Solution ofNonlinear Equations

*
algebraic and transcendental
Numerical solution of equations.
Consider the expression
a,x" a,n"" (*)
f(x) = +
+... +
an, x an + -

where as are some real constants (9070), and neX+

called ofnth
Then () is a
polynomial in se
degree.
called
algebraic equation of
The
~
polynomial fcus = 0 is an

nth
degree.
flu) contains other functions,
↑ If some
namely, Argonometric,
logarithmic, exponential, etc.,then the
equation flus to is

called a transcendental equation.

Examples:
*
0;x" x(1 + 2n)
Algebraic equations: n 3x+
+
2 = =

Transcendental
*
equations:e"=x; +
10g,4 6 0;2x 0sa 1 0.
= -
a
+
- =

*
Properties ofPolynomial Equations:

If f(n) then said to be

1. is
exactly divisible
by x-C, a is

a root off(x) = 0.

2. The total number ofroots of

the
algebraic equation f(x) =
0

is same as its
degree.
3. If f(x) a,x" a,n""= +
+ ...
an x + an have roofs a, cy...,ah,
then the
following hold:

2 ai
a
=

j(,2ixj
0
=

;π1Xi 11) =
 In
solving algebraic equations, thatis, finding its roots, one uses

the methods
direct
only
ifthe
degree ofequation is four or lower.
Thereis no direct
method available for
solving higher order

polynomial equations or transcendental equations. In such cases,

numerical methods can be used to solve them.
Numerical methods are iterative in nature. An iterative method
solution ofnumerical
gives an
approximate by repeated application
In this method, start with initial solution and
process.
we an

the method
improves
this solution until it is
improved to acceptable
accuracy.

Initial Approximation to a Root:A common

requirement in all the numerical
methods that
one has to make initial the root.
guess for
is an

To locate the ofthe equation f(x)

root = 0 one has the well-known
theorem in calculus stated below:
as

*
Intermediate Value Theorem:

continuous the interval [a,b] and f (a) and f(b)

If fins
in if
is

are of oppositive i.e., f(a).f(b) co, then the interval will

signs,
contain at least one root of
the
equation flu) = 0. In other

words, there exists some constant & E f(x)

(a,b) such that =
0.

Assumptions:
and to sufficient
~ f(x) is continuous
continuously differentiable up

number oftimes.
vfcns =0 has no multiple roots, thatis, if a is a real root of
f(x) 0, = then f(x) 0.
=
 %3-ti n y xx
=

root

f(a)

value <t(a,b)
Any can be taken as an initial
approximation to the root.

1. Bisection Method (Bolzano Method)

This is one ofthe simplest method and is based on the repeated application
ofthe intermediate value theorem. The method used
ofthe root of may
be
conveniently
roughapproximations given equation.
in the
method. The interval within which
This method is a slowly convergent

the root lies is bisected each time until the obtained root has
achieved the desired level of
accuracy.

function fans continuous [a,b] such that fal

~Suppose the be on

and f(b) are ofopposite signs. Then one root of

the
equation
between and b the Intermediate value Thm)
f(x) = 0 lies a
(by
For definiteness one assumes f(a)
that so and f(b) >0.

Bisection method
* is used to find the root between a and b to the

desired approximation as follows:

Let ab be the firstapproximation ofthe root,

(1) x i.e.,

x, is the midpoint ofendpoints a and b.

 2) If f(x) 0
=
then x, is a root of
fix). Otherwise, the root
lies between and and b whether f(Ro
depending
or
x, or s, on

or f(x) < 0.

the interval
(3) Bisect which the root lies and continue the process
in

until the root is found to the desired accuracy.

Note:After bisections, the ofthe subinterval which contains

a
length
xa is
ba
If the to be made less than
error is a small
quantity, say, a,
2")b
ba
then 2 =>
a

at
a
nz
=>

re
This
provides the number ofiterations required for
achieving
5.
accuracy
an

Examples:
real root of
the
(1) Find the positive equation us-sn
=

lying between
1 and I correct to three decimal places the bisection method.
using
Soln:Let f(x) x3 =
-
3x + 1.

Sincef (1) = - 1 c0 and f(2) 3/0, f(x)

=
= 0 has one root

between and Let and b

lying
1 2. a = 1 2.
=

a b
approximation
+

Then the first 1.5

=

is x, =
z

Now, f (15) -0.125

=
<0. Also f(z) > 0. Hence the root lies between
and 2.
1.5
2

The second approximation is s

a
=

b 1
=

5 1.75
=
Now, f(1.15) = 11094> 0. We have f (1.5) 20. So, the root lies

between
a =1.5 and b 115.
=

Third approximation is x,
a
=

b 1115 =
1625
=

Now, f(1625) 0.416070.

= Since f(15) <0 the root lies between
a 15
= and b 1625.
=

The
process
is
repeated until the desired
accuracy
and the calcul-
ation of successive
approximation is
given
below:

Iteration a b xi a
= b +

f(xi)
flas<0 f(b) > 0 z

↓
&
I 2 1.5 -

0.125

2. 1.5 2 1.75 1.10938

3. 1.5 1.75 1.625 0.41602

4. 1.5 1.625 1.5625 0.1272

5. 1.5 1.5625 1.53125 -

000339

6. 1.53125 15625 1.54688 0.06079

7. 1.53125 1.54688153907 0.02844

8. 1/
1.53907 1.53516 0.01246

9. 1/
1.53516 1.53321 0.00454

10. 1/ 1.53321 153223 0.00057

11. / 153223 153174

-

00014 |

It is observed that
20= x1 1.532 correct
to
places of
=
a

decimals.
decimal
required root,
to
correct
Therefore, the a
places is, x 1.532.
=
(2) Find a root of re-cosc 0
= correct to two decimal places
Disection method.
using
Soln.:Lat
f(x) =
x-cosc

f( -
1) 1
=
1 -
105( 1) 11-
=
-
0.5403 -
= 1.5403<0

f (0) 0
= -
10 (0) =
-
120

f(x) 1 -cos(1)
=
1.05403 0.4597>0
=
=

:The root lies between 0 and 1.

First approximation is x, =0
1
=0.5

f(0.5) -0.37750.
=
Also f(1)>0, so the root lies between

and
0.5 1.

0=
second approximation is m
=
0.75

f (0.75) 0.0183
=
> 0. Also f (0.5) 10, so the root lies between

and 0.75.
0.5

0.5 0.75
+

Third approximation is us = 0.625

=

f(0.625) =- 0.186 <0.

Now the root lies between 0.625 and 0.75.

0.75 0.6875
Fourth 0.625
+

approximation is xp = =

f(06875) = -
008534<0

Now the root lies in between and

0.6875 0.25

The
process is
repeated until desired follows:
accuracy
as

Iteration a +b
b
(i) flaico f(b) >8
xi= z
f(xi)

I o I 0.5 -
0.3775

2 0.5 1 0.75 0.0183

3 0.5 0.75 0.625 - 0.186

4 0.625 0.75 0.6875 -

0.08534
 06875075 0.7188 0.03388
5
-

6 0.7188 0.75 0.7344 -

0007833

7 0.7344 0.75 0.7422 0.005217

⑧ 0.7344 0.7422 0.7383 -

0001314

q 0.738307422 0.74025 0.00195

we observe that
x 10 x9
= =

0.74
=
correct to two decimal

places.
Therefore, the
required root is 0.24.

Note:If fins contains Trigonometric functions, then keep the

Calculator on Radian mode.