110–116
tion of Hebei Province, China (Grant No. A2023203019), and the (2008)

123
6214 G. Li et al.

11. Budanur, N.B., Dogra, A.S., Hof, B.: Geometry of transient 28. P. Glendinning, T. Jäeger, G. Keller. How chaotic are strange
chaos in streamwise-localized pipe flow turbulence. Phys. nonchaotic attractors. Nonlinearity.19, 2005–2022 (2006)
Rev. Fluids 4, 102401 (2019) 29. Glendinning, P.: Global attractors of pinched skew products.
12. Grebogi, C., Ott, E., Pelikan, S., Yorke, J.A.: Strange attrac- Dyn. Syst. 17, 287–294 (2002)
tors that are not chaotic. Physica D 13, 261–268 (1984) 30. U. Feudel, S. Kuznetsov, A.S. Pikovsky, Strange Non-
13. Li, G., Duan, J., Li, D., et al.: Generation mechanisms of chaotic Attractors: Dynamics between Order and Chaos in
strange nonchaotic attractors and multistable dynamics in a Quasiperiodically Forced Systems. World Scientific. Singa-
class of nonlinear economic systems. Nonlinear Dyn. 111, pore, Chapter 2, pp. 9–27
10617–10627 (2023) 31. Emans, J., Wiercigroch, M., Krivtsov, A.M.: Cumulative
14. Senthilkumar, D.V., Srinivasan, V., Thamilmaran, K., et al.: effect of structural nonlinearities: chaotic dynamics of can-
Birth of strange nonchaotic attractors through formation and tilever beam system with impacts. Chaos Solitons Fractals
merging of bubbles in a quasiperiodically forced Chua’s 23, 1661–1670 (2005)
oscillator. Phys. Rev. E 78, 066211 (2008) 32. Grebogi, C., Ott, E., Yorke, J.: Chaotic attractors in crisis.
15. Zhang, Y., Luo, G.: Torus-doubling bifurcations and strange Phys. Rev. Lett. 48, 1507–1510 (1982)
nonchaotic attractors in a vibro-impact system. J. Sound Vib. 33. Grebogi, C., Ott, E., Yorke, J.: Crises, sudden changes in
332, 5462–5475 (2013) chaotic attractors and chaotic transients. Physica D 7, 181–
16. Lindner, J.F., Kohar, V., Kia, B., et al.: Strange nonchaotic 200 (1983)
stars. Phys. Rev. Lett. 114, 054101 (2015) 34. Place, C.M., Arrowsmith, D.K.: Control of transient chaos
17. Prasad, A., Mehra, V., Ramaswamy, R.: Intermittency route in tent maps near crisis. I. Fixed point targeting. Phys. Rev.
to strange nonchaotic attractors. Phys. Rev. Lett. 79, 4127– E 61, 1357–1368 (2000)
4130 (1997) 35. Datta, S., Ramaswamy, R., Prasad, A.: Fractalization route
18. Zhang, Y., Shen, Y.: A new route to strange nonchaotic to strange nonchaotic dynamics. Phys. Rev. E 70, 046203
attractors in an interval map. Int. J. Bifurc. Chaos 30, (2004)
2050063 (2020) 36. Li, G., Yue, Y., Xie, J., Grebogi, C.: Strange nonchaotic
19. Duan, J., Zhou, W., Li, D., et al.: Birth of strange nonchaotic attractors in nonsmooth dynamical system. Commun. Non-
attractors in a piecewise linear oscillator. Chaos 32, 103106 linear Sci. Numer. Simul. 78, 104858 (2019)
(2022) 37. Pikovsky, A.S., Feudel, U.: Correlations and spectra of
20. Prasad, A., Ramaswamy, R., Satija, I.I., et al.: Collision and strange non-chaotic attractors. J. Phys. A 27, 5209–5219
symmetry breaking in the transition to strange nonchaotic (1994)
attractors. Phys. Rev. Lett. 83, 4530–4533 (1999) 38. Pikovsky, A.S., Feudel, U.: Characterizing strange non-
21. Romeiras, F., Ott, E.: Strange nonchaotic attractors of the chaotic attractors. Chaos 5, 253–260 (1995)
damped pendulum with quasiperiodic forcing. Phys. Rev. A 39. Nayfeh, A.H., Balachandran, B.: Applied Nonlinear
35, 4404–4413 (1987) Dynamics: Analytical, Computational, and Experimental
22. Shen, Y., Zhang, Y.: Mechanisms of strange nonchaotic Methods. John Wiley & Sons. New York, Chapter 3, pp.
attractors in a nonsmooth system with border-collision bifur- 147–219 (1995)
cations. Nonlinear Dyn. 96, 1405–1428 (2019)
23. Zhao, Y., Zhang, Y.: Border-collision bifurcation route to
strange nonchaotic attractors in the piecewise linear normal Publisher’s Note Springer Nature remains neutral with regard
form map. Chaos Solitons Fractals 171, 113491 (2023) to jurisdictional claims in published maps and institutional affil-
24. Zhang, Y.: Strange nonchaotic attractors with Wada basins. iations.
Physica D 259, 26–36 (2013)
25. Ditto, W.L., Spano, M.L., Savage, H.T., et al.: Experimen- Springer Nature or its licensor (e.g. a society or other partner)
tal observation of a strange nonchaotic attractor. Phys. Rev. holds exclusive rights to this article under a publishing agreement
Lett. 65, 533–536 (1984) with the author(s) or other rightsholder(s); author self-archiving
26. Kim, S.Y., Lim, W., Ott, E.: Mechanism for the intermit- of the accepted manuscript version of this article is solely gov-
tent route to strange nonchaotic attractors. Phys. Rev. E 67, erned by the terms of such publishing agreement and applicable
056203 (2013) law.
27. Aravindh, S.M., Venkatesan, A., Lakshmanan, M.: Strange
nonchaotic attractors for computation. Phys. Rev. E 97,
052212 (2018)

123