Contents

1 Introduction 4

2 A giving up smoking model 6

2.1 Representation of smoking model as system of ODEs . . . . . . . . . . . . . . . . 6
2.2 Positivity of solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Boundedness of solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Smoking free and smoking present equilibrium points . . . . . . . . . . . . . . . . 10
2.5 Reproduction number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.6 Local stability analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.7 Global stability analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3 Numerical solution of mathematical model of smoking 22

3.1 Fourth order Runge-Kutta method for initial value problem . . . . . . . . . . . . 22
3.2 RK4 method for Smoking model . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.3 Graphical representation and discussion on numerical solution . . . . . . . . . . 25
3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

1
 Abstract
In this project, we represent the analysis of a giving up smoking model [25]. First, we
present the positivity of the solution of smoking model. Then, we show boundedness of solution
of smoking model. We present the smoking free and smoking present equilibrium of the smoking
model. We calculate the basic reproduction number R0 of the smoking model and discuss the
local stability analysis for smoking free and smoking present equilibrium of the model. We also
consider the global stability analysis of smoking model. We calculate the numerical solution of
smoking model by using RK4 method. We present the numerical results of di¤erent classes of
smoking model with the help graphs.

3
Chapter 1

Introduction

Among the natural disasters epidemic diseases kill many people in di¤erent eras. Some of these
diseases are now curable/controllable and scientists have made enormous e¤orts to control such
diseases. For example, for the polio control free vaccination is provided to kids of …ve years.
These epidemics are natural and could be treated up to some extent. Dengue, malaria, small
pox, etc. are the epidemic for which the pharmacists have manufactured some medicines.
Furthermore, scientists are struggling continuously to formulate more economical and e¤ective
antidotes.
In the thick of these widespread diseases, tobacco epidemic is one of the huge health haz-
ards the world has ever faced. This is among those threats which are generated by humans
themselves. Moreover, it is under the control of humans and hence by suitable strategies and
e¤orts it is possible to curb this intimidation.
Smoking kills up to 50 percent of its users. Nearly, each year 6 million people die from
smoking, of whom more than 5 million are users and ex-users and more than 0:6 million are
nonsmokers exposed to passive smoking. People who die prematurely deprive their families of
income, consumes family budget, raise the cost of health care and retard economic development.
Death statistic reveal, there is a death caused by tobacco every second; ten percent of the adult
population die because of smoking related diseases. Due to the increase in smokers community,
tobacco use is also a fetal threat which need to be curbed.
Mathematical modeling is a way to study the dynamics of many real world problems. It
is the pro…ciency of converting problems from an application area into tractable mathemati-

4
cal formulations whose theoretical and numerical analysis provide insight and guidance which
are useful for the originating application. A mathematical model is an abstract model that
uses mathematical language to describe the behavior of a system. In mathematical modeling
recently many people have contributed signi…cantly. Various infectious diseases are modeled
mathematically. Similarly the dynamics of smoking has modeled mathematically. Furthermore,
in order to perceive the spread of smoking and to predict the impact of smoking on society and
to reduce the number of smokers, it is worth noticing to study the dynamically aspects of
smoking in a community.
In order to identify the spreading rate of smoking using mathematical tools, researchers
have introduced di¤erent classes of smokers. In 1997, Castillo-Garrsow at all.[5]presented a
general epidemiological model to describe dynamically aspect of giving up smoking model. The
Arthur assumed PSQ model with a total constant population divided into three categories:
potential smoker P., smokers S. and quite smokers Q. Moreover, they considered the aspect of
peer pressure, relapse, counseling and treatment. Subsequently, in 2008 Sharomi and Gumel
[28], introduced a new category Qr of smokers who temporarily give up smoking.
In epidemiology, incidence is the probability of occurrence of a speci…c medical condition in a
population within a speci…ed period of time. In the transmission of smoking epidemics incidence
rate plays a vital role. In order to explain the dynamical aspects of various individual in the
community, researchers have incorporated bi-linear and saturated incidence rates [5],[4],[18]. In
a …nite time these aforementioned incidence rates do not help to show the credibility of having
the smokers ers population tends to extinction. For this purpose the authors in [7] incorporated
square root type incidence rate, which was introduced by Mickens [18].
In the project, we review the study of Rahman etal [25]. In this chapter 2 of this project,
we present the positivity of the solution of smoking model [25].We show the disease free and
endamic equilibrium points of the model. We discuss the local and global stability analysis of
the smoking model at disease free and endamic equilibrium points. In chapter3, we use RK4
method to calculate the numerical solution of the smoking model. We present the solution by
ploting the graphs. All the results of this project report are taken from [25].

5
Chapter 2

A giving up smoking model

In this chapter, we consider and study the analysis of a smoking model [25]. Assume that the
total population size at any time t is represented by N (t). The population N (t) is partitioned
into four classes P (t) (potential smokers), L(t) (occasional smokers), S(t) (chain smokers) and
Q(t) (quit smokers). All the results of this chapter are taken from the model [25].

2.1 Representation of smoking model as system of ODEs

In this section, we present the smoking model [25] as system of ODEs.

The governing equations for the given up smoking model are given as:

dP
= W (L; P ) (d + )P;
dt
dL
= W (L; P ) ( + d + )L;
dt
dS
= L ( + d + )S; (2.1)
dt
dQ
= S ( + d)Q:
dt

With the initial conditions

P (0) = P0 0; L(0) = L0 0

6
 S(0) = S0 0; Q(0) = Q0 0

Furthermore, W (L; P ) is a function. This function relates the incidence of potential smoker
with occasional smoker. In this model, the interaction of potential and occasional smokers in
the following form:

2P L
W (P; L) := : (2.2)
P +L

Moreover, the function de…ned in (2.2) represents the harmonic mean of potential and
occasional smokers. For many investigation the uptake function W (P; L) is taken to be the
product of potential and occasional smokers, which is known as bi–linear incidence rate [13]:The
authors of [7] discussed the dynamics of giving up smoking model while considering the uptake
function to be the square root of P and L. Beside these, it is very fascinating to consider the
2P L
uptake function to be W (P; L) = P +L : Since it is known that if P and L are two quantities
such that P; L 0, then the following relation holds true:

2P L p P +L
PL :
P +L 2

2.2 Positivity of solution

In this section, we present the positivity of solution of the giving up smoking model (2.1).
Remark 2.1 [25]
It is worth noticing that the system ODEs (2.1) satis…es the following conservative law:

dn
= ( + d)N: (2.3)
dt

where N (t) denotes total population size at any time t with N (t) = P (t)+L(t)+S(t)+Q(t):
Eq.(2.3) is an exact di¤erential equation and has exact solution

N (t) = + (N0 )e ( ( + d)t) : (2.4)

( + d) ( + d0

with the initial condition

7
 N (0) = P (0) + L(0) + S(0) + Q(0): (2.5)

In addition to the above, we have

P (0) > 0; L(0) > 0; S(0) > 0; Q(0) > 0; (2.6)

! P (t) > 0; P (t) > 0; P (t) > 0; P (t) > 0:

Therefore, it is observed that the solution is positive de…nite.

Remark 2.2:[25]
It must be noted that the characteristic of the system are essentially described by the
P
nonlinear term U (P ) = P +L: In the literature this function U : [0; 1) ! [0; 1); is know as
uptake function and satis…es some properties, which are stated as [24]:
1:U (0) = 0; U (P ) > 0, for P > 0;
P (t)
2: limp!1 P (t)+L(t) = L1 ; 0 < L1 < 1; where L1 = 1;
3:U (P ) is continuously di¤erentiable;
dU L
4: dt = (P +L)2
> 0;implies that U is monotonically increasing.
Hence, the function U makes sense.
Governing equations of the model. Taking into account Eq.(2.2) system (2.1). we obtain
the following form:

dP 2P L
= (d + )P; (2.7)
dt P +L
dL 2P L
= ( + d + )L;
dt P +L
dS
= L ( + d + )S;
dt
dQ
= S ( + d)Q;
dt

along the conservative law

8
 dN
= ( + d)N; (2.8)
dt

N (t) = P (t) + L(t) + S(t) + Q(t): (2.9)

2.3 Boundedness of solution

In this section, we show the boundedness of solution of the giving up smoking model.
Theorem 2.1 [25]
All feasible solutions of the system (2.7) is bounded for non-negative initial condition that
are not identically zero on any interval and enter the region

4
= f(P; L; S) 2 R+ :P +L+S+Q g:

Proof. Assume 4 g be any solution of the system (2.7) with non-

= f(P; L; S; Q) 2 R+
negative initial conditions. As N (0) = P (0) + Q(0) + S(0) + L(0) as well as N (0) > for any
non-negative initial values of P; L; S; Q the solution is bounded below by 0, Moreover, adding
four equations of system (2.7) we have

d
(P (t) + L(t) + S(t) + Q(t)) = ( + d)N N:
dt

It follows

t
0 N (t) + N0 e : (2.10)

where N0 represents initial value of the total population. Thus, 0 N (t) ; as t ! 1:

Therefore, it is inferred that all feasible solutions of the system (2.7) enter the region =
4 : P +L+S+Q
f(P; L; S; Q) 2 R+ g: Hence, is positively invariant. In order to study
the system (2.7) it is su¢ cient to consider the set :

9
2.4 Smoking free and smoking present equilibrium points

In this section, we present the smoking free and smoking present equilibrium points of the
giving up smoking model.
By putting the values P = P 0 ; L = 0; S = 0; Q = 0; in the following equations:

dP
= W (0; ) (d + ) = 0;
dt d+ d+
dL
= W (0; ) ( + d + ) = 0;
dt d+
dS
= (0) ( + d + )(0) = 0;
dt
dQ
= S ( + d)(0) = 0:
dt

We get the smoking-free equilibrium E 0 = (P 0 ; L0 ; S 0 ; Q0 ) = ( d+ ; 0; 0; 0):

The system of ordinary di¤erential equation given by system (2.7) has non-trivial solution.
Equating the right hand side of each equation of the system (2.7) equal to zero, we obtain the
following smoking present equilibrium (SPE) point, by putting the values P = P ; L = L ; S =
S ; Q = Q , in the following equations:

dP
= W (0; ) (d + ) = 0;
dt d+ d+
dL
= W (0; ) ( + d + ) = 0;
dt d+
dS
= (0) ( + d + )(0) = 0;
dt
dQ
= S ( + d)(0) = 0:
dt

We get the smoking-present equilibrium point E 0 = (P ; L ; S ; Q ); with

P = ; (2.11)
(2 )

10
 (R0 1)( + d + )P
L = ; (2.12)
(d + + )

(R0 1)( + d + )P
S = ; (2.13)
( + d)2 + ( + )( + d) +

and
(R0 1)( + d + ) P
Q = : (2.14)
( + d)[( + d)2 + ( + )( + d) + ]

2.5 Reproduction number

In this section, we calculate reproduction numbers of the giving up smoking model (2.7).
The smokers generation number R0 is obtained by the method, described in [8]. Assume
x = (L; S), then system (2.7) will yield.

dx
=F V;
dt

where 0 1
2 PL
F =@ P +L A;
0

and

0 1
(& + d + )L
V =@ A;
&L + ( + d + )S

here F and V are Jacobians of F and V at E 0 ;respectively, therefore,

2 3
2 0
F =4 5;
0 0

and

2 3
+d+
V =4 5;
0 +d+

11
 Furtherrmore, we have

2 3
1
V 1
=4 +d+ ( +d+ )( +d+ ) 5
: (2.15)
0 0

thus, one can obtain the next generation matrix as

1
G = FV ;

2 3
2 2
G=4 +d+ ( +d+ )( +d+ ) 5
;
0 0

Now we …nd the eigen values of matrix G:Consider

2 3
1 0
I=4 5;
0 1

denotes the identity matrix, and be the we have Now, we multiply with identity matrix

2 3 2 3
1 0 0
I= 4 5=4 5;
0 1 0

Now, subtract the matrix G from I

2 3
2 2
G I=4 +d+ ( +d+ )( +d+ ) 5
;
0

Now, we taking determinant of the following matrix

2 2
p( ) = det(G I) = + = 0;
+d+

2
2 + ( +d+ )
= 0;
+d+

2
2 + ( + d + ) = 0;

12
 ( 2 + ( + d + ) = 0:

Case 1:

1 = 0;

Case 2:

2 + ( + d + ) = 0;

( +d+ )=2 ;

2
2 = ;
( +d+ )

2
R0 = ( ) = max(j1 j; j 2 j) = max j0j; ;
( +d+ )

Hence the value of R0

2
R0 = (G) = :
( +d+ )

Now and hence it is concluded that the smoking generation number R0 is given by R0 =
2
(G) = +d+ ;where denotes spectral radius of the matrix G.

2.6 Local stability analysis

In this section, we will calculate the local stability analysis at smoking free equilibrium of giving
up smoking model [25]
Theorem 2.2.[25] (Local stability of E 0 )
If R0 < 1; then the smoking -free equilibrium of system (2.7) is locally asymptotically stable.

13
If R0 = 1; E 0 is locally stable and unstable if R0 > 1:
Proof:
In order to investigate the local stability of E0 ;linearization method is used to evaluate the
Jacobian matrix at E 0 ; which is given by

02 31
d 2 0 0
B6 7C
B6 7C
B6 0 2 d 0 0 7C
J =B
B6
6 7C ;
7C (2.16)
B6 0 & d 0 7C
@4 5A
0 0 d

Thus, the eigenvalues of J(E 0 ) are 1 = d < 0; 2 = (R0 1)( + d + ) and

3 = 4 = d < 0:
Since 1; 3 and 4 are negative.
Now if R0 < 1; then 2 is also negative and thus all eigenvalues of (2.16)are negative i,e E 0
2
is sink and hence E 0 is locally asymptotically stable. Now if R0 = 1; then = 0 and E 0 is
locally stable. Now if R0 > 1 then 2 > 0 which means that there exist a positive eigenvalue of
(2.16), hence E 0 is unstable in this case.

2.7 Global stability analysis

In this section, we study the global stability analysis of giving up smoking model (2.7). The
method of Castillo-Chavez is used to prove the global asymptotically stable at smoking free
equilibrium. First, we provide a brief analysis of the Castillo-Chavez method and geometrical
approach to prove the global stability of the model (2.7) smoking free and smoking present
equilibrium points. The smoking model (2.7) will be reduced into the following two subsystems:

dx1
= G(x1 ; x2 );
dt

dx2
= H(x1 ; x2 ): (2.17)
dt

In the system (2.17), x1 and x2 represent the number of nonsmokers and smokers individuals,

14
that is, x1 = (P (t); Q(t)) 2 R2 and x2 = (L(t); S(t)) 2 R2 . The smoking free equilibrium
point E0 is de…ned as E0 = (x01 ; 0): Thus, the existence of the global stability at disease free
equilibrium point depends on the following two conditions
dx1 0
C1 . If dt G(x1 ; 0); x1 is globally asymptotically stable.
C2 .H(x1 ; x2 ) = Bx1 H(x1 ; x2 ); where H(x1 ; x2 ) 0 for (x1 ; x2) 2 :
In second condition, B = Dx2 H(x01 ; 0) is an M-matrix that is the o¤ diagonal entries are
positive and represents the required feasible region. Then the following statement hold;
Lemma 2.1.[25]
If R0 < 1 then the equilibrium point E0 = (X 0 ; 0 ) of the system (2.7) is said to be globally
asymptotically stable, if the conditions C1 and C2 are satis…ed.
Theorem 2.3.[25]
Let R0 < 1, then model (2.7) is globally asymptotically stable at smoking free equilibrium
E0 .
Proof.
To prove the results it need to check the condition, C1 and C2 : Let X1 = (P (t); Q(t))
represents the numbers of nonsmokers classes and X2 = (L(t); S(t)) represent the number of
smokers classes. De…ne E0 = (X0 ; 0), where

X10 = ( 0): (2.18)

+d

Now using the proposed model (2.7), we have

dx1
= G(x1 ; x2 );
dt

dx1 2 PL
=( ( + d)P ): (2.19)
dt P +L

For P = P0 ; L = 0; S = 0; Q = 0 and G(X1 ; 0) = 0; Eq (3:17) becomes

G(X1 ; 0) = ( ( + d)P0 ): (2.20)

Therefore, if t ! 1, then X1 ! X10 . Hence, X1 ! X10 is globally asymptotically stable.

15
For the second condition, …rst it need to express H(X1 ; X2 ) in the form, H(X1 ; X2 ) = BX1
H(X1 ; X2 ):
So,

2 32 3 2 3
2 P0
( +d+ ) 2 P0 L 2 P0 S
BX1 H(X1 ; X2 ) = 4 54 5 4 P +L 5
; (2.21)
( +d+ ) S 0

The total population is bounded by P0; that is P; L; S; Q P0 ; soP P0 S; implies that

2 PL 2 [P P0 + L(P0 S P]
2 P0 S = > 0:
P +L P +L

Therefore, H(x1; x2 ) is positive de…nite. Also, it is clear that the matrix B is M-matrix
that is the o¤ diagonal element are non-negative. Thus condition C1 and C2 are satis…ed, so
by Lemma 3:1, the smoking free equilibrium E0 is globally asymptotically stable. To prove
the global stability of the model [?] at smoking present equilibrium E1 , we use the geometrical
approach 2[16]0. According to this method, we will investigate su¢ cient condition through
which the smoking present equilibrium point is globally asymptotically stable. To do this, we
consider the following

x_ = f (x) (2.22)

where f : U ! Rn ; U Rn is an open set simply connected and f 2 C 1 (U ). Let us

assume that the solution to Eq. (2.22) is f (x ) = 0 and for x(t; x0) the following hypothesis
hold there exist a compact absorbing set K 2 U . System (2.22) has a unique equilibrium.
The solution x is said to be globally asymptotically stable in U , if it is locally asymptotically
stable and all trajectories in U converges to the equilibrium x . For n 2, a condition satis…ed
for f , which precludes the existence of non-constant periodic solution of Eq. (2.22) known is
Bendixson criteria. The classical Bendixson criteria div f (x) < 0 for n = 2 is robust under C 1
[?]. Furthermore, a point x0 2 U is wandering for Eq. (2.22) , if there exist a neighborhood

16
N of x0 and > 0, such that N \ x(t; N ) is empty for all t > . Thus the following global
stability principle established for autonomous system in any …nite dimension.

Lemma 2.2.[25]
If the conditions (C1 ) (C2 ) and Bendixson criterion satis…ed for Eq. (2.22) that is robust
under C1 local perturbation of f at all non equilibrium and non wandering point for Eq (2.22).
Then x is globally asymptotically stable in U provided it is stable. Now to prove the robustness
required for Lemma 3:2: Let us de…ne a function, such that

0 1
n n
P (x) = @ A: (2.23)
2 2

Eq (2.23) is a matrix valued function on U . Further, assume that P 1 exist and is continuous
for x 2 K. Now we de…ne a quantity,

Z t
1
q = lim sup sup [ (B(x(s; x0 )))]ds; (2.24)
t!1 t 0

where B = Pf P 1 + P J [2] P 1 and J [2] is the second additive compound matrix of the
Jacobian matrix J, that is J(x) = U f (x).
Let
l(B) be the Lozinski measure of the matrix B with respect to the norm jj:jjinRn [19] de…ned
by

jI + Bxj 1
l(B) = limx !0 : (2.25)
x

Hence, q < 0;shows the presence of any orbit that give rise to a simple closed recti…able
curve, such as periodic orbits and hetero cyclic cycles.
Lemma 2.3.[25]
Let U is simply connected and the condition (C1 ) (C2 ) are satis…ed, then the unique
equilibrium x of Eq. [?] is globally asymptotically stable inU , if q < 0:
Theorem 2.4.[25]
If R0 > 1, then the smoking present equilibrium of the model (2.7) is globally asymptotically

17
stable in .
Proof.
Let us assume a subsystem of (2.7) such that

dP 2 PS
= (d + )P; (2.26)
dt P +L
dL 2 PS
= ( + d + )L;
dt P +L
dS
= L ( + d + )S;
dt

;Let J be the Jacobian matrix of the system (2.26) given as

2 3
2 L2 2 P2
(P +L)2
( + d) (P +L)2
0
6 7
6 2 L2 2 P2 7
6 (P +L)2
( +d+ )+ (P +L)2
0 7: (2.27)
4 5
0 ( +d+ )

The second additive compound matrix of J is denoted by J j2j and de…ne as.

2 3
P L
2( + d) +2 0 0
6 P +l 7
6 2 L2 2 p2 7
J j2j = 6 2( + d) (P +L)2 (P +L)2
7:
4 5
2 L2 2 p2
0 (P +L)2 (P +L)2
2( + d) :
(2.28)
Choose a function P (x) = P (P; L; S) = diagf1; L
S:
L
S g, which implies that P 1 (x) =
S S
diagf1. L ; L g , then taking the derivative with respect to time, that is Pf (x);we get

L_ LS_ L_ LS_
P f (x) = diagf0; ; ; g: (2.29)
S S2 S S2
_ S_ L_ S_
Now Pf (P ) 1 = diagf0; L
S S; L Sg and P J j2j (P ) 1 = J j2j : Thus we take B =
Pf (P ) 1 + P J j2j (P ) 1, which can be written as

2 3
B11 B12
B=4 5; (2.30)
B21 B22

18
 where the entries of B is given as;

2 (P L)
B11 = 2( + d);
P +L

B12 = (0 ; 0);

2 3
L
B21 = 4S 5;
0
2 3
W11 W12
B22 =4 5:
W21 W22

L_ S_ 2 L2 2 P2 L_ S_ 2 P2
with w11 = L S 2( + d) (P +L)2
; W12 = (P +L)2
W21 = W22 = L S + (P +L)2

2( + d)
Let (a1; a2 ; a3 ) be a vector in R3 and its norm jj:jj de…ned by

jj(a1; a2 ; a3 )jj = maxfjja1 jj; jja2 jj + jja3 jjg:

Moreover, assume l(B) to be the Lozinski measure with respect to the norm given by (??)
and described by Martin [19], then we get

l(B) supfg1 ; g2 g = supfl(B11 ) + jjB12 jj; l(B22 ) + jjB21 jjg; (2.31)

where jjB12 jj jjB21 jj are matrix norms,then

g1 = l(B11 ) + jjB12 jj; g2 = l(B22 ) + jjB21 jj; (2.32)

_ S_ _
where l(B11 ) = 2( + d) +2 (P L)
P +L ; jjB12 jj = 0; l(B22 ) =maxf L
L S 2( + d) L
;L
S_
S 2( + d) g
and

19
 L L
jjB21 jj = maxf ; 0g = :
S S

Putting these values of l(B11 ); jjB12 jj; ljjB22 jj and jjB21 jj in (2.32) , we have

2 (P L)
g1 = 2( + d) + ;
P +L

L_
g2 = ( + d):
L

Using the second equation of system (2.7) we have the following relation between g1 and g2 .

L_ 2 (P L)
g2 g1 = ( + d) 2( + d) + ;
L P +L

2 L 2 (P L)
= 2( + d) + 2( + d) + ;
P +L P +L

2 L
= > 0:
P +L

Hence, g2 g1 > 0: Thus, the Lozinski measure of B is estimated as

L
l(B) supfg1 ; g2 g = ( + d);
L

along every solution (P (t); L(t); S(t)) of the system with (P (0); L(0); S(0)) 2 K1 , where K1
is compact absorbing set. Now one can obtain that

Z t
1 1 L(t)
l(B)dL log ( + d)
t 0 t L(0)

Z t
1
q = lim supsup l(B)dt < ( + d) (2.33)
t!1 t 0

Thus, by Theorem3:1 [16] , we obtain the global asymptotic stability at equilibrium (P1 ; L1 ; S1 ).
Now consider the last equation of the model (2.7), such that

20
 dQ(t)
= S ( + d)Q(t);
dt

and its limit equation is

dQ(t)
= S ( + d)Q(t): (2.34)
dt

Solving (2.34) we have

S Q(0)
Q(t) + +d ;
+d e t

which implies that

S
Q(t) ! = Q ast ! 1:
+d

Thus, we obtain that the system (2.7) is globally asymptotically stable at smoking present
equilibrium point.

2.8 Summary

In this chapter, we have considered the analysis of a giving up smoking model [25]. First, we have
considered the positivity of the solution of smoking model. Then, we showed boundedness of
solution of smoking model. We have presented the smoking free and smoking present equilibrium
of the smoking model. We calculated the basic reproduction number R0 of the smoking model
and discussed the local stability analysis for smoking free and smoking present equilibrium of
the model. We also considered the global stability analysis of smoking model.

21
Chapter 3

Numerical solution of mathematical

model of smoking

In this chapter, we calculate the numerical solution of smoking model (2.7) by using RK4
method.

3.1 Fourth order Runge-Kutta method for initial value problem

Consider a …rst order di¤erential equation

dx(t)
= (x(t)); x(t0 ) = x0 : (3.1)
dt

Thus, the approximation of the Eq (3.1) with RK4 (Runga Kutta) method of 4th order is
given by the following formula

h
xm+1 = xm + (k1 + 2k2 + 2k3 + k4); m = 0; 1; 2; :::;
6

where

tm+1 = tm + h;

22
 k1 = f (tm ; xm );

k2 = f (tm + h=2; xm + h=2k1 ;

k3 = S(tm + h=2; xm + h=2

k4 = f (tm + h; xm + hk3 )

The parameter’s values are taken in such a way that they are biologically feasible. Precisely,
the following set of parameters
is considered.

3.2 RK4 method for Smoking model

Consider the system of ODEs given in smoking model (2.7).

dP
= W ((0; )) (d + )P;
dt d+
dL
= W ((0; )) ( + d + )L;
dt d+
dS
= L ( + d + )S;
dt
dQ
= S ( + d)Q:
dt

Thus, the approximation of the with RK4 (Runga Kutta) method of 4th order is given by
the following formula

h
xm+1 = xm + (k11 + 2k12 + 2k13 + k14 ); m = 0; 1; 2; :::;
6

23
 h
xm+1 = xm + (k21 + 2k22 + 2k23 + k24 ); m = 0; 1; 2; :::;
6

h
xm+1 = xm + (k31 + 2k32 + 2k33 + k34 ); m = 0; 1; 2; :::;
6

h
xm+1 = xm + (k41 + 2k42 + 2k43 + k44 ); m = 0; 1; 2; :::;
6

where

tm+1 = tm + h;

2 3
k11 = W (L; P ) (d + )P (tm ; xm );
6 7
6 7
6k12 = W (L; P ) (d + )P (tm + h=2; xm + h=27
6 7
6 7
6k13 = W (L; P ) (d + )P (tm + h=2; xm + h=27
4 5
k14 = W (L; P ) (d + )P (tm + h; xm + h)

2 3
k21 = W (L; P ) ( + d + )L(tm ; xm )
6 7
6 7
6 k22 = W (L; P ) ( + d + )L(tm + h=2; xm + h=2 7
6 7
6 7
6k23 = W (L; P ) ( + d + )L(tm + h=2; xm + h=2k2 )7
4 5
k24 = W (L; P ) ( + d + )L(tm + h; xm + hk3 )
2 3
k31 = L ( + d + )S(tm ; xm )
6 7
6 7
6k32 = L ( + d + )S(tm + h=2; xm + h=27
6 7
6 7
6k33 = L ( + d + )S(tm + h=2; xm + h=27
4 5
k34 = L ( + d + )S(tm + h; xm + h)
2 3
k41 = S ( + d)Q(tm ; xm )
6 7
6 7
6k42 = S ( + d)Q(tm + h=2; xm + h=2)7
6 7
6 7
6k43 = S ( + d)Q(tm + h=2; xm + h=2)7
4 5
k44 = S ( + d)Q(tm + h; xm + h)

24
3.3 Graphical representation and discussion on numerical so-
lution

we use the following parameter values:

= 0:25; = 0:006; d = 0:00004; = 0:08; = 0:02 & = 0:000274:
Graphical solution between potential smoker and time(days), shows when time is 0 the
value of potential smoker is approaches to 10. In the beginning the value of potential smoker
is maximum. With the passage time(days) the value of potential smoker is decreasing. After
30 days value of potential smoker is not change much.

10

8
Potential Smokers P(t)

2
0 50 100 150 200
Time(days)

Figure 2.1. Graph of potential smokers P(t) for 200 days

The plot shows the population of potential smokers both with and without control.
Graphical solution between occasional smoker and time(days), shows when time is 0 the
value of occasional smoker is approaches to 80. In the beginning the value of occasional smoker
is maximum. With the passage time(days) the value of occasional smoker is decreasing. After
50 days value of occasional smoker becomes zero and remain constant.

25
 80

70

60

Occasional Smokers L(t)

50

40

30

20

10

0
0 50 100 150 200
Time(days)

Figure 2.2. Graph of occasional smokers L(t) for 200 days

Graphical solution between chain smokers and time(days), shows when time is 0 the value
of chain smoker is approaches to 24. In the beginning the value of chain smoker is minimum.
With the passage time(days) the value of chain smoker is increasing. After 50 days value of
chain smoker does not change and remain constant.

28.5

28

27.5

27
Chain Smokers S(t)

26.5

26

25.5

25

24.5

24
0 50 100 150 200
Time(days)

Figure.2.3. Graph of chain smokers S(t) for 200 days

Graphical solution between quite smokers and time(days), shows when time is 0 the value
of quite smokers is approaches to 180. In the beginning the value of quite smoker is maximum.
With the passage time(days) the value of quite smoker is decreasing. After 50 days value of

26
quite smoker becomes zero and remain constant.

3.4 Summary

In this chapter, we calculated the numerical solution of smoking model by using RK4 method.
We presented the numerical results of di¤erent classes of smoking model with the help graphs.

27
Bibliography

[1] Alkhundhri, Z., Al-Sheikh, S. & Al-Tuwairqi, S. (2014). Global dynamics of a mathematical
model on smoking, ISRN Appl. Math. 2014, 847075.

[2] Ahlbom, A. (1993). Bio-statistics for Epidemiologists, CRC Press.

[3] Castillo-Garsow, .C, Jordan-Salivia, .G & Herrera, A.R. (1997). Mathematical models for
the dynamics of tobacco use, recovery and relapse, Technical Report Series, BU-1505,
Ithaca NY, USA.

[4] Capasso, V. & Serio, G. (1978). A generalization of the Kermack–Mckendrick deterministic

epidemic model, Math. Biosci. 42, 43–61.

[5] Castillo-Chavez, C., Feng, Z. & Huang, W. (2002). Mathematical approaches for emerging
and reemerging infectious diseases: an introduction, in: Proceedings of the IMA vol. 125,
pp. 229–250, Springer-Verlag, Berlin-Heidelberg New York.

[6] Durkin, S., Brennan, E. & Wake…eld, M. (2012). Mass media campaigns to promote smok-
ing cessation among adults: an integrative review, Tob. Control, 21 (2) 127–138, doi:
10.1136/tobaccocontrol- 2011- 050345.

[7] Din, Q., Ozair, M., Hussain, T. & Saeed , U. (2016). Qualitative behavior of a smoking
model, Adv. Di¤ er. Equ. 2016, 96.

[8] Driessche, P.V.D. & Watmough, J. (2002). Reproduction numbers and sub-threshold en-
demic equilibria for compartmental models of disease transmission, Math. Biosci. 180,
29–48.

28
 [9] Fleming, W.H. & Rishel, R.W. (1975). Deterministic and Stochastic Optimal Control,
Springer-Verlage, New York Inc.

[10] Flynn, B.S. & al , E .(1994). Mass media and school interventions for cigarette smoking
prevention: e¤ects 2 years after completion, Am. J. Public. Health 84, 1148–1150.

[11] Gumel, A.B., Shivakumar, P.N. & Sahai, B.M. (2001). A mathematical model for the
dynamics of HIV-1 during the typical course of infection, Nonlinear Anal. 47, 1773–1783.

[12] Hethcote, H.W. (2000). The mathematics of infectious diseases, SIAM Rev. 42, 599–653.

[13] Huo, H.F. & Zhu, C.C. (2013). In‡uence of relapse in a giving up smoking model, Abstr.
Appl. Anal. 525461 . 12 pages.

[14] Kumar, D., Agarwal , R.P., & Singh, j. (2018). A modi…ed numerical scheme and con-
vergence analysis for fractional model of Lienards equation, J. Comput. Appl. Math. 339,
405–413.

[15] Li, M.Y. & Muldowney, J.S. (2006). A geometric approach to global-stability problems.
SIAM J. Math. Anal. 27,1070–1083.

[16] Li, L. & al, E. (2010). Predictors of smoking cessation among adult smokers in malaysia and
thailand: …ndings from the international tobacco control southeast asia survey, Nicotine
Tob. Res. 12 , S34–S44.

[17] Luo, B., Wan, L., Liang, L. & Li, T.S. (2015). The e¤ects of educational campaigns and
smoking bans in public places on smokers’ intention to quit smoking: Findings from 17
cities in china , Biomed. Res. Int. 853418 . 9 pages.

[18] Mickens, R.E. (2010). SIR model with square-root dynamics: an NSFD scheme, J. Di¤ er.
Equ. Appl. 16, 209–216.

[19] Martin, R.H. (1974). Logarithmic norms and projections applied to linear di¤erential sys-
tems. J. Math. Anal. Appl. 45 432–454.

[20] Pang, L., Zhao, Z., Liu, S. & Zhang, X. (2015). A mathematical model approach for tobacco
control in china, Appl. Math. Comput., 259, 497–509.

29
[21] Rahman, G.U., Agarwal, R.P., Liu, L. & Khan, A. (2018). Threshold dynamics and optimal
control of an age-structured giving up smoking model, Nonlinear Anal. RWA. 43, 96–120.

[22] Rakocevi, M.M. (2004). A harmonic structure of genetic code , J. Theor. Biol. 229, 221–
234. doi: 10.1016/j.jtbi.03.017.

[23] Rakocevic, M.M. (2013). Harmonic mean as a determinant of the genetic code.

[24] Rao, V.S.H. & Rao, P.R.S. (2009). Dynamic Models and Control of Biological Systems,
Springer-Verlang.

[25] Rahman, G.U., Agarwal, R.P. & Din, Q. (2019). Mathematical Analysis of a Giving Up
Smoking model via harmonic mean type incidence rate. Appl. Math. Comput. 354 , 128-148.

[26] Singh, J., Kumar, D. & Baleanu, D. (2017). On the analysis of chemical kinetics system
pertaining to a fractional derivative with Mittag-Le- er type kernal. Chaos. 27, 103-113.

[27] Singh, j., Kumar, D., Hammouch, Z. & Atangana,A. (2018). A fractional epidemiological
model for computer viruses pertaining to a new fractional derivative. Appl. Math. Comput.
316, 504–515.

[28] Sharomi, O. & Gumel, A.B. (2008). Curtailing smoking dynamics: a mathematical mod-
eling approach, Appl. Math. Comput. 195 , 475–499.

[29] Wake…eld, M.A., Loken, B. & Hornik, R.C. (2010). Use of mass media campaigns to change
health behavior, Lancet 376 (9748) 1261–1271, doi: 10.1016/ S0140- 6736(10)60809- 4.

[30] Zaman, G. (2011). Qualitative behavior of giving up smoking models, Bull. Malays. Math.
Sci. Soc. 34, 403–415.

30