Chapter 3

Quick Review

Hamza J. Magnier
Taylor series expansion

Who created it
What is the equation
75
The physical meaning
 Review
formulated by the Formally introduced by
Scottish mathematician the English mathematician
James Gregory Brook Taylor in 1715
Chapter 3

Hamza J. Magnier
If the Taylor series is centered at zero, then that series is
also called a Maclaurin series, named after the Scottish
mathematician Colin Maclaurin, who made extensive use
of this special case of Taylor series in the 18th century. 76
 Review (continued…)
A question:
Can you calculate sin(31.56), (angle in degrees) using a
blank paper and pencil?
Chapter 3

Think it over a while.

Let's work it out:
We would like to have, at this point, a table with the exact

Hamza J. Magnier
values of the sines of angles at intervals of 0.02 degrees
which would give us values of sin(x). We probably learnt
the sines of standard angles from such a table.
Doesn't seem very promising if you want to extend the
level of accuracy.

We cannot, of course sit down to draw an exact 31.56

degree angle and hope to be able to calculate ratios
by construction.
77
 What else is possible? Sit down
with a high resolution graph
Chapter 3

(drawn somehow)?
Any other suggestion?

Hamza J. Magnier
Enough of nonsense?
By now you are either irritated of the answer and ask me
to come to the point unless you're very patient and have
guessed I am already pointing at the Taylor series
 Chuck the exact value - give
me a good approximate
answer.
Chapter 3

Why would you like anyone to calculate the value

of sin(31.56)? Probably you have come up with an
expression of some quantity and the angle that you

Hamza J. Magnier
need to calculate turns out to be this. You would
probably require it get the idea of what is the value
keeping which in mind you could get your required
dimension.
Well, not always do you want the answer correct up to
the last decimal place, for you might not require to
measure any length upto that accuracy. You want a
good enough workable value, right?
 Well in such situations, we we need to approximate the
value of non-polynomial function. In such a scenario
Chapter 3

you would resort to a Taylor series approximation, to a

level where the error, which is the left over part of the
series, is very very... small.

Hamza J. Magnier
 Chapter 3
A physical
Interpretation

Hamza J. Magnier
 Express f(a+x) as a power series in
x, i.e.
Chapter 3

Hamza J. Magnier
 Mauclarin Series
Chapter 3

f a x x2 xn
x f (a)  f ʹa   f ʹa   !  (n)  a   n1
 1! 2! n! f )
 O(x

Hamza J. Magnier
Thus the coefficients depend on derivatives of the function at x = a.

f n a x n


 n!
f a  x n0
 Single variable Taylor series:
Let f be an infinitely differentiable function in some open
interval around x = a.
Chapter 3

(k)
Σf f ′′
f (x) (x − a) = f (a)+ f (a)(x −
k ′
(x − a)2 + · · ·
∞ (a) (a

Hamza J. Magnier
= a)+
k=0
k! )
2!

Linear approximation in one variable:

Take the constant and linear terms from the Taylor series.
In an open interval around x = a,

f (x) ≈ f (a)+ f ′(a)(x − a) linear approximation

 Quadratic approximation in one variable: Take the constant,
linear, and quadratic terms from the Taylor series.
In an open interval around x = a, quadradic
Chapter 3

′′ ( a)
approximation
f (x) ≈ f (a) + f ′(a)(x − (x − a)2
f
a) +
2!

Hamza J. Magnier
Multi-variable Taylor series
Let f be an infinitely differentiable function in some open
neighborhood around (x, y) = (a, b).
f (x, y) = f (a, b) + fx (a, b)(x − a) + fy (a, b)(y − b)
1Σ 2fxy
fxx (a, b)(x − a) + (a, b)(x − a)(y −
2
+
2!
b) + fyy Σ
(y − b) 2
+ ···
 Linear approximation in multiple variables:
Chapter 3

Take the constant and linear terms from the Taylor series.
In a neighborhood of (x, y) = (a, b),

Hamza J. Magnier
f (x, y) ≈ f (a, b) + fx (a, b)(x − a) + fy (a, b)(y − b)
 Chapter 3

End of Review

Hamza J. Magnier
Chapter 3
Approach

y is the output and u is the input

A linear approximation of this equation can be obtained by

Hamza J. Magnier
using a Taylor series expansion and truncating after the
first-order terms.
The reference point for linearization is the nominal steady-
state operating point

88
 Approach (continued…)
By definition, the steady-state condition corresponds to
- -
f(y , u ) = 0.
Chapter 3

In addition, note that deviation variables arise naturally out

of the Taylor series expansion- namely,

Hamza J. Magnier
y’ = y - -y and u'=u - u- .

Hence, the linearized differential equation in terms of y' and

u' is (after substituting, dy'/dt = dy/dt)

89
 Chapter 3

Hamza J. Magnier