Unit 5: The Laplace Transform

Introduction

The Laplace Transform is a powerful mathematical tool used to convert signals and systems from the time-domain to the frequency-domain (specifically, the complex frequency domain or s-domain). It is a generalization of the Continuous-Time Fourier Transform (CTFT).

Why use the Laplace Transform?

1. Broader Class of Signals: The CTFT integral does not converge for many common signals that are not absolutely integrable (e.g., u(t), e^at for a>0). The Laplace Transform includes a convergence factor, e^(-σt), which allows it to handle a much wider range of signals, including unstable ones.
2. Analysis of LTI Systems: It simplifies the analysis of Linear Time-Invariant (LTI) systems described by linear constant-coefficient differential equations. It transforms these differential equations in the time-domain into algebraic equations in the s-domain, which are much easier to solve.
3. System Stability and Causality: The properties of the Laplace Transform and its Region of Convergence (ROC) provide a direct and insightful way to determine the stability and causality of a system from its transfer function.

The Laplace Transform can be seen as the Fourier Transform of the signal x(t) multiplied by a real exponential e^(-σt):
X(s) = X(σ + jω) = ∫[-∞ to ∞] [x(t)e^(-σt)] e^(-jωt) dt = F{x(t)e^(-σt)}
where s = σ + jω is the complex frequency variable.

The Laplace Transform

Definition

The bilateral (or two-sided) Laplace Transform of a continuous-time signal x(t) is defined as:

X(s) = L{x(t)} = ∫[-∞ to ∞] x(t)e^(-st) dt

where s is a complex variable given by s = σ + jω.

• σ (sigma) is the real part, representing the damping or growth rate.
• ω (omega) is the imaginary part, representing the angular frequency.

The unilateral (or one-sided) Laplace Transform is used for causal signals (signals that are zero for t < 0) and is particularly useful for solving differential equations with initial conditions. It is defined as:

X(s) = L{x(t)} = ∫[0- to ∞] x(t)e^(-st) dt

Note: In many textbooks and contexts, "The Laplace Transform" implicitly refers to the unilateral version when dealing with system analysis.

Common Laplace Transform Pairs

The Region of Convergence (ROC) for Laplace Transforms

The Region of Convergence (ROC) is the set of values of s in the complex s-plane for which the Laplace Transform integral converges. The ROC is a crucial part of the transform; the algebraic expression X(s) alone does not uniquely specify the time-domain signal x(t).

Properties of the ROC

1. Structure: The ROC consists of strips parallel to the jω-axis in the s-plane.
2. Poles: The ROC does not contain any poles. A pole is a value of s for which |X(s)| becomes infinite.
3. Finite-Duration Signals: If x(t) is of finite duration and is absolutely integrable, the ROC is the entire s-plane.
4. Right-Sided Signals: If x(t) is a right-sided signal (i.e., x(t) = 0 for t < T₁), the ROC is a right-half plane of the form Re{s} > σ_max, where σ_max is the real part of the rightmost pole.
5. Left-Sided Signals: If x(t) is a left-sided signal (i.e., x(t) = 0 for t > T₂), the ROC is a left-half plane of the form Re{s} < σ_min, where σ_min is the real part of the leftmost pole.
6. Two-Sided Signals: If x(t) is a two-sided signal, the ROC is a vertical strip in the s-plane between two poles, σ₁ < Re{s} < σ₂.
7. Relation to Fourier Transform: If the ROC includes the jω-axis (Re{s} = 0), then the Continuous-Time Fourier Transform (CTFT) of x(t) exists and can be found by evaluating X(s) at s = jω. X(jω) = X(s)|_(s=jω).

Example: Consider X(s) = 1 / (s + 2).

• If the ROC is Re{s} > -2, it corresponds to the right-sided signal x(t) = e^(-2t)u(t).
• If the ROC is Re{s} < -2, it corresponds to the left-sided signal x(t) = -e^(-2t)u(-t).

The Inverse Laplace Transform

The inverse Laplace Transform converts X(s) back to the time-domain signal x(t).

Definition

The formal definition is given by the complex contour integral (also known as the Bromwich integral):

x(t) = (1 / 2πj) ∫[σ - j∞ to σ + j∞] X(s)e^(st) ds

where the integration is performed along a vertical line Re{s} = σ that lies within the ROC of X(s). This formula is rarely used for direct computation in undergraduate courses.

Inverse Transform by Partial Fraction Expansion (PFE)

For rational functions X(s) = N(s) / D(s), the most common method is Partial Fraction Expansion.

Step 1: Ensure the fraction is proper (degree of N(s) < degree of D(s)). If not, use long division.
Step 2: Find the roots of the denominator D(s) (i.e., the poles of X(s)).
Step 3: Expand X(s) into a sum of simpler terms based on the poles.

Case 1: Distinct Real Poles
If X(s) has distinct poles p₁, p₂, ..., p_n, the expansion is:
X(s) = A₁ / (s - p₁) + A₂ / (s - p₂) + ... + A_n / (s - p_n)
The coefficient A_k can be found using the residue method:
A_k = [(s - p_k) X(s)] |_(s = p_k)

Case 2: Repeated Real Poles
If X(s) has a pole p₁ with multiplicity r, the expansion includes terms:
X(s) = A₁ / (s - p₁) + A₂ / (s - p₁)² + ... + A_r / (s - p₁)^r + ...
The coefficients are found by:
A_{r-k} = (1/k!) * d^k/ds^k [ (s - p₁)^r X(s) ] |_(s = p₁) for k = 0, 1, ..., r-1

Case 3: Complex Conjugate Poles
If X(s) has a complex conjugate pole pair a ± jb, they are often grouped together:
(As + B) / (s² + 2as + a² + b²) = (As + B) / ((s+a)² + b²)
This form can be manipulated to match the Laplace transforms for damped sine and cosine functions.

Step 4: Find the inverse Laplace Transform of each simple term using the standard transform table, considering the ROC to determine if the resulting signals are right-sided (u(t)) or left-sided (u(-t)).

Geometric Evaluation of the Fourier Transform from the Pole-Zero Plot

The pole-zero plot is a graphical representation of a transfer function in the s-plane, where 'x' denotes a pole and 'o' denotes a zero. The Fourier Transform X(jω) is X(s) evaluated on the jω-axis.

Let the transfer function be in factored form:
H(s) = K * ( (s-z₁) (s-z₂) ... ) / ( (s-p₁) (s-p₂) ... )

To evaluate H(jω) for a specific frequency ω:

1. Draw vectors from each pole and zero to the point jω on the imaginary axis.
2. The magnitude |H(jω)| is calculated as:
   |H(jω)| = |K| * (Product of lengths of zero vectors) / (Product of lengths of pole vectors)
|H(jω)| = |K| * ( |jω-z₁| |jω-z₂| ... ) / ( |jω-p₁| |jω-p₂| ... )
3. The phase ∠H(jω) is calculated as:
   ∠H(jω) = (Sum of angles of zero vectors) - (Sum of angles of pole vectors) + ∠K
∠H(jω) = ( ∠(jω-z₁) + ∠(jω-z₂) + ... ) - ( ∠(jω-p₁) + ∠(jω-p₂) + ... )

This geometric view provides powerful intuition. For example:

• If a pole is close to the jω-axis, the magnitude of the frequency response will have a peak at that frequency.
• If a zero is close to the jω-axis, the magnitude will have a dip or null at that frequency.

Properties of the Laplace Transform

Let x(t) ↔ X(s) with ROC R_x and y(t) ↔ Y(s) with ROC R_y.

Analysis and Characterisation of LTI Systems using the Laplace Transform

The Transfer Function

For an LTI system, the ratio of the Laplace Transform of the output y(t) to the Laplace Transform of the input x(t) is called the transfer function, denoted by H(s).

H(s) = Y(s) / X(s)

Assuming zero initial conditions. The transfer function H(s) is the Laplace Transform of the system's impulse response h(t).
H(s) = L{h(t)}

The system's behavior is completely characterized by H(s).

Causality and Stability

The properties of an LTI system are directly related to the poles of its transfer function H(s) and the associated ROC.

Causality

• An LTI system is causal if its output at any time depends only on present and past inputs (h(t) = 0 for t < 0).
• Condition: A continuous-time LTI system is causal if and only if the ROC of its transfer function H(s) is a right-half plane. For a rational H(s), this means the ROC is to the right of the rightmost pole.

Stability

• An LTI system is BIBO (Bounded-Input, Bounded-Output) stable if every bounded input produces a bounded output.
• This is equivalent to the impulse response being absolutely integrable: ∫[-∞ to ∞] |h(t)| dt < ∞.
• Condition: An LTI system is BIBO stable if and only if the ROC of its transfer function H(s) includes the jω-axis (Re{s} = 0).

Causal and Stable Systems

Combining these two conditions leads to a fundamental result for system design:

A causal LTI system with a rational transfer function H(s) is stable if and only if all of its poles lie in the left-half of the s-plane (LHP).

• If all poles are in the LHP, the ROC (which must be a right-half plane for causality) will be Re{s} > σ_max, where σ_max < 0. This ROC will necessarily include the jω-axis, implying stability.

Software Simulation of System Representation and Pole-Zero Analysis

Software tools like Python (with SciPy/Matplotlib) and MATLAB are essential for modern system analysis. They allow for quick representation, visualization, and simulation.

Here's an example using Python to analyze a second-order system described by the differential equation:
d²y(t)/dt² + 3 dy(t)/dt + 2 y(t) = dx(t)/dt + 4 x(t)

Taking the Laplace Transform (assuming zero initial conditions):
s²Y(s) + 3sY(s) + 2Y(s) = sX(s) + 4X(s)
Y(s)(s² + 3s + 2) = X(s)(s + 4)

The transfer function is:
H(s) = Y(s)/X(s) = (s + 4) / (s² + 3s + 2)

Python Implementation

import numpy as np
import matplotlib.pyplot as plt
from scipy import signal

# 1. Define the Transfer Function
# H(s) = (s + 4) / (s^2 + 3s + 2)
# Numerator coefficients: [1, 4] for s + 4
num = [1, 4]
# Denominator coefficients: [1, 3, 2] for s^2 + 3s + 2
den = [1, 3, 2]

# Create a transfer function object
system = signal.TransferFunction(num, den)

print("System Transfer Function:")
print(system)

# 2. Find Poles and Zeros
zeros, poles, gain = signal.tf2zpk(num, den)
print(f"\nZeros: {zeros}")
print(f"Poles: {poles}")
print(f"Gain (k): {gain}")

# 3. Plot the Pole-Zero Map
plt.figure(figsize=(8, 6))
# Plot Zeros
plt.scatter(np.real(zeros), np.imag(zeros), s=100, marker='o', 
            facecolors='none', edgecolors='b', label='Zeros')
# Plot Poles
plt.scatter(np.real(poles), np.imag(poles), s=100, marker='x', 
            color='r', label='Poles')

# Add unit circle for reference (though more relevant for z-transform)
# Add axes
plt.axhline(0, color='black', lw=0.5)
plt.axvline(0, color='black', lw=0.5)

plt.grid(True, which='both')
plt.title('Pole-Zero Plot')
plt.xlabel('Real Part (σ)')
plt.ylabel('Imaginary Part (jω)')
plt.legend()
plt.axis('equal')
plt.show()

# System Analysis from Pole Locations:
# The poles are at -1 and -2. Both are in the Left-Half Plane (LHP).
# Therefore, the system is STABLE.
# Assuming causality, the ROC is Re{s} > -1, which includes the jw-axis.

# 4. Simulate System Response
# Time vector for simulation
t = np.linspace(0, 10, 500)

# Impulse Response h(t)
t_imp, h = signal.impulse(system, T=t)

# Step Response s(t)
t_step, s_t = signal.step(system, T=t)

plt.figure(figsize=(12, 5))

# Plot Impulse Response
plt.subplot(1, 2, 1)
plt.plot(t_imp, h)
plt.title('Impulse Response h(t)')
plt.xlabel('Time (s)')
plt.ylabel('Amplitude')
plt.grid(True)

# Plot Step Response
plt.subplot(1, 2, 2)
plt.plot(t_step, s_t)
plt.title('Step Response')
plt.xlabel('Time (s)')
plt.ylabel('Amplitude')
plt.grid(True)

plt.tight_layout()
plt.show()