Chaotic dynamics is an active area of research in biology, physics, sociology, psychology, physiology, and engineering. This interest in chaos is also expanding to the social scientific fields such as politics, economics, and argument of prediction of societal events. In this paper, we propose a dynamic model for addiction of tobacco. A proposed dynamical model originates from the dynamics of tobacco use, recovery, and relapse. In order to make an addiction model of tobacco, we try to modify and rescale the existing tobacco and Lorenz models. Using these models, we can derive a new tobacco addiction model. Finally, we obtain periodic motion, quasiperiodic motion, quasichaotic motion, and chaotic motion from the addiction model of tobacco that we established. We say that periodic motion and quasiperiodic motion are related to the preaddiction or recovery stage, respectively. Quasichaotic and chaotic motion are related to the addiction stage and relapse stage, respectively.
1. Introduction
In the last two decades, chaotic dynamics have been widely used in realworld applications such as biological systems
[1]
, brain modeling
[2]
, weather modeling
[3]
, vibration modeling
[4]
, mechanical and electrical engineering
[5
,
6]
, control and synchronization
[7
,
8
,
19]
, robotics
[9
,
10]
, and more
[12

15
,
20
,
21]
. Currently, it is an active area of research by many in biology, physics, sociology, psychology, physiology, and engineering. However, this research is not only performed in natural sciences but also in the social sciences, including politics, economics, and the prediction of societal events composed of various addiction problems
[12
,
18
,
19]
.
Addictions caused by tobacco, drugs, games, shopping, and the Internet are serious problems in our current society and many researchers are interested in treating and curing these addictions. Due to the different nature of each addiction, each symptom differs dramatically from one to the other. Therefore, it is not so easy to define encompassing symptoms for all the addictions mathematically
[16]
. Tobacco addiction has been subject to intense investigation by medical research groups. However, they mainly approach the tobacco addiction epidemiologically rather than mathematically. Therefore, it is necessary to develop a mathematical model of nonlinear phenomena in addiction of tobacco.
In this paper, we propose a dynamic model of tobacco addiction that originates from the dynamics of tobacco use, recovery, and relapse
[17]
. In order to make an addiction model of tobacco, we try to modify and rescale from the existing chaotic tobacco model and Lorenz model. By using this information, we derive a new tobacco addiction model mathematically. Finally, we obtain periodic and chaotic motion from the addiction model of tobacco. We say that periodic motion and quasiperiodic motion are related to the preaddiction stage or recovery stage, respectively. Quasiperiodic and chaotic motion are related to the addiction stage and relapse stage, respectively.
2. Dynamic Model of Lorenz System
The mathematical models of the Lorenz system proposed by Edward L. Lorenz describe various dynamical models of weather systems. The formulation of the Lorenz system is given by the following Eq. (1).
where sigma is called the Prandtl number, which is the ratio of momentum diffusivity (kinematic viscosity) and thermal diffusivity;
ρ
is the Rayleigh number, which is determined by the heat transfer, primarily in the form of conduction or convection; and
β
is a geometric factor. The typical values of these three parameters are
σ
= 10,
ρ
= 28,
β
= 8/3;
x
is proportional to the intensity of convection motion;
y
is proportional to the temperature difference between the ascending and descending currents; and
z
is proportional to the distortion of the vertical temperature profile from linearity.
3. Dynamic Model of Tobacco Use, Recovery, and Relapse
Carlos CastilloGarsow et al.
[17]
proposed mathematical models for the dynamics of tobacco use, recovery, and relapse.
They described the general model for drug abuse, vital dynamics, and nonlinear relapse. From these mathematical models, they found the parameter value by using stability theory.
 3.1 General Mathematical Model for Drug Abuse
From the model proposed by Carlos CastilloGarsow et al.
[17]
, a general mathematical model for drug abuse is shown as
Figure 1
and can be represented as s Eq. (2).
A general mathematical model for drug abuse.
where
S
is the total number of individuals in a given population that are susceptible to becoming regular drug users.
D
is a habitual drug user in that population, and
R
is the number of people recovered from habitual use. The parameters
β
,
γ
, and
δ
are per capita influence rate, recovery rate per habitual drug user per unit of time, and relapse rate per recovered individual per unit of time, respectively.
 3.2 General Mathematical Model for Drug Abuse with Vital Dynamics
Because Eq. (2) assumes a constant population, we need to consider constant mortality rate, following as Eq. (3).
where
μ
is the constant mortality rate.
The previous two models did not provide the mechanism for the relapse process. We consider a similar peer pressure mechanism for relapse of recovered individuals. Eq. (4) represents relapse of recovered individuals.
 3.3 Parameter Rescaling
In order to derive chaotic or nonlinear phenomena from Eqs. (2) and (3), we need to modify these equations. From Eqs. (2) and (3), we obtain a solution, which is shown in
Figure 2
and
3
.
Figures 2
and
3
show timeseries and a phase portrait for Eqs. (2)(4), respectively. Here, we use
N
= 100,
δ
= 0.89,
γ
= 0.15, and
β
= 0.625 for Eq. (2),
N
= 100,
μ
= 0.1,
δ
= 0.89,
α
= 0.4,
γ
= 0.15, and
β
= 0.625 for Eq. (3), and
N
= 100,
μ
= 0.1,
δ
= 0.89,
α
= 0.4,
γ
= 0.15, and
β
= 0.625,
R
= 2.5 for Eq. (4). We recognize that there is no vibration or fluctuation because Eqs. (2) and (3) lack a nonlinear term.
Time series and phase protrait from Eq. (2).
Time series and phase portrait from Eq. (3).
Time series and phase portrait from Eq. (4).
To make a similar Lorenz equation such as Eq. (1) from Eq. (2), we need parameter rescaling and some substitution. To do this, we rescale each parameter
x
,
y
, and
z
, as described in Eqs. (5)(7).
When we take derivatives of Eqs. (5)(7), we obtain Eqs. (8)(10).
We can set Eq. (1) equal to Eqs. (8)(10). The new equation can be written as Eqs. (11)(13).
From Eqs. (11)(13), we can rearrange a new dynamic equation for tobacco use, recovery, and relapse, as shown in Eqs. (14)(16).
In order to display nonlinear phenomena from Eqs. (14)–(16), we define each parameter in the previously shown equations to find periodic motion and chaotic motion.
 3.4 Preaddiction, Addiction Recovery, and Relapse Stages
From Eqs. (14)(16), after assigning parameter values for
α
,
γ
, and
δ
, we varied
σ
,
ρ
, and
β
. As a result, we obtained various time series, phase portraits, and Poincaré for the preaddiction stage of tobacco, as shown in
Figure 5
and
6
. The preaddiction stage has a periodic motion, as with the following:
Time series, phase portrait, and Poincaré map of preaddiction stage of tobacco model (periodic motion).
Time series, phase portrait, and Poincaré map of preaddiction stage of tobacco model (periodic motion).
We obtain various time series, phase portraits, and Poincaré maps for the addiction stage of tobacco as the parameters vary. These are shown in
Figures 7
and
8
. These addiction stages show chaotic motions such as quasichaotic and chaotic attractor.
Time series, phase portrait, and Poincaré map of addiction stage (chaotic motion).
Time series, phase portrait, and Poincaré map of addiction stage (chaotic motion).
We also obtain another time series, phase portrait, and Poincaré map for the recovery stage of tobacco, as shown in
Figure 9
. The recovery stage has a quasiperiodic motion unlike the chaotic attractors of
Figures 7
and
8
. We also varied parameters of
σ
,
ρ
, and
β
, and obtained another phase portrait for the readdiction stage of tobacco, as shown in
Figure 10
. The readdiction stage displays chaotic motion well.
Time series, phase portrait, and Poincaré maps of recovery stage of tobacco model (quasiperiodic motion).
Time series, phase portrait, and Poincaré map of readdiction stage of tobacco model (chaotic motion).
When we change the parameter values
σ
,
ρ
, and
β
accordingly, we obtain the relapse stage of the tobacco model, as shown in
Figure 11
. We obtain recovery and complete the recovery stage of tobacco, as shown in
Figures 12
and
13
, as we adjust more parameter values accordingly.
Figures 12
and
13
represent periodic motion of the complete recovery stage for an addiction.
Time series, phase portrait, and Poincaré map of relapse stage of tobacco model (quasichaotic motion).
Time series, phase portrait, and Poincaré map of recovery stage of tobacco model (quasiperiodic motion).
Time series, phase portrait, and Poincaré map of complete recovery stage of the tobacco model (periodic motion).
4. Conclusions
We proposed a tobacco addiction model by rescaling parameters from the existing tobacco model and the Lorenz system. The mathematical addiction model of tobacco is organized by a thirdorder differential ordinary system. It is similar to the Lorenz system because it is derived from the system.
From the tobacco addiction model, we obtained periodic motion, quasiperiodic motion, quasichaotic motion, and chaotic motion. These motions correspond to preaddiction or complete recovery, recovery or rerecovery, and the relapse and addiction stages, respectively.
In the future, we need to find detailed roles for each parameter in the addiction model of tobacco that we devised in order to control and treat the addiction status in humans.
Conflict of Interest No potential conflict of interest relevant to this article was reported.
BIO
Youngchul Bae received his B.S., M.S., and Ph.D. from the Department of Electrical Engineering, Kwangwoon University, Korea, in 1984, 1985, and 1997, respectively. He worked at Korea Electric Power Company (KEPCO) during 1986–1991, and also worked at Korea Institute of Science and Technology Information (KISTI) during 1991–1997 as a senior researcher. He is currently a professor at the Division of Electrical·Electronic Communication and Computer Engineering, Chonnam National University, Yeosu, Korea. His research interests include nonlinear dynamics, chaos dynamics, robot control, intelligent systems, and motor control. He is member of KIEE, KIECS, and KIIS.
Tel: +82616597315, Fax: +82616597310
Email: ycbae@chonnam.ac.kr
Garfinkel A.
,
Spano M.
,
Ditto W.
,
Weiss J.
1992
“Controlling cardiac chaos, ”
Science
http://dx.doi.org/10.1126/science.1519060
257
(5074)
1230 
1235
DOI : 10.1126/science.1519060
Faure P.
,
Korn H.
2001
“Is there chaos in the brain? I. Concepts of nonlinear dynamics and methods of investigation,”
Comptes Rendus de l’Academie des Sciences Serie III, Sciences de la vie
324
(9)
773 
793
DOI : 10.1016/S07644469(01)013774
May R. M.
1976
“Simple mathematical models with very complicated dynamics,”
Nature
http://dx.doi.org/10.1038/261459a0
261
(5560)
459 
467
DOI : 10.1038/261459a0
Bae Y. C.
2001
“A study on chaotic phenomena in rolling mill bearing, ”
Journal of Korean Institute of Intelligent Systems
11
(4)
315 
319
Bae YoungChul
,
Kim JuWan
,
Kim YiGon
,
Shon YoungWoo
2003
“Secure communication using embedding drive synchronization,”
Journal of Korean Institute of Intelligent Systems
13
(3)
310 
315
DOI : 10.5391/JKIIS.2003.13.3.310
Bae Y.
,
Kim Y. G.
,
Tinduka M.
2005
“A study on generalized synchronization in the statecontrolled cellular neural network(SCCNN),”
International Journal of Fuzzy Logic and Intelligent Systems
5
(4)
291 
296
DOI : 10.5391/IJFIS.2005.5.4.291
Yu S. H.
,
Hyun C. H.
,
Park M.
2011
“Backstepping control and synchronization for 4D LorenzStenflo chaotic system with single input,”
International Journal of Fuzzy Logic and Intelligent Systems
http://dx.doi.org/10.5391/IJFIS.2011.11.3.143
11
(3)
143 
148
DOI : 10.5391/IJFIS.2011.11.3.143
Yu S. H.
,
Hyun C. H.
,
Park M.
2011
“Control and synchronization of new hyperchaotic system using active backstepping design,”
International Journal of Fuzzy Logic and Intelligent Systems
http://dx.doi.org/10.5391/IJFIS.2011.11.2.077
11
(2)
77 
83
DOI : 10.5391/IJFIS.2011.11.2.077
Bae Y. C.
2014
“Synchronization of dynamical happiness model,”
International Journal of Fuzzy Logic and Intelligent Systems
http://dx.doi.org/10.5391/IJFIS.2014.14.2.91
14
(2)
91 
97
DOI : 10.5391/IJFIS.2014.14.2.91
Bae Y. C.
,
Kim Y.
,
Koo Y.
2007
“The target searching method in the chaotic mobile robot embedding BVP model,”
Journal of Korean Institute of Intelligent Systems
17
(2)
259 
264
DOI : 10.5391/JKIIS.2007.17.2.259
Bae Y. C.
2011
“Chaotic phenomena in MEMS with duffing equation,”
The Journal of the Korea Institute of Electronic Communication Sciences
6
(5)
709 
716
Bae Y. C.
,
Kim Y.
,
Kim C.
2004
“Obstacle avoidance method in the chaotic unmanned aerial vehicle,”
Journal of Korean Institute of Intelligent Systems
14
(7)
883 
888
DOI : 10.5391/JKIIS.2004.14.7.883
Bae Y. C.
2013
“Chaotic phenomena in addiction model for digital leisure,”
International Journal of Fuzzy Logic and Intelligent Systems
http://dx.doi.org/10.5391/IJFIS.2013.13.4.291
13
(4)
291 
297
DOI : 10.5391/IJFIS.2013.13.4.291
Sprott J. C.
2005
“Dynamical models of happiness,”
Nonlinear Dynamics, Psychology, and Life Sciences
9
(1)
23 
36
Satsangi D.
,
Sinha A. K.
2012
“Dynamics of love and happiness: a mathematical analysis,”
International Journal of Modern Education and Computer Science
4
(5)
31 
37
DOI : 10.5815/ijmecs.2012.05.05
Lykken D.
,
Tellegen A.
1996
“Happiness is a stochastic phenomenon,”
Psychological Science
http://dx.doi.org/10.1111/j.14679280.1996.tb00355.x
7
(3)
186 
189
DOI : 10.1111/j.14679280.1996.tb00355.x
Kim S.W.
,
Choi S. K.
,
Bae Y. C.
,
Park Y. H.
2014
“Chaotic phenomena in MEMS with duffing equation,”
Journal of the Korea Institute of Electronic Communication Sciences
http://dx.doi.org/10.13067/JKIECS.2014.9.6.711
9
(6)
711 
717
DOI : 10.13067/JKIECS.2014.9.6.711
Bae Y. C.
2013
“Diagnosis of power supply by analysis of chaotic nonlinear dynamics,”
Journal of the Korea Institute of Electronic Communication Sciences
http://dx.doi.org/10.13067/JKIECS.2013.8.1.113
8
(1)
113 
119
DOI : 10.13067/JKIECS.2013.8.1.113
Kim M. M.
,
Bae Y. C.
2014
“Mathematical modelling and chaotic behavior analysis of cyber addiction,”
Journal of Korean Institute of Intelligent Systems
http://dx.doi.org/10.5391/JKIIS.2014.24.3.245
24
(3)
245 
250
DOI : 10.5391/JKIIS.2014.24.3.245
ROSSI C.
2002
“The role of dynamic modeling in drug abuse epidemiology,”
Bulletin on Narcotics
54
(12)
33 
44
CastilloGarsow C.
,
JordnSalivia G.
,
RodriguezHerrera A.
“Mathematical models for the dynamics of tobacco use, recovery, and relapse [BU1505M]
Available