VIETNAM NATIONAL UNIVERSITY HO CHI MINH CITY
HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY
REPORT OF SUBJECT: INDUSTRIAL MOTION CONTROL SYSTEM
TOPIC: SLIDING MODE CONTROL
Class: CC01 - Semester: 241
Instructor: Phd. Nguyễn Duy Ánh
Name Student ID
Nguyễn Ngọc Khoa 2053139
Vũ Hoàng Tùng 2152338
Phạm Quang Trường 2153077
Nguyễn Vũ Thành An 2152376
HO CHI MINH CITY, 2024
1
TABLE OF CONTENT
CHAPTER 1: THEORY BASIS ...................................................................... 3
1.1 Introduction to the selected control Method .............................................. 3
1.2 Detailed explanation of its working principle ............................................ 4
1.3 Pratical application of Sliding Mode Control ............................................ 8
1.4 Advantages and Limitations of Sliding Mode Control ............................ 11
CHAPTER 2: CASE STUDIES ANALYSIS ................................................ 14
2.1 Sliding Mode Controller for DC Motor ................................................... 14
2.1.1 DC Motor Diagram ............................................................................ 14
2.1.2 Simulation set up in Matlab and Result ............................................. 15
2.2 Sliding Mode Controller for Mass – Spring – Damper system ............... 16
2.2.1 Mathematical Model .......................................................................... 16
2.2.2 Simulate Controller without Disturbances ........................................ 17
2.2.3 Simulate System with Disturbances .................................................. 19
2.3 Adaptive Sliding Mode Control for Uncertain Wheeled Mobile Robot .. 22
2.3.1 Mathematical Model .......................................................................... 22
2.3.2 Sliding Mode tracking controller design for robot ............................ 23
2.3.3 Simulation result ................................................................................ 25
2.4 Sliding mode robot manipulators control ................................................ 28
2.4.1 Mathematical Model .......................................................................... 28
2.4.2 Simulation setup & result .................................................................. 29
CONCLUSION ................................................................................................ 32
REFERENCE .................................................................................................. 33
2
CHAPTER 1: THEORY BASIS
1.1 Introduction to the selected control Method
1.1.1 Sliding Mode Control
Figure 1.1: Phases of Sliding Mode Control: Reaching Mode and Sliding Mode
Control systems play a critical role in ensuring the efficient and safe operation of many
devices, machines, and processes. One of the most robust and effective control
techniques that has gained popularity in recent years is sliding mode control (SMC).
Sliding mode control (SMC) is a feedback control technique that uses a discontinuous
control law to drive the system state to a sliding surface, where the dynamics of the
system are simplified. The control law is designed to maintain the system state on the
sliding surface, which is typically defined as a hyperplane in the state space.
The sliding surface is defined by a switching function, which is a continuous function
that changes sign when the system state crosses the surface. The control law is
discontinuous because it switches between two modes of operation depending on the
sign of the switching function.
3
1.1.2 Modes of operation
The two modes of operation are the sliding mode and the reaching mode.
- In sliding mode, the control law is designed to keep the system state on the
sliding surface. This is achieved by applying a high control effort that drives the
system towards the sliding surface.
- In the reaching mode, the control law is designed to bring the system state to the
sliding surface in a finite time. This is achieved by applying a lower control effort
that drives the system towards the sliding surface but does not keep it on the
surface.
The goal of SMC is to design the switching function and the control law such that the
system state remains on the sliding surface for all time. This ensures that the system is
robust to uncertainties, disturbances, and nonlinearities that may affect the system
dynamics.
1.2 Detailed explanation of its working principle
1.2.1 Closed loop schematic diagram of Sliding mode control
Figure 1.2: Closed loop schematic diagram of Sliding mode control
• Reference r: Firstly, the reference r is the desired ouput value that the system
aim to achieve is fed into the summation block. This bock will compare the
different between reference r and the actual output y of the system. The error e
is:
𝑒 =𝑟 −𝑦
• Sliding surface: Next, The error e is fed into the sliding surface block. The
sliding surface defines the condition that the system states need to satisfy for
4
ideal behavior. The sliding surface is usually defined as a function of the system
states (error and the error’s derivative).
o It outputs a sliding variable sss, which measures how far the system is
from the desired sliding surface. The goal of the controller is to drive the
sliding variable sss to zero.
• Switching Function: After that, the sliding variable s is then processed through
the switching function, which generates the discontinuous control signal 𝑈𝑑 . The
switching function applies high-frequency switching to drive the system toward
the sliding surface
o The control signal 𝑈𝑑 is designed to force the system states onto the
sliding surface and maintain them there.
• Continuous Control: To mitigate issues such as chattering, a continuous control
signal 𝑈𝑒 is generated alongside the discontinuous control. This control signal
helps smooth the control action and may apply a small continuous correction to
prevent the control input from oscillating too rapidly.
o The final control input U is a combination of both the discontinuous
control 𝑈𝑑 and the continuous control 𝑈𝑒 :
𝑈 = 𝑈𝑑 + 𝑈𝑒
• Feedback Loop: The output y is fed back to the summation block where it is
compared to the reference r. This feedback is essential for the controller to
continuously adjust the control input and maintain accurate tracking of the
reference
1.2.2 Implementation Steps for Sliding Mode Control
To provide a clearer understanding of the working principles of Sliding Mode Control
(SMC), we will now explore the implementation steps for applying this control method
to any system. This detailed breakdown will guide you through each phase, from system
modeling to the final control law execution, illustrating how SMC ensures robust
performance and stability.
5
a) System Modelling
The first step is to establish the dynamic model of the system to be controlled. This
model includes the system's states, inputs, outputs, and any disturbances that may affect
its behavior.
The model is typically represented in state-space form:
• Where A, B and D are system matrices.
b) Define the Error
Define the tracking error, which is the difference between the desired system behavior
and the actual system state. The tracking error e(t) is defined as:
• This error serves as the basis for designing the sliding surface, with the goal of
minimizing the error over time.
c) Design the Sliding Surface
In the this step, a custom-made surface should be designed. While on the sliding surface,
the plant's dynamics is restricted to the equations of the surface and is robust to match
plant uncertainties and external disturbances.
• The sliding surface ensures that, when s(t) =0, the system error e(t) tends to zero,
meaning that the system is tracking the desired reference perfectly.
• S(t): The sliding surfaces, which provides a function of time that the system
states forced to follow.
𝑑
• Differential Operator ( + 𝜆)𝑛 : The parameter λ and exponent n allow
𝑑𝑡
flexibility in how aggressively the controller responds to the error.
𝑡
• Integral ∫0 𝑒(𝑡)𝑑𝑡: This integral represents the cumulative error over time from
t = 0 to t. In other words, it captures how much error has accumulated over time.
6
a) Control Law Design
In this, a feedback control law should be designed to provide convergence of a system's
trajectory to the sliding surface. The control law that drives the controlled variable to
its reference value and satisfies equation:
𝑈(𝑡) = 𝑈𝑐 (𝑡) + 𝑈𝑑 (𝑡)
• 𝑈𝑐 (𝑡 ): The continuous control part, responsible for controlling the system in
the sliding phase.
• 𝑈𝑑 (𝑡): The discontinuous control part, used during the reaching phase to bring
the system to the sliding surface.
The Reaching Law (Discontinuous control part):
• This part is responsible for driving the system towards the sliding surface, and it
includes terms like sin(𝜎) to create the switching behavior.
• 𝑏 −1 : Represents the inverse of a system parameter or control gain.
• 𝑎1 , 𝑎2 : Coefficients or control gains that weight the contributions of the state
variables 𝑥1 , 𝑥2 .
• 𝑥1 , 𝑥2 : State variables of the system.
The Sliding Law (Continuous control part):
• Once the system reaches the sliding surface s(t) = 0, the control switches
to the continuous part. This law ensures that the system "slides" along the
surface toward the desired state.
• 𝑥 (𝑡 ): The current state of the system (e.g., position, velocity).
• 𝑟(𝑡): The reference signal or desired trajectory.
b) Switching Control Mechanism
7
Implement a switching mechanism to alternate between discontinuous control
during the reaching phase and continuous control during the sliding phase.
The control law is designed to switch based on the value of the sliding surface
s(t):
• If s(t) ≠ 0, the system is in the reaching phase, and the control law
aggressively pushes the system toward the surface using discontinuous
control.
• Once s(t) = 0, the system enters the sliding phase, and the control law
smoothly drives the system along the sliding surface.
c) Chattering Mitigation
This section highlights an important aspect of Sliding Mode Control (SMC),
specifically regarding how aggressiveness in reaching the sliding surface affects
the system's behavior and potential for chattering.
• This formua expresses the relationship between 𝑈𝐷 (𝑡) and the sliding
surface 𝑠(𝑡). Specifically, it represents how 𝑈𝐷 (𝑡) is influenced by the
current state of the sliding surface 𝑠(𝑡) and a control gain 𝐾𝐷 .
• 𝐾𝐷 : The control gain, which controls the strength or aggressiveness of the
discontinuous control.
• 𝑠(𝑡): The sliding surface or sliding variable, which represents the
difference between the current state and the sliding surface.
• 𝛿: A small constant used to smooth the control signal and reduce
chattering
1.3 Pratical application of Sliding Mode Control
Sliding Mode Control (SMC) has been widely adopted across various fields due to its
robustness against system uncertainties and external disturbances. The practical
application of SMC can be found in many domains, including mechanical systems,
8
power electronics, robotics, and aerospace. Below are some key areas where SMC has
been effectively implemented:
a) Mechanical Systems (Mass-Spring-Damper Systems)
In mechanical systems, such as the mass-spring-damper system, SMC is applied to
control the position or velocity of a mass subjected to spring forces and damping effects.
The robustness of SMC is especially useful in this domain, where variations in mass or
unknown damping can affect system performance.
• In real life, the very famouse car - Mercedes-Benz S Class uses Sliding Mode
Control to mange the suspension system. The controller adjusts the suspension
to maintain ride stability and comfort even when road conditions change,
ensuring robust performance under various driving conditions.
Figure 1.3: The 2021 Mercedes Benz S Class Suspension System
b) Robotics
In robotics, SMC is applied to control the motion of robotic arms and autonomous
vehicles. It provides precise trajectory tracking, ensuring that robotic systems follow
desired paths even in environments with uncertainties, such as varying payloads, joint
friction, or external disturbances.
• KUKA Industrial Robots: The KUKA KR series of industrial robots use SMC to
control their arm movements in environments with external disturbances and
payload variations.
9
• Beside that, a Japanese Robotic company, Fanuc also use SMC dor precise joint
control in their indutrial robotic arms.
c) Aerospace Systems
In aerospace systems, such as flight control systems for aircraft and spacecraft, SMC
is utilized to ensure robust performance under varying atmospheric conditions and
mechanical uncertainties.
• Boeing 787 Dreamliner: The Boeing 787 Dreamliner implements advanced
flight control systems, which include SMC to manage the aircraft's trajectory and
stability under different flight conditions, such as turbulence, wind gusts, or rapid
altitude changes.
• NASA Mars Rover: NASA’s Mars Rover uses SMC to control its attitude and
orientation as it navigates the Martian surface. The SMC-based controller helps
the rover maintain stability while executing various maneuvers on uneven and
unpredictable terrain.
Figure 1.4: The NASA Mars Rover Robot
In addition to the three real-life examples, Sliding Mode Control (SMC) is widely used
across various industries, including automotive control systems, renewable energy
systems, etc. Its popularity is due to its robustness against uncertainties and
10
disturbances, making it ideal for complex, real-world applications where other control
methods may struggle.
1.4 Advantages and Limitations of Sliding Mode Control
1.4.1 Advantages
a) Robustness to System Uncertainties and External Disturbances
One of the most significant strengths of SMC is its inherent robustness. SMC maintains
system stability and performance despite uncertainties in system parameters and
external disturbances. This robustness stems from the control law's design, which forces
the system states to reach and stay on a predefined sliding surface, effectively rejecting
disturbances and compensating for model inaccuracies.
For example, in robotics, SMC is used to control the precise movement of robotic arms
even in environments with variable loads or disturbances. In a 3-DOF robotic arm, SMC
ensures accurate trajectory tracking despite uncertainties in the model, such as friction
or payload variations
b) Effective Control of Nonlinear Systems
SMC is particularly effective for controlling nonlinear systems. Unlike linear control
methods, which may require linear approximations or simplifications, SMC can handle
inherent nonlinearities directly. This capability allows for precise control in systems
where linear models are insufficient to capture the dynamics accurately.
For example, in wind turbine control, SMC is applied to the pitch control of wind
turbine blades, which involves highly nonlinear dynamics due to fluctuating wind
conditions. SMC effectively handles these nonlinearities, maintaining optimal
performance by adjusting the blade angle to maximize energy output.
• This nonlinear function represents the mechanical power Pm extracted from the
wind turbine.
c) Simplified Control Law for Complex Systems
The control law in SMC is often simpler compared to other robust control techniques,
especially for high-dimensional systems. This simplicity arises from the focus on
11
maintaining the system states on the sliding surface rather than addressing all system
dynamics simultaneously.
In a robotic arm, with uncertainties in friction or load, SMC focuses on maintaining
trajectory by forcing the state 𝑆(𝑥, 𝑥̇ ) simplifying the control by ignoring minor
dynamics while ensuring robustness. Unlike other methods (e.g., LQR or H-infinity),
which require addressing all system dynamics and disturbances explicitly, SMC only
requires maintaining the sliding surface. This reduces computational complexity while
ensuring stability.
There are several more advantages of implementing this method in real-life
applications. However, in this report, I have highlighted the three most commonly
recognized benefits.
1.4.2 Limitations of the method
a) Chaterring Phenomenon
Chattering refers to high-frequency oscillations in the control signal caused by the
discontinuous switching nature of SMC. These oscillations can lead to wear and tear in
mechanical systems, induce noise in electrical systems, and degrade overall system
performance.
For robotic manipulators, Sliding Method Control can cause both benefits and
drawback. The rapid switching of the control signal can cause high-frequency
oscillations, which lead to mechanical wear in the actuators and joints. Moreover, this
can introduce undesirable vibrations in the system, reducing precision and efficiency.
Mitigation techniques, such as boundary layer methods, can reduce this chattering
effect.
b) Complexity in Controller Design and Tuning
While the basic concept of SMC is straightforward, designing and tuning an effective
SMC can be complex, particularly for high-order or highly nonlinear systems. Selecting
appropriate sliding surfaces and ensuring stability through Lyapunov-based methods
require significant expertise and computational effort.
Consider the aerospace industry, where SMC is used for attitude control of satellites.
The satellite's rotational dynamics are highly nonlinear due to external disturbances like
12
gravitational torques. Designing an SMC for such systems involves defining the sliding
surface, such as:
Ensuring system stability requires applying Lyapunov-based methods, which involve
complex mathematical formulations and tuning. For high-order systems like satellites,
the computational effort increases significantly to ensure optimal performance.
c) Sensitivity to Measurement Noise
The switching control law in SMC is highly sensitive to measurement noise. High-
frequency noise can induce rapid switching, exacerbating the chattering problem and
potentially destabilizing the system.
In unmanned aerial vehicles (UAVs), precise altitude and position control are crucial.
Sensors like barometers or GPS units provide real-time data, but these are prone to
noise. The switching control law in SMC look like this:
This responds to even small sensor inaccuracies, causing frequent switching. This
induces chattering, leading to instability in the UAV's flight path, affecting
maneuverability and potentially causing erratic behavior
13
CHAPTER 2: CASE STUDIES ANALYSIS
2.1 Sliding Mode Controller for DC Motor
2.1.1 DC Motor Diagram
DC motors are essential in vehicles, robotics, aircraft, and industry. A common
controller for DC motors is the Proportional Integral Derivative (PID) controller, but it
has inconsistent performance, working well only at one set point. The Fuzzy Logic
Controller (FLC) is another option, but it requires experience, experimental data, and
more memory due to its complexity. The state feedback controller is also used, but it
requires all states to be measurable, limiting its hardware implementation. Most
research on these controllers focuses on simulations.
In the DC motor diagram, variables include the DC power supply (vDC), back EMF
voltage (vEmf), armature resistance (R), inductance (L), friction torque (fk), inertia (J),
angular speed (ω), motor torque (T), and armature current (i). The motor parameters
used in simulation and controller design are in Figure 1.
Figure 2.1 DC Motor Diagram
Parameter Value
R 0.4
L 2.7
J 0.0004
Kb 0.0022
Kt 0.015
14
Ke 0.05
2.1.2 Simulation set up in Matlab and Result
Based on the state-space model, the transfer function model can be obtained as
Figure 2.1 Simulation set up in Matlab
Figure 2.2 Response system
15
Comparing system response with PID, the result for the step reference signal is shown
in Figure 3. The detailed system response is shown in Table II. It can be seen that the
SMC response did not have an overshoot. Meanwhile, the PID response had a 22.48%
overshoot. Thus, the SMC response was better than the PID Controller.
Controller System Response Performance
Rise Time Settling Time Overshoot
SMC 0.6070 1.0899 0
PID 0.1830 1.0014 22.4793
2.2 Sliding Mode Controller for Mass – Spring – Damper system
2.2.1 Mathematical Model
In SMC, you define a sliding surface that the system state trajectory converges to and
remains on. This sliding surface is designed such that it is insensitive to disturbances
and uncertainties in the system. Once the system state trajectory is on the sliding surface,
the controller uses a feedback control law to drive the system state trajectory to the
desired state along the sliding surface.
Consider a mass-spring-damper system with mass, M, attached to a spring with stiffness
coefficient, K, and a damper with damping coefficient, D. The dynamics of the system
are also subject to an external control force, F, exerted on the mass. This force serves
as a mechanism to influence the system dynamic response.
The differential equation for the system is:
𝑀𝑥̈ + 𝐷𝑥̇ + 𝐾𝑥 = 𝐹
16
Define the state variables 𝑥1 = 𝑥 and 𝑥2 = 𝑥̇ , which produces the following set of first-
order differential equations describing the system.
𝑥1̇ = 𝑥2
{
𝑥2̇ = 𝑓(𝑥) + 𝑔(𝑥). 𝑢(𝑡)
𝐷𝑥̇ +𝐾𝑥
Here, 𝑓(𝑥) = − represents the inherent forces of the damper and spring acting
𝑀
1
on the mass, and 𝑔(𝑥) = signifies the influence of the control force 𝐹 on the mass,
𝑀
normalized by its magnitude.
Define the mass-spring-damper system parameters, where the mass 𝑀 is 1 kg, the
damping coefficient 𝐷 is 0.1 𝑁𝑠/𝑚, and the spring constant 𝐾 is 1 N/m.
2.2.2 Simulate Controller without Disturbances
For this example, define the sliding mode function as 𝑠(𝑡) = 𝑐𝑒(𝑡) + 𝑒̇ (𝑡), where 𝑥𝑑 is
the desired position and 𝑒 = 𝑥𝑑 − 𝑥 is the tracking error.
Define the SMC control law as follows.
1
𝑢= (𝜂sign(𝑠) + 𝑘𝑠 + 𝑐𝑒̇ + 𝑥𝑑̈ − 𝑓)
𝑔
Here:
• 𝑢(𝑡) uses the sign of 𝑠(𝑡) to produce the discontinuous control structures
(𝑢𝑘+ (t) and 𝑢𝑘− (t)).
• 𝜂 is the amplitude of the discontinuous control action, which enhances
robustness against disturbances.
• 𝑘 is the proportional gain in the continuous part of the control law, which reduces
steady-state error and improves response time.
• 𝑐 is the same coefficient used in the sliding surface, which affects the rate of
convergence to the sliding surface.
Set the SMC parameters, specifying the amplitude of the control action η, the
proportional gain k, and the sliding surface coefficient c for the controller design. For
the initial controller design, the SMC controller assumes that there is no disturbance in
the system.
𝑐=1
{𝑘 = 1.1
𝜂 = 0.1
17
Figure 2.3 State Response without Disturbances
Figure 2.5 Position Error without Disturbances
18
Figure 2.6 Control Input without Disturbances
2.2.3 Simulate System with Disturbances
In real-world scenarios, systems are often influenced by external disturbances that can
affect their performance. To test the robustness of the SMC strategy, you can introduce
a disturbance into the mass-spring-damper system simulation. The bounded disturbance
is modeled as a time-varying function 𝑑(𝑡) that directly affects the dynamics of the
system, where ∣ 𝑑(𝑡) ∣≤ 𝐷.
For this example, consider a sinusoidal disturbance 𝑑(𝑡) given by:
𝑑(𝑡) = 𝐷𝑠𝑖𝑛(𝑡)
The SMC control law is designed to counteract the effects of disturbances and drive the
system state towards the desired sliding surface. The control law is defined as:
1
𝑢= (𝜂sign(𝑠) + 𝑘𝑠 + 𝑐𝑒̇ + 𝑥𝑑̈ − 𝑓 + 𝐷sign(𝑠))
𝑔
where the term D sign(s) is an additional term that accounts for the upper bound of the
disturbance.
Update the SMC controller parameters.
𝑐=5
𝑘=3
{
𝜂 = 0.1
𝑑 = 0.6
19
Figure 2.4 System Response with Disturbances
Figure 2.5 Position error with Disturbances
20
Figure 2.6 Control input with Disturbances
Reduce Chattering Using Quasi-Sliding Mode Control
A common issue in SMC is chattering, which is the high-frequency switching of the
control input when the system state is close to the sliding surface. To mitigate
chattering, quasi-sliding mode control can be employed. In this approach, the
discontinuous sign function in the control law is replaced with a continuous
approximation, such as a saturation function.
For this example, use the following saturation function.
1 if 𝑠 > 𝜙
sat(𝑠, 𝜙) = {𝑠/𝜙 if − 𝜙 ≤ 𝑠 ≤ 𝜙
−1 if 𝑠 < −𝜙
Here, 𝜙 is the boundary layer thickness. This saturation function smoothly interpolates
between −1 and 1 when the sliding variable s is within the boundary layer [−𝜙, 𝜙],
which reduces the high-frequency switching.
The modified control law with the saturation function is given by:
1
𝑢= (𝜂sat(𝑠, 𝜙) + 𝑘𝑠 + 𝑐𝑒̇ + 𝑥𝑑̈ − 𝑓 + 𝐷sat(𝑠, 𝜙)))
𝑔
21
The choice of 𝜙 is critical as it defines the thickness of the boundary layer around the
sliding surface. A larger 𝜙 results in less chattering but can increase the steady-state
error. Conversely, a smaller 𝜙 can reduce steady-state error but increase chattering.
Figure 5.7 Control input after employed quasi-sliding mode control
2.3 Adaptive Sliding Mode Control for Uncertain Wheeled Mobile Robot
In this example, I will presents a technique for developing a sliding mode controller
(SMC) using the state model of wheel mobile robot (WMR).
2.3.1 Mathematical Model
In the article, the study focuses on the WMR concerning the robot's motion on a
horizontal surface. The body of the system incorporates two rear-wheel drive wheels
for propulsion and a front steering wheel. The movement frames of the WMR are
illustrated in Figure 2.11 below
22
Figure 2.11: Wheel mobile robot model
On the scope of this report, I will not dive into the steps of developing the mathematical
model of this mobile robot. Here is the state model function of the robot, which is built
based on the kinematic and dynamic equations.
Where:
• The disturbances 𝐷1 = 𝜆𝑥1 and 𝐷2 = 𝜆𝑥2
• 𝑥1 represents the position error (the error between the actual position and the
target position of the wheeled mobile robot)
• 𝑥2 = 𝑥1̇ + 𝜆𝑥1
• 𝜆 is the positive constant with unknow exact value.
2.3.2 Sliding Mode tracking controller design for robot
In this section, an adaptive SMC will be introduced for tracking control of the WMR.
The control scheme is designed through two steps: the first one, the chattering free SMC
is built based on the real disturbances. Next, the unknown disturbances will be estimated
by a disturbance observer then the real disturbances will be replaced by estimated ones.
The detail of each step is illustrated in the following subsections
a) Disturbance observer based stable sliding surface design
Before designing the SMC, the sliding surface should be designed first. The sliding
surface is defined as:
23
As the sliding surface is reached, 𝑠 = 𝑠̇ = 0. Under this condition, replace this equation
into state model function of the robot above we have:
However, mismatched disturbances can affect this stability, so a nonlinear disturbance
observer is introduced to estimate and compensate for these disturbances. Estimation
errors are then used to refine the control, allowing the errors to converge and ensuring
the system remains stable
Finally, the sliding surface is redefined to incorporate the disturbance estimates,
providing robust performance even in the presence of external disturbances, ensuring
the system can handle uncertainties effectively.
• This equation represents the redefined sliding surface s in a sliding mode control
(SMC) system that takes into account disturbances
• 𝑘1 , 𝑘2 is the control gain associated with the state 𝑥1 and 𝑥2 respectively.
• 𝜉2 is the estimated disturbance associated with the state 𝑥2 . In sliding mode
control, disturbances need to be compensated to ensure stability.𝜉2 helps in
minimizing the effect of any disturbances acting on the system.
b) Adaptive sliding mode control design
After ensuring that the sliding surface is stable, in this section, the reaching controller
will be designed for closed loop stable purpose. Here is the control law:
• 𝜏 here is the overall control signal used to drive the motors.
• 𝜏1 : A part of the control signal that deals with state feedback and disturbance
compensation.
• 𝜏2 : Responsible for the sliding mode controller's error correction and
adaptive control.
Next, the Lyapunov function is chosen to prove the system stability:
24
The dirivative of Lyapunov function must be proven to negative. Since this step is verry
difficult for us to under stand so we will only include the conlusion here:
𝑉̇ < −𝜇|𝑠| ≤ 0
This implies that the introduced I show above guarantees the reaching to the sliding
surface of the system states as time goes to finite.
2.3.3 Simulation result
In order to verify the effectiveness of the proposed control scheme, the simulations are
setup in MATLAB/Simulink for a WMR with the following parameters:
Parameters of the controller are given by 𝜆 = 10, 𝑘1 = 𝑘2 = 10, 𝛼 = 100, 𝜇 = 10, 𝜀 =
5
The WMR is controlled to move with the circular shape to demonstrate the effectiveness
of the proposed adaptive sliding mode control (ASMC).
Circle shape with the following equations of reference trajectory:
The initial condition of the mobile robto are given by [𝑥𝑀 (0), 𝑦𝑀 (0), 𝜃(0)] =
𝜋
[5; −3.2; ]
2
25
Figure 2.12: Trajectory tracking of the WMR in circular shape
Here, we will compare the simulation result with the DOSM control method
(Department of Statistics Malaysia) under similar conditions.
It can be seen from Figure 2.12 that the responses of both ASMC and DOSM are good
with similar trajectory at steady state time.
26
Figure 2.13: The tracking error in x-axis and y-axis
However, at initial time, the response of the DOSM takes a long time to track the
reference curve and has little oscillation at 3 seconds where the disturbances act on the
system. This result is expressed thoroughly in the figure 2.13 above which presents the
tracking error in x-axis and y-axis. In Figure 2.13, the tracking errors of two controllers
are almost the same in the y-axis with nearly zero steady state error. In the x-axis, the
maximum tracking error of the proposed ASMC is about 0.05 m while this is bigger for
DOSM (about 0.65 m).
27
Figure 2.14: The control torques of generated by the 2 method
The advantages of the proposed ASMC and the DOSM are illustrated clearly in Figure
2.14, the control torques. In 2.14 the control torques of the ASMC have a big pulse at
initial, but they are smooth during the working time. Meanwhile, the control torques
generated by the DOSM scheme are affected by chattering phenomena of the SMC, so
they are highly oscillated in both transient time and steady state. This chattering is not
good for power electronic devices as well as the mechanical systems. Also, at the time
acting on the system, the control torques of the DOSM are too higher than those of the
proposed ASMC.
2.4 Sliding mode robot manipulators control
2.4.1 Mathematical Model
The dynamics of an n-link robotic manipulator can be expressed as
𝑀(𝑞)𝑞̈ + 𝐶 (𝑞, 𝑞̇ )𝑞̇ + 𝐺 (𝑞) = 𝜏 + 𝜏𝑑
where
𝑞, 𝑞̇ , 𝑞̈ ∈ 𝑅𝑛 represent the position, velocity and acceleration of the joints respectively
28
𝑀(𝑞) = 𝑀0 (𝑞) + Δ𝑀 (𝑞) ∈ 𝑅𝑛×𝑛 stands for the inertia matrix
𝐶 (𝑞, 𝑞̇ )𝑞̇ = 𝐶0 (𝑞, 𝑞̇ )𝑞̇ + Δ𝐶 (𝑞, 𝑞̇ )𝑞̇ ∈ 𝑅𝑛×𝑛 is the centripetal Coriolis matrix
𝐺 (𝑞) = 𝐺0 (𝑞) + Δ𝐺(𝑞) ∈ 𝑅𝑛 is the gravitational vector
𝜏 ∈ 𝑅𝑛 is the joint torque vector
𝜏𝑑 ∈ 𝑅𝑛 is the disturbance torque vector
The controller input is defined as
𝜏̇ = 𝑢̇ 0 + 𝑢̇ 1
where
𝑑 𝑑
𝑢̇ 0 = 𝑀0 (𝑞) 𝑞̈ 𝑑 + [𝐶0 (𝑞, 𝑞̇ )𝑞̇ + 𝐺0 (𝑞)] − (𝑞/𝑝𝐾 −1 𝑀0 𝑠̇ 2−𝑝/𝑞 ) − 𝛽𝑀0 𝑒̈
𝑑𝑡 𝑑𝑡
− 𝑀0 𝑀̇0−1 (𝜏 − 𝐶0 (𝑞, 𝑞̇ )𝑞̇ − 𝐺0 (𝑞))
˙
𝑢̇ 1 = −𝐾 † 𝑀0 (𝑞)𝜎 − 𝑀0 (𝑞)(𝐵0 + 𝐵1 ‖𝑞‖ + 𝐵2 ‖𝑞 ‖2 )sgn(𝜎)
2.4.2 Simulation setup & result
The simulations are carried out in the MATLAB – Simulink platform by using ODE 4
solver with a fixed step size of 0.005 s.
The dynamic equation of the manipulator model is given by:
(𝑚1 + 𝑚2 )𝑙12 + 𝑚2 𝑙22 + 2𝑚2 𝑙1 𝑙2 𝑐2 + 𝐽1 𝑚2 𝑙22 + 𝑚2 𝑙1 𝑙2 𝑐2
𝑀(𝑞) = [ ]
𝑚2 𝑙22 + 𝑚2 𝑙1 𝑙2 𝑐2 𝑚2 𝑙22 + 𝐽2
−𝑚2 𝑙1 𝑙2 𝑠2 𝑞̇ 22 − 2𝑚2 𝑙1 𝑙2 𝑠2 𝑞̇ 1 𝑞̇ 2
𝐶(𝑞, 𝑞̇ )𝑞̇ = [ ]
𝑚2 𝑙1 𝑙2 𝑠2 𝑞̇ 22
29
(𝑚1 + 𝑚2 )𝑙1 𝑔𝑐1 + 𝑚2 𝑙2 𝑔cos (𝑞1 + 𝑞2 )
𝐺(𝑞) = [ ]
𝑚2 𝑙2 𝑔cos (𝑞1 + 𝑞2 )
Physical parameters of the two=link manipulator
𝑙1 1𝑚
𝑙2 0.85 𝑚
𝐽1 5 𝑘𝑔𝑚
𝐽2 5𝑘𝑔 𝑚
𝑚1 0.5 𝑘𝑔
𝑚2 1.5 𝑘𝑔
𝑚
̂1 0.4 𝑘𝑔
𝑚
̂2 1.2 𝑘𝑔
g 9.81 𝑚/𝑠 2
The adaptation update laws used are given by:
̂̇ = 1.74‖𝐾‖‖𝑠̇ (𝑝/𝑞)−1 𝜎‖
𝐵 0
̂̇ = 1.45‖𝐾‖‖𝑠̇ (𝑝/𝑞)−1 𝜎‖‖𝑞‖
𝐵 1
̂̇ = 3.30‖𝐾‖‖𝑠̇ (𝑝/𝑞)−1 𝜎‖‖𝑞̇ ‖2
𝐵 2
All initial conditions are 0.
30
It is observed that both the joints 1 and 2 track the reference trajectory faithfully. The
control inputs applied to both the joints show no chattering.
31
CONCLUSION
In this report, we have explored Sliding Mode Control (SMC) through a structured
analysis, beginning with the theoretical foundations and progressing into case studies
and practical applications. The selected control method was thoroughly introduced,
emphasizing its robust performance under various dynamic conditions. A detailed
explanation of the working principles highlighted the key aspects of switching surfaces
and reaching laws, which are integral to the success of SMC in rejecting disturbances
and uncertainties.
The practical applications discussed revealed SMC’s adaptability across various control
systems, such as DC motors, mass-spring-damper systems, and robotic manipulators.
Each case study provided insight into the method's versatility and its response under
both ideal and perturbed conditions. Specifically, the analysis of adaptive sliding mode
control for uncertain wheeled mobile robots showcased SMC’s strength in handling
systems with varying parameters.
The simulation results presented in various case studies provided strong evidence of
SMC’s effectiveness, particularly in systems with uncertainties and external
disturbances. These results not only validate the theoretical framework but also
demonstrate SMC’s real-world applicability.
In conclusion, Sliding Mode Control proves to be a powerful tool for a wide range of
control applications, offering significant advantages in terms of robustness and
disturbance rejection. However, as with any control method, it comes with challenges
that must be addressed for practical implementation.
Finally, I would like to express my deepest gratitude to my advisor, Ph.D. Nguyen Duy
Ánh, for his invaluable guidance, support, and encouragement throughout the
completion of this report.
32
REFERENCE
[1] A. Ma’arif, "Simulation and Arduino Hardware Implementation of DC Motor
Control Using Sliding Mode Controller," Journal of Robotics and Control, vol. 2, no.
6, pp. 578-583, Nov. 2021.
[2] MathWorks, "Sliding Mode Control Design for Mass-Spring-Damper System,"
available online: https://www.mathworks.com/help/slcontrol/ug/sliding-mode-control-
design-for-mass-spring-damper-system.html.
[3] S. Mondal and C. Mahanta, "Adaptive second order terminal sliding mode
controller for robotic manipulators," Journal of the Franklin Institute, vol. 351, no. 4,
pp. 2356–2377, 2014.
[4] J.-M. Yang and J.-H. Kim, "Sliding mode control for trajectory tracking of
nonholonomic wheeled mobile robots," IEEE Transactions on Robotics and
Automation, vol. 15, no. 3, pp. 578-583, 1999.
[5] B. Bandyopadhyay, S. Janardhanan, and S. Spurgeon, Advances in Sliding Mode
Control: Concept, Theory and Implementation, vol. 440, London: Springer, 2013.
[6] X. Zhang, Y. Li, X. Lin, Z. Zhang, and H. Zhang, "Improved adaptive terminal
sliding mode control with extended state observer for uncertain systems," IEEE
Access, vol. 8, pp. 135073-135084, 2020.
[7] M. Nasiri, M. Moradi, and M. M. Saidi, "Sliding mode control for position
tracking of robot manipulators with model uncertainties," IEEE Transactions on
Control Systems Technology, vol. 28, no. 5, pp. 2037-2044, Sept. 2020.
[8] H. Chen, L. Li, Z. Hou, and Z. Lu, "Adaptive sliding mode control for wheeled
mobile robot tracking based on disturbance observer," IEEE Access, vol. 7, pp.
169998-170008, Dec. 2019.
[9] B. Bandyopadhyay, S. Janardhanan, and S. Spurgeon, Advances in Sliding Mode
Control: Concept, Theory and Implementation, Springer, 1st ed., 2013.
33
