roots of algebraic equations
TRANSCRIPT
-
7/27/2019 Roots of Algebraic Equations
1/27
Chapter one: Roots of algebraic andtranscendental equations
1.1 equations
In this chapter methods of finding roots to various equations are
explored. Roots of an equation are defined as values of x where
the solution of an equation is true.
Equations are generally grouped into two main categories,
algebraic equations and transcendental equations.
The first type of equation, algebraic, is defined as an equation that
involves only powers of x. The following are examples of
algebraic equations:
!" # $ %x x x + =
&! %x
x+ =
&.!# %x =
'n the other hand, transcendental equations are non(algebraic
equations or functions that transcend, or cannot be expressed in
terms of algebra. Examples of such are exponential functions,
trigonometric functions, and the inverses of each. The following
are examples of transcendental equations:
cos sin %x x+ =
%x
e + =
Roots of )lgebraic and Transcendental Equations &(&
-
7/27/2019 Roots of Algebraic Equations
2/27
Transcendental functions may have an infinite number of roots or
may not have any roots at all. *or example, the function sinx has
an infinite number of roots with x k= , and %,&,!,...,etc.k=
The solution of algebraic equations is rarely carried out from thebeginning to the end by one method. The roots of the equation are
generally determined by one method with some small accuracy,
and then made more accurate by other methods. *or the intent and
purpose of this text, only a handful of the available methods are
discussed. They include: +escartes Rule, -ynthetic +ivision,
Incremental -earch, Refined Incremental -earch, isection, *alse
/osition, -ecant, 0ewton(Raphson, 0ewtons -econd 'rder,
1raeffes Root -quaring, and airstows methods.
1.2 Polynomials
) polynomial is defined as an algebraic equation involving only
positive integer powers of x. /olynomials are generally expressed
in the following form:
& ! &
& ! &.... %n n n n
n nx a x a x a x a x a
+ + + + + + =
*or these polynomials, the following apply:
2 The order or degree of the polynomial is equal to the highest
power of x and the number of roots is directly equal to the
degree or 3n4, where na is not equal to %. *or example, a
sixth degree polynomial, or a polynomial with $n = has six
roots.
2 The value of 3n4 must be a non(negative integer. In other
words, it must be whole number that is equal to 5ero or a
positive integer.
Roots of )lgebraic and Transcendental Equations &(!
-
7/27/2019 Roots of Algebraic Equations
3/27
2 The coefficients 6 % & ! &, , ,.... ,n na a a a a 7 are real numbers.
2 There will be at least one real root if 3n4 is an odd integer.
2 It is possible that equal roots exist.
2 8hen complex roots exist, they occur in con9ugate pairs, for
example:
&== vuviux
1.3 Descartes Rule
+escartes Rule is a method of determining the maximum number
of positive and negative real roots of a polynomial. This rule states
that the number of positive real roots is equal to the number of sign
changes of the coefficients or is less than this number by an even
integer. *or positive roots, start with the sign of the coefficient of
the lowest 6or highest7 power and count the number of sign
changes from the lowest to the highest power 6ignore powers
which do not appear7. The number of sign changes proves to be the
number of positive roots. sing &x= in evaluating 6 7 %f x = is theeasiest way to loo; at the coefficients.
*or negative roots, begin by transforming the polynomial to6 7 %f x = . The signs of all the odd powers are reversed while the
even powers remain unchanged. 'nce again, the sign changes can
be counted from either the highest to lowest power, or vice versa.
The number of negative real roots is equal to the number of sign
changes of the coefficients, or less than by an even integer. sing&x= in evaluating 6 7 %f x = is the easiest way to loo; at the
coefficients.
8hen considering either positive or negative roots, the statement
3less than by an even integer4 is included. This statement accounts
Roots of )lgebraic and Transcendental Equations &(
-
7/27/2019 Roots of Algebraic Equations
4/27
for complex con9ugate pairs that could exist. +escartes rule is
valid as long as there are no 5ero coefficients. If 5ero coefficients
exist, they are ignored in the count. )lso, one could find a root
and divide it out to form a new polynomial of degree 3n(&4 and
apply +escartes rule again.
Example &.(&
Example &.(!
Example &.(
Example &.( , second interval contains the root
& 6 7 6 7 %f x f x = , x is the root
If the first interval contains the root, let the following be assigned:
&x and &6 7f x remain unchanged
! x x=
! 6 7 6 7f x f x=
If the second interval contains the root, let the following be
assigned:
!x and !6 7f x remain unchanged
& x x=& 6 7 6 7f x f x=
Bontinue with the process until the desired accuracy is obtained.
Example &."(&
Roots of )lgebraic and Transcendental Equations &(F
http://var/www/apps/conversion/tmp/scratch_3/false%20position_ex1.dochttp://var/www/apps/conversion/tmp/scratch_3/false%20position_ex1.doc -
7/27/2019 Roots of Algebraic Equations
10/27
1., !ecant %ethod
This method is similar to the *alse /osition Aethod except that the
two most recent values ofx
6 !x
and x
7 and their correspondingfunction values 6 !6 7f x and 6 7f x 7 are used to obtain a new
approximation to the root instead of chec;ing values that bound the
root. In the renaming process for iteration, use the following:
& ! ! ,x x x x= =
& ! ! 6 7 6 7, 6 7 6 7f x f x f x f x= =
In some instances interpolation occurs, while in others,extrapolation occurs.
x!
f6x&7
f6x!7
x&x
*igure &.F
Example &.F(&
1.1- e/ton0Raphson %ethod e/tonsangent
Roots of )lgebraic and Transcendental Equations &(&%
http://var/www/apps/conversion/tmp/scratch_3/secant_ex1.dochttp://var/www/apps/conversion/tmp/scratch_3/secant_ex1.doc -
7/27/2019 Roots of Algebraic Equations
11/27
'nce an approximate root nx has been found, the slope of the
function at that point, G6 7nf x , is incorporated to converge to the
root more rapidly. The slope intersects the x(axis at &nx + .
xn
f6xn7
fG6xn7
xn@&
f6x7
*igure &.&%
&
6 7G6 7 nn
n n
f xf x
x x+
= 'R &
6 7
G6 7
nn n
n
f xx x
f x+ =
Repeat the process using a new value until convergence occurs.
Bonvergence may not occur in the following two cases:
2 GG6 7f x 6curvature7 changes the sign near a root.
2 Initial approximation is not sufficiently close to the true root
and the slope at that point has a small value.
Example &.&%(&
1.11 e/tons !econd 4rder %ethod
Roots of )lgebraic and Transcendental Equations &(&&
http://var/www/apps/conversion/tmp/scratch_3/newtonraphson_ex1.dochttp://var/www/apps/conversion/tmp/scratch_3/newtonraphson_ex1.doc -
7/27/2019 Roots of Algebraic Equations
12/27
0ewtons -econd 'rder Aethod is often a preferred method to
determine the value of root due to its rapid convergence and
extremely close approximation.
The equation 6 7 %f x = is considered once again.
f6x7x
xnxn@& x
*igure &.&&.&
The following is a Taylor series expansion of 6 7f x about nx x= :
!
&
GG6 76 7 GGG6 76 76 7 6 7 G6 76 7 ...
!? ?
n nn n n
f x x f x xf x f x f x x
+
= + + + +
*or a means of determining a value of x that will ma;e theTaylor series expansion go to 5ero, the first three terms of the right
hand side of the equation is set equal to 5ero to obtain an
approximate value:
GG6 76 76 7 G6 7 %
!
nn n
f x xf x x f x
+ + =
The exact value of x can not be determined from this equationsince only the first three terms of the infinite series were used in
Roots of )lgebraic and Transcendental Equations &(&!
-
7/27/2019 Roots of Algebraic Equations
13/27
the calculation. Dowever, a close approximation of the root is a
result.
8hen using this equation to calculate x , a quadratic must besolved yielding two possible roots. In order to avoid this problem,
6 7 C G6 7n nx f x f x = from 0ewtons Tangent may be substitutedinto the brac;eted term only:
GG6 7 6 76 7 G6 7 %
! G6 7
n nn n
n
f x f xf x x f x
f x
+ =
-olving for x we obtain the following:
6 7
GG6 7 6 7G6 7
! G6 7
n
n nn
n
f xx
f x f xf x
f x
=
*rom *igure &.&&, &n nx x x+ = . -ubstituting into the previousequation, the formula can be written as follows:
&
6 7
GG6 7 6 7G6 7
! G6 7
nn n
n nn
n
f xx x
f x f xf x
f x
+
=
If the first derivative is small, use 9ust the second derivative termas follows:
G6 7 %nf x
Roots of )lgebraic and Transcendental Equations &(&
-
7/27/2019 Roots of Algebraic Equations
14/27
GG6 76 76 7 %
!
nn
f x xf x x
+ =
!
GG6 76 7!
nn x f xf x =
! !6 7 6 7 %GG6 7 GG6 7
! !
n n
n n
f x f xx x
f x f x
= + =
-olving by the quadratic equation where,
6 7&, %,
GG6 7
!
n
n
f xa b c
f x= = =
!
6 7 6 7
-
7/27/2019 Roots of Algebraic Equations
15/27
+ouble roots occur when the first derivative is 5ero. Triple roots
occur when the first and second derivatives are 5ero, etcH
*igure &.&&.!
-ynthetic division may be used with all the previously discussed
methods.
Example &.&&(&
1.12 5raeffes Root !quaring %ethod
1raeffes Root -quaring Aethod is a root(finding method which
was among the most popular methods for finding roots of
polynomials in the &Fth and !%th centuries. The 1raeffes Root
-quaring Aethod is especially effective if all roots are real. The
method proceeds by multiplying a polynomial 6 7f x by 6 7f x using
the following equations:
& !6 7 6 76 7......6 7nf x x a x a x a=
& !6 7 6 &76 76 7......6 7nf x x a x a x a = + + +
so the result is:
Roots of )lgebraic and Transcendental Equations &(
http://var/www/apps/conversion/tmp/scratch_3/newtonsecondorder_ex1.dochttp://var/www/apps/conversion/tmp/scratch_3/newtonsecondorder_ex1.doc -
7/27/2019 Roots of Algebraic Equations
16/27
! ! ! !
&6 7 6 7 6 &7 6 7.....6 7n
nf x f x x a x a =
*or example, use a
rd
degree polynomial with root & !,x x
and x
asfollows:
!
& ! 6 7 %f x x a x a x a= = + + +
) polynomial with roots & !,x x and x would be as follows:
!
& ! 6 7 %f x x a x a x a = = + +
Aultiplying the two equations together yields the following:
$ ! < ! ! !
& ! ! & 6 7 6 7 % 6 ! 7 6 ! 7f x f x x a a x a a a x a = = + + + +
etting!y x= , this equation may be written as follows:
! ! ! !
& ! ! & % 6 ! 7 6 ! 7y a a y a a a y a
= + + +
This equation has roots of! !
& !x x and!
x . If the procedure was
applied again, another polynomial would be derived with roots of