Instrumentation

Lecture 3 PPT

Quiz ChatGPT conversation

Controller Design — Full Educational Summary

Page 1 — Introduction

Controller design This lecture block consists of 6 lectures/exercises.

This means you are starting a new course module focused on automatic control systems.

The core idea:

How do we make a system automatically behave the way we want?

Examples:

keep temperature constant
keep pH constant
keep liquid level stable
control pump speed

This is one of the most important engineering concepts.

Page 2 — Course outline

The lecture gives the roadmap:

1) Structuring a control problem

Before solving, we must define:

What should be controlled?
What can we manipulate?
What disturbs the system?

This is system thinking.

2) Description / modelling

This is the mathematical part.

We convert the real physical system into equations:

differential equations
transfer functions
block diagrams

3) Controller design

Once we know system behavior, we can design a controller.

Examples:

on-off controller
proportional controller
PID controller

4) Lab exercise

Practical implementation.

Very important for exam projects.

Page 3 — Outline for today

Today’s topics:

Introduction to control
Feedforward control
Feedback control
Block diagrams
On-off control

This is basically the control theory foundation lecture.

Page 4 — Control and supervision system

This slide shows a real industrial control system.

The system contains:

Inputs (measured variables)

Sensors measure things like:

temperature
level
oxygen concentration
pH

These are called process variables (PV).

Outputs / actuation

The system can change:

power
valve position
mixing speed
pump velocity

These are manipulated variables (MV).

Signal chain

The flow is:

sensor → computer/controller → amplifier → actuator

This is extremely important.

The controller never directly changes the process.

It changes actuators.

Example:

controller changes valve opening

which changes

hot water flow

which changes

room temperature

Page 5 — Main purposes of a controller

This slide is extremely important.

1) Stationary conditions

This means steady state behavior.

Question:

When the system settles, is it close to the desired value?

Example: Wanted temperature = 25°C Actual = 24.9°C

Very good steady state.

2) Dynamic conditions

This means:

How fast does the system recover after change?

Example: Someone opens window → room cools

How quickly does it return?

Fast recovery = good controller

3) Disturbance and noise reduction

Real systems are constantly disturbed.

Examples:

room window opens
external temperature changes
noisy sensors

Good controller minimizes their effect.

4) Stability

This is one of the most important ideas.

A bad controller can cause oscillation.

Example: temperature: 24 → 27 → 23 → 28 → 22

This is unstable / oscillatory.

A good controller converges smoothly.

Page 6 — Why use control?

Two main reasons.

Follow a reference

Reference = desired target value.

Examples:

temperature profile
fixed pH
nutrient addition

The controller makes output follow this target.

Reject disturbances

Very important.

Example: Outside temperature suddenly drops

The controller compensates.

This is classic disturbance rejection.

Page 7 — Controlled vs uncontrolled

This is a conceptual slide.

Three cases:

A) Poor control

Large variation.

Because of large fluctuation, you must stay far from system limits.

Example: If maximum safe temp = 50°C you operate at 40°C

for safety.

B) Better control

Less variation.

Safer and more accurate.

C) Excellent control

Very small variation.

Now you can move reference closer to optimal limit.

This improves:

efficiency
productivity
safety margins

Very important in industry.

Page 8–10 — Indoor temperature example

This is a fantastic beginner example.

Feedforward control

This means:

act before measuring output

Example: You know night is coming → room will cool

So you increase heater before temperature drops

This is predictive control.

Feedback control

This means:

measure output and correct error

Example: Room too cold → open valve more

Room too hot → close valve

This is the most common control strategy.

Important difference

This is highly exam-relevant.

Feedforward

Uses prediction / knowledge.

No direct error correction.

Fast, but sensitive to modelling errors.

Feedback

Uses measured output.

Corrects actual error.

More robust.

Pages 11–17 — Block diagrams

This is one of the most important sections.

Why block diagrams?

They simplify complicated systems.

Instead of drawing real hardware, we draw functional blocks.

Example:

Reference → Controller → Plant → Output

Standard feedback loop (VERY IMPORTANT)

Page 17 is fundamental.

The standard setup is:

Reference → Error → Controller → Plant → Output

with sensor feedback loop.

Definitions

Reference

Desired value

Example: 25°C

Error

Difference:

e = reference - output

This is central.

e = r - y

If output too low → positive error

Controller

Computes action from error.

Example: increase heater power

Plant / System

The physical thing.

Example: room + heater

Sensor

Measures output.

Example: thermometer

Disturbance

External unwanted input.

Example: open window

Pages 18–21 — Heat exchanger

Very important industrial example.

This is the real-world application of the theory.

The goal:

control outlet temperature

Manipulated variable:

valve opening / hot water flow

Measured variable:

output temperature

Classic process engineering example.

Very exam-relevant.

Pages 23–26 — On-off control

This is extremely important.

What is on-off control?

The controller only has two states:

No intermediate values.

Exactly like home thermostats.

Example

If temperature < setpoint → heater ON

If temperature > setpoint → heater OFF

Important consequence

Output oscillates around setpoint.

This is normal.

It does NOT settle perfectly.

This is explicitly mentioned.

pH example

Very good example.

Target: pH = 4.8 \pm 0.05

If pH too high → add acid

If pH low enough → stop

This is discrete switching control.

Timer logic

Very important practical design.

The acid is not continuously added.

Instead:

dose for some time
wait
measure again

This avoids overshoot.

Very realistic industrial logic.

Pages 27–31 — Dynamic modelling

This is where math starts.

Very important.

Liquid heating model

This is classic first-order system.

The equation:

cM\frac{dT_L}{dt}=P_-hA(T_L-T_s)

This is simply:

accumulation = input heat − heat loss

Physical meaning

Left side

Temperature change over time

Thermal storage

First term

Heat supplied by heater

Second term

Heat loss to surroundings

Bigger temperature difference → bigger loss

Very intuitive.

Time constant

This is one of the most important concepts.

\tau = \frac{cM}{hA}

This tells how fast system responds.

Large τ

Slow system

Example: large tank

Small τ

Fast system

Example: small cup of water

The famous 63% rule

t=\tau

the system reaches 63% of final value

y(\tau)=0.63K

This is extremely important in control engineering.

You should absolutely remember this.

Pages 32–35 — Transfer functions

This is a major concept.

Instead of differential equations, we transform them into Laplace domain.

This makes calculations algebraic.

Why useful?

Differential equations are difficult.

Laplace transform converts:

\frac{dy}{dt}

into

sY(s)

Much easier.

Transfer function

General form:

G(s)=\frac{Y(s)}{U(s)}

This describes system response.

Extremely important for controller design.

Pages 36–40 — Block diagram algebra

These pages are very important for problem solving.

Series / cascade

Two blocks in sequence:

A then B

Equivalent:

G_=AB

Parallel

Two paths added:

C + D

G_=C+D

Block transformations

These slides teach how to simplify complex diagrams.

Very likely exam topic.

Page 41 — Detailed real implementation

Excellent practical slide.

This shows real digital control.

Flow:

sensor
ADC
controller algorithm
DAC / PWM
amplifier
actuator
system

This is how embedded systems work.

Very relevant for engineering.

Page 42 — Mini project

This slide is extremely exam relevant.

Project requires:

control concept
dynamic model
theoretical controller
implementation + test

This is basically the entire course workflow.

Big-picture understanding

This whole lecture builds the foundation:

real process → model → block diagram → controller

That is the control engineering workflow.

One-sentence takeaway

A controller continuously compares the actual output with the desired reference and changes the system input to minimize error while remaining stable.

Quiz

Score: 0/30 (0%)

Q0. In a standard feedback control loop, what is the main role of the error signal?

To represent the difference between the reference value and the measured or actual output

To directly measure the physical disturbance entering the system

To convert analog sensor signals into digital values

To describe the energy stored in the system

Q1. Which statement best describes feedback control?

The controller acts only from a pre-programmed schedule without measuring the output

The controller uses measured output information to correct deviations from the reference

The controller completely eliminates all disturbances before they enter the system

The controller only works when the plant has no dynamics

Q2. Which statement best describes feedforward control?

It waits for the output to deviate and then corrects the error

It uses measured output only and ignores predicted disturbances

It acts based on prior knowledge or predicted conditions before directly observing an output error

It always produces perfect control regardless of model accuracy

Q3. Why can good control allow a reference value to be placed closer to a system limit?

Because good control increases random variation around the reference

Because good control reduces output variation, leaving more predictable safety margins

Because good control removes the need for sensors

Because good control makes system limits physically disappear

Q4. In the indoor temperature example, what is the plant or system?

Only the thermometer

The physical room and heating dynamics being controlled

Only the reference temperature

Only the computer algorithm

Q5. What is the purpose of a sensor in a feedback control loop?

To define the desired reference value

To calculate the mathematical transfer function

To measure the output so it can be compared with the reference

To remove all noise from the controller

Q6. In the heat exchanger example, what is typically the controlled output?

The temperature of the outlet flow

The color of the valve

The clock frequency of the controller

The total number of blocks in the diagram

Q7. What does an on-off controller do?

It continuously varies the actuator signal over all possible values

It switches between two discrete states, typically on and off

It uses only Laplace transforms and no sensors

It eliminates the need for a setpoint

Q8. Why does on-off control usually cause output variation around the setpoint?

Because the controller cannot apply intermediate actuator values and the process has reaction time

Because the reference value is always unknown

Because sensors cannot be used in on-off control

Because all physical systems are unstable

Q9. In the pH-adjustment example, why is acid added?

Because the uncontrolled pH is always below 4.8

Because the uncontrolled pH is always above 4.8 and must be lowered

Because the pH sensor cannot measure values below 7

Because acid addition increases the pH toward neutrality

Q10. In the pH on-off controller, what is the practical purpose of using dosage and break timers?

To replace the pH sensor entirely

To control acid addition in pulses while allowing time for the process to react

To make the acid flow uncontrollable

To calculate the Laplace transform of pH

Q11. In the liquid heating model, what does the term hA(TL - Ts) represent?

The heat supplied by the electrical heater

The heat loss from the liquid to the surroundings

The mass of liquid in the tank

The reference temperature of the controller

Q12. For the first-order liquid heating system, what does the time constant tau describe?

The maximum possible temperature of the liquid

The speed of the system response to a change in input

The exact power supplied by the heater

The number of sensors in the loop

Q13. For a first-order system responding to a unit step input, approximately how much of the final response has occurred after one time constant?

10%

37%

63%

100%

Q14. Why are transfer functions useful in controller design?

They allow linear differential equations to be handled using algebraic expressions in the Laplace domain

They remove all physical disturbances from real systems

They eliminate the need to understand system input and output

They only apply to manual control systems

Q15. In block diagrams, arrows represent physical pipes rather than signals.

True

False

Q16. A major purpose of feedback control is to follow a reference value and reduce the effect of disturbances.

True

False

Q17. Feedforward control requires perfect measurement of the output error before it can act.

True

False

Q18. Feedback control is generally more robust than pure feedforward control when unexpected disturbances occur.

True

False

Q19. In the indoor temperature example, opening a window can be considered a disturbance.

True

False

Q20. The controller directly changes the physical output, such as temperature or pH, without using an actuator.

True

False

Q21. A block diagram helps structure a control problem by identifying elements, inputs, outputs, and their connections.

True

False

Q22. In steady state, the temperature in the liquid heating example is no longer changing with time.

True

False

Q23. The gain K in the liquid heating example describes the steady-state change in liquid temperature per change in heater power input.

True

False

Q24. A larger mass of liquid generally increases the time constant of the heating system, making the system slower to respond.

True

False

Q25. Transfer functions are mainly used because they make it possible to replace algebraic equations with more complicated differential equations.

True

False

Q26. In a cascade or series block diagram, the equivalent transfer function is found by multiplying the transfer functions of the blocks.

True

False

Q27. In a parallel block diagram, signals from parallel branches are commonly added or subtracted at a summing point.

True

False

Q28. A real digital control system may include A/D conversion, a control algorithm, D/A conversion or PWM, an amplifier, an actuator, the system, and a sensor.

True

False

Q29. The mini-project workflow includes developing a control concept, making a dynamic model, designing a theoretical controller, and implementing and testing it.

True

False