A Reliable Solution of Nonlinear Time Dependent Fractional Model of Ebola Virus Disease with Arbitrary Order Derivative in Liouville-Caputo Sense.

In this article, the analysis of an arbitrary order Ebola virus disease model is conducted to find out its reliable solution. The fractional derivative is taken in Liouville-Caputo sense. The solution of this nonlinear model is achieved using fractional variational iteration scheme. The convergence analysis of the obtained solution is also presented which confirms that it is positive, bounded and convergent. The outcomes are discussed with figures explaining variation of susceptible, infected, recovered population and number of disease induced deaths with time. The negligible error in successive iterations of various population shows the competency of the presented scheme. The results endorse that FVIM is extremely effective, powerful and easy in usage.

Introduction

Ebola is a fatal and complex pathogen for a human being. It attacks healthy cells of the body of host and replicates itself in it [1, 2]. It is a rare virus that causes bleeding [3-8]. In 1976, it was first detected in two outbreaks, one in South Sudan, and the other in Congo near the river Ebola. In spite of wide search, its reservoir could not be recognized but it may contain bats [1]. The 2014–2016 outbreak of Ebola in West Africa was the biggest since 1976. After starting in Guinea, it moved to Liberia and Sierra Leone. The recent outbreak of Ebola virus disease (EVD) is found in the Republic of Guinea on 14 February 2021. However, on 19 June 2021, the outbreak was declared over. There were 16 confirmed and 7 probable cases, of whom 12 people died [9].

This virus transmits through physical connection with secretions, tissues, body fluids or semen of infected ones. Its early symptom are headache, fever, sore throat, joint ache, muscle ache, weakness etc. In the later stage, the symptom are vomiting, diarrhea, bleeding, stomach pain and organ failure. Its incubation period is between 2–21 days and infection period is about 4–10 days. Its mortality rate ranges from 25 to 90% [10]. Despite high mortality rate, some patients survive and few develop long-lasting manifestations. It is seen that the individual genetic differences have a major role in the deaths by EVD [11]. In 2014, the average incubation time was 11.4 days. The average period from the beginning of symptom to hospitalization is days. The average period to death after hospitalization was days [12]. Ebola virus remains in semen for a long time in survivors with a risk of transmission [13]. The antivirals like Favipiravir is found to be effective in delaying survival and dropping viral load. The study of Ebola virus is done in [14-16]. The transmission EVD Dynamics is studied by various researchers in [17-19].

Mathematical investigation is done to state several phenomena of biological [20-22], medical [23] and physical [24-28] significance. The fractional derivatives [29-34] are used to describe the real-life problems. The arbitrary order of derivative is an index of memory [35]. For details of work on the fractional EVD models (see, for details, [36-43]). Fractional calculus [44] is a latent tool in electromagnetic theory [45], bio-engineering [46], plasma physics [47], control theory [48], neurophysiology [49], electric technology [50], visco-elasticity [51], disease dynamics [52-54] and many others. Over the years, the researchers have worked on various techniques for the solution of nonlinear FDE such as homotopy perturbation method [55], homotopy analysis method [56], Adomian decomposition method [57], fractional order homotopy perturbation method [58], variational iteration scheme and its modifications [59-66] etc.

However, the consistency in the solution of schemes [67] is vital. Most of the FDE have no exact solution, therefore approximation by FVIM is needed. He [68] established VIM for the solution of nonlinear equations and then applied to FDE. Highly nonlinear equations cannot successfully be solved by general methods. FVIM [69-72] is remarkably competent to address the challenge. It offers solution in an infinite convergent series. It entails no discretization and linearization. In FVIM, initial solution is picked with unknown parameters as searched solutions. It reduces computation significantly and has meticulousness in finding Lagrange’s multiplier.

Our goal is to investigate an efficient solution of time fractional Ebola virus model by FVIM. This paper is planned as below. In Sect. 1, there is introduction. In Sect. 2, the definition along with properties of some fractional operators are specified. In Sect. 3, the fractional EVD model is discussed. In Sect. 4, the idea of FVIM is discussed with analysis of convergence and solution by FVIM is also found. In Sect. 5, we elaborate results and afterwards conclusion is found.

Preliminaries

We give certain properties with definition of Liouville–Caputo operator (see, for details, [73-76]).

Definition 2.1:

Liouville–Caputo -order derivative is defined as:where,

If is positive integer, it turns into ordinary derivative,

Definition 2.2:

Liouville–Caputo -order derivative ( on space is:

Model Description

Atangana and Goufo [19] developed a model describing the spread of Ebola fever in the West African countries. Following this, we have reconstructed a nonlinear time dependent EVD model to include the features using memory effect. Let be the total population of West African countries, be the susceptible, be the infected, be the number of recovery and be the number of disease induced deaths due to EVD in these countries. We consider that the people get susceptible at rate infect at rate recover at rate and the disease induced death rate is The death rate due to other reasons is assumed as

Rates of change in susceptible and infected with time are,

Here, are the people removed from the susceptible.

Rates of change in the recovered and the number of deaths with time are,

Hence the EVD model is written as,with conditions and The property of non-locality is main use of working with FDE. It means that the next state of system is dependent on prior states. Therefore, these models adhere to reality. The time dependent fractional derivative submits memory modulation [77] of a system. Clearly, the presented EVD model is impacted by memory in time that confirms aptness of fractional order modeling. Therefore, a systematic and exhaustive study of Eq. (1) is important. Equation (1) is nonlinear and it is visibly tough to find its exact solution. Therefore, it turned into our motivation to solve Eq. (1) by FVIM because it gives reliable results for nonlinear equations without finding special polynomials.

FVIM for the Fractional Ebola Virus Disease Model

Consider a fractional EVD model in Eq. (1). The correctional functionals for Eq. (1) are,

Here, is Lagrangian multiplier.

By variational theory,

We instantly get,

From Eq. (2),

Consecutive approximations are hence calculated. are the restricted variations so, .

We get .

Solution is,and

FVIM Procedure

Step I. find set

Step II. use and to get and from Eq. (4);

Step III. define := : = := and : =

Step IV. if max and

max stop, otherwise continue;

Step V. set : = : = := and : =

Step VI. Set return to step II ;

Convergence Analysis of FVIM

Consider operators :also, state the components

Equation (7) quickly converges [78].

Theorem 1

[79]: Let be the operators from Banach space BS to BS. Equation (7) converge [80, 81] if 0 <  < 1 exists in such a way that,

(i.e.)

Theorem 2

[79]: Let Eq. (7) converge to and . If is used as approximation to of Eq. (1), maximum error is,

If , we define,then of Eq. (7) converge if

From Theorem 2,

Implementation of FVIM

By Eq. (4),

Placing in Eq. (4),and so on.

Here, and,

Similarly, we can find the ensuing terms. Consequently, we get the solution as given in Eq. (5).

From Eqs. (6) and (7), the iterations for Eq. (1) are,and so on.

If we compute we get,when for instance, and

It approves that the solution of Eq. (1) by FVIM is bounded, convergent and positive.

Numerical Results and Discussion

Simulations are done for at distinct fractional values of and using Maple package The FVIM solution of Eq. (1) is achieved using value of parameters given in Table 1. Figure 1a depicts the behavior of susceptibles with time for distinct . It shows that the susceptibles increase with increase in value of the arbitrary order Figure 1b portrays the behavior of infected ones with time for distinct . In it, the infected increase with decrease in Fig. 1c illustrates the behavior of recovery with time for separate . In it, this number decreases as we increase the value of Fig. 1d shows behavior of number of deaths with time for diverse . In it, increases with decreasing values of . In Table 2 the absolute error between the successive iterations of susceptibles decrease sharply at the orders and of the fractional derivative. In Table 3 the absolute error between the consecutive approximations of infected also decrease sharply at orders and In Table 4, the absolute error between the succeeding estimates of recovered reduce at orders and while in Table 5 the absolute error between the subsequent iterates of disease induced deaths also decrease at the orders and

Table 1

Value of parameters for simulation [19]

S. no	1	2	3	4	5	6	7	8	9	10
Parameters	N	α	r	i	s	d	S0	I0	R0	D0
Values	1000	0.01	0.4	0.01	0.02	0.6	900	10	0	0

Fig. 1

Variation of a Susceptible b Infected , c Recovered , d Death D population with time at diverse values of order of fractional derivative

Table 2

Error between consecutive iterations for at order

t	β=1		β=0.50
	S_1-S_0	S_2-S_1	S_1-S_0	S_2-S_1
0	0	0	0	0
0.005	0.50	0.008870666667	7.978845608	3.522509382
0.010	1.00	0.03546933333	11.28379167	7.022576161
0.015	1.50	0.07977600000	13.81976597	10.50803302
0.020	2.00	0.1417706667	15.95769121	13.98167506
0.025	2.50	0.2214333333	17.84124116	17.44511768

Table 3

Error between consecutive iterations for at order

t	β=1		β=0.50
	I_1-I_0	I_2-I_1	I_1-I_0	I_2-I_1
0	0	0	0	0
0.05	4.00	0.1033208333	20.18506018	4.856180562
0.15	12.00	0.9898875000	34.96154978	16.44994277
0.25	20.00	2.916354167	45.13516668	29.57547985
0.35	28.00	6.042720833	53.40464942	43.86244602
0.45	36.00	10.52898750	60.55518053	59.12587515

Table 4

Error between consecutive iterations for at order

t	β=1		β=0.50
	R_1-R_0	R_2-R_1	R_1-R_0	R_2-R_1
0	0	0	0	0
0.001	0.004	0.00001596	0.1427299293	0.031920
0.002	0.008	0.00006384	0.2018506018	0.063840
0.003	0.012	0.00014364	0.2472154893	0.095760
0.004	0.016	0.00025536	0.2854598586	0.127680
0.005	0.020	0.00039900	0.3191538243	0.15960

Table 5

Error between consecutive iterations for at order

t	β=1		β=0.50
	D_1-D_0	D_2-D_1	D_1-D_0	D_2-D_1
0	0	0	0	0
0.2	3.20	0.1920	8.074024070	1.920
0.4	6.40	0.7680	11.41839434	3.840
0.6	9.60	1.7280	13.98461991	5.760
0.8	12.80	3.0720	16.14804814	7.680
1.0	16.00	4.8000	18.05406667	9.600

Ebola virus may be used for bio-terrorism so the research on producing effective vaccines is in progress. The detailed study of EVD model given by Eq. (1) is relevant. Prevention from EVD can be done by the early identification of infected and stopping its transmission. A good control of outbreak depends on applying a set of involvements, such as case management, observation and contact tracing, better laboratory facility, safe burials and social mobilization. Ebanga and Inmazeb are the two monoclonal antibodies which were approved for the treatment of EVD infection by the US Food and Drug Administration in the late 2020 [10]. A vaccine rVSVdeltaG-ZEBOV-GP is found to be safe and immunogenic till now but more studies are required to stop new outbreaks. The present EVD model exposes a new feature of which was absent in integer order () (Figs. 2, 3, 4 and 5).

Fig. 2

Variation of Susceptibles with time at different

Fig. 3

Variation of number of infected with time at various

Fig. 4

Variation of with time at distinct

Fig. 5

Variation of with time at some

Conclusion

Severe devastation of mankind was observed in Africa by EVD few years ago. A time-fractional EVD model of arbitrary order is examined with efficacious use of FVIM in finding its new reliable solution. FVIM scheme is capable of lessening the size of calculation. The results verify that FVIM scheme is competent even with low iterations. The convergence analysis shows that the FVIM solution is convergent, bounded, positive. Tables display that the absolute error between consecutive iterations is insignificant for As a future work, certain hybrid methods using integral transforms and numerical methods can also be applied to solve the nonlinear fractional EVD models. A comparative study of results can be performed to obtain more accuracy. It is concluded that FVIM is highly efficient and influential in finding the trustworthy solution to the non-integer order nonlinear models of extreme significance for society.