CC BY
105URAL MATHEMATICAL JOURNAL, Vol. 5, No. 1, 2019, pp. 83-90
DOI: 10.15826/umj.2019.1.008
A NEW ROOT-FINDING ALGORITHM USING EXPONENTIAL SERIES
Srinivasarao Thota
Department of Applied Mathematics, School of Applied Natural Sciences, Adama Science and Technology University, Post Box No. 1888, Adama, Ethiopia [email protected], [email protected]
Abstract: In this paper, we present a new root-finding algorithm to compute a non-zero real root of the transcendental equations using exponential series. Indeed, the new proposed algorithm is based on the exponential series and in which Secant method is special case. The proposed algorithm produces better approximate root than bisection method, regula-falsi method, Newton-Raphson method and secant method. The implementation of the proposed algorithm in Matlab and Maple also presented. Certain numerical examples are presented to validate the efficiency of the proposed algorithm. This algorithm will help to implement in the commercial package for finding a real root of a given transcendental equation.
Keywords: Algebraic equations, Transcendental equations, Exponential series, Secant method.
Introduction
The root finding algorithms are the most relevant computational problems in science, engineering. The applications of root finding algorithms arise in many practical applications of Biosciences, Physics, Engineering, Chemistry etc. As mentioned in [1], the finding of any unknown appearing implicitly in engineering or scientific formulas, gives rise to root finding problem. A root of a function f (x) is a number 'a' such that f (a) = 0. Generally, the roots of transcendental functions cannot be expressed in closed form or cannot be computed exactly. The root-finding algorithms give us approximations to the roots, these approximations are expressed either as small isolating intervals or as floating point numbers. In the literature, there are several root finding algorithms are available, see for example, [1-11]. The basic root-finding methods are Bisection, False position, Newton-Raphson, Secant methods etc. Most of the algorithms use iteration, producing a sequence of numbers that hopefully converge towards the root as a limit. The rates of converge of different algorithms are different. That is, some algorithms are faster converges to the root than others algorithms. The purpose of existing algorithms is to provide higher order convergence with guaranteed root. Many existing algorithms do not guarantee that they will find all the roots; in particular, if such an algorithm does not find any root, that does not mean that no root exists. There are many well known root finding algorithms available to find an approximate root of algebraic or transcendental equations., see for example, [1, 5, 6, 8, 9, 11].
In this work, we propose a new algorithm based on exponential series and secant method becomes a special case of the proposed algorithm. This algorithm provides faster roots in comparison of the previous methods. The new proposed algorithm will be useful for computing a real root of transcendental equations. The rest of the paper is as follows: Section 1 describes the proposed method, their mathematical formulation, calculation steps and flow chart; implementation of the proposed algorithm in Maple is presented in Section 2 with sample computations; and Section 3
discuss some numerical examples to illustrate the algorithm and comparisons are made to show efficiency of the new algorithm.
1. New algorithm using exponential series
The new iterative formula using exponential series is proposed as follows, for any two initial approximations x0, x1 of the root,
(xn— 1f \xn) xnj \xn) \ 0 ft
—Ti—n—7t-W' n= i1-1)
xnJ(xn) xnJ\xn—l) J
By expanding this iterative formula, one can obtain the standard secant method as in first two terms, and many methods are obtained based on series truncation. Indeed,
_ f(xn){xn ~ Xn-1) n 1?]
Xn+1-Xn f{xn)-f{xn. !)• (i-2j
/ (xn)(xn xn— l) I__( /(xra)(x n - xn— 1)N
Xn+l Xn f(xn) - f{xn-1) 2Xn V f{xn) - f{xn-1) /
^ _ ^ f{X'll){X1l Xn—1)_|_ ^ | f(Xn){Xn Xn-l)\ 1 | f{X1l){Xn Xn-l)\ ^ ^
Xn+l Xn f{xn) - f{xn-1) 2Xn V f{Xn) - f{xn-1) / 6.T.2 V f(xn) ~ f(xn-1) J This is shown in the following theorem.
Theorem 1. Suppose a — 0 is a real exact root of /(x) and d is a sufficiently small neighbourhood of a. Let f"(x) exist and f'(x) / 0 in 9. Then the iterative formula given in equation (1.1) produces a sequence of iterations {xn : n = 1,2,3,...} with order of convergence p > (1 + \/5)/2.
Proof. The iterative formula given in equation (1.1) can be expressed in the following form
xn—
1f(xn)- xn /(xn)
n— 1 n - n n
Xn+1 — Xn exp
Since
xn /(xn) - xn / (xn—1) / '
lim X (Xnf (xra) i _
xn^« \xnf (x n) xnf (xn—1),
and hence xn+1 = a.
Using the standard expansion of ex as
exp(.r) = 1 + x + \x2 + -U3 + -^-x4 + ■■■ (1.5)
2 6 24
and from equations (1.1) and (1.5), we have
n— 1 /( xn) -
f (xn)
n f (xn) - f (xn—1)
xn+1 — xn exp
n— 1 /( xn) - /(xn) 1 n— 1 /( xn) - /(xn)
— Xral,-L +1 „, \ „, / o
n /(xn) - /(x n— 1 n /(xn) - /(x n— 1)
n— 1/(xn) - n f(xnA . A 6 \xnf(x n) xnf(xn—1 ) /
__f (xn)(xn xn— l) J_ / f{Xn){xn - Xn-1)
~ n / (xn) - / (x n—1) 2xn V /(xn) - /(xn—1)
__1 / f{Xn){xn - Xra_i)\3 / 1 / f(xn)(xn — xra-l)S
2
6xn V / (xn ) — /(xn—0 / \ 24xn V /(xn) — /(xn—1)
Since f (xn) œ 0, when we neglect higher order terms, then the above equation gives secant method in first two terms. Indeed, we have the following formulae obtained from first two terms, three terms and four terms of the expansion respectively as given in equations (1.2)-(1.4).
xn+1 — xn
f l) / On) - /On-l) '
^ _ ^ /On) On xn— l) 1 //On) On ^n—l)
Xn+1 Xn /On) - /On-l) 2xn V /On) - /On-l)
xn+1 — xn
/(xn)(xn xn-1 ) . 1 / /On)On xn-1 )
+
/ On ) - / On—1 ) 2xn V / On ) - / On—1 ) / 6xn V / On ) - / On—1 )
1 ( / (xn )(xn xn—1 )
In the above equations, we obtained secant method having (1 + \/5)/2 convergence in first two terms. Therefore, the order of convergence of proposed methods (1.1), (1.3) and (1.4) are at least P > (1 + VE)/2. □
1.1. Steps for computing root
I. Select two approximations x0 and x\ = 0. II. Compute f (x0) and f (xi).
III. Compute the next approximate root using formula given in (1.1).
IV. Repeat Step II and III until we get desired approximate root.
Flow chat of the proposed algorithm is presented in Figure 1.
2
2
3
Figure 1. Flow chart for proposed algorithm
2. Implementation of proposed algorithm
2.1. Proposed algorithm in MATLAB
In this section present MATLAB implementation of the proposed algorithm as follows.
a=input('Given Function:','s'); f=inline(a); % Given function is storing in f
x(1)=input('Enter x0: '); % Initial approximation x0
x(2)=input('Enter x1: '); % Initial approximation x1
n=input('Enter Allowed Error: '); % n is allowed error iteration=0;
for i=3:1000
x(i) = x(i-1)*exp((x(i-2)*f(x(i-1))-x(i-1)*f(x(i-1)))/ (x(i-1)*f(x(i-1))-x(i-1)*f(x(i-2)))); % main eq (1.1) iteration=iteration+1; if abs((x(i)-x(i-1))/x(i))*100<n root=x(i)
iteration=iteration
break % breaking if abs error > n
end
end
Sample computations using the implementation of the proposed algorithm are presented in Section 3.
2.2. Proposed algorithm in MAPLE
In this section, we provide the implementation of the proposed method in Maple as follows. To execute the maple code, one should enter the required date at type text.
eps_step := type; eps_abs := type; f(x):= type; x[0] := type; x[1] := type; n:= type;
for i from 2 to n do
printf("Iteration No: %g", i-1);
x [i] := x [i-1]*exp((x [i-2]*f(x [i-1])-x [i-1]*f(x [i-1]))/
(x [i-1]*f(x [i-1])-x [i-1]*f(x[i-2]))); if abs(x [i]-x [i-1]) < eps_step and abs(f(x[i])) < eps_abs then break; end if; end do;
Sample computations using the implementation of the proposed algorithm are presented in Section 3.
3. Numerical examples
This section provides some numerical examples to discuss the algorithm presented in Section 1 and comparisons are taken into account to conform that the algorithm is more efficient than other existing methods.
Example 3.1. Consider an equation
x6 - x - 1 = 0. (3.1)
This equation has two real roots -0.7780895987 and 1.134724138. The following Table 1 shows the comparison between various existing methods and proposed method at accurate to within e = 0.00001 with initial approximations x0 = 1 and x1 = 1.5.
Table 1. Comparing approximate root using various existing methods
Ite Bisection Ite Secant Ite Regula-Falsi Ite Proposed
No. method No. method No. method No. method
1 1.25 1 1.05055292 1 1.05055292 1 1.111637022
2 1.125 2 1.08362707 2 1.08362707 2 1.121248067
3 1.1875 3 1.14718724 3 1.10430109 3 1.135602993
4 1.15625 4 1.13311087 4 1.11683267 4 1.134695420
5 1.140625 5 1.13467619 5 1.12428166 5 1.134724078
16 1.13472748 7 1.13472414 18 1.13471575 6 1.134724138
One can observe from the Table 1 that the proposed algorithm gives approximate root quicker than the other existing methods.
Example 3.2. Consider the following transcendental equations. We compare the number of iterations required to get approximation root. The numerical results are provided in Table 2.
(i) f (x) = ex — x — 2, with initial approximation 1 and 2 with accuracy of 10-5.
(ii) f (x) = 2x3 + 11x2 + 12x — 9, with initial approximations —5 and —1 with accuracy of 10-10.
(in) f(x) = 8 — 4.5(x — sinx), with initial approximations 2 and 3 with accuracy of 10-10.
(iv) f (x) = xe-x — 0.1, with initial approximation —0.9 and 0.9 with accuracy of 10-10.
The numerical results given in Table 2 shows that the proposed method is more efficient than other methods.
Example 3.3. Recall the Example 3.1 for the sample computations using Matlab and Maple implementation as described in Section 2.
f(x) = x6 — x — 1
Table 2. Comparing No. of iterations by different methods
Exact Bisection Regula-Falsi Secant Proposed
Fun. Root method method method method
1.146193221 18 14 7 6
(«) -3.00000000 37 56 8 8
(ill) 2.43046574 34 11 7 6
(iv) 0.11183256 36 55 15 11
with initial approximations 1 and 1.5.
Using Matlab implementation, we have the following computations. >> ExpSecant Given Function:xA6-x-1 Enter x0: 1.0 Enter x1: 1.5
Enter Allowed Error: 0.00001 root=
1.1347
iteration=
6
Using Maple implementation, we have the following computations.
> eps_step := le-5:
> eps_abs := le-5:
> f(x):= xA6-x-1:
> x[0] := 1.0:
> x[1] := 1.5:
> n:= 100:
> for i from 2 to n do
> printf("Iteration No: °/»gM, i-1);
> x [i] := x [i-1]*exp((x [i-2]*f(x [i-1])-x[i-1]*f(x[i-1]))/
(x [i-1]*f(x [i-1])-x [i-1]*f(x [i-2])));
> if abs(x [i]-x [i-1]) < eps_step and
> abs(f(x[i])) < eps_abs then
> break;
> end if;
> end do;
1
1.111637022
2
1.121248067
3
1.135602993
4
1.134695420
5
1.134724078
6
1.134724138
One can use the implementation of the proposed algorithm to speed up the manual calculations.
4. Conclusion
In this work, we presented a new algorithm to compute an approximate root of a given transcendental function better than previous existing methods as illustrated. The proposed new algorithm was based on exponential series having better convergence than previous existing methods (for example, Bisection, Regula-Falsi, Secant methods etc.). This proposed algorithm is useful for solving the complex real life problems. Implementation of the proposed algorithm in Matlab and Maple is also discussed presented sample computations.
Acknowledgment
The author is thankful to the editor and reviewers for providing valuable inputs to improve the present format of manuscript.
REFERENCES
1. Datta B.N. Lecture Notes on Numerical Solution of Root-Finding Problems. 2012. URL: http://www.math.niu.edu/ dattab/math435.2009/R00T-FINDING.pdf.
2. Chen J. New modified regula falsi method for nonlinear equations. Appl. Math. Comput., 2007. Vol. 184, No. 2. P. 965-971. DOI: 10.1016/j.amc.2006.05.203
3. Noor M.A., Noor K.I., Khan W. A., Ahmad F. On iterative methods for nonlinear equations. Appl. Math. Comput., 2006. Vol. 183, No. 1. P. 128-133. DOI: 10.1016/j.amc.2006.05.054
4. Noor M. A., Ahmad F. Numerical comparison of iterative methods for solving nonlinear equations. Appl. Math. Comput., 2006. Vol. 180, No. 1. P. 167-172. DOI: 10.1016/j.amc.2005.11.151
5. Ehiwario J. C., Aghamie S. O. Comparative study of bisection, Newton-Raphson and secant methods of root-finding problems. IOSR J. of Engineering, 2014. Vol. 4, No. 4. P. 1-7.
6. Hussain S., Srivastav V. K., Thota S. Assessment of interpolation methods for solving the real life problem. Int. J. Math. Sci. Appl., 2015. Vol. 5, No. 1. P. 91-95. http://ijmsa.yolasite.com/resources/12.pdf
7. Sagraloff M., Mehlhorn K. Computing Real Roots of Real Polynomials. 2013. arXiv: 1308.4088v2 [cs.SC].
Iteration No : Iteration No : Iteration No : Iteration No : Iteration No : Iteration No :
8. Thota S., Srivastav V. K. Quadratically convergent algorithm for computing real root of non-linear transcendental equations. BMC Research Notes, 2018. Vol. 11, art. no. 909. DOI: 10.1186/s13104-018-4008-z
9. Thota S., Srivastav V. K. Interpolation based hybrid algorithm for computing real root of nonlinear transcendental functions. Int. J. Math. Comput. Research, 2014. Vol. 2, No. 11, P. 729-735. URL: http://ijmcr.in/index.php/ijmcr/article/view/182/181
10. Abbasbandy S., Liao S. A new modification of false position method based on homotopy analysis method. Appl. Math. Mech, 2008. Vol. 29, No. 2. P. 223-228. DOI: 10.1007/s10483-008-0209-z
11. Srivastav V. K., Thota S., Kumar M. A new trigonometrical algorithm for computing real root of non-linear transcendental equations. Int. J. Appl. Comput. Math., 2019. Vol. 5, art. no. 44. DOI: 10.1007/s40819-019-0600-8
