Pure Appl. Biol.

, 9(1): 674-682, March, 2020

http://dx.doi.org/10.19045/bspab.2020.90073

Research Article

Analysis and numerical solution of

epidemic models by using nonstandard
finite difference scheme
Muhammad Farman1*, Aqeel Ahmad1, Muhammad Umer Saleem2 and
Muhammad Ozair Ahmad1
1. Department of Mathematics and Statistics, University of Lahore-Pakistan
2. Division of Science and Technology, University of Education, Lahore-Pakistan
*Corresponding author’s email: [email protected]
Citation
Muhammad Farman, Aqeel Ahmad, Muhammad Umer Saleem and Muhammad Ozair Ahmad. Analysis and
numerical solution of epidemic models by using nonstandard finite difference scheme. Pure and Applied Biology.
Vol. 9, Issue 1, pp674-682. http://dx.doi.org/10.19045/bspab.2020.90073
Received: 13/06/2019 Revised: 16/12/2019 Accepted: 17/12/2019 Online First: 26/12/2019
Abstract
In this work, nonlinear epidemic models for measles, dengue fever, smoking, and SVIR are
considered in order to understand the outbreaks of this epidemic disease. Developed
unconditionally convergent nonstandard finite difference scheme by applying Mickens approach
with ∅(ℎ) = ℎ + 𝑂(ℎ2 ) as this will provide a feasible solution to overcome its disease risk as these
are the main requirement for the dynamical systems in real-life solution. This method proved to
be a very efficient technique for solving epidemic models to control infection diseases. Numerical
simulations are carried out to support the analytic results.
Keywords: Epidemic model; Non-standard finite difference scheme; Qualitative analysis;
Stability analysis
Introduction types of diseased models are often called
In the human population, epidemiology infectious diseases (i.e. the disease which
studies play an important role to understand transferred from one person to another
the disease in the human population. Often person). Measles, rubella, chickenpox,
the work of mathematical epidemiology mumps, aids and gonorrhea syphilis are
consists of model building, estimation of examples of infectious disease [4, 5].
parameters and investigation of the The Nonstandard Finite Difference (NSFD)
sensitivity of models to change in the schemes preserve the main properties of the
parameters and numerical simulations. differential counterparts, such as positivity,
Epidemiologists use mathematical models to monotonicity, periodicity, stability, and some
understand previous outbreaks of diseases other invariant including energy and
and to better understand the dynamics of how geometrical shapes. It should be emphasized
infections spread through populations [1]. that NSFD schemes can preserve all
The research of this kind helps to understand properties of continuous models for any
the ratio of disease spread in the population discretization parameters. The discrete
and to control their parameters [2, 3]. These

Published by Bolan Society for Pure and Applied Biology 674

 Farman et al.

models with these properties are called susceptible is increased by birth at a rate 𝜋𝑁
dynamically consistent [6-9]. where both rate 𝜋 and death rate 𝜇 are the
SIR Model same in [7], so susceptible is increased by
Kermack and McKendrick are those persons 𝜇𝑁. It is decreased by infection following
who are the inventor of the disease models contacts with infected individuals I at rate 𝛼.
and they played an important role in This class is decreased by recovery from
Mathematical epidemiology. In the purposed infection at a rate 𝛾 and dimensioned by
model, the population is divided into three natural death rate 𝜇. This generates a class R
groups which denoted by S (susceptible), I of individuals who have complete protection
(infected) and R (recovered). The class S of against disease and can be seen in (Figure 1).

Figure 1. Flow chart of SIR model

Followings are the equations of the model 𝑅, with initial conditions 𝑆(0) =
𝑑𝑆
= 𝜇𝑁 − 𝛼𝑆𝐼 − 𝜇𝑆 (1) 990, 𝐼(0) = 10, 𝑅(0) = 0
𝑑𝑡
𝑑𝐼 Qualitative analysis of the model
= 𝛼𝑆𝐼 − (𝛾 + 𝜇)𝐼 (2) By simplifying the above equations we get
𝑑𝑡
𝑑𝑅
= 𝛾𝐼 − 𝜇𝑅 (3) disease-free equilibrium denoted by 𝐸0 i.e
𝑑𝑡
Where 𝜇, 𝛼 𝑎𝑛𝑑 𝛾 represents the per capita 𝐸0 = (𝑁, 0,0) and the endemic equilibrium
removal rate, transitivity rate ant per capita point denoted by 𝐸1 i.e.
recovery rate respectively and 𝑁 = 𝑆 + 𝐼 +

(𝛾+𝜇) 𝛼𝜇𝑁−𝜇(𝛾+𝜇) 𝛼𝛾𝜇𝑁−𝜇𝛾(𝛾+𝜇) 𝛼𝑁

𝐸1 = ( , , ) and the reproductive number is 𝑅0 = (𝛾+𝜇) > 1
𝛼 𝛼(𝛾+𝜇) 𝛼𝜇(𝛾+𝜇)

Nonstandard finite difference (NSFD) the approximation of 𝑋(𝑡𝑘 ) where 𝑡𝑘 =

scheme 𝑘∆𝑡 , 𝑤𝑖𝑡ℎ 𝑘𝜖𝑁, ℎ = ∆𝑡 > 0 be a step size
In recent years, NSFD discrete models have then:
been constructed and or tested for a wide 𝑆 𝑘+1 −𝑆 𝑘
= 𝜇𝑁 − 𝛼𝑆 𝑘+1 𝐼 𝑘 − 𝜇𝑆 𝑘+1
range of nonlinear dynamical systems. In this ∅
section, we design the NSFD scheme that
𝐼 𝑘+1 −𝐼 𝑘
replicates the dynamics of a continuous = 𝛼𝑆 𝑘+1 𝐼 𝑘 − (𝛾 + 𝜇)𝐼𝑘+1
∅
model (1-3). Let 𝑌𝑘 = (𝑆𝑘 , 𝐼𝑘 , 𝑅𝑘 )𝑡 denoted

675
Pure Appl. Biol., 9(1): 674-682, March, 2020
http://dx.doi.org/10.19045/bspab.2020.90073

𝑅 𝑘+1 −𝑅𝑘 has been presented. To observe the effects of

= 𝛾𝐼 𝑘+1 − 𝜇𝑅 𝑘+1
∅ the parameters using in this dynamics
𝜇𝑁∅+𝑆 𝑘
𝑆 𝑘+1 = 1+∅𝛼𝐼𝑘 +∅𝜇 (4) measles SIR model (1-3), conclude several
(𝛼∅𝑆 𝑘+1 𝐼 𝑘 +𝐼 𝑘 )
numerical simulations with parameters
𝐼 𝑘+1 = (5) values 𝛼 = 0.003 , 𝜇 = 0.05 , 𝛾 = 1 , 𝑁 =
1+∅(𝛾+𝜇)
𝛾∅𝐼 𝑘+1 +𝑅 𝑘 1000 are given in [7]. In (Figure 2) shows
𝑅 𝑘+1 = (6) that the model presented gradually
1+∅𝜇
Which is the purposed NSFD scheme for the approaches the steady states for different
given model, where values of h. It is clearly shows that infected
1 − 𝑒 −(𝛾+𝜇)ℎ individuals at the latent period are diagnosed
∅ = ∅(ℎ) = and treated, the number of susceptible
(𝛾 + 𝜇)
Numerical simulations individuals decreases significantly, the
The mathematical analysis of the epidemic infected decreases steadily while the
measles SIR model with non-linear incidence recovered increased steadily.

Figure 2. Numerical solutions for susceptible, infected individual, recovered population in a

time 𝒕 for ∅ = ∅(𝒉), where 𝒉 = 𝟏𝟎

Dengue model 𝑑𝑆ℎ 𝛽ℎ 𝑏

=𝐴− 𝑆ℎ 𝐼𝑣 − 𝜇ℎ 𝑆ℎ (7)
Esteva and Vargas [11] settled this model 𝑑𝑡 𝑁ℎ
𝑑𝐼ℎ 𝛽ℎ 𝑏
with the assumption that if a person gets the = 𝑆ℎ 𝐼𝑣 − (𝜇ℎ + 𝛾)𝐼ℎ (8)
𝑑𝑡 𝑁ℎ
immunity against the disease, then the 𝑑𝑅ℎ
= 𝛾𝐼ℎ − 𝜇ℎ 𝑅ℎ (10)
individuals will not be reinserted by this 𝑑𝑡
𝑑𝑆𝑣 𝛽𝑣 𝑏
disease. In this model host population =𝐵− 𝑆𝑣 𝐼ℎ − 𝜇𝑣 𝑆𝑣 (11)
𝑑𝑡 𝑁ℎ
assuming constantly that means natural death 𝑑𝐼𝑣 𝛽𝑣 𝑏
rate and birth rate are equal. The differential = 𝑆𝑣 𝐼ℎ − 𝜇𝑣 𝐼𝑣 (12)
𝑑𝑡 𝑁ℎ
equation of the dengue model as follows.

676
 Farman et al.

Where 1 − 𝑒 −(𝛾+𝜇ℎ )ℎ
𝑆ℎ , 𝐼ℎ , 𝑅ℎ , 𝑁𝑣 , 𝑆𝑣 , 𝐼𝑣 , 𝜇𝑣 , 𝛽ℎ , 𝛽𝑣 , 𝛾, 𝑏, 𝐴, 𝐵 are ∅ = ∅(ℎ) =
(𝛾 + 𝜇ℎ )
host population of susceptible individuals, The discrete method (13-17) is certainly
host population of infected individuals, host defined by using Mickens's rules.
population of permanent immune Numerical simulation
individuals, vector population, vector The mathematical analysis of the vector-
population of susceptible individuals, vector borne disease dengue fever model with non-
population of infected individuals, death rate linear incidence has been presented. To
of vector population, transmission observe the effects of the parameters using in
probability from vector to host, transmission this dynamics Dengue epidemic model (7-
probability from host to vector, rate of 12), conclude several numerical simulations
recovery from host population, biting rate of with parameters values 𝐴 = 0.457 , 𝐵 =
vector, recruitment rate of host population, 400, 𝛽ℎ = 0.75, 𝑏 = 0.5, 𝜇ℎ =
Recruitment rate of host population 0.0000457 , 𝛾 = 0.1428 , 𝑁ℎ = 1000, 𝛽𝑣 =
respectively. 1, 𝜇𝑣 = 0.25 are given in [11]. In (Figure3)
Nonstandard finite difference (NSFD) shows the convergence solution using the
scheme of the model NSFD scheme at ℎ = 1 for the true
In this section, we design the NSFD [8-10] equilibrium point. It can be easily seen from
scheme that replicates the dynamics of a (Figure3) which shows fast convergence for
continuous model (1)-(5). Let 𝑌𝑘 = each compartment. It clearly shows that
((𝑆ℎ )𝑘 , (𝐼ℎ )𝑘 , (𝑅ℎ )𝑘 , (𝑆𝑣 )𝑘 , (𝐼𝑣 )𝑘 )𝑡 denoted infected individuals at the latent period are
the approximation of 𝑋(𝑡𝑘 ) where 𝑡𝑘 = diagnosed and treated, the number of
𝑘∆𝑡 , 𝑤𝑖𝑡ℎ 𝑘𝜖𝑁, ℎ = ∆𝑡 > 0 be a step size susceptible individuals decreases
then; significantly, the infected decreases steadily.
𝑆ℎ𝑘+1 −𝑆ℎ𝑘 𝛽
= 𝐴 − 𝑁ℎ 𝑏𝑆ℎ𝑘+1 𝐼𝑣𝑘 − 𝜇ℎ 𝑆ℎ𝑘+1 Smoking Model
∅ ℎ
Smoking model is divided into five sub-
𝐼ℎ𝑘+1 −𝐼ℎ𝑘 𝛽ℎ
= 𝑏𝑆ℎ𝑘+1 𝐼𝑣𝑘 − (𝛾 + 𝜇ℎ )𝐼ℎ𝑘+1 compartment like potential smokers𝑃(𝑡),
∅ 𝑁ℎ
𝑘+1
𝑅ℎ 𝑘
−𝑅ℎ occasional smokers𝐿(𝑡), heavy smoker’s
= 𝛾𝐼ℎ𝑘+1 − 𝜇ℎ 𝑅ℎ𝑘+1 𝑆(𝑡) temporary quitters 𝑄(𝑡) and smokers
∅
𝑆𝑣𝑘+1 −𝑆𝑣𝑘 𝛽 who quit permanently 𝑅(𝑡) specified
= 𝐵 − 𝑁𝑣 𝑏𝑆𝑣𝑘+1 𝐼ℎ𝑘 − 𝜇𝑣 𝑆𝑣𝑘+1
∅ ℎ by𝑇(𝑡) = 𝑃(𝑡) + 𝐿(𝑡) + 𝑆(𝑡) + 𝑄(𝑡) +
𝐼𝑣𝑘+1 −𝐼𝑣𝑘 𝛽𝑣 𝑅(𝑡). The proposed smoking [12] in the form
= 𝑏𝑆𝑣𝑘+1 𝐼ℎ𝑘 − 𝜇𝑣 𝐼𝑣𝑘+1
∅ 𝑁ℎ of the system of the nonlinear differential
𝐴∅+𝑆ℎ𝑘
𝑆ℎ𝑘+1 = 𝛽 (13) equation is given as:
1+ ℎ 𝑏∅𝐼𝑣𝑘 +∅𝜇ℎ 𝑑𝑃
𝑁ℎ = 𝑎(1 − 𝑃) − 𝑏𝑃𝑆 (18)
𝛽ℎ 𝑑𝑡
𝑏∅𝑆ℎ𝑘+1 𝐼𝑣𝑘 +𝐼𝑣𝑘 𝑑𝐿
𝑁ℎ
𝐼ℎ𝑘+1 = (14) = −𝑎𝐿 + 𝑏𝑃𝑆 − 𝑐𝐿𝑆 (19)
1+∅(𝛾+𝜇ℎ ) 𝑑𝑡
𝑑𝑆
𝛾∅𝐼ℎ𝑘+1 +𝑅ℎ𝑘 = −(𝑎 + 𝑑)𝑆 + 𝑐𝐿𝑆 + 𝑓𝑄 (20)
𝑅ℎ𝑘+1 = (15) 𝑑𝑡
𝑑𝑄
1+∅𝜇ℎ
𝐵∅ = −(𝑎 + 𝑓)𝑄 + 𝑑(1 − 𝑒)𝑆 (21)
𝑆𝑣𝑘+1 = 𝛽 (16) 𝑑𝑡
𝑑𝑅
1+ 𝑣 𝑏∅𝐼ℎ𝑘 +∅𝜇𝑣 = −𝑎𝑅 + 𝑒𝑑𝑆 (22)
𝑁ℎ
𝑑𝑡
𝛽𝑣
𝑏∅𝑆𝑣𝑘+1 𝐼ℎ𝑘 +𝐼𝑣𝑘+1
𝑁ℎ
𝐼𝑣𝑘+1 = (17) Here 𝑏 indicates the contact rate between
1+∅𝜇𝑣
Which is the purposed NSFD scheme for the potential smokers and smokers who smoke
given model, where occasionally, 𝑐 represents the contact rate

677
Pure Appl. Biol., 9(1): 674-682, March, 2020
http://dx.doi.org/10.19045/bspab.2020.90073

between temporary quitters and smokers who ∅𝑑(1−𝑒)𝑆 𝑘+1 +𝑄𝑘

𝑄 𝑘+1 = (26)
smoke occasionally, 𝑑 represents the rate of 𝟏+∅(𝑎+𝑓)
𝑅 𝑘 +∅𝑒𝑑𝑆 𝑘+1
giving up smoking, (1 − 𝑒) stances for the 𝑅 𝑘+1 = (27)
1+𝑎∅
fraction of smokers who temporarily give up
Which is the proposed NSFD Scheme for the
smoking at a rate 𝑑, 𝑓indicate the contact rate given model, where
between smokers and temporary quitters who 1−𝑒 −𝑑(1−𝑒)ℎ
return back to smoking, 𝑎 denotes the rate of ∅ = ∅(ℎ) = (28)
𝑑(1−𝑒)
natural death, 𝑒 denotes the remaining Numerical simulation
fraction of smokers who give up smoking The technique creates a better impact to
permanently (at a rate 𝑑). control the smoking, it reduces the infected
Nonstandard finite difference (NSFD) rate and increases the potential smokers
scheme of the model during disease Free State. In (Figure 4) by
Let 𝑌𝑘 = (𝑃𝑘 , 𝐿𝑘 , 𝑆𝑘 , 𝑄𝑘 , 𝑅𝑘 )𝑇 denoted an decreasing the value of ℎ potential
approximation of 𝑋(𝑡𝑘 ) where 𝑡𝑘 = 𝑘∆𝑡 with smokers(𝑃)increases with time.In (Figure 5)
𝑘 ∈ 𝑁, ℎ = ∆𝑡 > 0 be a step size, then by decreasing the value of ℎ occasional
𝑃 𝑘+1 −𝑃𝑘 smokers (𝐿) decreases rapidly with time. In
= 𝑎 − 𝑎𝑃𝑘+1 − 𝑏𝑃𝑘+1 𝑆 𝑘
∅
𝐿𝑘+1 −𝐿𝑘
(Figure 6) by decreasing the value of ℎ heavy
= −𝑎𝐿𝑘+1 + 𝑏𝑃𝑘+1 𝑆 𝑘 − 𝑐𝐿𝑘+1 𝑆 𝑘 smokers (𝑆) decreases rapidly with time. In
∅
𝑆 𝑘+1 −𝑆 𝑘 (Figure 7), by decreasing the value of ℎ
= −(𝑎 + 𝑑)𝑆 𝑘+1 + 𝑐𝐿𝑘+1 𝑆 𝑘 + 𝑓𝑄 𝑘
∅ temporary quitters (𝑄) increases initially, but
𝑄 𝑘+1 −𝑄 𝑘 after some time decreases rapidly with time.
= −(𝑎 + 𝑓)𝑄 𝑘+1 + 𝑑(1 − 𝑒)𝑆 𝑘+1
∅
In (Figure 8) by decreasing the value of ℎ
𝑅 𝑘+1 −𝑅𝑘
= −𝑎𝑅 𝑘+1 + 𝑒𝑑𝑆 𝑘+1 permanent quitters (𝑅) increases initially, but
∅
𝑎∅+𝑃 𝑘 after some time decreases rapidly with time.
𝑃𝑘+1 = (23) It can be easily seen that by reducing the step
1+𝑎∅+𝑏∅𝑆 𝑘
𝑏∅𝑃 𝑘+1 𝑆 𝑘 +𝐿𝑘 size the system (18-22) converge rapidly to
𝐿𝑘+1 = (24)
𝟏+𝑎∅+𝑐∅𝑆 𝑘 the steady-state point [13].
∅𝑐𝐿𝑘+1 𝑆 𝑘 +∅𝑓𝑄 𝑘 +𝑆 𝑘
𝑆 𝑘+1 = (25)
1+(𝑎+𝑑)∅

Figure 3. Host Population of Susceptible individual, Host Population of Infected Individual,

Host Population of Permanent Immune individual, Vector Population of Susceptible, Vector
Population of Infected

678
 Farman et al.

Figure 4. Numerical solutions for potential smokers in a time 𝒕 with step size 𝒉 = 𝟏, 𝒉 = 𝟓
and𝒉 = 𝟏𝟎 for disease free equilibrium points

Figure 5. Numerical solutions for occasional smokers in a time 𝒕 with step size 𝒉 = 𝟏, 𝒉 = 𝟓
and 𝒉 = 𝟏𝟎 for disease free equilibrium points

Figure 6. Numerical solutions for heavy smokers in a time 𝒕 with step size 𝒉 = 𝟏, 𝒉 = 𝟓 and
𝒉 = 𝟏𝟎 for disease free equilibrium points

679
Pure Appl. Biol., 9(1): 674-682, March, 2020
http://dx.doi.org/10.19045/bspab.2020.90073

Figure 7. Numerical solutions for temporary quitters in a time 𝒕 with step size 𝒉 = 𝟏, 𝒉 = 𝟓
and 𝒉 = 𝟏𝟎 for disease free equilibrium points

Figure 8. Numerical solutions for permanently quitters in a time 𝒕 with step size 𝒉 = 𝟏, 𝒉 =
𝟓 and 𝒉 = 𝟏𝟎 for disease-free equilibrium points

SVIR model Here 𝑆(𝑡), 𝐼(𝑡) and 𝑅(𝑡) denote the densities
𝑑𝑆 of susceptible, infected and recovered
= 𝜇 − 𝜇𝑆 − 𝛽𝑆𝐼 − 𝛼𝑆 (29)
𝑑𝑡
𝑑𝑉 individuals, respectively. 𝑉(𝑡) denotes the
= 𝛼𝑆 − 𝛽1 𝑉𝐼 − 𝛾1 𝑉 − 𝜇𝑉 (30) density of vaccines who have begun the
𝑑𝑡
𝑑𝑆
= 𝛽𝑆𝐼 + 𝛽1 𝑉𝐼 − 𝛾𝐼 − 𝜇𝐼 (31) vaccination process. In the above model, all
𝑑𝑡
𝑑𝑆 parameters are positive and have the
𝑑𝑡
= 𝛾1 𝑉 + 𝛾𝐼 − 𝜇𝑅 (32) following features: 𝜇 is the recruitment rate
and natural death rate of the population; 𝛽

680
 Farman et al.

denotes the transmission coefficient between Qualitative analysis

compartments 𝑆 and 𝐼; 𝛼 is the rate at which 𝑑𝑆
For the equilibrium points, we take 𝑑𝑡 =
susceptible individuals are moved into the 𝑑𝑉 𝑑𝐼 𝑑𝑅
vaccination process; 1 denotes the disease 0, 𝑑𝑡 = 0, 𝑑𝑡 = 0, 𝑑𝑡 = 0 then we get
transmission rate when the vaccines contact 0 = 𝜇 − 𝜇𝑆 − 𝛽𝑆𝐼 − 𝛼𝑆
with infected individuals before obtaining 0 = 𝛼𝑆 − 𝛽1 𝑉𝐼 − 𝛾1𝑉 − 𝜇𝑉
immunity;  1 is the average rate for the 0 = 𝛽𝑆𝐼 + 𝛽1 𝑉𝐼 − 𝛾𝐼 − 𝜇𝐼
0 = 𝛾1 𝑉 + 𝛾𝐼 − 𝜇𝑅
vaccines move into recovered population; 𝛾
By solving the above model we get
is the recovery rate of infected individuals
[14].

𝜇 𝛼𝜇 𝛼𝛾1
𝑆𝐸 (𝑆, 𝐼, 𝑉, 𝑅) = 𝑆𝐸 ( , 0, , )
𝛼 + 𝜇 (𝛼 + 𝜇)(𝜇 + 𝛾1 ) (𝛼 + 𝜇)(𝜇 + 𝛾1 )

Nonstandard finite difference (NSFD) Recovered) model with non-linear incidence

scheme of the model has been presented. To observe the effects of
Followings are the scheme the parameters using in this dynamics SVIR
ℎ𝜇+𝑆 𝑘 model, conclude several numerical
𝑆 𝑘+1 = 1+ℎ𝜇+ℎ𝛽𝐼𝑘 +ℎ𝛼 (33)
simulations with the given parameter values.
(ℎ𝛼𝑆 𝑘+1 +𝑉 𝑘 )
𝑉 𝑘+1 = 1+ℎ𝛽 (34) In (Figure 9) shows that the model presented
𝑘
1 𝐼 +ℎ𝛾1 +ℎ𝜇 gradually approaches the steady states and
ℎ𝛽𝑆 𝑘+1 𝐼 𝑘 +ℎ𝛽𝑉 𝑘+1 𝐼 𝑘
𝐼 𝑘+1 = (35) infected individuals at the latent period are
1+ℎ𝛾+ℎ𝜇
ℎ𝛾1 𝑉 𝑘+1 +𝑅 𝑘 diagnosed and treated after vaccination, the
𝑅 𝑘+1 = (36) number of susceptible individual’s decreases
1+ℎ𝜇
significantly; the infected decreases steadily
The mathematical analysis of the epidemic while the recovered increased rapidly.
SVIR (Susceptible Vaccination Infected

Figure 9. Numerical solutions for susceptible, vaccination, infected individual and recovered
population in a time t for ∅ = ∅(𝒉)

681
Pure Appl. Biol., 9(1): 674-682, March, 2020
http://dx.doi.org/10.19045/bspab.2020.90073

Conclusion Islands in the year 1846, Delta Omega

Numerical simulations are performed and all Society, Cleveland.
the analytical results are verified numerically 4. WHO (2010). Vaccine-preventable diseases:
by using the Non-Standard Finite Difference monitoring system. Geneva: World Health
Scheme to reduce the infected rate in Organization.
5. Ministry of Health, Manatu Hauora (1998).
different epidemic models. This leads to the Predicting and Preventing Measles
practical implications for the control of Epidemic in New Zealand: Application of
measles infections, dengue fever, smoking Mathematical Model.
and SVIR by these models. It can easily be 6. Anguelov R, Dumont Y, Lubuma JMs &
seen from the developed scheme and their Shillor M (2014). Dynamically consistent
graphical representations that NSFD nonstandard finite differences schemes for
preserves all the mandatory properties of epidemiological models. J of Comp and
continuous models like positivity, Appl Maths 225: 161-182
boundednes etc. for epidemic models, while 7. Anguelov R & Lubuma JMS (2001).
all other numerical techniques fail to Contribution to the mathematics of the
maintain the properties of continuous models nonstandard finite difference method and
applications, Numer. Methods Partial
for the larger step sizes. NSFD Scheme also Differ. Equ 17: 518-543.
always converges to their steady-state points 8. Lubuma JMS & Patidar K (2007). Non-
while other numerical techniques like standard methods for singularity perturbed
Adomian Decomposition Method (ADM), problems possessing oscillatory/ layer
Differential Transform Method (DTM), solutions, Appl Maths Comp 187(2). 1147-
Runge Kutta Method and Euler’s Methods 1160.
etc. fails to do so. This can help to control the 9. Roeger LW (2014). Exact difference
spreading of infectious diseases in the schemes, in A. B. Gumel Mathematics of
community. continuous and discrete Dynamics Systems.
Authors’ contributions Contemp Maths 618: 147-161.
10. Mickens RE (1989). Exact solutions to a
Conceived and designed the experiments: M
finite difference model of a nonlinear
Farman & A Ahmad, Performed the reactions advection equation: Implications
experiments: MU Saleem & M Farman, for numerical analysis. Numer Meth for
Analyzed the data: A Ahmad & MO Ahmad, Partial Diff Equa 5: 313-325.
Contributed materials/ analysis/ tools: MU 11. Hamed AL, Moustafa ElS, Juan J & Nieto
Saleem & MO Ahmad, Wrote the paper: M WS (2014). On fractional Order Dengue
Farman & A Ahmad. Epidemic Model. Mathem Prob in Engin,
References doi.org/10.1155,456537, 1-6.
1. Abubakar S, Akinwande NI & Abdulrahman 12. Singh J, Kummar D, Qurashi M & Baleanu
(2012). A Mathematical Model of Yellow D (2017). A new fractional model for giving
Fever Epidemics, Afrika Mathematika 6: 56- up smoking dynamics. Adv in Diff Equa 88.
58. 13. Aqeel A, Farman M, Faisal Y & Ahmad M
2. Grenfell T (1992). Chance and chaos in O (2018). Dynamical transmission and
measles dynamics. J R Stat Soc B 54: 383- effect of smoking in society. Inter J of Adv
398. and Appl Sci 5(2): 71-75.
3. Panum PL (1940). Observations Model 14. Liu Y, Takeuchi & Awarni S (2008). SVIR
during the epidemic of Measles on the Faroe epidemic models with vaccination
strategies. J Theoret Biol 253.

682