THIRD ORDER ITERATIVE METHOD FOR SOLVING NON-LINEAR PARABOLIC PARTIAL DIFFERENTIAL EQUATION IN FINANCIAL APPLICATION1 Department of Mathematics, College of Natural and Computational Sciences, Ambo University, Ambo, Ethiopia |
|
||
|
|||
Received 10 February 2022 Accepted 02 March 2022 Published 27 March 2022 Corresponding Author Kedir Aliyi Koroche, kediraliyi39@gmail.com DOI 10.29121/IJOEST.v6.i2.2022.299 Funding: This research
received no specific grant from any funding agency in the public, commercial,
or not-for-profit sectors. Copyright: © 2022 The
Author(s). This is an open access article distributed under the terms of the
Creative Commons Attribution License, which permits unrestricted use, distribution,
and reproduction in any medium, provided the original author and source are
credited. |
ABSTRACT |
|
|
In this paper, a third-order iterative scheme is
presented for searching approximate solutions of a non-linear parabolic
partial differential equation to simulate the elaboration of interest rates
in the fanatical application. First, by using Taylor series expansion we gain
the discretization scheme for the model problem. Then, using the Gauss-Seidel
iterative scheme we solve the proposed model problems. To validate the
convergences of the proposed numerical techniques, three model illustrations
are considered. The convergent analysis of the present techniques is worked
by supporting the theoretical and fine numerical statements. The accuracy of
the present numerical techniques has been measured by using average absolute
error root mean square error and point-wise maximum absolute error. Then, we
compare these get crimes with the result attained in the literature. These
results are also presented in tables and graphs. The comparison of physical
behavior between present numerical versus its exact solutions is also
presented in terms of graphs. As we can see from the table and graphs, the
present numerical techniques approximate the exact result veritably well. So,
it is relatively effective for simulating fanatical application to the
non-linear parabolic partial differential equation. |
|
||
Keywords: Non-Linear Parabolic Equation, Third-Order
Iterative Scheme, Convergent Analysis 1. INTRODUCTION
Partial differential equations arise not only from subfields within
mathematics such as differential geometry and analysis but also from almost
every scientific and engineering field as mathematical models of various
application problems Feng et al. (2013). As the behavior of the solutions underlying, these
application problems depend on governing partial differential equations. So,
Solving, analyzing, and implementing the solution to these partial
differential equations has been critically important for the resolutions of
many scientific and engineering application problems. Concerning different
criteria, partial differential equations can be categorized into several
types.
However, using nonlinearity as a criterion, partial differential
equations can be divided into two categories: linear partial differential
equations and nonlinear partial differential equations. Evans (1998), Taylor (1996),Taylor (1996) . In the nonlinear category, partial differential
equations are further classified as semi-linear partial differential
equations, quasi-linear partial |
|
||
differential equations, and fully non-linear partial differential equations based on the degree of the non-linearity. Semi-linear partial differential equations are differential equation that is non-linear in the unknown function but linear in all its partial derivatives. The non-linear Poisson equation is a well-known example of this class of partial differential equations. A quasi-linear partial differential equation is non-linear in at least one of the lower order derivatives but linear in the highest order derivative of the unknown function Feng et al. (2013).
Fully non-linear second-order partial differential equations arise from many fields in science and engineering such as astrophysics, antenna design, differential geometry, geotropic fluid dynamics, image processing, materials science Ambrosio et al. (2001), mathematics and finance Ambrosio et al. (2001), Aliyi et al. (2021) mesh generation, meteorology, optimal transport, and stochastic control Ambrosio et al. (2001). Various phenomena and applications Pardoux (2007), Prevot and Rockner (2007) and the references therein with stochastic influence in natural or artificial complex systems can be modelled by Stochastic partial differential equations, including stochastic quantization of the free Euclidean quantum field, turbulence, population dynamics and genetics, neurophysiology, the evolution of the curve of interest rate, non-linear filtering, movement by mean curvature in random environment, the hydrodynamic limit of particle systems, fluctuations of an interface on a wall, and path wise stochastic control theory Li et al. (2021), Alharbi (2020). In these fundamental applications, several examples of canonical Stochastic partial differential equations arise, such as the Zakai equation, reflected stochastic heat equation, stochastic reaction-diffusion equations, stochastic Burger’s equation, stochastic Navier–Stokes equation, stochastic porous media equation Li et al. (2021), and non-linear advection-diffusion equation. A common example of diffusion is given by heat conduction in a solid body Aliyi et al. (2021), Ahmed (2017). Conduction comes from molecular collision, transferring heat by kinetic energy, without macroscopic material movement Ahmed (2017). The application of non-linear partial differential equations is also found as the Black–Scholes model, see in Alharbi (2020).
The stochastic discrimination equation for the CIR garrulousness satisfies the Yamada-Watanabe condition, so it admits a unique strong result Gatheral and Taleb (2013), Rouah (n.d.). In fine finance, the Cox – Ingersoll – Ross (CIR) model describes the elaboration of interest rates. It's a type of short-rate model as it describes interest rate movements as driven by only one source of request trouble Orlando et al. (2018), Orlando et al. (2019). The model can be used in the evaluation of interest rate derivatives. A CIR process is a special case of an introductory affine jump-long-windedness, which still permits an unrestricted- form expression for bond prices. Time-varying functions replacing portions can be introduced in the model to make it harmonious with the assigned term structure of interest rates and possibly volatilities Orlando et al. (2018),. Also, non-linear equations appear in condensed matter, solid-state medicines, fluid mechanics, chemical kinetics, tube medicines, non-linear optics, propagation of fluxions in Josephson junctions, the proposition of turbulence, ocean dynamics, biophysics star evidence, and multitudinous others Maher et al. (2013). This non-linear equation has its own either exact or numerical result and these results show the behaviour of governing equation in the result intervals.
In recent years, directly searching either exact or numerical solutions of non-linear parabolic equations has become more and more attractive parts in different branches of physics and applied mathematics. The majority of non-linear parabolic partial differential equations do not have analytical solutions. But also, some numerical methods have a slow rate of convergence, instability, low accuracy, and difficulty in applying it to implement the simulation of non-linear parabolic partial differential equations in complex geometries. Therefore, due to this reason, several numerical methods have been developed for investigating the simulation of a non-linear parabolic equation. For instance, in Nhan et al. (2021), the high-order iterative scheme was used for the study of the non-linear pseudo-parabolic equation. They apply the Faedo-Galerkin approximation method and use basic concepts of non-linear analysis, but grid generation is usually more automatic for the Galerkin approximation method, although not completely for complex geometries. Also, this iterative scheme is used to search the solution of stochastic parabolic equations and study the existence of these solutions in Ngoc et al. (2010), Truong et al. (2009), Ahmed (2017). Thus, they get a better approximation of their applied governing problem.
The stochastic parabolic original value problem also
mainly shops and freckled to the operation of financial models of the
stochastic volatility model. The nontrivial point of the equation appears from
the non-linear first-order term in spatial variable and a Holder, yet not
Lipchitz, too rough to be differentiable in space and time. Indeed, still, with
dropping the quadratic non-linearity from these parabolic partial
discriminative equations, it's reduced to the stochastic heat equation with the
mean-returning term, whose result is not differentiable Li et al. (2021),. This system does not always meet the exact results for
coarser step lengths. Every type of finite element system depends on the number
of grid points.
Numerical and analytical techniques for solving conformable parabolic partial differential equations and conformable initial boundary value problems also have been investigated in Yavuz and Ozdemir (2017), Yavuz (2018) respectively. But the conformability transform is not only useful to solve local conformable fractional nonlinear dynamical systems of problems. Kocacoban et al. solved the Burgers–Fisher equation by applying various numerical schemes Kocacoban et al. (2011) that showed relatively faster convergence than other plots Lima et al. (2021). The collocation method is also a numerical method that the researcher used to obtain approximate solutions of non-linear parabolic types of partial differential equations in Hepson (2021). Hence, the researchers were developing the high-speed computers allows and improvements for the algorithm of several numerical methods to solve non-linear parabolic types of partial differential equations on both complex domain and complicated boundary conditions in different applications. For instance, this powerful series approach was applied by several researchers to find the solution of the Burger–Fisher equation, which is called a non-linear Parabolic Partial differential equation, see Behzadi and Araghi (2011).
However, each class of methods offers numerous and, in many ways, complementary benefits. Considering previous studies, it has been perceived that either analytical or numerical solutions of non-linear parabolic equations are very scarce. In the ideal case, we seek a method defined on arbitrary geometries that behaves regularly in any dimension and avoids the cost of time-consuming and mesh generation. As a result, many investigators have decided to advance too accurate and efficient numerical methods. Among those numerical methods, the Finite Difference scheme produces potential outcomes and has been widely used despite some limitations, such as being unable to obtain the solutions at every single point between two grid points. Another drawback is the computational cost to obtain higher accuracy of the numerical solution. Motivated by all the above studies, we come up with the idea to study the non-linear parabolic partial differential equations in one-dimensional space. Therefore, the main goal of this paper is to apply Guess-seidel iterative method to approximate the solution of the Non-linear parabolic partial differential equation and searching the results with its accuracy increases through iterative steps. The convergence of the present numerical scheme has been measured in the sense of average absolute error (AAE), maximum point-wise absolute error (), and root mean square error (). The stability and confluence of the present techniques are also delved by using Von Neumann stability analysis techniques. In this paper, we consider the non-linear parabolic partial differential equation given by:
Subjected to both initial and boundary
conditions are given by:
,
where is
arbitrary constant and ,
,
& are smooth function in This smoothness of this function is used for the
existence of solutions in the domain. Moreover, the existence of solution of
non-linear partial differential equation in the solution domain is studded in
the references Ngoc et al. (2010), Truong
et al. (2009), Ahmed
(2017). To find the solution
in this paper, the rectangular domain can be partitioned into sub-intervals
given by:
, Equation 3
where &, where & . & are the maximum numbers of grid points respectively in the x and t direction. Therefore, this paper is organized as follows. Section two is a description of the numerical scheme, section three is confluence analysis, and section four is the results of numerical experiments. Section five is Discussions of numerical experiments; section six is the conclusion.
2. DESCRIPTION OF NUMERICAL SCHEME
Recall that non-linear parabolic partial differential equation in Equation 1 with their initial and boundary condition in Equation 2 and we want to approximate solution in the rectangular domain. Now to approximate this model problem, first, we want to discretize its derivative concerning in both temporal variable and spatial variable by using Taylor series expiation.
2.1. DISCRETIZATION OF TEMPORAL DERIVATIVE
Assume that has continuous
higher order partial derivative on its domain. Now, let us consider that, where which we call, order partial derivative of
concerning spatial variable. Now the Taylor series expansions of,, and about given by
Equation 6
The first and second-order
finite difference scheme for first, second, and third-order partial derivative
concerning with temporal variable is
,
Now combining Equation 4, Equation 5, and Equation 7 we obtain
Using Equation 8 in this difference in
terms of second and third-order partial derivative we obtain
This implies that:
Simplifying the difference
result, we obtain the third-order finite difference scheme for the first-order
finite difference scheme of the form
where is its maximum local truncation error term?
2.2. DISCRETIZATION OF SPATIAL DERIVATIVE
Whit-out losing generality, the discretization of temporally derivative by using Taylor series expiation given by:
Equation 12
Now using Equation 10, Equation 13, the
second-order finite difference scheme for first and second-order partial
derivative concerning spatial derivative is given by:
Equation
15
Equation
16
Equation 17
Without losing generality,
combining Eqs. Equation 10, Equation 11, and Equation 13, we obtain
the differential equation given by:
Substituting Equation 14 into this
difference equation, we obtain:
Now multiplying both sides and simplifying the result, we obtain:
By truncating the last
(truncation) error terms from this difference scheme, we obtain a third-order central
difference scheme for the second-order partial derivative of the model problem
is given by:
where is their maximum local truncation error term?
Now substituting Equation 9, Equation 18 into Equation 1 we obtain
the difference scheme:
This implies that:
where . Then we can rewrite the
iterative scheme by using the Gauss-seidel iterative scheme as a form of:
where and . From this scheme, we can rewrite this, into the embedded matrix
form of Gauss-seidel iterative scheme:
Equation 19
Where and are lower, diagonal, and an
upper triangular matrix is respectively given by:
, and
And
Equation 20
Where is a local truncation error? Hence, the
simplified form of Guess-seidel iterative scheme is:
where and
Thus, using Guess-seidel iterative scheme in Equation 18 by rewriting MATLAB program, we obtain the solution of model problem and validity of proposed numerical scheme.
3. CONSISTENCY AND CONVERGENCE ANALYSIS
3.1. THE CONSISTENCY OF THE PROPOSED SCHEME
Since from
the general iterative scheme, local truncation error is. Hence by using the definition
referenced in Morton
and Mayers (2005), we have:
Hence this indicated that, as simultaneously
both step-length and time step approach to zero (i.e),
the truncation error in difference scheme is approximate to zero (i.e.,,). So, this shows that, the above iterative scheme is consistent.
3.2. CONVERGENCE ANALYSIS
The convergence of the proposed numerical method is
investigated by using matrix form convergence analysis. Such an approach has been
used in many textbooks and different recent articles. As it worked in reference
Mohanty and Jha (2005), JAIN et al. (1984) assume that is the exact solution of the
problem in Equation 1 and it can
be a writer as:
Subtraction Eq. Equation 21 form Eq. Equation 22 and substituting ,
we obtain:
where and it follows that. . Since in our work we follow that error produced in this scheme is less-than principal local truncation error produced in sequences of the scheme. It means that where is maximum local truncation error term at point. This shows the stability of the scheme. Now assuming that matrix is irreducible and monotone Mohanty and Jha (2005),. This shows that the inverse of A exists, and its elements are nonnegative. Hence, using matrix norm, from Equation 23, we get
Thus, we define the norms of the matrix
as constant which is given by
where and
norm of truncation error.
Therefore, we have
However, all error in the scheme is
bounded. Thus, we have.
Theorem 1: Let in
Equation 23 be a square matrix and
is distinct Eigenvalue of matrix A.
Then if or where is the spectral radius of matrix A and.
Proof: If,
we have and.
For simplicity, assume that all the eigenvalues of A are distinct. Then, there
exists a similarity transformation P, such that where D is the diagonal matrix having the
eigenvalues of A on the diagonal. Therefore, and
This is implying if and only if all Eigenvalue of A satisfies.
Therefore.
Theorem 2: A
necessary and sufficient condition for convergence of an iterative method of
the form given in Equation 21 is that the
eigenvalues of the iteration matrix satisfy,
Proof: Prove of this theorem is given in JAIN et al. (1984).
Therefore, by supporting these two theorems with the above theoretical and numerical error bound, the present method is convergent with third-order convergence. Hence to measure the accuracy of the proposed method, we use norms of average absolute error (AAE), root mean square (RMS) error () and maximum point-wise absolute error (). These norms are calculated as follows:
,
Where and are the respectively exact and numerical solutions of the given model example.
4. RESULTS OF NUMERICAL EXPERIMENTS
To demonstrate the applicability of the methods, three model examples have been considered and they are bellowed.
Example 1. Consider the following nonlinear parabolic problem:
;
,
Initial condition ,
Boundary condition
Exact Solution:
Here source
function:
Example 2. Consider the following
nonlinear parabolic problem:
;
,
Initial condition ,
Dirichlet boundary condition
Exact Solution:
Here source function:
Example 3. Consider the
following nonlinear parabolic problem:
;
,
Initial condition ,
Robin boundary condition
Exact Solution:
Here source function:
Table 1 Displaying the efficiency of the proposed scheme by listing Exact
solution, Numerical solution, and point-wise absolute error, for problem give an example one for Mx=20
and at time t=1.154 when computations domain carried out until final time T=2 |
|||
Values
of x |
Numerical
and Exact solution |
Point-wise
Absolute Error |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Table 2 Displaying the efficiency of the proposed scheme by listing point
average absolute error, Root Mean Square Error, and pointwise maximum
absolute error for problem give an example one when computations domain
carried out until final time T=1 for different mesh size h and time step
∆t |
||||
Specified
Mesh size |
Estimated
errors at Specified Mesh size |
|||
h |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Figure 1 Solution profile for the solution of
example one on the uniform mesh of maximum number grid point is & time step is |
|
Figure 2 Graphical comparison of numerical versus
exact solution of example one using a uniform mesh with maximum number grid
point & time step is |
Table 3 Comparison of maximum
point-wise absolute error and Root Mean Square error for problem give an example two for
computations domain carried out until final time with different mesh size h and time step and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) |
Result by Present Methods |
|||
|
|
|
|
|
32 |
|
|
2.22E-06 |
3.92E-07 |
64 |
|
|
2.22E-06 |
2.77E-07 |
128 |
|
|
2.22E-06 |
1.96E-07 |
256 |
|
|
2.22E-06 |
1.39E-07 |
512 |
|
|
2.22E-06 |
9.80E-08 |
Table 4 Comparison
of maximum point-wise absolute error and Root Mean Square error for problem
give an example two for computations domain
carried out until final time with different mesh size h and time step, and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) |
Result by Present Methods |
|||
|
|
|
|
|
32 |
|
|
2.21E-05 |
3.91E-06 |
64 |
|
|
2.21E-05 |
2.76E-06 |
128 |
|
|
2.21E-05 |
1.95E-06 |
256 |
|
|
2.21E-05 |
1.38E-06 |
512 |
|
|
2.21E-05 |
9.77E-07 |
Table 5 Comparison of maximum
point-wise absolute error and Root Mean Square error for problem give an example two in
computations domain carried out until final time with different mesh size h and time step, and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) |
Result by Present Methods |
|||
|
|
|
|
|
32 |
|
|
6.71E-06 |
1.19E-06 |
64 |
|
|
1.39E-05 |
1.7356eE-06 |
128 |
|
|
2.86E-05 |
2.53E-06 |
256 |
|
|
5.77E-05 |
3.60E-06 |
512 |
|
|
1.16E-04 |
5.11E-06 |
|
Figure 3 Graphical representation of numerical solution for example two using
uniform mesh mesh-size & time step
and and and |
Table 6 Comparison of maximum
point-wise absolute error and Root Mean Square error for problem give an example three in
computations domain carried out until final time with different mesh size h, and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) time step |
Result by Present Methods time step |
|||
|
|
|
|
|
32 |
|
|
1.76E-05 |
3.12E-06 |
64 |
|
|
1.76E-05 |
2.20E-06 |
128 |
|
|
1.76E-05 |
1.56E-06 |
256 |
|
|
1.76E-05 |
1.10E-06 |
512 |
|
|
1.76E-05 |
7.79E-07 |
Table 7 Comparison of maximum
point-wise absolute error and root mean square error for problem give an example three in
computations domain carried out until final time with different mesh size h, and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) time step |
Result by Present Methods time step |
|||
|
|
|
|
|
32 |
|
|
7.18E-07 |
1.27E-07 |
64 |
|
|
7.15E-07 |
8.94E-08 |
128 |
|
|
7.15E-07 |
6.32E-08 |
256 |
|
|
7.14E-07 |
4.47E-08 |
512 |
|
|
7.14E-07 |
3.16E-08 |
Table 8 Comparison of maximum
point-wise absolute error and root mean square error for problem give an example three in
computations domain carried out until final time with the different mesh size of , and |
||||
Number of grid points |
Estimated norm of error with Specified Mesh size |
|||
Result by R. Alharbi in Alharbi (2020) |
Result by Present Methods |
|||
|
|
|
|
|
32 |
|
|
5.76E-07 |
1.02E-07 |
64 |
|
|
1.09E-06 |
1.36E-07 |
128 |
|
|
2.13E-06 |
1.88E-07 |
256 |
|
|
4.18E-06 |
2.61E-07 |
512 |
|
|
8.27E-06 |
3.65E-07 |
|
Figure 4 Graphical representation of the numerical solution of example three
using a uniform mesh with different mesh-size & time step is , and |
|
Figure 5 Solution profile for the solution of example three on the uniform
mesh of maximum number grid point is , the time step
is and |
5. DISCUSSION
In this paper, the third-order iterative scheme is presented to solve a one-dimensional non-linear parabolic partial differential equation. To demonstrate computation for the accuracy of percent of the method with the pre-existing method, three model examples are solved by taking different values for step size h, and time step k. The computation of numerical results obtained by the present method has been presented in terms of average absolute error, root means square error and maximum point-wise absolute error. Results presented in Table 1 and Table 2 show that average absolute error (AAE), roots mean square error and point-wise maximum absolute error norm decreases as mesh-size and are decreases. Again, as we can see the comparison of the numerical and exact solution of the model problem given in example one summarized in Table 1 at selected grid points shows that the numerical solution is in good agreement with its exact solution. Also, the results of example two, given in Table 3 up to Table 5 show that, the accuracy of the present iterative method increases, and it's superior to the accuracy of the scheme in Alharbi (2020). Figure 1,Figure 2, Figure 3,Figure 4, Figure 5 shows the physical background of the solution within the traditional form of future expected interest rates for each maturity. It means it shows that the forecasting rate of change of the total amount of interest rates is increasing to decreasing for each maturity of U in any sections of with a balance to the net inflow of interest across 0 to the 0.5-time interval. Also, in this case, the accuracy of the present method is rapidly increased and it’s superior to the accuracy of the scheme in Alharbi (2020). Further, as shown in Figure 3, the proposed method approximates the solution very well. The result presented in Table 6,Table 7,Table 8 also shows that; the root mean square error norm and maximum absolute error norm are decreased uniformly for solving example three with different values of step-length h, time-step, and constant f variation As we predict from this result, the accuracy of the present method is rapidly increased and it’s superior to the accuracy of the scheme in Alharbi (2020). Therefore, the accuracy of the present method confirmed the established numerical error bound. Hence in solving all applied three model examples, the results given in tables in terms of error norm and graphs of numerical versus exact solutions are further confirmed that the computational rate of convergence and theoretical estimates error bounds
6. CONCLUSION
A new approach, a third-order iterative scheme is used
to solve nonlinear parabolic partial differential equations numerically and the
result is presented in table and graph. The comparison of the accuracy
indicates that; the present method is the more convenient, reliable, and
effective scheme. As it can be seen from the table and graphs, the present
methods, improve the accuracy, by minimizing the number of grid points in a
time interval and an equal number of grid points in the spatial interval with
pre-existing methods. This shows that the present method avoids the cost of
time-consuming and meshes generation. In a summary, the third iterative scheme
is capable to solve nonlinear parabolic partial differential equations. Based
on the findings, this method is well approximate and gives better accuracy for the
numerical solution with a decreasing step size h, and fixed time step ∆t.
AUTHOR CONTRIBUTIONS
The author planned to work by this scheme, initiated
the Research idea, suggested their experiment, conducted the experiments, and
analyzed the empirical results. Also, the Author developed the mathematical
modeling and examined the numerical validation and was written their manuscript
properly. The author reviews its results and approves the final version of the
manuscript.
ACKNOWLEDGMENTS
First of all, the author wishes to express his thank to Allah, which gives him full health, and for granting him this opportunity to broaden his knowledge in this field. Next, the author wishes to express his thank to the authors of the literature for the provision of the initial idea for this work. Last but not least, the author wishes to express his thank to all his beloved friends; their kindness and bits of help be a great memory for him.
NOMENCLATURE
|
Root mean square error Maximum Pointwise error Average Absolute Error Maximum Number of the
grid point in temporal direction |
|
step length time step maximum number grid point
in the spatial direction, |
REFERENCES
Ahmed H., (2017) Faedo-Galerkin method for heat equation, Global J. of Pure and App. Math. ,13(4), 1195-1207.
Alharbi R. (2020), Nonlinear Parabolic Stochastic Partial Differential Equation with Application to Finance, A University of Sussex PhD thesis.
Aliyi K., Shiferaw A., and Muleta H. (2021). Radial Basis Functions Based Differential Quadrature Method for One Dimensional Heat Equation, American J. of Math. And Comp. Mod., 6(2), pp.35-42. Retrieved from https://doi.org/10.11648/j.ajmcm.20210602.12
Ambrosio L., Caffarelli L. A., Brenier Y., Buttazzo G., and Villani C. (2001). Optimal Transportation and Applications : Lectures from the C.I.M.E. Summer School held in Martina Franca.
Behzadi S. S. and Araghi M. A. F. (2011). Numerical solution for solving Burger's-Fisher equation by using iterative methods, Math. Comput. Appl. 16(2), 443-455. Retrieved from https://doi.org/10.3390/mca16020443
Evans L. C (1998). Partial Differential Equations, Grad. Stud. Math. AMS. Providence. RI.
Feng X, Glowinski R., and Neilan M. (2013), Recent developments in numerical methods for fully nonlinear second-order partial differential equations, Siam. rev. vol. 52, pp. 205-67. Retrieved from https://doi.org/10.1137/110825960
Gatheral J., and Taleb N. N. (2013). The Volatility surface A Practitioner's Guide. John Wiley & Sons, New York, NY. (2011). OCLC : 899182374. Retrieved from https://doi.org/10.1002/9781119202073
Hepson O. E. (2021). An exponential cubic B-spline algorithm for solving the nonlinear Coupled Burgers' equation, Comp. Methods for Dif. Eq.,9, (4), pp.1109-1127.
JAIN M. K., IYENGAR S. R. K., and JAIN R. K. (1984). Numerical Methods for Scientific and engineering Computation, Published in the Western Hemisphere by Halsted Press, A Division of John Wiley & Sons, Inc., New York, New Delhi October
Kocacoban D., Koc A. B., Kurnaz A., and Keskin Y., (2011). A better approximation to the solution of Burger-Fisher equation, in Proceedings of the World Congress on Engineering 1, 1-6.
Li Y., Shu C. W., and Tang S. (2021). A Local Discontinuous Galerkin Method For Nonlinear Parabolic SPDES, ESAIM Math. Mod., and Num. Anlys. Vol.55, pp. S187-S223. Retrieved from https://doi.org/10.1051/m2an/2020026
Lima A. S., Kamrujjaman M., and Islam M. S. (2021). Numerical solution of convection-diffusion-reaction Equations by a finite element method with error correlation, AIP Advances, vol.11, 085225. Retrieved from https://doi.org/10.1063/5.0050792
Maher A., El-Hawary H. M., and Al-Amry M. S. (2013). New Exact Solutions for New Model Nonlinear Partial Differential Equation, J. of Appl. Math., 2013. Retrieved from https://doi.org/10.1155/2013/767380
Mohanty R. K., and Jha N. (2005). A class of variable meshes spline in compression methods for singularly perturbed two-point singular boundary-value problems, Appl. Math. &Comp., Vol.168, 704-716. Retrieved from https://doi.org/10.1016/j.amc.2004.09.049
Morton K.W., and Mayers D. F (2005). Numerical Solution of Partial Differential Equations, An introduction, Second Edition, Cambridge University Press, New York. Retrieved from https://doi.org/10.1017/CBO9780511812248
Ngoc L. T. P., Son L. H. K., and Long N. T. (2017). An N-order iterative scheme for a nonlinear Carrier wave in an annular with Robin-Dirichlet conditions, Nonlinear Functional Analysis and applications, 22 (1), pp.147-169, Retrieved from https://digital.lib.ueh.edu.vn/handle/UEH/56185
Ngoc L. T. P., Truong L. X., and Long N. T. (2010). High-order iterative methods for a nonlinear Kirchhoff wave align, Demonstrations Mathematica, 43(3) ,pp. 605-634. Retrieved from https://doi.org/10.1515/dema-2010-0310
Nhan N. H., Dung T. T. M, Thanh L. T. M., Ngoc L. T. P., and Long N. T. (2021). A High-Order Iterative Scheme for a Nonlinear Pseudo-parabolic Equation and Numerical Results. Mathematical Problems in Engineering. Retrieved from https://doi.org/10.1155/2021/8886184
Orlando G., Mininni R. M., and Bufalo M (2018). A New Approach to CIR Short-Term Rates Modelling. New Methods in Fixed Income Modeling, Contributions to management Science. Springer International Publishing, pp. 35- 43. Retrieved from https://doi.org/10.1007/978-3-319-95285-7_2
Orlando G., Mininni R.M.,
and Bufalo M. (1 Jan. 2019), A new approach to
forecast market interest rate through the CIR model, Studies in Economics and
Finance, 267-292. Retrieved from https://doi.org/10.1108/SEF-03-2019-0116
Pardoux E. (2007), Stochastic Partial Differential Equations. Lecture notes for the course given at Fudan University, Shanghai.
Prevot C., and Rockner M. (2007). A Concise Course on Stochastic Partial Differential Equations. In : Vol. 1905 of Lecture Notes in Mathematics. Springer, Berlin.
Rouah F., (n.d.) Hesston Model and its Extensions in Matlab and C. Hoboken, Wiley.
Taylor M. E. (1996), Partial Differential Equations II : Qualitative Studies of Linear Equations, Appl. Math. Sci. Springer, New York.
Taylor M. E. (1996). Partial Differential Equations I : Basic Theory, Appl. Math. Sci. Springer, New York.
Truong L. X., Ngoc L. T.
P., and Long N. T. (2009). The n-order iterative
schemes for a nonlinear Kirchhoff-Carrier wave equation associated with the
mixed inhomogeneous conditions, Applied Mathematics and Computation, 215(5),
pp. 1908-1925. Retrieved from https://doi.org/10.1016/j.amc.2009.07.056
Truong L. X., Ngoc L. T. P., and Long N. T. (2009). High-order iterative schemes for a nonlinear Kirchhoff-Carrier wave align associated with the mixed homogeneous conditions, Nonlinear Analysis : theory, Methods & Applications, 71(1), pp. 467-484. Retrieved from https://doi.org/10.1016/j.na.2008.10.086
Yavuz M. (2018). Novel solution methods for initial boundary value problems of fractional order with conformable differentiation, An International J. of Opt. and Cont. Theory. And appl.,8(1), pp.1-7. Retrieved from https://doi.org/10.11121/ijocta.01.2018.00540
Yavuz M. and. Ozdemir N. (2017). New numerical techniques for solving fractional partial differential equations in conformable sense, in Non-integer Order Calculus and its Applications, pp. 49-62. Retrieved from
https://doi.org/10.1007/978-3-319-78458-8_5
This work is licensed under a: Creative Commons Attribution 4.0 International License
© Granthaalayah 2014-2022. All Rights Reserved.