Bi-Finite Difference Method to Solve Second-Order Nonlinear Hyperbolic Telegraph Equation in Two Dimensions

is study introduces a computational scheme by the bi-nite dierence method (Bi-FDM) to solve the hyperbolic telegraph equation in two dimensions.e proposed numerical method converts nonlinear two-dimensional hyperbolic telegraph equation of second order to dierence equations that can be solved by the Mathematica program. Consistency and stability of the proposed scheme are discussed and found to be accurate of O((hx) 2 + (hy) 2 + (Δτ)2) and conditionally stable, respectively. e eciency and accuracy of Bi-FDM have been shown by comparing the numerical results of the presented problems with the exact solutions and other numerical techniques.


Introduction
For δ > 0 and c > 0, we consider the following second-order nonlinear hyperbolic telegraph equation in two dimensions on region R (x, y): x ∈ [0, 1], y ∈ [0, 1] fV ττ V xx + V yy − 2δV τ − c 2 V + G(x, y, τ, V)or temporal coordinate with seconds τ > 0 [20]: with the initial conditions, Equation (1) is in spatial-temporal spaces and represents damped wave equation at δ > 0 and c 0 and telegraph equation at δ > 0 and c > 0. e two-dimensional nonlinear hyperbolic telegraph (1) of second-order under conditions (2) and (3) arises in electric voltage and current in a double conductor, signal analysis, wave propagation, random walk theory, and atomic physics [1][2][3][4]. Also, the hyperbolic partial di erential equations are useful in explaining and understanding structures vibrations such as buildings, beams, and machines [5]. So, most previous branches of science can be studied by telegraph equation which is more appropriate in modeling reaction-di usion than ordinary di usion equation. In most of the literature, the numerical schemes for hyperbolic telegraph equations of second order in one dimension and multidimensions recently developed it as follows. Mohanty et al. studied the stability of numerical schemes of telegraph equations [6][7][8]. Hyperbolic telegraph equations of one dimension are solved by cubic and quartic B-spline collocation methods [9][10][11]. In [12,13], Mohanty et al. used high-order approximation to solve two-dimensional quasilinear hyperbolic equations. Many authors solved two-dimensional linear hyperbolic equations [14][15][16][17][18]. Two-dimensional nonlinear hyperbolic equations of second order are studied using modi ed B-spline method [19], bicubic B-spline method [20], meshless local and stronge forms [21], and high-order di erence schemes [22]. Smith introduced nite di erence methods [23]. Raslan and Ali used the finite difference method to solve MHD-duct flow model [24]. Baleanu et al. applied a nonstandard finite difference scheme for fractional chaotic system [25]. Erturk et al. found some electrical engineering applications as capacitor microphone system [26] and description motion of beam on nanowire [27]. Hasan et al. [28] presented formulations of fractional optimal control of two spherical structures and one hollow cylindrical structure which represented fractional partial differential equations in two dimensions and can be solved via our presented scheme and will discuss it in the future work. e present study is organized as follows. In Section 2, we present the bi-finite difference method of the proposed problem. We discuss the consistency and stability of the proposed scheme as given in Sections 3 and Section 4, respectively. Section 5 investigates some numerical problems and explains them in some tables and figures. Finally, Section 6 provides the conclusion for this paper.

Bi-Finite Difference Method
We divide the region R into knots (x i , y j ), for i � 0, 1, . . . , n and j � 0, 1, . . . , m, in spatial directions with horizontal where n and m are + ve integers. Suppose that temporal direction with spacing Δτ > 0 is τ k � k * Δτ, for k � 0, 1, . . . , l, where l is +ve integer. e finite difference method in two dimensions to find the approximate solution v(x, y, τ) corresponding to the exact solution V(x, y, τ) can be written as and its partial derivative up to second order in spatial and temporal directions at various grid points (x i , y j , τ k ) is as follows: Equation (1), at points (x i , y j , τ k ), for i � 0, 1, . . . , n, j � 0, 1, . . . , m, and k � 0, 1, . . . , l, where V(x i , y j , τ k ), is written as V k i,j : Conditions (2) and (3) become Substituting (5) into (6), we obtain as follows: where (8) with conditions (7) can be solved.

Local Truncation Error
In this section, we introduce the local truncation error of our scheme.

Theorem 1.
e local truncation error of finite difference scheme (8) Proof. By using Taylor's expansion in (8), we investigate the local truncation error in two-dimensional space and time as follows: Hence, T(c k i,j ) ⟶ 0 as (h x ) 2 , (h y ) 2 , (Δτ) 2 ⟶ 0. en, the local truncation error of finite difference scheme (8) is □

Stability of Bi-Finite Difference Scheme
In this section, we test the stability of the difference scheme.

Theorem 2. e difference scheme (8) is conditionally stable.
Proof. Firstly, suppose c k i,j � ζ k e Iϑ(ih x ) e Iφ(jh y ) , and substituting with it in (8), we use the following linearization of the nonlinear term where , and we divided equation (12) by where en, If |B|⩽1 + δ * Δτ and δ * Δτ ⩽ 1, so, the bi-finite difference scheme is conditionally stable under the condition |B| ⩽ 2.

Numerical Problems
In this section, we test some numerical experiments for several two-dimensional hyperbolic telegraph problems using the bi-finite difference scheme (Bi-FDM). We calculate L ∞ -errors, L 2 -errors, root mean square errors (RMSE), and relative errors (RE) to show the efficiency of the proposed numerical scheme by comparing the exact solutions and the numerical solutions in our presented method and other different methods. e formulas of the used errors in twodimensional space at different levels τ are given by [16] Problem 1. We consider two-dimensional second-order nonlinear hyperbolic telegraph equation [20]: Subject to the initial and boundary conditions, for x, y ∈ [0, 1] and τ ⩾ 0, V(1, y, τ) � sin 1 cos y cos τ, x cos τ, V(x, 1, τ) � sin x cos 1 cos τ. (18) and the exact solution is V(x, y, τ) � sin x cos y cos τ, where δ � c � 1.

Mathematical Problems in Engineering
Keeping Δτ � 0.01, Table 1 presents L 2 , L ∞ errors and relative error for equal grid sizes h x � h y � 0.1, while Table 2 shows the same errors for different grid sizes h x � 0.1 and h y � 0.2 at different time levels. Our results are compared with the obtained results in [19,20]. Obviously, our numerical results are preferable. Figure 1 and 2 show the surface plots of numerical and exact solution at τ � 1, 3 for h x � h y � 0.05 and Δτ � 0.001, respectively. e maximum absolute error plots are observed that, with increase in time, the errors are decreasing, as shown in Figure 3.

Problem 2.
We consider two-dimensional second-order nonlinear hyperbolic telegraph equation [20]:  Root mean square error and relative error for equal grid sizes h x � h y � 0.1 and Δτ � 0.01 are calculated at different time levels, as shown in Table 3. Comparison of obtained results with results of other methods in [19,20] display that our numerical results are effective. Figure 4 appears the surface plots of numerical and exact solution at τ � 2 for h x � h y � 0.05 and Δτ � 0.001. Figure 5 presents error plots at various time levels. We find the maximum absolute errors are increasing with time increasing.
We tabulate root mean square errors and relative errors for h x � h y � 0.05 and Δτ � 0.001 at different time levels in Table 4. e acquired results are better than acquired results in [19,20]. Surface plots of and exact and numerical solution at τ � 1 for h x � h y � 0.05 and Δτ � 0.001 are as observed in Figure 6.
Problem 4. We consider two-dimensional second-order nonlinear hyperbolic telegraph equation [20]: Subject to the initial and boundary conditions, for x, y ∈ [0, 1] and τ ⩾ 0, We compute L 2 , L ∞ norms and relative errors for equal grid sizes h x � h y � 0.05 and Δτ � 0.001 at different time levels, as shown in Table 5. Table 5 clarifies our obtained results are in good agreement with results in [19,20]. Figure 7 illustrates surface plots of numerical solution with exact solution at τ � 1, for h x � h y � 0.1 and Δτ � 0.01. Figure 8 shows error plots are decreasing versus time increasing at various time levels for h x � h y � 0.1, Δτ � 0.01 and h x � h y � 0.05, Δτ � 0.001, respectively. Problem 5. We consider two-dimensional second-order nonlinear hyperbolic telegraph equation [20]: with the initial and boundary conditions, for x, y ∈ [0, 1] and τ ⩾ 0,

Conclusion
is study employed the bi-finite difference method (Bi-FDM) as a collocation method to solve a nonlinear hyperbolic telegraph equation of second order in two dimensions. e local truncation error of the proposed scheme is derived. e numerical scheme has conditionally stable. Five different forms of the equation under study were studied to verify the accuracy of the proposed method, and their L 2 , L ∞ norms were calculated. According to the obtained results, the proposed numerical scheme is efficient and in good agreement with other works in [19,20]. In addition, the Bi-FDM allows us to produce an extremely better estimation of the hidden aspects for a system of the different applications under study.

Data Availability
All data included in this study are available from the corresponding author upon request.