Chapter 1

1.1 Background of the Project

1.1.1 Introduction

The ball and plate system is a classic control engineering example that demonstrates key
principles of dynamics and control in a practical. It involves controlling the position of a ball on
a tilting beam by adjusting the plate's angle, showcasing fundamental concepts like stability,
feedback, and system response. This project explores the modeling and automated control of the
ball and plate system, aiming to enhance precision and reliability compared to traditional manual
control methods.

1.1.2 Historical Context

The Ball-on-Plate balancing system has its roots in the Ball-on-Beam system, a classic control
engineering experiment introduced in the mid-20th century. Early systems relied on manual
control, where operators visually monitored the ball's position and adjusted the beam’s tilt, but
these methods were prone to human error and lacked precision. By the 1970s, advancements in
control theory enabled the adoption of automated techniques like Proportional-Integral-
Derivative (PID) control, significantly improving performance and reliability. The transition
from a beam to a plate in later years added complexity, allowing for two-dimensional control and
paving the way for modern applications in robotics and automation.

1.1.3 Development of the Project Idea

This project is inspired by the Ball-on-Plate system's ability to demonstrate key control
principles like stability and feedback. By addressing the limitations of manual systems, the focus
is on developing automated control strategies to enhance performance and reliability. Through a
review of existing literature, challenges in manual operation were identified, emphasizing the
need for innovative approaches. This project aims to create a mathematical model and implement
control techniques to effectively manage the ball's position on the plate.
1.1.3 General Description

The ball and plate balance project are an engaging exploration of controlling a dynamic system
where a ball rolls along a tilted plate. Our main goal is to keep the ball at a specific position by
adjusting the plate’s angle in real time, effectively demonstrating key principles of control
engineering.

We will start by developing a mathematical model to describe the system's dynamics, including
the equations of motion that govern the ball's behavior when the palate is tilted. This model will
help us design effective control strategies.

Next, we’ll implement control methods, including Proportional-Integral-Derivative (PID) control

for stability These strategies will be tested through simulations using MATLAB/Simulink to
visualize the system's performance under different conditions.

1.1.4 Expected Outcomes

By the end of the project, we expect to achieve a reliable mathematical model, effective control
algorithms, and a MATLAB and Solidwork simulation that demonstrates automated control
principles. Ultimately, this project aims to bridge theoretical concepts with practical applications,
providing valuable insights into dynamic systems in a hands-on manner.

By engaging with this hands-on project, we aim to deepen our understanding of automated
control systems and their implications across various engineering domains, enriching both
academic knowledge and practical skills.

1.2 problem statement

The problem statement for the Ball-on-Plate Balancing System involves developing a control
mechanism to maintain the position of a ball on a flat surface, which is inherently unstable due to
external disturbances. The primary challenge lies in accurately controlling the plate's tilt using
motors to counteract these disturbances and keep the ball centered.

The problem under study focuses on designing an effective control strategy and utilizing
techniques such as PID control and simulation tools like MATLAB to achieve precise trajectory
tracking and stabilization of the ball on the plate while addressing issues of system non-linearity
and instability.

1.3 Objectives

1.3.1 General Objective

The general objective of the "Ball-on-Plate Balancing System Simulation and Control Design
Using MATLAB Simscape project is to develop a comprehensive understanding of dynamic
systems and control theory through the design and simulation of a ball-on-plate balancing
system. This system serves as a practical example of how to apply theoretical concepts in a real-
world scenario.

1.3.2 Specific Objectives

1. Modeling the System

 Develop a mathematical model of the ball-on-plate system, incorporating the dynamics of

the ball, plate, and the forces acting on them.

 Utilize MATLAB Simscape to create a detailed simulation model that accurately represents
the physical characteristics and behaviors of the system.

2. Control Design

 Design a control strategy (PID) to achieve desired performance specifications, such as

stability, speed of response and robustness against disturbances.

 Implement the control algorithm within the MATLAB Simscape environment and ensure it
can effectively balance the ball on the plate.

3. Simulation Analysis

 Conduct simulations to visualize the ball's movement and the plate's response to control
inputs.

 Analyze the system's behavior under various initial conditions and disturbances to
understand stability and performance metrics.
4. Performance Evaluation

 Evaluate the performance of the control system through simulations, assessing metrics such
as settling time, overshoot and steady-state error.

 Compare different control strategies to identify the most effective approach for maintaining
the ball's position on the plate.

5. Documentation and Reporting

 Document the entire process, including modeling, simulation results, control design, and
performance evaluations to provide a comprehensive report.

 Present findings and insights gained from the project, highlighting the importance of control
design in dynamic systems.

1.4 Literature review

The ball and beam balance system has become an essential tool in control engineering, providing
a hands-on way to explore various control strategies and dynamic behaviors. This literature
review will look at key developments related to modeling, control techniques, educational
applications, and some of the limitations of the ball and beam system—all of which are relevant
to our project.

1.4.1 System Modeling

Accurate modeling is crucial for effectively controlling the ball and beam system. Early studies
focused on deriving equations of motion based on Newtonian mechanics, examining the forces
acting on the ball and the torque applied to the beam. For instance, Ogata (2009) highlights how
important it is to develop precise mathematical models to ensure that controllers perform well,
especially when dealing with nonlinear dynamics. Creating state-space representations is also
vital, as they provide a comprehensive view of how the system behaves.

Recent research has refined these models by incorporating factors like friction, variations in
mass, and external disturbances, which leads to more robust control designs. However, it's
important to note that many models rely on idealized assumptions that might not hold true in
real-world scenarios. This can create gaps between predicted and actual performance, making it a
challenge for engineers to achieve the desired results.

1.4.2 Control Strategies

The ball and beam system has been widely used to test various control strategies. One of the
most commonly applied methods is Proportional-Integral-Derivative (PID) control, known for its
simplicity and effectiveness. Research by Astrom and Wittenmark (1997) shows how well PID
controllers can stabilize the ball’s position, demonstrating their robustness in different
conditions. However, while PID control works well for many applications, it has its limitations,
particularly when it comes to handling nonlinearities and time delays.

State-space control methods have gained traction for their ability to provide better performance
through full state feedback. Lewis and Syrmos (1995) discuss how these methods can manage
the dynamic nature of the ball and beam system more effectively. Additionally, nonlinear control
strategies, highlighted by Khalil (2002), have shown promise in managing the system's
complexities, especially during transient responses. Yet, these advanced techniques can
introduce new challenges, such as increased computational demands and the need for precise
modeling.
Chapter Two

2.1 Introduction

The Ball-on-Plate Balancing System serves as a benchmark for testing classical control
techniques due to its inherently unstable and nonlinear dynamics. This system requires precise
control to maintain the ball's position on a flat plate by adjusting the plate's inclination. The study
of such systems is crucial in robotics, aerospace, and industrial automation, where stability and
dynamic control are essential.

This paper focuses on the simulation and control design of a Ball-on-Plate system using
Solidwork and MATLAB, a powerful platform for modeling multidomain physical systems. By
leveraging Simscape's capabilities, the study aims to provide an intuitive understanding of the
system's dynamics and evaluate control strategies such as PID. The development process
highlights the integration of mechanical and computational models, enabling a comprehensive
approach to system analysis and controller design.

2.2 User/Customer Requirement

The Ball-on-Plate Balancing System must maintain the ball's position at a desired target point on
the plate with minimal error and rapid response. The system should exhibit stability under
disturbances, ensuring precise and reliable operation. Key user requirements include real-time
control, intuitive visualization of system performance, and the flexibility to test various
disturbance signals.

2.3 Target Specification

The Ball-on-Plate Balancing System is designed to achieve precise ball positioning within a
tolerance of ±2 mm from the target point. The system should respond to position changes within
1 second, maintaining stability under disturbances such as external forces. The simulation model
must accurately represent real-world dynamics, including friction and inertia, while supporting
the integration of control algorithms like PID. The MATLAB Simscape environment should
allow for adjustable parameters to optimize performance and support iterative testing for
controller refinement.

2.4 Concept generation

2.4.1 Concept 1 Three-Servo Motor Configuration with camera Sensor

Figure 2.1 Three-Servo Motor Configuration with camera Sensor

2.4.1 Concept 2: Two-Servo Motor Configuration with Resistive Touch Sensor

Figure 2.2: Two-Servo Motor Configuration with Resistive Touch Sensor

2.5 Concept Selection

The three-servo motor configuration and a camera for position sensing offers more degrees of
freedom, enabling precise and versatile control over the plate's movement. This setup is
particularly beneficial for applications requiring advanced, multidirectional adjustments. The
camera sensor provides high-resolution feedback and flexibility in detecting the ball's position
across the entire plate. However, the system's complexity increases significantly, demanding
higher computational power. Additionally, the reliance on image processing may introduce
latency, making it less ideal for scenarios where real-time responsiveness is critical.

The two-servo motor configuration with Resistive Touch Sensor excels in simplicity, efficiency,
and reliability. This design achieves real-time control with minimal latency due to the near-
instantaneous feedback provided by the Resistive Touch Sensor. It is highly cost-effective and
requires less computational power, making it accessible for practical and standard balancing
tasks. The two-servo system is robust, easy to implement, and requires minimal maintenance,
making it a practical choice for applications.

Three- Servo Motor Design Two-Servo Motor Design

Degree of freedom 3degree of freedom 2 degree of freedom
Sensor Camera Resistive Touch Sensors
Complexity High Moderate
Feedback Latency Potential delays (image Minimal (direct analog
processing) feedback)
Response Time Slow Fast
TABLE 2.1 CHARACTERISTICS OF THREE SERVO MOTOR AND TWO SERVO MOTOR
DESIGN
2.5.1 NON-WEIGHTED DECISION MATRIX (PUGH CONCEPT SELECTION)

CONCEPT

SELECTION A B

Three- Servo Motor Two-Servo Motor Design

Design

EASE OF HANDLING 0 +

EASE OF USE - +

DURABILITY 0 +

RESPONSE TIME + 0

SYSTEM COMPLEXITY + 0

EASE OF - +
MANUFACTURING

PORTABILITY - +

SUM +’S 2 5

SUM 0’S 2 2

SUM –‘S 3 0

NET SCORE -1 5

RANK 2 1

CONTINUE NO YES

TABLE 2.2 NON-WEIGHTED DECISION MATRIX (PUGH CONCEPT SELECTION)

2.5.2 WEIGHTED DECISION MATRIX

CONCEPT

SELECTION A B

Three- Servo Two-Servo

Motor Design Motor Design

WEIGHT RATING WEIGHTE RATING WEIGHTE

D SCORE D SCORE

EASE OF 5% 2 0.1 4 0.2

HANDLING

EASE OF USE 15% 2 0.3 4 0.6

DURABILITY 10% 3 0.3 4 0.4

RESPONSE TIME 25% 5 1.25 4 1

SYSTEM 15% 5 0.75 3 0.45

COMPLEXITY

EASE OF 20% 3 0.6 5 1

MANUFACTURIN
G

PORTABILITY 10% 3 0.3 5 0.5

TOTAL 3.6 4.15

SCORE
2 1
RANK

CONTINU NO YES
E

TABLE 2.3 WEIGHTED DECISION MATRIX

2.6 System block diagram and operation workflow

Figure 2.3 Operational flowchart

Figure 2.4 System block diagram
3D model of the selected concept
Chapter Three

Mechatronic Design: System Level Design

3.1 Mechanical Design

3.1.1 System overview/components

Component Purpose
Plate Ball rolling surface
Supporting Frame Structural stability
Servo Motor Arms Connects motor to plate
Pivot Joint Enables smooth multi-
directional tilt
Ball Rolling object
Bearings Smooth rotation of moving
parts
Base Platform Stability and foundation
Fasteners Assembly of components
Spring Dampers Reduces vibrations

Couplers Connects servo motors and

arms

3.1.2 Material selection

When selecting a material different parameters must be considered those considerations include
several factors including their mechanical properties which contain strength stiffness weight and
wear resistance also their functional requirements and manufacturability also machinability
regarding to their cost and safety must be in consideration in order to achieve a reliable, efficient,
and cost-effective design.
3.1.2.1 Mechanical Properties
 A. Strength
The material must withstand the forces and stresses applied during operation without
deforming or breaking.

Key properties to evaluate:

 Tensile Strength: Ability to resist stretching or pulling forces (important

for servo arms and frame).
 Compressive Strength: Ability to resist squeezing forces (relevant for the
plate).
 Shear Strength: Resistance to sliding forces (important for joints and
fasteners).
A. Stiffness (Rigidity)
 Stiffness is crucial for components like the plate and servo arms to prevent
excessive bending or deformation under load.
 Young’s Modulus (E): A material with a high Young’s Modulus resists
deformation better.
B. Weight (Density)
 Lightweight materials reduce the overall load on actuators (servo motors) and
improve system responsiveness.
 Components like the plate, servo arms, and frame must balance strength and
weight.
C. Wear Resistance
 Materials exposed to friction or rolling (e.g. the ball and the plate) must resist
wear over time to maintain functionality and precision.

3.1.2.2 Functional Requirements

A. Plate
 The plate must be flat, rigid, and lightweight to provide a stable rolling surface for
the ball.
 It should resist bending under the ball’s weight and external disturbances.
 B. Servo Arms
 Servo arms must be lightweight and strong to minimize inertia and maximize the
system’s speed and efficiency.
 They must also resist bending and torsional stresses.
C. Frame
 The frame must provide stability and durability, ensuring that vibrations or
external forces do not compromise the system’s operation.
D. Pivot Joint
 The pivot joint must allow smooth, low-friction motion while being strong
enough to handle the forces from the tilting plate.
E. Ball
 The ball must have a smooth surface for consistent rolling and minimal friction
but should also resist wear and deformation from repeated impacts.
F. Base Platform
 The base must be rigid and stable to prevent the system from tipping over during
operation.

3.1.2.3 Environmental Factors and Manufacturability

I. Environmental conditions can influence material performance, so we must consider the

followings;
A. Corrosion Resistance
 If the system will be exposed to moisture or humidity, materials like aluminium,
stainless steel, or corrosion-resistant coatings are recommended.
B. Temperature Resistance
 If the system operates in environments with significant temperature changes,
materials with stable mechanical properties under varying temperatures (e.g. steel
or aluminium alloys) should be selected.
II. Manufacturability Consider how easily the material can be manufactured, shaped, or
modified for the system which include Machinability;
 Materials like aluminium are easy to cut, drill, and machine into precise shapes.
3.1.2.4 Material Compatibility

We must ensure that the selected materials for different components work well together:

I. Thermal Expansion
 Different materials expand and contract at different rates when exposed to
temperature changes. This must be considered to avoid stress or misalignment.
II. Friction and Wear
 Components in contact (e.g. the ball and plate) should have low friction and wear
properties to avoid degradation over time.

3.1.3 Analysis of each mechanical part

1. Plate (Ball Support Surface)

Material Selected: Aluminium

 Why Aluminium?
 Lightweight: Reduces the load on servo motors, allowing faster and more
efficient plate tilting.
 High Strength-to-Weight Ratio: Ensures the plate can withstand the
weight of the ball and external forces without deforming.
 Corrosion Resistance: Suitable for long-term use and reduces wear.
 Ease of Machining: Can be easily cut and shaped to precise dimensions
using CNC or manual tools.
Analysis of the Plate:

1.1 Strength Analysis (Bending Stress):

 The plate will experience bending stress due to the load applied by the
ball.

Mc
Formula: σ=
I

Where;

 M: Is the Maximum bending moment (force from the ball ×

distance from the pivot)
 c: Is the thickness of the plate
 I: Is the Moment of inertia of the plate’s cross-section.

Dimensions of the plate:

2
s p=1×1 m

m p=0 .1 kg

Material properties = Young's Modulus (E)= 70GPa=70∗109 pa

Yield Stress (σ y ¿ = 95MPa

Allowable Stress( σ all¿ ¿ ¿ =60MPa

3
I =W t /12

I =1∗¿

−9 4
I =8 . 333∗10 m
Bending Moment (Mmax): Mmax =W b *L/4

Mmax= 1.0791*1/4 = 0.2698Nm

Bending stress(𝛔): 𝛔=Mmax*C/I

Where C=t/2=0.005m

𝛔=0.2698*0.005/8.333*10−9=161.88MPa

Comparison with Aluminium Allowable Stress (( σ all¿ ¿ ¿ =60MPa

𝛔=161.88MPa¿ σ all¿ ¿=60MPa

The plate fails under this load. To make it safe, we need to increase the thickness of the plate.

Updated Plate Thickness: To ensure the stress does not exceed 60 MP, we adjust the
thickness t. the bending stress is:

𝛔=Mmax*c/I = (Mmax*t/2)/Wt 3/12 = 6Mmax/Wt 3

Rearranging for t= t=
√ 6 Mmax
Wσall
=
√
6∗0 . 2698
1∗60∗10 6 = √ 2 .698∗10
−8
= 0.000164m

The required thickness is t=0.0164m = 16.4mm

1.2 Deflection Analysis:

 To ensure the plate does not deform excessively

Formula:
3
WL
δ=
48 EI

Where; W= Load from the load

L= length of the plate
 E=young’s modules of aluminium
I=moment of inertia
Substitute values:
W= 1.0791 N
L= 1.0m
E= 70 GPa = 70∗109 Pa
3
Wt
I= = 8.333*10−9 m4
12

𝞉max=
3
1.07091∗1 1.0971
9 −9 = 2 = 0.385 mm
48.70∗10 ∗8.333∗10 28.0∗10

The deflection is acceptable.

2. Supporting Frame

Material Selected: Aluminium or Steel

 Why Aluminium/Steel?
 High Durability: Steel provides excellent structural rigidity; aluminium
offers lightweight alternatives.
 Stability: The frame must resist vibrations and external forces.
 Cost-Effectiveness: Both materials are widely available and affordable.

Analysis of the Frame:

2.1 Structural Stability:

 Analyze the frame for static and dynamic loads. We Use Finite Element
Analysis (FEA) to simulate how the frame responds to the weight of the plate
and ball.

Stress in Frame Legs

The total load per leg remains Wleg= 1.7413 N. For aluminium legs with a cross-sectional area:

A= 𝜋r2leg = 𝜋 (0.02)2= 1.256*10-3m2

Stress in each leg:

𝛔leg=
Wleg 1.7413
= −3 = 1.386 KPa
A 1.256∗10

This is much less than 𝛔all= 60 MPa. The frame is safe.

The frame supports the following components:

Weight of the ball: Wb=Mb⋅g=0.11⋅9.81=1.0791 N

Weight of the plate: Wp=mp⋅g=0.1⋅9.81=0.981 N

Weight of the servo motor: (Assume mservo=0.5 kg)

Wservo= mservo* g= 0.5*9.81= 4.905N

Then total static load on the frame is.

Wtotal= Wb+Wp+Wservo= 1.0791+0.981+4.905= 6.9651N

The load is evenly distributed over the frame’s legs.

3. Servo Motor Arms (Linkages)

Analysis of the Arms:

3.1 Shear Stress:

 The arms experience shear stress from the torque applied by the servo motors.

F
τ=
A
 Where; τ = is shear stress from the torque applied
F= is the Force applied by the weight of the plate and ball
A= is Cross-sectional area of the arm.
Torque Analysis
The torque required to tilt the plate is unchanged:
T= F*r = 1.0791*0.5 = 0.5396 Nm
The servo motor must provide at least 0.5396 Nm of torque.
Vibration Analysis
The natural frequency of the servo motor arm is:

fn=
√
1 3 EI
2 π m L3
For aluminium:
 E=70 GPa =70*109 Pa
 I= π ¿ ¿ = 4.91*10-9m4
 M= 0.05 kg
 L= 0.1 m

√
9 −9
1 3∗70∗10 ∗4.91∗10 1
fn= =
2π
√ 6
2.058∗10 = 229.1 Hz
2π 0.05∗0.1
3

The servo motor’s operating frequency must be 229.1 Hz.

4. Pivot Joint (for the Plate)

Material Selected: Hardened Steel or Stainless Steel

 Why Hardened Steel?

 High Durability: Withstands rotational and shear forces over long periods.
 Low Friction: Allows smooth tilting motion.
 Corrosion Resistance: Stainless steel ensures long-term reliability.

Analysis of the Pivot Joint:

4.1 Rotational Stress

 T
τ=
J

Where; T= Is Torque applied by motors.

J= Is Polar moment of inertia.

4.2 Shear Stress

The shear stress due to the force of the plate and ball is given by;

F pivot
τ=
A pin

2 2
π d pin π (0.01)
Apin= = = 7.854*10-5m2
4 4

Substituting Fpivot= 2.0601 N:

𝜏=
2.0601
−5 = 26.23 kPa
7.854∗10

This is much less than aluminum’s allowable shear stress (𝜏all= 60MPa). The pivot joint
is safe.

4.3 bearing selection for the pivot joint

 given parameters, step one determines the loads acting on the pivot joint.
mass of the plate(m p=¿¿ 0.1kg)
gravitational acceleration(g=9.81)
then: weight of the plate (w p=¿m ¿*g) =0.981N
p

weight of the ball(w b=¿m b∗¿¿ ¿ g)=0.11*9.81=1.0791

weight of servo motor(mservo *g)=0.5*9.81=4.905
total load acting on the pivot(w p+w b+w servo )=6.9651
 step two=calculating the radial and axial loads on the bearing
torque applied(t=0.5396)
 shaft radius(r=5mm/0.005m)
then Fr=T/rshaft, 0.5396/0.005=107.92N
axial force( F a=Wtotal=6.9651)
 step three=select suitable bearing type
since the pivot joint experiences both axial and radial loads a deep groove
bearing is selected
 Calculation of expected life:
60 n Lh
L= 6
10
Where n - speed (rpm)
Lh- life in hours
So, for our system, we want it to rotate with 100 rpm and to work for 12 hours
for 3 years based on this specification:
n= 100 rpm
Lh= 12 hr/day * 7 days/week * 52 week/year * 3 year
Lh = 13,104 hr
60∗100∗13,104
L= 6
=78.624 million revolution
10
 Selection of bearing: from our shaft diameter we will select bearings from the SKF
datasheet; Co = 7100 C =10400
 Calculation of equivalent load

P = (XFr + YFa)S

Where Fr is the radial force acting on the bearing due to the rotation of a shaft

Fa is the axial force acting on the bearing

S is a service factor

Load ratios, Fa/Co (6.9651/7100)=0.00098

Fa/Fr (6.9651/107.92)=0.0645
Then from SKF table e=0.22 since Fa/Fr<e, the load is mostly radial so lets use X=1
and Y=0 then P=129.50N

 Verifying bearing life calculated as.

L10=¿ ¿ *(10¿¿ 6) ¿/60*n

Where L - life in Million revolutions

L10 – 518,000 million revolutions
N=100rpm
P=129.50
C=10400
The bearing must last 131.04 million revolutions since the calculated life of 518 million
is far greater than the expected life, so the design is safe and reliable.
5 Ball (Rolling Object)

Material Selected: Hardened Steel or Acrylic

 Why Hardened Steel or Acrylic?

A. Hardened Steel:
 Excellent wear resistance for consistent rolling.
 Smooth surface for minimal friction.
B. Acrylic
 A lightweight alternative for low-cost systems.
 Suitable for educational setups or low-precision applications

Analysis of the Ball:

5.3 Rolling Friction

 Ensure the ball rolls smoothly by considering the coefficient of rolling
friction.
F r = μr * N
 Fr=0 . 0015∗1. 0791=0 .00162

Where = μr Is the Coefficient of rolling friction. (By taking an average of hardened

steel on aluminium)

N Is Normal force (weight of the ball).

Summary of Analysis Techniques

Component Material Analysis

Plate Aluminium Bending,

deflection
Base Platform Aluminium/Steel Static load

Arms Aluminium Shear, bending

Pivot Joint Hardened Steel Rotational and

shear stress
Ball Hardened Steel/Acrylic Rolling
friction

3.2 Mathematical Modeling

In this section we are going to derive the motion equations of system. In these part we assume
following simplifications:

• There is no slipping for ball.

 • The ball is completely symmetric and homogeneous.

• All frictions are neglected.

• The ball and plate are in contact all the time.

Here we derive dynamical equations of ball-on-plate system by the help of Lagrangian.

Side view of ball on plate system.

The Euler-Lagrange equation of ball-plate system is as followings:

Where qi stands for i-direction coordinate, T is kinetic energy of the system, V is potential
energy of system and Q is composite force. The system has 4 degree of freedom; two in ball
motion and two in inclination of plate. Here we assume the generalized coordinates of system to
be xb and yb ball’s position on plate and α and β the inclination of the plate. It is important to
note that we assume the center of x − y coordinates be at center of plate. The kinetic energy of
ball consists of its both rotational with respect to its center of mass and translational energy

Where mb is mass of the ball and Ib is moment of inertia of the ball; ˙x b and Yb are ball’s
translation velocities along x-axis and y-axis; ωx and ωy are ball’s rotational velocities along x-
axis and y-axis.
The following relations between translational velocities and rotational velocities:

In which rb denotes ball’s radius. By substituting equations 3 into equations 4 we will have:

The kinetic energy of the plate, by considering ball as a point mass which is placed in (x b, yb),
consists of the its rotational energy with respect to its center of mass:

Where α and β are plate’s angle of inclination along x-axis and y-axis, respectively. Therefore ά
and β֠ are plate’s rotational velocity. Here we can calculate the kinetic energy of system as
followings:

Where α and β are plate’s angle of inclination along x-axis and y-axis, respectively. Therefore ά
and β֠ are plate’s rotational velocity. Here we can calculate the kinetic energy of system as
followings

The potential energy of the ball relative (the relative potential) to horizontal plane in the center of
the inclined plate can be calculated as:

Here we can derive the system’s equation by Lagrangian and equations 7:

We use L to derive system’s equations:

We assume generalized toques as τ x and τy which are exerted torques on the plate. From
Lagrange-Euler equation we can write:

So the non-linear differential equations for the ball-plate-system as followings:

The equations (18)-(17) shows the relation between ball’s state and plate’s state that is plate’s
inclination. The equations (19)-(20) shows the effect of external torque on the ball-on-plate
system. It is obvious that working with such complex equations is too hard. So in order to do
Laplace analysis on system’s transfer function, we derive the linearised model of system in
neighbourhood of working state.

Plant value

In this project we use the values that are shown in following table

Parameter Description Value

mb Mass of the ball 0.11kg

rb Radius of the ball 0.02m

Sp Dimension of the plate 1.0 x 1.0m2

Vmax Maximum Velocity of the ball 4mm/s

mp Mass of the plate 0.1kg

Ip Mass moment of inertia of the plate 0.5kg.m2

Ib Mass moment of inertia of the ball 1.76 x 10-5kg.m2

g Gravitational acceleration 9.81

Model Linearisation

Linearised Model of a Ball-on-Plate

The approximate value for a solid ball’s moment of inertia is I b = 2/5mbrb2 Therefore equations
(17) and (18) can be written as:

We can linearise above equation by assuming

 Small angle of inclinations for the plate(up to 5o )

 Slow rate of change for the plate

By linearising the above equations, we find two separate differential equations

for each of x and y axis.

State-space Model of System

One can write state equations of ball-on-plate system by considering Equations (17)-(20)[3]. As
mentioned τx and τy are toques exerted on the plane

Because of the low velocity and acceleration of the plate rotation(|α˙ | << 1 and to sub-systems
and control them independently. (|β˙| < < 1) we can omit the coupling term in f(x, u) and divide
the system into
Motor Model

Here, we consider Armature-Controlled DC motor controlled by an armature current, The motor

uses the armature voltage (Ia) as the control input. The stator field is established by a field coil
current, which remains constant throughout the operation. A detailed description of the
governing equations can be found. Under the assumption of a constant field coil current, the
system dynamics can be described by the following postulates.

Transfer function

Complete Linearisation of Ball-on-Plate and DC Motor.

Transfer function of decoupled equations are

If we plot the root-locus plot of the above transfer function, we will get the

Fig pole zero map of system transfer function

Chapter Four

4. Sensor, Actuators and Signal conditioning element

4.1 selection a necessary sensor for a particular application

Resistive Touch Screen Sensor for Ball Position Feedback

The position of the ball on the plate is detected using an analog resistive touch screen sensor,
which enables direct interaction with the surface. This sensor operates by varying its resistance
through a straightforward design, typically composed of two layers made of glass or acrylic.
Both layers are coated with indium tin oxide (ITO). In its default state, a small gap exists
between these layers.
 Fig

Fig

Working principle

When pressure is applied, the layers make contact, allowing the sensor to determine the
coordinates at the point of contact. Four-wire resistive touch screens utilize a single pair of
electrodes on each Indium Tin Oxide( ITO) layer. The electrodes, known as busbars, are
arranged perpendicular to one another, with connections established through a four-wire flexible
cable. These wires are designated as X+ (left), X- (right), Y+ (top), and Y- (bottom). However,
simultaneous reading of both X and Y axes is not possible. Instead, the system alternates by
applying a voltage across one layer and measuring the resulting voltage on the other. These
voltages are processed through an analog-to-digital (A/D) converter, which interprets them as
coordinate values.

4.2 Selection of Actuators

Servo motor

A servomotor is a type of motor equipped with a gearbox and a shaft transmission, which
provides higher torque and greater precision in motion. These motors are capable of rotating 180
degrees, and certain models can rotate up to 360 degrees. In this prototype, we utilized a 180-
degree servomotor.

Regarding its control mechanism, a servomotor contains an internal control circuit, often
managed by microcontrollers. The motor responds to pulses delivered at specific time intervals,
causing it to rotate to an angle that corresponds to the duration of these pulses. A widely used
technique for generating analog signals from digital signals is called Pulse Width Modulation
(PWM).

Servomotors can be classified as either analog or digital, with the type influencing the
characteristics of the control signal required. Analog servos typically use signals with a 20-
millisecond (ms) period, which corresponds to a frequency of 50 Hz. As shown in Figure 8, the
pulse width, or duty cycle duration, ranges between 600 microseconds (μs) and 2400
microseconds (μs).

The pulse width determines the position of the servo motor. For example:

 600 μs corresponds to a -90° position.

 1500 μs represents the neutral (0°) position.
 2400 μs sets the motor to a +90° position.

Since the relationship between pulse width and position is linear, pulse durations for intermediate
positions can be calculated easily. However, it is important to note that these values may differ
for other servomotor models, as the mapping between pulse width and angular position can vary
depending on the manufacturer and design.
 Servo motor

Chapter Five

5. Design the control system

In control system design, selecting the appropriate controller is critical to achieving desired
performance, stability, and precision. Various control strategies, such as PID, Fuzzy Logic, and
Linear Quadratic Regulator (LQR), are commonly used, each offering unique advantages and
challenges. The choice of controller depends on factors like system dynamics, implementation
complexity, cost, and real-time performance requirements. For a ball-on-plate balancing system,
which involves stabilizing a dynamic, nonlinear system, comparing these controllers can help
identify the most effective approach for achieving balance with minimal error and smooth
operation. The table below highlights this comparison.

PID LQR Fuzzy

Ease of Design Simple Advanced math required Complex rules

Implementation Easy Complex Moderate

Performance High with tuning Excellent but costly Good but slower

Cost Low computational load High Moderate

Precision High High with accurate model Moderate

Suitability Best for this project Too complex for scale Unnecessary
Based on the comparison above, we selected the PID controller for its simplicity, cost-
effectiveness, and high performance. We will now proceed to design the PID controller to ensure
precise and stable control of the ball-on-plate system.

PID Controller

The PID (Proportional-Integral-Derivative) controller automatically applies corrections based on

a feedback loop. It is widely used for systems requiring continuous control to stabilize and
achieve the desired response. The controller evaluates the error e(t)e(t)e(t), which is the deviation
from the setpoint, and uses proportional (P), integral (I), and derivative (D) terms to calculate the
control signal. Any combination of these terms, such as P, PI, or PD, can be implemented. The
control signal is given by:

Proportional Control

The proportional controller reduces steady-state error by adjusting the proportional gain K p. The
proportional term is:

Increasing Kp reduces error but excessive gain may cause instability, while too little gain may
hinder the system's ability to handle disturbances.

Integral controller

The integral controller eliminates steady-state error by summing past errors over time. The
integral term is:
However, excessive KI can cause oscillations and worsen system response, leading to instability.

Derivative controller

The derivative controller predicts error changes by calculating the slope of the error over time,
reducing overshoot and improving stability. The derivative term is:

While KD improves transient response, excessive derivative gain can slow the system
unnecessarily.

Over all system design

The ball's position on the plate is measured using a touch panel and fed back to the
microcontroller. Based on the desired setpoint and the measured position, the microcontroller
calculates the error and determines the necessary corrections using the PID control algorithm. It
generates two control signals, corresponding to the angles for Servo 1 and Servo 2, to adjust the
plate's inclination along the X and Y axes. These servos act to tilt the plate accordingly, thereby
influencing the ball's position. The system operates in a closed-loop configuration, continuously
updating the ball's position to maintain stability and ensure it reaches the setpoint.
Controloblity

Stability

Obsevablity

Chapter Six

6. System Simulation

6.1 Simulation of the closed loop control system

The Simulink diagram illustrates a PID controller. The error signal, representing the difference
between the desired and actual output, is processed by three main components: a proportional
term that provides immediate response to the error, an integral term that eliminates steady-state
errors by accumulating the error over time, and a derivative term that anticipates future error to
dampen oscillations. These individual contributions are summed to generate the final control
signal, which is then typically limited by saturation and min/max blocks to prevent actuator
saturation or undesired behavior. This control signal is subsequently sent to the plant (not
depicted in the diagram) to regulate the system. The effectiveness of the PID controller heavily
relies on the tuning of the proportional, integral, and derivative gains to achieve the desired
system response, such as minimizing overshoot, settling time, and steady-state error.

The image illustrates a Simulink model designed to simulate the dynamics of a ball balancing on
a tilting plate. The model incorporates various components, including a World Frame defining
the inertial reference, a Solver to numerically integrate the equations of motion, and a
Mechanism Configuration block to set initial conditions for the ball and plate. The Ball and Plate
blocks represent their respective dynamics, while a Universal Joint connects the plate to the
world frame, allowing for rotation in two axes. The Ball-Plate Interaction block accounts for the
contact forces between the ball and the plate. Control signals, represented by Fx_command and
Fy_command, are applied to tilt the plate in the x and y directions. Finally, sensors measure the
ball's position (x, y) providing feedback for control purposes.

6.2 Results and Discussion

 The plot depicts the response of a system to a step input. The "x" signal exhibits a
characteristic step response with an initial overshoot followed by oscillations that gradually settle
down. This indicates that the system initially overreacts to the sudden input change but
eventually stabilizes. The "x" signal appears to be a smoothed version of the measured
acceleration, possibly representing a filtered or estimated value. Notably, the "x Failure
Detected" signal remains inactive throughout the observation period, suggesting that the system
did not encounter any failures during the step input response.

Conclusion

The step input response demonstrates the dynamic behavior of the system with an initial overshoot and
subsequent oscillations in the "x", “Y” signal, indicating inertia and a requirement for settling time. The
"x", “Y” signal likely represents a filtered or estimated version of the measured acceleration, providing a
smoother representation of the system's behavior. Notably, the absence of any "x Failure Detected"
signals during the step input suggests that the system functioned without encountering any significant
issues. Further analysis, including parameter tuning, robustness analysis, and the quantification of
performance metrics, is recommended to optimize the system's performance and ensure reliable
operation under various conditions.

Chapter seven
7. Conclusion and Recommendation

7.1 Conclusion

The ball-on-plate system is the integration of dynamic modeling, control design, and simulation
to address a challenging control engineering problem. By developing a mathematical model and
implementing a PID controller, the project successfully stabilized the ball on the plate, ensuring
precise trajectory tracking and robustness against disturbances. Simulations conducted in
MATLAB/Simulink demonstrated the system's performance and provided insights into its
dynamics, stability, and response under various conditions. This work highlights the importance
of combining theoretical and practical approaches in control engineering, offering a foundation
for future advancements in dynamic and nonlinear systems.

7.2 Recommendation

1. To enhance the system's functionality and broaden its applications, future work should
include incorporating a Kalman filter for improved state estimation and noise reduction.

2. Increasing the degrees of freedom to three-dimensional(3d) control would allow the system
to handle more complex tasks in advanced robotics and automation.

3. exploring and implementing alternative control algorithms, such as Fuzzy Logic Controllers
and Linear Quadratic Regulators (LQR), could optimize performance by addressing system
nonlinearities more effectively.
4. real-world prototype testing and robustness analysis under diverse conditions are
recommended to validate the system's reliability and expand its practical utility.

References
Appendices (optional)