Chapter 1: Introduction to Control Systems

Objectives of the Chapter

 Understand the definition and purpose of control systems.

 Differentiate between open-loop and closed-loop systems.
 See real-world examples of control systems in engineering.
 Learn basic components: controllers, actuators, sensors, plants.
 Introduction to feedback and system modeling.

1.1 Control System: Definition

A control system is an arrangement of physical components that regulate or command the

behavior of other systems to achieve a desired output.

1.2 Types of Control Systems

Open-Loop Control System

 No feedback.
 Output has no influence on control action.
 Simple and inexpensive.
 Example: Electric toaster, washing machine (timer-based).

Input → Controller → Actuator → Plant → Output

2. Closed-Loop Control System (Feedback System)

 Uses feedback to compare output with the input (desired response).

 More accurate and stable.
 Self-correcting behavior.

1.3 Examples of Control Systems

System Controlled Variable Controller Actuator

Cruise Control (Car) Vehicle Speed Microprocessor Throttle
Room Temperature Temperature Thermostat Heater
 System Controlled Variable Controller Actuator
Aircraft Autopilot Altitude Flight Computer Elevator control

1.4 Basic Terminology

 Plant: The part of the system to be controlled.

 Actuator: Device that converts control signal to physical action.
 Sensor: Measures output (feedback).
 Set Point: Desired value of the output.
 Error: Difference between set point and measured output.

1.5 Advantages of Feedback Control

 Better accuracy and stability.

 Reduces sensitivity to disturbances and changes.
 Can stabilize unstable systems.
 Facilitates automatic operation.

🔹 1.6 Disadvantages of Feedback Systems

 Complex design.
 Potential instability due to poor tuning.
 Expensive components.

Key Takeaways

 Open-loop = simple, no feedback.

 Closed-loop = feedback-based, more accurate.
 Real systems are often hybrid or use hierarchical control.
 Control engineering is essential in all modern automation fields.

Chapter 2: Mathematical Modeling of Physical Systems

Objectives of the Chapter

 Develop mathematical models (primarily differential equations and transfer functions)

for physical systems.
 Understand modeling of mechanical, electrical, thermal, and fluid systems.
 Apply the concept of analogous systems (e.g., electrical-mechanical analogies).

2.1 Introduction to Modeling

Modeling is the process of writing mathematical equations that describe the dynamic behavior
of a system.

 A physical system can be mechanical, electrical, hydraulic, thermal, etc.

 The dynamic behavior is typically described by linear, time-invariant differential
equations.
 These equations can be converted to transfer functions using Laplace transform.

2.2 Modeling of Mechanical Systems

There are two types:

 Translational systems
 Rotational systems

EXAMPLE 2.1: Translational System

A mass-spring-damper system with mass M, damping B, and stiffness K. Find the differential
equation.

Diagram:
External force F(t) applied to mass M, opposed by spring and damper.

Equation:

This is the governing equation of motion.

EXAMPLE 2.2: RLC Series Circuit

Find the differential equation for a series RLC circuit with input voltage V(t)V(t)V(t).

Diagram: Series R → L → C

Apply KVL:
2.4 Analogous Systems (Mechanical ↔ Electrical)

There are two types of analogies:

1. Force-Voltage Analogy
2. Force-Current Analogy

Mechanical Electrical (Force–Voltage)

Force (F) Voltage (V)

Velocity (x˙) Current (i)

Mass (M) Inductor (L)

Damper (B) Resistor (R)

Spring (K) Capacitor (1/C)

These analogies help model mechanical systems using easier electrical equivalents.

2.6 Transfer Function

The transfer function is:

2.7 Thermal and Hydraulic Systems (Briefly Discussed)

These are modeled using:

  Heat flow = Temperature difference / Resistance
 Flow rate = Pressure difference / Resistance

The approach is similar to electrical analogies.

🧠 Key Takeaways

 All physical systems can be modeled using differential equations.

 These equations can be transformed into transfer functions using Laplace transform.
 Understanding of analogy between mechanical and electrical systems simplifies analysis.
 Mastering modeling is essential for simulation and controller design.
 Chapter 3: Block Diagrams and Signal-Flow Graphs

Objectives of the Chapter

 Understand how to represent complex systems using block diagrams.

 Learn block diagram algebra for simplification.
 Analyze systems using signal-flow graphs (SFG).
 Apply Mason’s Gain Formula to compute overall system transfer function.

3.1 Block Diagrams

A block diagram is a graphical representation of a system showing the flow of signals and how
subsystems (blocks) interact.

Each block represents a transfer function, and arrows represent the direction of signal flow.

Basic Elements of a Block Diagram:

 Block: Represents a system with transfer function G(s)

 Summing point (⊕): Adds/subtracts signals
 Take-off point: Duplicates a signal to send to multiple blocks
3.2 Block Diagram Reduction Rules

There are a set of 10 rules Ogata provides for reducing complex diagrams:

1. Blocks in series → multiply

2. Blocks in parallel → add
3. Eliminate feedback → apply feedback formula
4. Move summing points
5. Move take-off points
6. Repeated application helps reduce complex block diagrams.

EXAMPLE 3.2: Reduction with Feedback and Parallel

A system has:

 G1=1/ s+1
 G2=2/ s
 Feedback H=1

Block diagram has G1and G2 in series with negative unity feedback.

State Space Representation of Dynamic System
 Chapter 4: Mathematical Modeling in State Space
Chapter Objectives

 Understand the state-space approach to modeling dynamic systems.

 Learn how to derive state equations from physical models.
 Represent systems using state variables, state-space matrices, and first-order ODEs.
 Convert between transfer function and state-space models.
 Understand canonical forms and their role in control design.

4.1 Introduction to State-Space Analysis

The state-space approach is:

 A modern method used especially for MIMO systems (multi-input, multi-output).

 Describes system behavior using first-order differential equations.
 Handles nonlinear, time-varying, and initial conditions effectively.

4.2 State, State Variables, and State Vector

✅ Definitions:

 State: Smallest set of variables that describe the complete condition of a system.
 State Variables: The variables used to represent the state.
 State Vector: A column vector containing all state variables.

If a system has state variables x1(t), x2(t),…,xn(t) then the state vector is:

4.3 General State-Space Model

A linear time-invariant (LTI) system can be written as:

Where:

Symbol Description
x(t) State vector (n×1)
u(t) Input vector (m×1)
y(t) Output vector (p×1)
A State matrix (n×n)
B Input matrix (n×m)
C Output matrix (p×n)
Feed forward (direct
D
transmission) matrix (p×m)
🔹 4.6 Canonical Forms

Three common forms:

1. Controllable Canonical Form

Built from transfer function, useful for control design.

2. Observable Canonical Form

Dual of controllable form, useful for observer design.

3. Diagonal (Jordan) Form

Simplifies analysis when eigenvalues are distinct.

Transfer Function into a State-Space Representation

Mechanical System
Inverted Pendulum
Electrical System
Liquid Level System
Resistance and Capacitance in Liquid Level

change ∈level difference m

R=
change ∈flow rate m 3/s

Q=KH

Q = Steady state rate flow m^3/s

H = Steady State head m

K = Coefficient m^2/s
TRANSISENT RESPONSE SYSTEM

Objectives of the Chapter

 Analyze how control systems behave immediately after a disturbance or input — called
the transient response.
 Understand how time-domain specifications (like rise time, overshoot) define
performance.
 Study the response of first-order and second-order systems.
 Learn how damping and natural frequency affect system dynamics.

5.1 Introduction

The total response of a system =

Transient Response + Steady-State Response

This chapter focuses on the transient part, which occurs before the system settles into its final
value.

 Transient response = how the system reacts initially to a change.

 It's important for designing fast, stable, and precise systems.
Ramp Input Response
Unit Impulse Input Response

Example
For unit step function
Second Order System
Neglecting S1 as effect of S1 is much less as compared to S2.
Definition of transient response specifications
Time-Domain Specifications (Expanded)

These are standard metrics used to quantify the shape of the time response.

Term Definition

Delay Time (td ) Time to reach 50% of final value

Rise Time (tr ) Time to go from 10% to 90% (or 0% to 100%) of final value

Peak Time (tp ) Time to first maximum (peak overshoot)

Maximum Overshoot (Mp) % by which response exceeds final value

Settling Time (ts ) Time to reach and stay within 2% or 5% of final value

Steady-State Error Final error value as t→∞

 Chapter 5: Basic Control Action & Response of Control System

Objective

 Understand the basic types of controllers: Proportional (P), Integral (I), Derivative
(D), and their combinations.
 Analyze how these affect the transient and steady-state behavior of control systems.
 See how each control action changes overshoot, rise time, settling time, and steady-
state error.

Proportional Controller
Final Value Theorem (FVT)
The Final Value Theorem (FVT) is a powerful tool in Laplace Transform analysis, especially in
control systems, to quickly determine the steady-state (final) value of a time-domain signal without
computing its full inverse Laplace transform.

Conditions for Valid Use

You can use FVT only if:

1. All poles of sF(s) (i.e., the system) are in the left-half of the s-plane (i.e., stable system)
2. No poles of F(s) are at the origin (s = 0) or in the right-half s-plane

If the system is unstable or oscillatory, FVT gives incorrect results.

Concept Final Value Theorem

Finds steady-state value directly from
Purpose
Laplace
Formula Lim t→∞f(t)=lim⁡s→0s⋅F(s)
Use Case Step/ramp response, steady-state error
Warning Only valid for stable systems

Summary of Rules
Situation Use FVT?
F(s)=1/s(s+1) ✅ Yes
F(s)=1/s2 ❌ No
Poles in right-half plane (e.g. +2) ❌ No
Poles at origin but not repeated ✅ Sometimes OK
sF(s) has pole at origin ❌ No
Background: Steady-State Error and System Type

In control theory, the steady-state error ess measures how well a system tracks a desired input
in the long run.

This error depends on:

 The type of input (step, ramp, parabola, etc.)

 The type of system (Type 0, 1, 2… based on number of integrators or poles at origin)

System Type Review

System Type Number of Poles at Origin Behavior for Inputs

Step: finite error

Type 0 0
Ramp: ∞

Step: zero
Type 1 1
Ramp: finite

Step: zero
Type 2 2 Ramp: zero
Parabola: finite

Inputs in Laplace Domain

 Step: R(s)=1/s
 Ramp: R(s)=1/s2
 Parabola: R(s)=1/s3
Definition of Velocity Error Constant
This is called the velocity error constant. It represents how well the system tracks a ramp
input.

🚫 If Kv=0:

 Ramp error becomes infinite.

 Happens in Type 0 systems (no integrator).

✅ If Kv>0:

 Ramp error is finite.

 Happens in Type 1 systems (1 integrator, e.g., PID controller).

Summary

 ess=1/Kv is a standard result for ramp input in unity-feedback systems.

 Kv is called the velocity error constant.
 It depends on how many poles are at the origin (system type).
We can apply FVT if the system is stable and sF(s) does NOT have any pole at the origin or in the right-
half plane
What is "System Type" in Control Engineering?

The type of a system refers to the number of integrators (i.e., number of poles at the origin) in
the open-loop transfer function:

G(s)H(s)

System Types (based on open-loop transfer function):

Poles at
Type Definition Example
Origin
Type 0 No integrators 0 G(s)=1/s+1
Type 1 One integrator 1 G(s)=1/s(s+1)
Type 2 Two integrators 2 G(s)=1/s2(s+1)

Effect of Type on Steady-State Error

Input Type Type 0 Type 1 Type 2

Step Finite error Zero error Zero error

Ramp Infinite error Finite error Zero error

Input Type Type 0 Type 1 Type 2

Parabolic ∞ ∞ Finite error

Why Does This Happen?

Because integrators (poles at s=0s = 0s=0) accumulate error over time and work to eliminate it:

 Type 0 has no integrators → cannot eliminate steady-state error for ramp or parabolic
inputs
 Type 1 has one integrator → eliminates step error
 Type 2 has two integrators → eliminates both step and ramp error

How Controllers Affect System Type

P Controller:

 Only proportional action

 Does not change the type
 So system type = plant type

PI Controller:

 Has an integrator Kis\frac{K_i}{s}sKi

 Increases system type by 1
 So:
o Type 0 → Type 1
o Type 1 → Type 2
 Great for eliminating steady-state error

PD Controller:

 Adds derivative term

 Does not change system type
 Only improves transient response

PID Controller:

 Combines P + I + D
 Increases system type by 1 (due to the I term)
 Also improves speed and damping (due to D term)
Summary: How System Type Affects Error Behavior
Controller Adds Integrator? Changes Type? Effect on Error

P ❌ No Some error remains

PI ✅ Yes (+1) Steady-state error reduced/eliminated

PD ❌ No No error change, better speed/damping

PID ✅ Yes (+1) Combines best of both worlds

1. What is a P Controller (Proportional Controller)?

A Proportional (P) controller adjusts the output control signal proportional to the current
error.

What it does:

 If the error is large, it reacts strongly.

 If the error is small, it gives a small correction.
 If error is zero, control output is zero (i.e., no action).

Example: First-Order System with P Controller

System:

Let’s control a system (like a heater, motor, or tank level) modeled as:

This is a first-order plant with a time constant of 1.

🧪 Step Input:

Set point r(t)=1 (unit step input)

New Gain: Kp=5
Step 5: Interpretation
Quantity Value

Final value 5/6≈0.833

Time constant τ=1/6≈0.167

Rise time Faster (due to larger Kp)

Steady-state error 1−5/6=1/6≈0.1671

Summary Table
Method Steady-State Error Overshoot Risk

Increase Kp ↓ (but ≠ 0) ↑ Overshoot, instability

Kp→∞ →0 ↑↑↑ Actuator saturation, noise sensitivity

Use PI or PID controller = 0 Tunable Best choice in practice

Conclusion:

 The steady-state error for a ramp input is infinite with a P controller

 The output will always lag behind the input in a linearly increasing way

➤ Even very large Kp only reduces the slope of error, but never eliminates it

Summary Table
Controller Input Type Steady-State Error
P Step 1/1+Kp
P Ramp infinity
PI Step 0
PI Ramp 1/Ki

PI Controller with Ramp Input — Complete Explanation with Example

Step 1: Understanding the PI Controller

A PI controller adds an integrator, which helps eliminate steady-state error even for ramp
inputs.

📘 Control Law:
Final Results
Controller Ramp Input Response Steady-State Error

P only Output lags linearly ∞

PI Output catches up 1/Ki

PD Controller with First-Order System

PROBLEM!

There’s a pole at the origin in sE(s)

This violates the FVT condition:

FVT is NOT valid if sF(s) has a pole at the origin

But Why Did We Still Get a Finite Answer?

Good question.

Here’s the catch:

 Mathematically, the limit lim⁡s→0s⋅E(s) exists and is finite → that’s why the result (like
ess=1/6) makes sense in this special case.
 But formally, since there is a pole at the origin, FVT should not be trusted blindly.

So:

 You can still compute the limit manually in such cases (as we did).
 But you should NOT invoke FVT unless you're sure that sF(s)) has no pole at s=0

Summary Table
Input Type System Type Controller Steady-State Error Behavior

Step Type 0 PD Finite Faster response

Ramp Type 0 PD Finite Cannot eliminate error

Important Notes:

 PD does not increase system type → cannot eliminate steady-state error

 PD is best for transient improvement:
o Faster rise time
o Less overshoot
o More damping
 For zero steady-state error, use PI or PID

What Is a PID Controller?

A PID controller combines:

 P (Proportional): reacts to present error
 I (Integral): reacts to past error (removes steady-state error)
 D (Derivative): predicts future error (improves stability, damping)
Summary Takeaways

 PID greatly improves both transient and steady-state performance

 Integral term adds memory — removes error
 Derivative term adds foresight — improves damping
 With proper tuning, even a Type 0 first-order system can behave like a Type 1 or Type
2 system
Objective

We are analyzing how various controllers affect:

 Speed of response
 Accuracy (steady-state error)
 Damping (overshoot and oscillations)

For a second-order plant given by:

This is a standard second-order system with:

 No pole at the origin ⇒ Type 0

 Poles: at complex conjugate roots of s2+2s+4
 Damping ratio: ζ=2/2(4)05 =0.5
 Natural frequency: ωn=sqrt{4} = 2
Intuition

 A P controller only boosts the system response proportionally to error.

 It can’t accumulate the error over time → steady-state error remains.
 Since there is no integrator (no pole at origin), the system is unable to fully eliminate
the error, especially for non-constant inputs (like ramps).

Final Summary for P Controller

Input Type Output Behavior Error

Step Reaches 0.555 Finite error (≈ 0.445)

Ramp Diverges Infinite error

Left Plot: Step Input Response

 The black dashed line is the step input (ideal response).

 The blue curve is the system output:
o Rises quickly due to gain Kp=5K_p = 5Kp=5
o Overshoots slightly, then settles
o Final value ≈ 0.555, not 1

This confirms:

 The response is fast, but

 There is finite steady-state error (~0.445)

📉 Right Plot: Ramp Input Response

  The black dashed line is the ideal ramp input.
 The red curve is the system output:
o It lags behind the ramp input significantly
o Error grows larger over time

This confirms:

 A P controller alone cannot track a ramp input

 There is infinite steady-state error

PI CONTROLLER
System Type
System Type

Original 0

With PI 1

How this changes performance:

Input Type Type 0 System Type 1 System

Step Finite error ✅ Zero error
Ramp Infinite error ❌ Small error
Parabola ∞ ∞

Summary: PI Controller on Second-Order Plant

Feature Result
System Type Increases from 0 → 1
Poles Added 1 pole at origin (integrator)
Step Response Final value = 1 (zero error)
Ramp Response Small but finite error
Transient Behavior Slower than P, possible overshoot
Stability Stable if tuned carefully

PD CONTROLLER
Transient Behavior:

 Faster rise time due to derivative term

 Less overshoot (damping increased)
 More stable and smoother transition

Summary of PD Controller Effects

Feature Result
System Type Remains Type 0
Poles Added No pole at origin (no integrator)
Derivative Action Predicts future error → better damping
Step Response Finite error (~0.445), faster, smoother
Ramp Response ❌ Infinite error (can’t track ramp)
Use Case Improving speed & reducing overshoot
When to Use PD

 You want to reduce overshoot and oscillation

 You're okay with finite steady-state error
 You're working with noisy signals (careful—derivative amplifies noise)

PID CONTROLLER
This controller includes:

 Proportional (P): Kp=5

 Integral (I): Ki=4→ Eliminates steady-state error
 Derivative (D): Kd=2→ Improves speed and damping
Summary of Step Input Behavior

 Rises fast
 Minimal overshoot (derivative effect)
 Zero error (integrator effect)
Key Takeaways
Feature PID Behavior on 2nd Order Plant
System Type ✅ Type 1 (pole at origin added)
Step Input Error ✅ Zero
Ramp Input Error ✅ Small (almost zero)
Speed of Response ✅ High (due to derivative)
Overshoot Can occur if not tuned well
Stability ✅ Good with proper gains

Step Input Response

 ● P Controller: Fast rise, but doesn't reach full output (steady-state error).
 ■ PI Controller: Slower rise, but eliminates steady-state error.
 ▲ PID Controller: Fastest and accurate — reaches desired output with low overshoot.

Ramp Input Response

 ● P Controller: Fails to track ramp input — output lags far behind.

 ■ PI Controller: Tracks ramp better, small steady-state error.
 ▲ PID Controller: Closely follows the ramp — best tracking.
Interpretation of Graphs

🔷 Step Input:

 P controller rises quickly but settles below 1 → finite error.

 PI reaches 1 but slower → integrator improves accuracy.
 PID reaches 1 quickly and smoothly → best of both.

🔶 Ramp Input:

 P: Large growing error (can’t track slope).

 PI: Small error, lags but catches up slowly.
 PID: Best ramp tracking among all — very close to ideal.

Important Notes

 This is linear simulation with ideal controllers (no saturation, noise, etc.).
 The results assume:
o Controllers are properly tuned (Kp=5, Ki=4, Kd=2)
o Unity feedback
o No external disturbances

Varying Kp Values in 2nd order system

Varying the proportional gain Kp affects the behavior of a second-order control system — especially
in time domain (step response).
Summary Table
Kp Rise Time Overshoot Settling Time Steady-State Error

Low Slow None Long High

Moderate Faster Some Shorter Low

High Very Fast High May increase Very Low

Varying the derivative gain Kd affects a second-order system, while keeping proportional gain Kp
constant.
Key Observations
Parameter Effect of Increasing Kd

Damping Increases → reduces overshoot and oscillation

Rise Time Slightly increases (response slows a bit)

Settling Time Decreases due to better damping

Overshoot Reduces significantly

Steady-State Error Unchanged (no integrator present)

So, derivative gain Kd adds predictive control, damping the system to make it less oscillatory
without affecting steady-state error.

Qualitative Summary

Kd Response Characteristics
0 More oscillation, overshoot, slower settling
1 Less overshoot, damping starts to improve
2 Well-damped, minimal overshoot
5 Very smooth response, longer rise time but fast settling
10 Overdamped, slower rise, no overshoot

The derivative term acts like a brake: it slows down sudden changes and reduces oscillations.

Varying integral gain Ki (while keeping Kp constant) affects the behavior of a second-order system
Effect of Varying KiK_iKi
Parameter Behavior as Ki↑
System Type Increases from Type 0 → Type 1 (adds integrator)
Steady-State Error (Step) Becomes zero (with any non-zero Ki)
Steady-State Error (Ramp) Decreases with larger Ki
Overshoot Tends to increase (integrator adds phase lag)
Settling Time Can increase (slower decay)
Stability May degrade (larger Ki can destabilize the system)
Summary Table
Ki Rise Time Overshoot Settling Time Steady-State Error
0 Fast Minimal Moderate High
1 Slightly slower Some Longer Improved
5 Moderate Noticeable Long Zero (step)
May oscillate or become
10 Slower High Zero
unstable

Key Takeaway

 Low Ki : System improves accuracy slowly.

 Moderate Ki : Eliminates steady-state error with acceptable stability.
 High Ki Too aggressive — may cause instability or overshoot.

What Are Poles and Zeros?

Given a transfer function:

 Zeros are the roots of the numerator → z1,z2,…

 Poles are the roots of the denominator → p1,p2,…

In other words:

 Zeros are values of s that make the output zero

 Poles are values of s that make the system go to infinity

Where:

 Y(s): output in Laplace domain

 R(s): input in Laplace domain
 N(s): numerator polynomial
 D(s): denominator polynomial
Definition
Term Definition Mathematically

Zero Values of s that make the numerator N(s)=0→ output = 0 Solve N(s)=0

Pole Values of s that make the denominator D(s)=0→ system becomes infinite Solve D(s)=0

Physical Meaning

🔸 Poles:

 Represent natural response of the system.

 Determine stability, oscillation, and transient response.
 Related to time constants and damping.

🔹 Zeros:

 Influence the shape of the response.

 Can cause:
o Inverse response (initial dip)
o Suppression of certain frequencies
o Faster/slower rise times
General Rule:

 Poles in Left Half Plane (LHP): exponential decay → stable

 Poles in Right Half Plane (RHP): exponential growth → unstable
 Poles on Imaginary Axis (jω): sustained oscillation → marginally stable

Graphical Representation

We plot poles and zeros on the complex plane (s-plane):

 Real axis (horizontal): Decay or growth

 Imaginary axis (vertical): Oscillations
Importance in Control Design
Concept Role of Poles/Zeros

Stability Depends only on pole locations

Performance Rise time, overshoot, damping from pole positions

Root Locus Follows path of poles as KKK increases

Bode/Nyquist Frequency behavior influenced by zeros/poles

Controller Tuning Placing zeros (via controller) to modify system

Let's now theoretically analyze how changing zeros affects system response through a step-by-step
solved example with detailed explanation.
Because the initial slope is higher (1.0 vs 0.167), the system with zero at -1 responds faster
initially.
Summary Table (All 3 Cases)
Case Zero Location Time Response Initial Slope Effect
Base None Smooth ≈ 0.167 Normal
Zero at -1 LHP Faster rise 1 Faster
Zero at +1 RHP Initial dip 1 Inverse

Final Insight

The non-minimum phase behavior (from RHP zero) leads to unexpected initial output, even
though the system settles properly later. This is undesirable in most control applications — such
systems are harder to stabilize and tune.
POLES BEHAVIOR ON SYSTEM RESPONSE

What Are Poles?

Given a transfer function:

 The roots of the denominator D(s)=0D(s) = 0D(s)=0 are the poles.

 Poles describe the natural modes of the system — how it responds without input.

Effect of Pole Location on System Behavior

1. 📍 Real Poles in Left-Half Plane (LHP)

Summary Table
Pole Location Response Equation Behavior Settling Speed

s=−5 1−e−5t Fast, stable Fastest

s=−1 1−e−t Smooth, stable Moderate

s=−0.1 1−e−0.1t Slow rise Very slow

s=0 t Ramp (unbounded) No settling

s=+1 et−1 Unstable growth Diverges

Complex poles
Where:

 ωn= : natural frequency

 ζ: damping ratio

We'll analyze step responses for various values of ζ to observe:

 Oscillations
 Overshoot
 Settling time

We'll assume ωn=1 for simplicity.

CASES:
Case Damping Ratio ζ Nature of Poles
Underdamped 0<ζ<1 Complex conjugate (oscillatory)
Critically damped ζ=1 Repeated real pole (fastest no overshoot)
Overdamped ζ>1 Two real poles (slower, no oscillation)
Undamped ζ=0\= Pure imaginary poles (sustained oscillations)
Final Summary Table
ζ System Type Overshoot Oscillates? Settling Time Stability
0 Undamped 100% Yes Never Marginally stable
0.5 Underdamped ~16% Yes Moderate Stable
Critically
1 0% No Fastest Stable
damped
>1 Overdamped 0% No Slower Stable

Let's numerically analyze how adding controllers (P, PI, PD, PID) changes the pole
locations and step responses of a second-order plant
Steady-State Error (SSE)

Steady-state error is the difference between the input and output of a system as time
approaches infinity, for a given input.

It is a key performance measure of control systems and depends on:

 Type of input (Step, Ramp, Parabolic…)

 System Type (Type 0, 1, 2…) — determined by number of poles at origin
 Controller design (P, PI, PD, PID…)
So, higher system type = better tracking for faster-changing inputs

🔹 Position Error Constant (Kp)

🔹 Velocity Error Constant (Kv )

🔹 Acceleration Error Constant (Ka )

These are called Static Error Constants, used to determine the steady-state error (ess ) of a
system for standard inputs like step, ramp, and parabolic.

Objective of Static Error Constants

They tell us how well a system can track a reference input as time t→∞ (i.e., in steady-state).
Conclusion for Plant Alone:
Input Error Constant SSE Interpretation
Step Kp=∞ 0 Perfect tracking
Ramp Kv=0.5 2 Poor tracking (lagging)
Parabolic Ka=0 ∞ Cannot follow curvature at all
Final Comparison
Input Without Controller (Type 1) With PI Controller (Type 2)
Step 0 0✅
Ramp 2 0✅
Parabolic ∞ Finite ⚠️

Interpretation:

 Adding PI controller increases system type → improves SSE.

 BUT, it can decrease stability margin (more prone to oscillations) → needs tuning!

We calculate error constants (like Kp,Kv,Ka ) using the open-loop transfer function, but those
constants determine the error in the closed-loop system.
Final Takeaway:

 We design and analyze the controller using the open-loop transfer function
 But the actual system performance (like SSE) is observed from the closed-loop
behavior

PD Controller Structure

The PD controller has the transfer function:

What does Derivative Control do?

The derivative term acts on the rate of change of error, i.e., how quickly the error is changing.

Key role:

 Anticipates future error by observing current trend

 Acts as a predictor
 Damps oscillations and speeds up settling

Effects of Derivative Term

Effect Explanation

🌊 Reduces Overshoot Damping increases, so output does not overshoot as much

⏱ Faster Settling Time Reaches final value quicker, less oscillation

🚫 No change to SSE Derivative has no effect on steady-state error (SSE) — only affects transient

⚠️Sensitive to noise Amplifies high-frequency noise due to the "s" term in Kds
Why Ramp Error is Infinite?

Because PD controller doesn't include an integrator (no pole at origin). Only integrators
eliminate ramp error. That’s why PI or PID are used for ramp or parabolic tracking.

Why Use Routh-Hurwitz?

Stability Conditions

The system is stable if and only if:

 1. All coefficients of the characteristic equation are positive and non-zero.
2. All elements of the first column of the Routh array are positive (no sign changes).

If either fails, the system is unstable.

Special Cases

1. Zero in First Column

If a zero appears in the first column but other elements in the row are non-zero:

 Replace 0 with a small number ϵ\epsilonϵ

 Complete the array
 Take the limit ϵ→0

Used to detect marginal stability or oscillations.

2. Entire Row Becomes Zero

This means symmetrical root pair on imaginary axis (pure imaginary roots):

 Go to the row above the zero row

 Take its auxiliary polynomial (even power terms)
 Differentiate it
 Replace the zero row with derivative coefficients

Summary

 No need to calculate roots!

 Just construct the Routh array
 Count sign changes in the first column:
o 0 → stable
o 1+ → unstable
 Use special rules when encountering zeros

What Routh Tells You:

Feature Interpretation
All signs positive Stable system
One sign change One unstable pole in RHP
Row of zeros Imaginary axis poles (marginally stable)

What is Root Locus

The Root Locus is a graphical method that shows how the roots of the closed-loop transfer function
move in the complex s-plane as a system parameter (usually gain K) varies from 0 to ∞.

Why Use Root Locus?

 To determine stability of the system for various gains KKK

 To design controllers by choosing a gain that yields desired poles
 To visualize how pole locations affect transient response
 Helps in tuning PID controllers

2. Total Number of Branches = Number of Poles

Each branch starts at a pole and ends at a zero (or ∞ if more poles than zeros).

3. Real Axis Rule:

A point on real axis is part of the root locus if it lies to the left of an odd number of real poles
and zeros.
Interpretation of Root Locus

 Far-left poles → faster and more stable

 Poles near jω-axis → slower and potentially unstable
 Poles crossing into RHP → unstable system
 Adding a zero pulls root locus toward that zero (improves speed but may increase
overshoot)
 Adding a pole pushes root locus away (can reduce stability)

Sensitivity Analysis Using Root Locus

Sensitivity refers to how much the system's performance (e.g. poles) change with small changes
in parameters (like gain):

 High sensitivity → poles move significantly with slight gain change (less robust)
 Low sensitivity → poles move slowly (more robust)

Root locus allows you to visually see this sensitivity.

Summary Table
Feature Effect
Adding Zero Pulls root locus left (better speed)
 Feature Effect
Adding Pole Pushes root locus right (can destabilize)
Gain ↑ Moves poles along the locus
Pole in RHP Always leads to instability
Root near jω Slower oscillatory response
Symmetry Root locus is always symmetric about real axis

Step 1: Identify Poles and Zeros

This is an open-loop transfer function.

It has:

 Poles: at s=0,−4,−6
 No zeros

So, number of branches = number of poles = 3

Step 2: Root Locus Rules

1. Real Axis Segments:

To determine where the root locus lies on the real axis:

 It lies to the left of an odd number of poles and zeros.

From the pole locations:

 Between -−∞ to −6: one pole to the right → ✅ locus exists here.
 Between −6to −4: two poles → ❌ no locus.
 Between −4to 0: three poles → ✅ locus exists.
 Right of 0: zero poles → ❌ no locus.
What Happens As Gain KKK Increases

1. At K=0: Closed-loop poles are at open-loop poles → s=0,−4,−6

2. As K increases:
o Two poles (at s=−4and s=−6) move toward each other and then split into complex
conjugates → moving up and down in the complex plane
o One pole (at s=0) moves left toward infinity along the real axis

This is typical of systems with more poles than zeros → asymptotes guide them.

Interpretation of Pole Locations

Location Behavior
Left-half plane (LHP) Stable
Farther left (−10 or −20) Faster response
Near jω-axis (imaginary) Slower, oscillatory response
Right-half plane (RHP) Unstable

Sensitivity

Root locus gives visual sensitivity of pole locations to gain K:

 Where the root locus is steep or clustered, poles are sensitive to small changes in gain
→ poor robustness.
 Where root locus is spaced out smoothly, poles are less sensitive → better control.