Abstract
In this article, we design a novel fourthorder and derivative free rootfinding algorithm. We construct this algorithm by applying the finite difference scheme on the wellknown Ostrowski’s method. The convergence analysis shows that the newly designed algorithm possesses fourthorder convergence. To demonstrate the applicability of the designed algorithm, we consider five reallife engineering problems in the form of nonlinear scalar functions and then solve them via computer tools. The numerical results show that the new algorithm outperforms the other fourthorder comparable algorithms in the literature in terms of performance, applicability, and efficiency. Finally, we present the dynamics of the designed algorithm via computer tools by examining certain complex polynomials that depict the convergence and other graphical features of the designed algorithm.
1. Introduction
The role of computers in the fields of applied Mathematics cannot be denied in the modern age. Using different computer programs such as Mathematica, Matlab, and Maple, a plethora of different types of complex problems can be solved easily. In recent years, mathematicians employed the excessive use of computers in different branches of Mathematics especially in the determination of approximated roots of the transcendental and nonlinear algebraic equations which have played an important role in different branches of computational and applied mathematics. In many engineering disciplines, a lot of problems exist which can be easily converted into nonlinear forms by employing different mathematical techniques. Analytical methods cannot find the solution needed for these problems, and therefore, we need iterative algorithms for solving out these problems. To execute an iterative algorithm, we always need a starting point (initial guess) which is refined after every iteration, and we find the approximated root up to the required accuracy after some finite iterations. The convergence rate and convergence order of an iterative algorithm are relied upon the selection of that starting point. Some of the most popular and ancient iterative algorithms are given in [1–8] and the references are cited therein. In the 15^{th} century, Newton [1, 2] introduced a quadraticorder rootfinding algorithm which has been used successfully for many years. Over the time, many experts worked on iterative algorithms and brought several modified versions of Newton’s algorithm with higherorder convergence which involve predictor and corrector steps and are often referred to as multistep iterative algorithms. For more information, one can see [9–20] and the references are cited therein. In general, the convergence order of a multistep algorithm is higher because of predictor and corrector steps, but it results in a higher computational cost which is the downside of these algorithms. It is really difficult to handle the cost of computing and the convergence rate of an algorithm because these two terms are inversely proportional to each other.
Over the past few years, mathematicians have focused on the aforementioned issues and have tried to design some new iterative algorithms with higher convergence and low cost of computing by employing several mathematical methods. In [21], the authors introduced a new twostep Halley’s method with sixth order convergence and then replaced its second derivative for reducing computing cost and proposed a new fifth order second derivative free algorithm. In [22], Hafiz and AlGoria established a novel family of optimal eighth order iterative algorithms and then studied their dynamics. In [23], the authors introduced seventh and ninth orders novel iterative algorithms with the help of the predictorcorrector technique and Simpson quadrature formulae. By employing the Newton interpolation technique along with weight functions, Salimi et al. [24] introduced a new family of eighth order optimal rootfinding algorithms. In [25], the authors constructed some novel optimal iterative algorithms with higher convergence and demonstrated the applicability of the suggested methods by solving some engineering problems. Recently, Chu et al. [26] proposed a novel family iteration scheme and discussed the dynamics of the presented methods with the help of computer tools.
In the present research article, we introduce a new fourthorder and derivative free algorithm for solving engineering problems in the form of scalar nonlinear functions. The construction of this algorithm is based upon the finite difference scheme on Ostrowski’s method. We also certify that the designed algorithm has fourthorder convergence. The designed algorithm is then applied to some realworld engineering problems for certifying its better performance and applicability among the other fourthorder algorithms in the literature. The dynamical comparison of the designed algorithm with the other comparable ones has been also presented via the computer program Mathematica 12.0.
2. Main Results
Consider the nonlinear problem of the following form:where is a realvalued function with an open interval domain.
Suppose that is a root of (1) with as an initial guess near to the exact root , then the implication of Taylor’s series around for (1) gives us
If is nonzero, then the above expression implieswhich is Newton’s rootfinding algorithm [1, 2] for scalar nonlinear functions.
By taking it as a predictor, Ostrowski designed the following twostep iterative algorithm:which is wellknown Ostrowski’s rootfinding algorithm [11] for scalar nonlinear functions.
By including Newton’s algorithm, the above twostep method may be converted to threestep in the following form:which is a threestep iteration scheme for calculating zeros of nonlinear scalar equations. The main drawback of the above algorithm is its high computational cost per iteration as it requires six evaluations for its execution. To lower its computational cost make it more effective, we approximate its first derivatives and make it derivative free, so that it can be easily applied on those nonlinear scalar functions whose first derivative becomes infinite or does not exist. To approximate in the predictor step, we employ the forward difference approximation as
To approximate , we utilize the finite difference scheme as
Using (6) and (7) in (5), we can write Algorithm 1.
Algorithm 1. For a given , compute the approximate solution by the following iterative schemes
Algorithm 1 is a new iteration scheme for calculating the approximated roots of scalar nonlinear equations and needs only four evaluations per iteration. The main characteristic of the suggested algorithm is that it is derivative free and easily applicable to all those scalar functions whose derivatives become undefined within the domain. In this sense, the proposed algorithm’s computing cost is minimal which results in a higher efficiency index.
3. Convergence Analysis
In the present section, we shall discuss the convergence criterion of the newly designed algorithm, i.e., Algorithm 1.
Theorem 1. Suppose that is the root of the equation . If is sufficiently smooth in the neighborhood of , then the order of convergence of Algorithm 1 is at least four.
Proof. To analyze the convergence criterion of the iteration scheme (1), we assume that is a root of equation and be the error at ^{th} iteration; then, , and by using Taylor’s series expansion, we havewhereWith the help of equations (8) and (9), we getUsing equations (8)–(14) in Algorithm 1 gives us the following equality:which implies thatThe above equation shows that the designed algorithm is of fourthorder convergence.
4. RealLife Applications
In this section, we take five realworld problems in the form of scalar nonlinear functions to exhibit the applicability, validity, and efficiency of the newly designed fourthorder algorithm. We compare it with other wellknown fourthorder algorithms, namely, Ostrowski’s method (OM) [11], Traub’s method (TM) [12], and Zhanlav method (ZM) [27].
Example 1. Fluid permeability problem:
The hydraulic permeability is actually the measurement of the flow resistance. It relates the pressure gradient to fluid velocity and may be expressed aswhere denotes the specific hydraulic permeability, stands for the radius, and is the porosity. For further details see [28] and the reference cited therein. By taking the values of and in (14), we obtain the above problem in the following nonlinear function:To solve the initial guess has been chosen as for starting the iteration process, and the results are given in Table 1.
Example 2. Blood rheology model.
Blood rheology is a branch of science that works to study the physical and flow properties of blood [29]. Blood is actually a nonNewtonian fluid and treated as Caisson fluid. The model of Caisson fluid shows that the flow of simple fluids in a tube is such a way that the center core of the fluids will move as a plug with little deformation and velocity gradient occurs near the wall.
To study the plug flow of Caisson fluids flow, we consider the following function in the form of nonlinear equation aswhere flow rate reduction is computed by . Using in (19), we haveTo solve the initial guess has been chosen as for starting the iteration process, and the results are given in Table 2.
Example 3. Van Der Wall’s equation.
The wellknown equation for examining the behaviour of real and ideal gas was introduced by Van Der Wall’s [30], with the following expression:Equation (21) may be easily transformed into the following nonlinear function by taking the particular values of the parameters:where is the gas volume that may be easily determined by solving . Because the polynomial’s degree is three, it must have three roots. There is only one positive real root 1.9298462428 among these which is physically possible since the gas volume can never be negative. To solve the initial guess has been chosen as for starting the iteration process, and the results are given in Table 3.
Example 4. Plank’s radiation law.
The energy density within the black isothermal body is calculated using Planck’s radiation law [31] given as follows:Suppose we want to calculate wavelength for the peak value of the energy density . To transform (23) in nonlinear form, we assume and obtain the following nonlinear expression:One of the estimated roots of is 0.0000000000000000 which represents the maximum amount of the wavelength of the radiation. To solve the initial point has been chosen as for starting the iteration process, and the results are given in Table 4.
Example 5. The problem of beam designing.
In Physics and Engineering sciences, the beam designing problem [32] regarding the embedment of a sheet pile wall in the form of scalar nonlinear function is expressed asTo solve the initial guess has been chosen as for starting the iteration process, and the results are given in Table 5.
Here, we choose the accuracy in the following stopping criterion of the computer program:We used the computer application Maple 13 to solve all numerical problems.
Tables 1–5 exhibit the numerical comparison of the designed fourthorder algorithm with Ostrowski’s method (OM), Traub’s method (TM), and Zhanlav’s method (ZM). In the columns of the above tables, stands for the number of iterations, indicates the positive value of the function , indicates the estimated root, indicates the absolute difference of the consecutive estimations , and and represent the approximated computational order of convergence given aswhich was introduced by Weerakoon and Fernando [33].
5. Dynamical Analysis via Computer Technology
In this section, we give a detailed graphical comparison of the newly designed fourthorder algorithm with the other fourthorder algorithms via computer technology by considering some complex polynomials in the form of polynomiographs. A polynomiograph is a graphical object generated in a process known as polynomiography, introduced by Dr. Bahman Kalantri in 2005 [34]. It is defined as “the algorithmic visualization of polynomial equations by employing different iterative techniques” [35].
To draw dynamics by employing computer technology using various iterative algorithms, an initial rectangle which includes the root of the investigated complex polynomial has been chosen. Then, for every point in , we perform the process of iteration. The image’s quality is usually correlated with the discretization of, i.e., if the rectangle has been discretized into a grid, then the quality of the produced image will be better.
Typically, the colors of produced polynomiographs are fully associated with the number of iterations required to find the approximated roots with a given precision and a selected iterative algorithm. The main algorithm for the production of a polynomiograph is given in Algorithm 1.

A stopping criterion is always required for an iterative algorithm that includes the repetition of steps, since it informs us about the convergence or divergence of the investigated iterative algorithm. Such a criterion is commonly referred to as a convergence test with the following mathematical expression:where and denote the successive iterations, and >0 stands for the accuracy in the stopping criterion. The convergence test is considered TRUE if the iterative algorithm under consideration is converged and FALSE if it is diverged. The abovementioned stopping criterion (28) is also used in this study. The variety of polynomiographs’ colors is correlated with the performed iterations to find out the root with given precision . Using various iterative algorithms, a variety of aesthetically pleasant polynomiographs can be produced by altering the parameter K, where K specifies the upper limit of the number of iterations. For further information regarding polynomiography along with its applications in different fields, one can see [36–44] and the references cited therein.
For drawing polynomiographs through different iterative algorithms, we consider the following four complex polynomials:
The colormap used for the coloring of iterations in the generation of polynomiographs is shown in Figure 1:
Example 6. Polynomiographs for the polynomial through different fourthorder algorithms.
In the first example, we consider a cubicdegree polynomial having three distinct roots , and .We used a computer program to run all the methods to get the simple roots of the under consideration polynomial , and the results are shown in Figure 2.
(a)
(b)
(c)
(d)
Example 7. Polynomiographs for the polynomial through different fourthorder algorithms.
In the second example, we take a sexticdegree polynomial which has three unique roots , and with multiplicity two. We perform the process of iteration for all iterative algorithms for drawing polynomiographs, and the results are shown in Figure 3.
(a)
(b)
(c)
(d)
Example 8. Polynomiographs for the polynomial through different fourthorder algorithms.
In the third example, we consider a quarticdegree polynomial which has four unique roots We created the graphical objects by executing all iterative algorithms, and the results are shown in Figure 4.
(a)
(b)
(c)
(d)
Example 9. Polynomiographs for the polynomial through different fourthorder algorithms.
In the fourth example we take an eighthdegree complex polynomial with four unique roots of multiplicity two. We used a computer program to run all methods for drawing polynomiographs, and the results in the form of visually attractive pictures are shown in Figure 5.
In above examples, we compared the developed algorithm to various fourthorder iterative algorithms using a computer program by taking into account different degrees complex polynomials. Two key features may be identified from the produced graphics. The first is the iteration scheme’s speed of convergence, and the second feature is the iteration scheme’s dynamics. Low dynamics are seen in places with little color variation, and high dynamics are found in areas with a lot of color variety. The black coloring in the graphics denotes areas where the solution cannot be reached in the specified number of iterations. The darker zone in the abovepresented pictures indicates that the iterative algorithm under consideration requires fewer iterations for finding the solution of the given problem. The samecolored regions in the graphical objects represent the same number of iterations necessary to find the required solution with the given accuracy. Note that the polynomiographs created using our proposed iterative algorithm have considerably brighter and darker regions and no black areas as compared to other similar order algorithms in the literature. Furthermore, the polynomiographs of the proposed iterative algorithm show larger convergence areas than the other comparable techniques which demonstrate the better efficiency of the suggested algorithm.
We drew all graphical objects with the computer program Mathematica by using the values of parameters as and , where and K indicate the accuracy and the upper bound of the number of iterations, respectively.
(a)
(b)
(c)
(d)
6. Conclusion
By employing the finite difference scheme on Ostrowski’s method, we designed a new derivative free algorithm for calculating the approximate zeros of nonlinear scalar equations that possesses the fourthorder convergence. To analyze the applicability of the designed algorithm, we took some reallife engineering problems and solved them via computer tools. The numerical results given in Tables 1–5 proved the better performance and applicability of the designed algorithm against the other fourthorder algorithms. We have also presented the dynamics of the designed algorithm and gave a detailed comparison with the other comparable fourthorder algorithms in the literature via computer tools that revealed the convergence and other graphical characteristics of the designed algorithm. A new family of derivative free rootfinding algorithms can be constructed by applying the finite difference scheme to the existing methods in the literature.
Data Availability
The data used to support this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Authors’ Contributions
All authors contributed equally to this study.