key: cord-1003056-gsedfb36
authors: Nisar, Kottakkaran Sooppy; Ahmad, Shabir; Ullah, Aman; Shah, Kamal; Alrabaiah, Hussam; Arfan, Muhammad
title: Mathematical analysis of SIRD model of COVID-19 with Caputo fractional derivative based on real data
date: 2020-12-29
journal: Results Phys
DOI: 10.1016/j.rinp.2020.103772
sha: f060a6fa4b994e36ac3fbbb21050ae490a3db8f3
doc_id: 1003056
cord_uid: gsedfb36

We discuss a fractional-order SIRD mathematical model of the COVID-19 disease in the sense of Caputo in this article. We compute the basic reproduction number through the next-generation matrix. We derive the stability results based on the basic reproduction number. We prove the results of the solution existence and uniqueness via fixed point theory. We utilize the fractional Adams–Bashforth method for obtaining the approximate solution of the proposed model. We illustrate the obtained numerical results in plots to show the COVID-19 transmission dynamics. Further, we compare our results with some reported real data against confirmed infected and death cases per day for the initial 67 days in Wuhan city.

It is very important to study the mathematical models of infectious diseases for the better understanding of their evaluation, existence, stability and control [1] [2] [3] [4] . As the classical approaches of mathematical models do not determine high degree of accuracy to model these diseases, fractional differential equations were introduced to handle such problems, which have many applications in applied fields like production problems, optimization problems, artificial intelligence, medical diagnoses, robotics, cosmology, and many more. Fractional differential equations are used in mathematical models of biological phenomena during the last few decades [5] [6] [7] [8] [9] [10] [11] [12] . One of the infectious diseases is COVID-19 caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The disease was first reported in December 2019 in Wuhan city, China, which spread throughout the world and led to a continuing pandemic outbreak in 2020 [12] . It is declared that the COVID-19 pandemic represents a great global threat 2019-2020, which has affected 212 countries and different territories around the world. According to the data reported by Worldometer and WHO [13] [14] [15] [16] , as of May 03, 2020, it has been noticed that more than 3.5 million people were infected with 0.247 million deaths. Even in some countries like Italy and Spain, the death rate is as high as almost 0.066. It is reported that most people infected with COVID-19 will experience mild to moderate lung infections, such as trouble breathing, fatigue, vomiting, cough, and others physical signs. However, other symptoms of varying severity, such as gastroenteritis and neurological disorders have also been identified [15] . If an infected person coughs or sneezes, the COVID-19 transmits mostly by droplets from the nose, and once a person inhales the droplets in the air, he will be exposed to the danger of getting the infection. As a result, the best way to prevent the virus is to avoid mixing up with the people. The severity of this pandemic attracted the researchers and scientists throughout the world [16] [17] [18] [19] [20] [21] [22] [23] [24] . It was observed that more and more countries started to put ban on international traveling, close schools, businesses, and even shopping malls. The COVID-19 has caused considerable financial damage around the globe. A significant number of doctors and researchers dedicated themselves to the struggle against pandemics and carried out work in their areas of expertise. They studied COVID-19 from various viewpoints, such as virology, infectious diseases, public environmental occupational safety, microbiology, sociology, veterinary sciences, political economics, and media studies, etc. China, Spain, and the USA are the leading countries on COVID-19 research because the early outbreak of the virus urged them to start relevant research immediately. A group of researchers studied the origin of COVID-19. Recently, Hasan et al. [25] proposed a new compartmental SIRD model of COVID-19 as follow

In this model they divided the total population N into four epidemiological classes: S (susceptible class), I (infected class), R (recovered class), and D (death class). The description of the parameters are as follow:

• β is the average number of contacts per person per time, • γ is the recovery rate, • κ is the death rate.

In the last few decades, researchers have given significant attention to the study of fractional calculus. That is because the fractional calculus can more effectively describe and process the preservation and inheritance properties of different structures than the integer-order models. We include [26] [27] [28] [29] [30] for further applications about fractional calculus. Hence, the aforementioned area has been investigated from various angles such as qualitative theory, numerical analysis, etc. (see [28] [29] [30] ). Researchers, therefore, expanded the classical calculus to the fractional order via fractional order modeling in [31] [32] [33] using different mathematical techniques. For example, in 1980, Adomian devised a popular method of decomposition to handle nonlinear problems analytically. Before that, the aforementioned method was gradually introduced as an important tool for calculating analytical or approximate solutions to many specie problems. The mathematical models were extensively analyzed using decomposition, homotopy and variation techniques [34] [35] [36] [37] . The methods mentioned were extensively used to treat both linear and nonlinear FODEs [38] [39] [40] [41] [42] [43] . Recently, the "residual power series method", "Fourier transform method", "spectral methods" and "method of collocation" as well as several new ways of computational methods have been used to manage differential equations in both fractional and classical order and their systems (see for details [45, 44, [46] [47] [48] [49] ).

Since mathematical modeling of dynamical real-world processes/ phenomena has been well-predicted by using fractional-order equations. The mentioned operators have appeared in various fields such as physics, chemistry, and engineering. Because in most situations, the memory and hereditary characteristics of materials and processes have not been well-predicted via ordinary differential operators. They can be explained comprehensively with real acceptable remarks by using fractional order operators. Further since the most notable definitions among the fractional differential operators are those given by Reimann-Liouville, Caputo, and recently those given by Caputo-Fabrizo and Atangana-Baleanu type fractional operators. These operators have their characteristics in using. Since in the past two decades, Caputo derivatives have adopted mostly to handle biological models of infectious disease. Therefore by using fractional differential operator, we state that in the fields of continuous-time modeling, various researchers have pointed out that fractional derivatives are better tools in describing acoustics, rheology, polymeric chemistry, linear viscoelasticity, and many more such type of sciences [50] . Therefore, motivated from the above mentioned work, here we study the model (1) under Caputo fractional-order derivative. For 0 < α ≤ 1,

along with initial conditions

It is remarkable to develop some effective numerical methods to compute the approximate solutions to both linear and nonlinear problems based on fractional-order differential equations. In this regard, various methods and algorithms have been proposed for the analytical and numerical solutions of the mentioned problems. In several research articles, Euler's method and RK4 method have been utilized to investigate numerical solutions to various classes of fractional order differential equations. Therefore, finding the numerical solution of (2) with an appropriate technique is very important. In this paper, we use the Adams-Bashforth method [51] of fractional order to find the numerical solution of (2). The advantage of this method is that it converts the considered system into a system of Volterra type integral equations. In this way, one can apply the numerical schemes for Volterra type integral equations to find the numerical solution of (2). Keeping these points in view, we first establish some existence results, boundedness, and computation of the basic reproductive numbers for checking the stability of the considered model. Then, by utilizing fixed point theory, we derive some necessary conditions for the existence of at least a solution and its uniqueness.

We organized the paper as follows: Section 1 presents the introduction of the COVID-19 pandemic disease and the development of the fractional calculus. Section 2 includes basic notions of fractional calculus. Section 3 provides equilibrium points and stability analysis. It also deals with the existence and uniqueness results of the solution, and presents the numerical scheme used to find the approximate system solution of the model (2) . Section 4 gives the graphical representation of the model, while Section 5 presents the conclusion of the article.

. Definition 1. [25] The Caputo fractional derivative of a continuous function Φ on [0, T] is defined as:

where 0 < α ≤ 1, n = [α] + 1, and [α] represents the integers part of α. Definition 2. [28] The fractional integral of a continuous function Φ on L 1 ([0, T], R) of order 0 < α ≤ 1 corresponding to t is defined as: 

where Φ 0 is any real number given by

Here, we discuss the feasibility region of solution and stability analysis of equilibrium points on the basis of the basic reproduction number.

and the pandemic will occur if S 0 > γ+κ β .

Proof. Let

this implies that

Now adding all equation of (1), we get

Solving (3), we have

From this we can write as N (t)⩽N 0 . This shows the boundedness of the considered population dynamical model. Next, from first equation of (1)

therefore, S (t) is always decreasing and hence no pandemic will occur. Second equation of (1) is

where γ+κ β is called threshold phenomenon or critical community size for pandemic.

So infection class is decreasing and hence no pandemic will occur. If S 0 > γ+κ β which yields S (t) > γ+κ β . Hence, we have dI (t) dt > 0, so infection class is increasing and the pandemic will occur. Theorem 1. The disease-free equilibrium point of (2) is

Proof. For this, we write (2) as

From second equation of (4) and using (2), we take I (t) = 0, then

and putting values of I = 0 in 3rd and 4th equations of (4), we get

Thus the required disease-free equilibrium point is

Hence the theorem is proved. (2) is determined as

Proof. To find the basic reproduction number, we take the second equation of (2) as X = I,

The non-linear term is F and the linear term is V. Now we will find the next generation matrix FV − 1 , where

then

.

Since R 0 is the leading eigen-value of the next generation matrix FV − 1 at disease-free equilibrium point E 0 = ( γ+κ β , 0, 0, 0), which can be written as

Therefore, the basic reproduction number is given by

Thus the requisite result is demonstrated.

Theorem 3. If R 0 < 1, the pandemic free equilibrium point of (2) is locally asymptotically stable and unstable if R 0 > 1.

Proof. One can measure the Jacobian matrix for (2) as

where Φ i , i = 1, 2, 3, 4 represents right hand side of (2). Apply partial derivatives, we get

After using the values of E 0 , we obtain

Now, the characteristics equation of (7) can be determined as

After evaluation of the determinant our interested eigenvalue is λ, and is given by

or λ = R 0 − 1, implies that λ must be negative if R 0 < 1, hence proved our required result.

We demonstrate the existence of unique solution of the model (2) through fixed point theory. Finally, we present the numerical scheme of the proposed model via fractional Adams-Bashforth method. Let B = C a,b be the Banach space of all continuous and bounded real-valued function on a closed interval [a, b]. To study the existence of unique solution, we consider (2) as follows:

where

Applying fractional integral to both sides of Eq. (8), we have 

where

Now, we define a compact Banach space C a,b as

Also, we assume that

Define the infinite norm as:

Now, we define an operator Ψ : C a,b →C a,b by

First, we show that the operator Ψ is well defined. For this, we need to show that

Similarly, we can derive the above inequality for the remaining terms. Thus, if

then the operator Ψ is well-defined. Second, we have to show that the operator Ψ satisfies the Lipschitz condition, i.e

For the first component, we have

where

Similarly for the other terms, we have

Hence Ψ is a contraction on a compact Banach space. It implies that Ψ has a unique solution. Consequently, the model (2) has a unique solution.

Here the general method to finds the solution of the considered system by fractional Adams-Bashforth method has been discussed. Consider the model (8)

with the initial conditions

Applying fractional integral to both sides of (13) and using initial conditions, we get

Here, we will derive the numerical scheme for the first equation of the system (14) .

Let us denote t s = sh, s = 0, 1, 2, …, J, where h = T J is the step size and J > 0 is an integer and T > 0. At t = t s+1 , s = 0,1,2,…, the Eq. (15) gets the form

and

By subtracting (17) from (16), we obtain

We can write the above equation as

where

Using Lagrangian polynomial interpolation, we approximate the function F 1 (S , I , R , D , t) as

thus,

Solving the integrals on the right hand side of the above equation, we have

Substituting the values of A α,1 and A α,2 in Eq. (18), we get the approximate solution of the first equation of the model (2) .

Similarly for the remaning equations, we have

We conduct numerical simulations to represent the effects of our model with data from various world meter and WHO reports; Starting on 4 January 2020, when the Chinese authorities have confirmed 6 Cases a day. Slowly and gradually the number to 1460 on 20 January 2020 followed by 26 deaths. In next day the number confirmed cases increased to 1739 followed by 38 deaths. This number is repidly increased to 3892 with 254 deaths on 4 February 2020 data followd by worldometer. To control the disease Chinese authority lockdown in Wuhan city; the spread of the disease was reduced. As the total population of the Wuhan city is 11 million, therefore take, N = 11000000/250. In the first days of the outbreak, this denominator was determined and proved later to be a realistic value. It is an appropriate value for the restriction of individual movements according to the actual data recorded by the WHO. As for the preconditions, fixed theses values: S 0 = N − 6, I 0 = 1, R 0 = 0, and D 0 = 0.

In Figs. 1-4 , we have provided the approximate solutions of different classes against the data given and corresponding to various fractional orders. We see that the population are initially assumed to be uninfected (susceptible). Once the epidemic began the susceptible population continued to drop as seen in Fig. 1 . Hence, as they were exposed to infection, the population density of the infected class was rapidly increasing as shown in Fig. 2 . This rise resulted in an increase in the death class, and many people got rid of the infection that contributed to an increase in the recovered population. Due to the fractional-order the rate of decay and growth is different. The smaller the order the quicker the process concerned, and vice versa. Furthermore, if the fractionalorder α→1 the subsequent solutions always meet the integer-order solution. Since the fractional differential operator has a greater degree of freedom and offers a full geometry spectrum, we took just a few fractional orders to analyze the model's dynamic behaviors under consideration. Further, in Fig. 5 , we compare our simulated results with the available real data published in [5] from 4th January 2020 to 8th March 2020 for 67 days as [6, 12, 19, 25, 31, 38, 44, 60, 80, 131, 131, 259, 467, 688, 776, 1776, 1460, 1739, 1984, 2101, 2590, 2827, 3233, 3892, 3697, 3151, 3387, 2653, 2984, 2473, 2022, 1820, 1998, 1506, 1278, 2051, 1772, 1891, 399, 894, 397, 650, 415, 518, 412, 439, 441, 435, 579, 206, 130, 120, 143, 146, 102, 46, 45, 20, 31, 26, 11, 18, 27, 29, 39, 39] (see Table 1 ).

We see that the graphs of the curves of simulated data and real data are very close to each other at the order of 0.97. Hence, α = 0.97 is the best suitable fractional-order value. Further, the confirmed reported death in [5] Fig. 6 .

Again from Fig. 6 , we see that at fractional-order α = 0.97, the simulated data and real data are very close to each other; hence, the best choice of the fractional-order value is α = 0.97.

In this paper, we investigated the fractional-order SIRD model of COVID-19 in the case of Caputo. Using the next-generation matrix, we calculated the basic reproduction number and proved the stability conditions based on the basic reproduction number. We established the results of the proposed model's existence and uniqueness of the solution. We derived a proper numerical scheme using the Adams-Bashforth fractional method to find an approximate solution to the proposed model. Lastly, we presented the numerical simulation to illustrate the better dynamics of the SIRD fractional-order model COVID-19. Also, we have compared our simulated results of infected and death class with real data for reported cases of infection and death. We see that numerical results are close to real data solutions. The curve very well coincides with real data at α = 0.97. So this is the best value of fractional order.

We will study the SIRD model in the future under new operators called fractal-fractional operators.

All the authors have equal contribution in this work.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 

A stochastic SICA epidemic model for HIV transmission

Mathematical modeling of Zika disease in pregnant women and newborns with microcephaly in Brazil

Dynamics and optimal control of Ebola transmission

Study of a fractional-order epidemic model of childhood diseases

A model of Dengue fever

Competitive exclusion in a vectorhost model for the Dengue fever

The role of mathematical models in explaining recurrent outbreaks of infectious childhood diseases

A mathematical study of an imprecise SIR epidemic model with treatment control

Mathematical modeling for the impacts of deforestation on wildlife species using Caputo differential operator

A new third order convergent numerical solver for continuous dynamical systems

Fractional modeling of blood ethanol concentration system with real data application

Mathematical models in epidemiology

COVID-19 Coronavirus Pandemic

Epidemiol ogica Nuevo coronavirus (nCoV)

Actualizaci textasciiacute on Epidemioltextasciiacute ogica Nuevo coronavirus (2019-nCoV)

Investigating the impact of asymptomatic carriers on COVID-19 transmission

Analysis of Caputo fractionalorder model for COVID-19 with lockdown

Molecular mechanisms of coronavirus RNA capping and methylation

Effective containment explains subexponential growth in recent confirmed COVID-19 cases in China

One world, one health: the novel coronavirus COVID-19 epidemic

the role of super-spreaders in infectious disease

The characteristics of Middle Eastern respiratory syndrome coronavirus transmission dynamics in South Korea

Modeling of a super-spreading event of the Mers-corona virus during the Hajj season using simulation of the existing data

Mathematical modeling of COVID-19 transmission dynamics with a case study of Wuhan

A new estimation method for COVID-19 time-varying reproduction number using active cases

A fractional calculus approach to the dynamic optimization of biological reactive systems. Part I: fractional models for biological reactions

Fractional integrals and derivatives (theory and applications)

Theory and application of fractional differential equations North Holland mathematics studies

Fractional calculus in bioengineering

Applications of fractional calculus in physics

Fractional differential equations, mathematics in science and engineering

Theory of fractional dynamic systems

Applications of fractional calculus to dynamic problems of linear and nonlinear hereditary mechanics of solids

Solutions of the SIR models of epidemics using HAM

Solution of the epidemic model by Adomian decomposition method

Solution of the epidemic model by homotopy perturbation method

Variational iteration method for solving the epidemic model and the prey and predator problem

Existence of solution for a fractional order Lotka-Volterra reaction diffusion model with Mittag-Leffler kernel

Admoian decomposition method: convergence analysis and numerical approximations

Convergence of the Adomian decomposition method for initial value problems

Application of the Laplace adomian decomposition method and implicit methods for solving Burger, s equation

Solution of the fractional epidimic model by LADM

Numerical solution of fractional order smoking model via laplace Adomian decomposition method

Numerical analysis of fractional order model of HIV-1 infection of CD4+ T-cells

Analytical solutions of time-fractional wave equation by double Laplace transform method

Exact solutions of nonlinear time fractional partial differential equations by sub-equation method

A novel analytical method for time fractional differential equations

Residual power series method for fractional diffusion equations

Solution of fractional ordinary differential equations by natural transform

On the appearance of the fractional derivatives in the behaviour of real materials

A numerical method for delayed fractional-order differential equations

We are thankful to the reviewers for their careful reading and suggestions.