Why Laplace Transform?

Lecture 6 Frequency-domain analysis: Laplace Transform

Laplace transform is the dual (or complement) of the time-domain analysis. In time-domain analysis, we break input x(t) into impulsive component, and sum the system response to all these components. In frequency-domain analysis, we break the input x(t) into exponentials components of the form est, where s is the complex frequency:

s = + j

Laplace transform is the tool to map signals and system behaviour from the time-domain into the frequency domain.

Peter Cheung Department of Electrical & Electronic Engineering Imperial College London URL: www.ee.imperial.ac.uk/pcheung/teaching/ee2_signals E-mail: [email protected]
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 1

X (s) x (t )
Time-domain analysis h(t)

Y (s)
frequency-domain analysis H(s)

y (t )

x (t )

L-1

y (t )

PYKC 24-Jan-11

E2.5 Signals & Linear Systems

Lecture 6 Slide 2

Definition of Two-sided Laplace Transform

Definition of One-sided Laplace Transform

For a signal x(t), its Laplace transform is defined by:

The signal x(t) is said to be the inverse Laplace transform of X(s). It can be shown that

X ( s) = x(t )e dt

st

For the purpose of the 2nd year curriculum, let us assume that all signals are causal. For this the Laplace transform is defined as:

x(t ) =

2 j c j

c + j

X ( s) = L[ x(t )] = x(t )e st dt
0

X ( s)e st ds

where c is a constant chosen to ensure the convergence of the first integral. Note that this definition is slightly more complex than what you have seen in Dr Jaimoukas 2nd year Control course. This general definition does not assume casuality. This general definite is known as two-sided (or bilateral) Laplace Transform. L4.1 p340
E2.5 Signals & Linear Systems Lecture 6 Slide 3

This is the same as that defined on the 2nd year Control course, and is known as one-side (or unilateral) Laplace transform. Remember that the Laplace transform is a linear tranform (see Jamoukas notes, p15):

L4.1 p345
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 4

PYKC 24-Jan-11

A few examples

A few examples (2)

Find the Laplace transform of (t) and u(t).

Find the Laplace transform of eat u(t) and cos 0t u(t).

L[ (t )] = (t )e st dt = 1
0

for all s

L[eat u (t )] = eat e st dt
0

L[ (t )]

= e ( s a )t dt =
0 0

1 sa

L[eat u (t )]

1 sa

L[u (t )] = u (t )e st dt = e st dt
0

1 = e st s

1 s

Re s > 0

L[u (t )]
PYKC 24-Jan-11

1 s

1 L[cos 0t u (t )] = L[e j0t u (t ) + e j0t u (t )] 2 s 1 1 1 = + = 2 s j0 s + j0 s 2 + 0 2

L[cos 0tu (t )]

s s + 0 2
2

L4.1 p347
E2.5 Signals & Linear Systems Lecture 6 Slide 6

E2.5 Signals & Linear Systems

Laplace transform Pairs (1)

Laplace transform Pairs (2)

Finding inverse Laplace transform requires integration in the complex plane beyond scope of this course. So, use a Laplace transform table (analogous to the convolution table).

L4.1 p344
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 7 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

L4.1 p344
Lecture 6 Slide 8

Laplace transform Pairs (3)

Examples of Inverse Laplace Transform (1)

7s 6 . s2 s 6

(use partial fraction)

To find k1 which corresponds to the term (s+2), cover up (s+2) in X(s), and substitute s = -2 (i.e. s+2=0) in the remaining expression:

Similarly for k2: Therefore

L4.1 p344
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 9 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

Lecture 6 Slide 10

Examples of Inverse Laplace Transform (2)

Examples of Inverse Laplace Transform (3)

Easy to make mistake with partial fraction. Method to check correctness of:

Finding the inverse Laplace transform of

2s 2 5 . ( s + 1)( s + 2)

Substitute s = 0 into the equation (could use other values, but this is most convenient): 6 4 3 X (0) = =1= + (+2)(3) 2 3 Therefore, using Pair 5 from table:

The partial fraction of this expression is less straight forward. If the power of numerator polynomial (M) is the same as that of denominator polynomial (N), we need to add the coefficient of the highest power in the numerator to the normal partial fraction form:

Solve for k1 and k2 via covering: Therefore Using pairs 1 & 5:

L4.1 p349
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 11

PYKC 24-Jan-11

E2.5 Signals & Linear Systems

Lecture 6 Slide 12

Time Shifting Property of the Laplace transform

Application of Time Shifting

Find the Laplace transform of x(t) as shown:

[u (t 2) u (t 4)]

Delaying x(t) by t0 (i.e. time shifting) amounts to multiplying its transform X(s) by e st0 . Remember that x(t) starts at t = 0, and x(t - t0) starts at t = t0. Therefore, the more accurate statement of the time shifting property is:
(t 1)[u (t 1) u (t 2)] x(t ) = (t 1)[u (t 1) u (t 2)] + [u (t 2) u (t 4)]

= (t 1)u(t 1) (t 1)u(t 2)
= (t 1)u (t 1)

u(t 2) u(t 4)
u(t 4)

(t 2)u(t 2)

Time shift

L4.2 p360
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 13 PYKC 24-Jan-11

L4.2 p361
E2.5 Signals & Linear Systems Lecture 6 Slide 14

Frequency Shifting Property

Application of Frequency Shifting

Frequency Shifting property:

Given

, show that

Frequency shifting the transform X(s) by s0 amounts to multiplying its time signal by e s0t . Observe symmetry (or duality) between frequency-shift and time-shift properties.

Apply frequency-shifting property with frequency shift

Replace s with (s+a) means frequency shift by -a. This yields the RHS of the equation. By frequency-shifting property, we need to multiply the at LHS by e .

L4.2 p363
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 15 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

L4.2 p364
Lecture 6 Slide 16

Time-Differentiation Property

Proof of Time-Differentiation Property

Time-differentiation property:

Integration by parts gives:

Repeated application of this property yields:

For the Laplace integral to converge, it is necessary that Therefore we get:

x(t )e st 0 as t

Frequency-differentiation property:
L4.2 p364 L4.2 p365
E2.5 Signals & Linear Systems Lecture 6 Slide 17 PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 18

PYKC 24-Jan-11

Application of Time-Differentiation

Time-Integration Property

Find the Laplace transform of the signal x(t) using time differentiation and time-shifting properties.

Time-integration property:

d/dt

The dual property of time-integration is the frequency-integration property:

d/dt

Time-differentiation & time shifting properties

L4.2 p365
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 19 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

L4.2 p366
Lecture 6 Slide 20

Scaling Property

Time-Convolution & Frequency-Convolution Properties

Scaling property:

Time-convolution property:

Time compression of a signal by a factor a causes expansion of its Laplace transform in s-scale by the same factor.

Convolution in time domain is equivalent to multiplication in s (frequency) domain. Frequency-convolution property:

L4.2 p367
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 21

Convolution in s (frequency) domain is equivalent to multiplication in time domain. L4.2 p368

PYKC 24-Jan-11

Application of the convolution Properties

Relationship with time-domain analysis

Use the time-convolution property of the Laplace transform to determine

c(t ) = e u (t )* e u(t ).

at

bt

If h(t) is the impulse response of a LTI system, then we have see in lectures 4 & 5 that the system response y(t) to an input x(t) is x(t)*h(t). Assuming causality, and that h(t ) H ( s) and x(t ) X ( s ) then The response y(t) is the zero-state response of the LTI system to the input x(t). It follows that the transfer function H(s):

Since

eat u (t )

1 ( s a)

ebt u (t )

1 ( s b)

Therefore

eat u (t )* ebt u (t )

1 ( s a)( s b)

Perform inverse Laplace transform gives:

L4.2 p368
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 23 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

L4.2 p368
Lecture 6 Slide 24

Summary of Laplace Transform Properties (1)

Summary of Laplace Transform Properties (2)

L4.2 p369
PYKC 24-Jan-11 E2.5 Signals & Linear Systems Lecture 6 Slide 25 PYKC 24-Jan-11 E2.5 Signals & Linear Systems

L4.2 p369
Lecture 6 Slide 26

Relating this lecture to other courses

You have done Laplace transform in maths and in control courses. This lecture is mostly a revision, plus emphasis on the convolution multiplication properties for the two domains. Many of the properties are deliberately stated without proofs. It is more important on this course to understand the actual interpretations of Laplace transform (and more importantly the duality of time and frequency domains) than the mathematic proofs.

PYKC 24-Jan-11

E2.5 Signals & Linear Systems

Lecture 6 Slide 27