Robust control using Disk margin and H

infinity technique for uncertain system

A project report on the partial fulfilment of the requirement for the

degree of
BACHELOR OF TECHNOLOGY
in
ELECTRICAL
ENGINEERING
BY

Shivang Pandey Reg no. 20211047

Priyadarshiny Pandey Reg No. 20212064
Abhishek Kumar Reg No. 20212079

UNDER THE SUPERVISION OF

DR. SATNESH SINGH
ELECTRICAL ENGINEERING DEPARTMENT

DEPARTMENT OF ELECTRICAL ENGINEERING

MOTILAL NEHRU NATIONAL INSTITUTE OF TECHNOLOGY
PRAYAGRAJ- 211004

OCTOBER-2024
 UNDERTAKING

We, Shivang Pandey (20211047) ,Abhishek Kumar(20212079), Priyadarshiny Pandey (20212064)

declare that the project report titled "Robust control using Disk margin and H infinity technique
for uncertain system " submitted to the Department of Electrical Engineering, Motilal National
Institute of Technology, Allahabad(India) is an original work carried by us under supervision of Dr.
Satnesh Singh in Electrical Engineering Department.

Place: MNNIT Allahabad

Date:
Shivang Pandey Reg. No. 20211047
Priyadarshiny Pandey Reg. No.20212064
Abhishek kumar Reg. No.20212079

i
MOTILAL NEHRU NATIONAL INSTITUTE OF TECHNOLOGY ALLAHABAD
DEPARTMENT OF ELECTRICAL ENGINEERING
PRAYAGRAJ-211004, U.P.

CERTIFICATE
We certify that the work which is being presented in this project report entitled “Robust control
using Disk margin and H infinity technique for uncertain system”, is submitted by Shivang
Pandey(20211047), Priyadarshiny Pandey (20212064) and Abhishek Kumar (20212079) in
partial fulfilment for the award of the degree of Bachelor of Technology in Electrical
Engineering at Motilal Nehru National Institute of Technology Allahabad, Prayagraj. This is an
authentic record of our work carried out during a period from July 2021 to October 2021 under the
guidance of Dr. Satnesh Singh, Professor, Electrical Engineering Department, MNNIT
Allahabad.

Shivang Pandey Priyadarshiny Pandey Abhishek Kumar

Reg No. 20211047 Reg No. 20212064 Reg No.20212079

This is to certify that the above statement made by the student is correct to the best of my
knowledge.

Date: 14/10/2024 Dr. Satnesh Singh

Electrical Engineering Department

MNNIT, Allahabad

ii
 ACKNOWLEDGEMENT

It is a great pleasure to express our sincere gratitude and profound regards to our
project work supervisor Dr. Satnesh Singh, Electrical Engineering Department,
MNNIT Allahabad, for his constant encouragement, valuable guidance motivation,
support, advice, and supervision during the entire course of the work. His meticulous
guidance, constructive, and valuable suggestions, timely discussion and clarification
of my doubts increase my cognitive awareness and helped me for making a deeper
analysis of the subject under study. Our association with him throughout the thesis
activity was a great process of learning.
Thank is also due to, Prof. Richa Negi, Head of Electrical Engineering Department,
MNNIT for providing facilities and advice needed to complete the work.
We also express my heartfelt gratitude to the Department of Electrical Engineering
MNNIT Allahabad for giving me this opportunity, which has enriched my knowledge
and experience immensely.
Last but not the least my head bows with reverence before Almighty GOD, who has
given us strength, wisdom and will to complete the work.

Shivang Pandey Reg. No. 20211047

Priyadarshiny Pandey Reg. No.20212064
Abhishek kumar Reg. No.20212079

iii
Dedicated to God
&
My Family

iv
 Abstract
The main objective of this work is to design an H∞ optimal controller for very
essential applications in electrical circuits, i.e., DC-DC buck converter in the
presence of the system uncertainties, parameter variations and disturbance.
Here, H∞ method is applied to DC-DC buck converter circuit for robust control
against parameter variation and disturbance.
H∞ control method is a systematic and allows one to deal with model
uncertainty, directly which is not possible in classical control. Therefore this
work emphasis to give a clear understanding of the working procedure and how
to apply it to practical problems like DC-DC buck converter.
The model with its robust controller is simulated in MATLAB to verify its
performance.

v
 TABLE OF CONTENTS

Undertaking i
Certificate ii
Acknowledgement iii
Dedication iv
Abstract v
List of Contents vi
List of Figures vii
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 LITERATURE SURVEY 5
CHAPTER 3 MOTIVATION & OBJECTIVE 7
3.1 MOTIVATION 7
3.2 OBJECTIVE 8
CHAPTER 4 METHODOLOGY 10
CHAPTER 5 RESULTS 25
CHAPTER 6 WORK PLAN 27
CHAPTER 7 REFERENCES 28

vi
LIST OF FIGURES

1. Experimental Frequency response From many hard disk drives

2. MINI MUTT
3. MATLAB Code for varying finding gain and phase margin for a
given transfer function
4. MATLAB Code for varying f in SISO loops
5. Output Graph for Variation of f
6. Disk Visualization on SIMULINK
7. Variation of Skew in disk
8. Range of Gain and Phase Variation
9. Output Disk for SISO Loop
10.Uncertainty Disk in the Nyquist Plane

Vii
 CHAPTER-1
1.1Introduction:
In classical control, there are very developed techniques for the analysis and control of the
different Linear Time Invariant (LTI) systems. But actual systems are not linear and time
invariant. These conventional classical techniques are not completely effective under the
influence of the model uncertainty and disturbance. And conventional controllers can make
the system stable under certain conditions i.e. need of accurate mathematical model (without
uncertainty). However, in actual system, noise and disturbances and parameter uncertainties
are always present, so application of control law to these kind of system becomes difficult.

1.2. Outline of the dissertation:

In this dissertation, a method is proposed to design a controller in the presence of the system
uncertainties (parameter variations) and disturbance. For designing a controller, it is required
to present how to capture uncertainties and any control system into generalized plant frame.
Before discussing further, we will consider some robust control importance.

1.3. Robust Control and its Importance [1, 2] :

A primarily definition of robust control could be stated as: `Design a controller such that
some level of performance of the controlled system is guaranteed irrespective of changes in
the plant dynamics within a predefined class.'
The main goal of robust control is to take uncertainties into account when analyzing a control
system or when designing a controller for it. In order to do so, one has to arrive at a
mathematical description of the uncertainties. Sometimes it is obvious what to call an
uncertainty (such as parameter variations in physical model), but sometimes one just has to
consider certain structure of the uncertainty. These dynamics were analysed and controllers
were designed such that the closed loop system was at least stable and showed some desired
performance. So, with some margins, stability was effectively made robust against small
perturbations in the process dynamics.
The controller is to be designed such that the following goals and constraints can be realised
in some optimal form stability.
(1) The closed loop system should be stable.
(2) Tracking: The real output should follow the reference signal.
(3) Disturbance rejection: The output should be free of the influences of the disturbing noise.
(4) Sensor noise rejection: The noise introduced by the sensor should not affect the output.
 1
Robustness has been an important issue in control system design. Robustness means system
should be always able to maintain stability and performance level in spite of uncertainties in
system dynamics and in the working environment. And of course robustness has crucial
importance in control system because real engineering systems are always affected by
external disturbance and measurement noise and there are always difference between
mathematical models and the actual model. So our main goal is to stabilize the original plant
in the presence of disturbance, unmodelled plant dynamics and plant parameter variations.
So, from past few years robust control technique becomes popular. And many techniques
came for robust control like Linear Quadratic Gaussian (LQG), Internal Model Control,
Sliding Mode Control, H infinity optimal control theory, µ synthesis/analysis method.

1.4. Review of H infinity Technique:

The investigation of Hoo optimization of control systems began in 1979 with a conference
paper by Zames [3,4], who considered the minimization of the infinity-none of the sensitivity
function of a single-input-single-output linear feedback system. The work dealt with some of
the basic questions of "classical" control theory, and immediately caught a great deal of
attention. It was soon extended to more general problems, in particular when it was
recognized that the approach allows dealing with robustness far more directly than other
optimization methods.
The name " H infinity -optimization"[5], is somewhat unfortunate. H infinity is one member
of the family of spaces introduced by the mathematician Hardy. It is the space of functions on
the complex plane that are analytic and bounded in the right-half plane. The space plays an
important role in the deeper mathematics needed to solve Hoo optimal control problems.
Obviously, it is difficult to meet above specifications using conventional PID controllers. PID
controllers cannot make the system stable under certain conditions [6]. If applying linear
system theory can solve the control problem, then the accurate mathematical model of the
system is required. Another problem with analog controllers is that for a MIMO system with
m inputs and n outputs , m x n individual controllers have to be tuned for each input-output
couple making analysis and design complicated and quite involved. In this, an H infinity
control strategy is proposed to overcome the above-mentioned disadvantages. In H infinity
control system, the exact model is not required, and some uncertainty is allowed

2
1.5. KEY CONCEPTS OF ROBUST CONTROL

1. Uncertainty: In many real-world systems, uncertainties can arise from modelling

errors, external disturbances, or changes in system dynamics. Robust control aims to
handle these uncertainties effectively.
2. Performance Specifications: Robust controllers are designed to meet specific
performance criteria, such as stability, tracking accuracy, and disturbance rejection,
even when the system operates outside nominal conditions.
3. Mathematical Framework: Robust control often uses techniques from linear and
nonlinear control theory. Common approaches include H-infinity (H∞) control, which
minimizes the worst-case effects of disturbances, and mu-synthesis, which deals with
structured uncertainties.
4. Controller Design: The design process usually involves deriving a controller that can
stabilize the system and ensure performance across a range of scenarios. This may
include techniques like gain scheduling, where the controller parameters are adjusted
based on operating conditions.
5. Applications: Robust control is widely used in aerospace, robotics, automotive
systems, and many other fields where system reliability is critical despite
uncertainties.
Overall, robust control seeks to create systems that are resilient and dependable, capable of
handling real-world variability effectively.

1.6. Limitations of Classical Margins:

1) Real systems differ from their mathematical models in both magnitude and phase: The
Bode plot in Figure 2 shows a collection of frequency responses obtained from input– output
experiments on hard disk drives (blue). A low-order model used for control design is also
shown (yellow). The model accurately represents the experimental data up to 2–3 rad/s.
2) Small plant perturbations may cause robustness issues even if the system has large
gain/phase margins: The key point is that some care is required when using classical gain and
phase margins. This did not present itself as an issue when controllers were designed
primarily with graphical techniques. These classical controllers were typically of limited
complexity and did not have enough degrees of freedom to get into this corner.
3) Margin requirements must account for the increase in model uncertainty at higher
frequencies: Consider again the hard disk drive frequency responses shown in Figure 2. The
design model (yellow) loses fidelity at high frequencies. As a result, the margins must
necessarily be larger at higher frequencies to ensure stability
 3
4) There are alternative robustness margins that provide more useful extensions to MIMO
systems: A typical extension of classical margins for MIMO systems is to assess stability with
a gain or phase perturbation introduced into a single channel. This analysis is repeated for
each input and output channel. This “loop-at-a-time” analysis fails to capture the effect of
simultaneous perturbations occurring in multiple channels

Figure 1 Experimental Frequency Responses from hard disk drives

 4
CHAPTER-2
LITERATURE SURVEY

Classical Control Theory and Margins

Ziegler-Nichols (1942): Early work on stability analysis in feedback control systems focused
on methods like Ziegler-Nichols, which helped develop tuning rules for Proportional-
Integral-Derivative (PID) controllers. These methods introduced gain margin (GM) and phase
margin (PM), crucial metrics in ensuring system stability.
Nyquist and Bode (1930s-1940s): The Nyquist stability criterion, developed by Harry
Nyquist, became a cornerstone in control systems design, allowing engineers to assess
stability by observing how the frequency response encircles the critical point (-1). Classical
GM and PM methods emerged as natural extensions.
Robust Control and the Limits of Classical Margins
H∞ Control (1980s): In the 1980s, the development of H∞ control theory by researchers like
Doyle, Glover, and Stein shifted the focus to robustness analysis for systems with
uncertainties. H∞ introduced techniques that focused on worst-case performance and
robustness, which led to a rethinking of traditional gain and phase margins.
Classical gain and phase margins assume independent gain and phase variations, but in real-
world systems, both vary together due to model inaccuracies, external disturbances, or sensor
noise. This limitation motivated the development of more sophisticated robustness measures
like disk margins.
Disk Margins: Development and Theoretical Foundations
Doyle and Stein (1981): In their pioneering work, John Doyle and Gary Stein introduced
methods to analyze robustness in multivariable control systems, acknowledging that gain and
phase often vary simultaneously. This laid the groundwork for more robust analysis
techniques, eventually leading to disk-based stability measures.
Vinnicombe (2001): In "Uncertainty and Feedback: H∞ Loop-shaping and the ν-gap Metric",
Glenn Vinnicombe introduced the concept of the ν-gap metric, a framework for comparing
different control system designs. The ν-gap and disk margin analysis are closely related, as
both assess how much uncertainty a system can tolerate, but the ν-gap metric specifically
accounts for model variation in frequency-domain design.
Practical Applications of Disk Margins
Multivariable Systems (MIMO Control): Disk margins were initially introduced to address
robustness issues in Multivariable Input, Multivariable Output (MIMO) systems, where the
interaction between multiple inputs and outputs can cause both gain and phase to vary
unpredictably. Skogestad and Postlethwaite (2005) discuss disk margin techniques in their
work on multivariable feedback control, emphasizing its utility in managing real-world
uncertainties.
 5
Aerospace and Automotive Systems: Disk margin analysis found early applications in
industries requiring high levels of robustness, such as aerospace and automotive systems.
Systems in these fields often face simultaneous gain and phase uncertainties due to variable
environmental conditions and aging components, making traditional margins insufficient.
NASA and Boeing have been notable users of disk margins in the design of robust flight
controllers for aircraft, ensuring that control systems can tolerate significant variations in
dynamics, fuel burn, and atmospheric conditions without compromising stability.
Computational Tools for Disk Analysis
MATLAB Control Toolbox (1984–Present): As disk analysis became more widely used,
computational tools began to integrate disk-based robustness analysis. MATLAB introduced
functions like diskmargin to compute disk margins for control systems, allowing engineers to
evaluate robustness quickly and effectively. These tools have become a staple in modern
control system design.
Robust Control Toolbox (2005–Present): MATLAB’s Robust Control Toolbox allows users to
compute classical margins, disk margins, and more complex robustness metrics like H∞ and
µ-analysis. This has greatly enhanced the accessibility of advanced disk margin techniques in
industry and academia.
Recent Advances in Disk Analysis
Nonlinear and Adaptive Control Systems: Disk analysis techniques are increasingly being
applied to nonlinear and adaptive control systems, where uncertainties are more complex and
harder to model. Recent research explores disk-based methods for handling large parameter
variations, with promising results in fields like robotics and renewable energy systems.
Chen et al. (2020) published studies on using disk margin techniques in nonlinear adaptive
controllers, highlighting improved robustness in uncertain environments, such as those found
in renewable energy and unmanned aerial vehicle (UAV) control.
Model Predictive Control (MPC): Disk margins have also been integrated into Model
Predictive Control (MPC), especially in systems with constrained inputs and outputs. Recent
research by Rawlings and Mayne (2017) shows how disk margins can be used in predictive
control to maintain robust stability despite the inherent model uncertainties in MPC
frameworks.
Challenges and Future Directions
Time-Delay Systems: A challenge in disk margin analysis is addressing time-delay systems,
where the interaction between time delays and uncertainties can complicate robustness
analysis. Ongoing research aims to extend disk margin methods to handle time delays
effectively.
Data-Driven Control: With the advent of machine learning and big data, there is growing
interest in using data-driven approaches to improve disk margin analysis. Future work may
involve using AI algorithms to predict system uncertainties and adjust disk margins
dynamically in real-time systems.
 6
Quantum Control Systems: A cutting-edge area of research is exploring how disk margin
techniques could be adapted for the control of quantum systems. These systems exhibit
probabilistic behaviors and uncertainties that classical control frameworks struggle to handle,
and disk-based robustness may provide a solution for quantum control stability.

7
 CHAPTER-3

MOTIVATION & OBJECTIVE

3.1 Motivation

The proposed method of disk margin and H infinity control can be used to evaluate the LPV
robustness margins of a flutter suppression controller for a flexible aircraft. The airframe is a
small, radio-controlled aircraft denoted mini-MUTT, as shown in Figure The design is based
on Lockheed Martin's Body Freedom Flutter vehicle.

Figure 2 MINI MUTT

3.2 OBJECTIVE
Primary Goal
The main aim of this project is to design a robust H∞ controller specifically for a DC-DC
buck converter. This controller should be able to maintain stable and reliable output despite
the presence of uncertainties such as parameter variations and external disturbances.
Why Robust Control?
In real-world scenarios, electronic circuits are often exposed to unpredictable changes in
input and environmental conditions. This makes it crucial for control systems to handle these
variations effectively, ensuring that the device operates correctly and consistently.

8
Role of H∞ Control
Traditional control methods often fall short when faced with complex uncertainties. The H∞
method allows us to directly tackle these issues by minimizing the worst-case effect of
disturbances on the system. This approach is particularly useful for systems like DC-DC
converters that require high reliability.

Project Focus
By applying the H∞ technique, the project aims to develop a controller that not only
stabilizes the buck converter but also optimizes its performance in uncertain conditions. This
involves modeling the uncertainties, designing the controller, and validating the design
through MATLAB simulations.

9
 CHAPTER-4
METHODOLOGY

4.1 DISK MARGIN

4.1.1. Definition of Disk Margin
Disk margin refers to the measure of robustness of a control system against uncertainties.
Specifically, it quantifies the maximum amount of structured uncertainty that a system can
tolerate while still maintaining stability. The concept is often represented visually as a disk in
the complex plane, where the radius of the disk indicates the amount of allowable
uncertainty. The disk margin refers to the amount by which the loop gain and phase of a
system can change before causing instability. It is typically represented as a range of
permissible gain and phase variations that form a disk on the Nyquist plot.
If a control system has a large disk margin, it means the system can tolerate a wider range of
changes or uncertainties in its parameters without becoming unstable. A smaller disk margin
indicates that the system is more sensitive to parameter changes.
4.1.2. Uncertainty Representation
In robust control, uncertainties can be represented in several ways:
 Additive Uncertainty: The actual system can be viewed as the nominal system plus
some perturbation.
 Multiplicative Uncertainty: This representation involves scaling the nominal model.

4.1.3 The Complex Plane and Stability

To analyze stability, we often use the Nyquist criterion. The Nyquist plot shows how the
open-loop transfer function behaves as frequency varies.
 Stability Criterion: According to the Nyquist criterion, for the closed-loop system to
be stable, the Nyquist plot must not encircle the point −1+j0-1 + j0−1+j0 in the
complex plane. The goal is to ensure that the phase and gain conditions hold true
across a range of uncertainties.

10
4.1.4. Visualization of Disk Margin
In the context of disk margin:
 Center of the Disk: The center represents the nominal system without uncertainties
(often at the origin).
 Radius of the Disk: The radius indicates the maximum perturbation that can be added
to the system without leading to instability.If the Nyquist plot of the system remains
within this disk, the system is considered robust against the modeled uncertainties
4.1.5. Computation of Disk Margin
To compute disk margin:
1. Determine the Nominal System: Analyze the open-loop transfer function.
2. Construct the Nyquist Plot: Plot the frequency response of the open-loop transfer
function.
3. Evaluate Stability: Identify how close the Nyquist plot comes to the critical point −1.
The distance from the nearest point on the plot to −1 gives an indication of how much
uncertainty can be added before instability occurs.
4. Calculate Disk Radius: The radius of the disk is defined as the minimum distance
from −1 to the Nyquist plot. This distance reflects the disk margin.
4.1.6. Practical Implications
 Designing Controllers: In robust control design, engineers can use disk margins to
ensure that their controllers are capable of maintaining stability under the worst-case
scenarios of parameter variations or external disturbances.
 Performance Trade-offs: A larger disk margin generally indicates better robustness,
but it might come at the cost of performance. Engineers need to balance robustness
with performance metrics such as transient response and steady-state error.
 Application Areas: Disk margin analysis is particularly useful in fields such as
aerospace, robotics, and automotive systems, where safety and reliability are
paramount.

11
12
 13

4.2 Varying of f in the open loop

Figure 3 MATLAB Code for varying finding gain and phase margin for a given transfer function

14
 Now lets vary f = 4-3i

15
  MATLAB Code for Varying f

Figure 4 MATLAB Code for Varying f in SISO Loop

Figure 5 Output Graph for Variation of f

16
Why disk margin ?

Disk margin is a way to measure how much uncertainty a closed-loop system can tolerate
before becoming unstable. It's used to quantify the stability of a system against variations in
phase or gain in the open-loop response. Disk margin is a more robust way to analyze
stability than classical gain and phase margins because it considers all frequencies and loop
interactions.
Here are some things to know about disk margin:
What it computes?
Disk margin computes the smallest variation that destabilizes the feedback loop. It also
computes the gain and phase margins by finding the largest disk for which the closed-loop
system is stable.
Why it's useful? Disk margin is used to investigate the feedback loop for robust stability
against gain and phase uncertainty. It's a design tool that helps ensure a system is robust to
variations and uncertainty that might occur in the real system.

17

4.3. Robust Control Toolbox provides tools to:

  Analyze system stability against gain and phase variations. Use diskmargin to
compute the disk-based gain and phase margins of SISO and MIMO feedback
loops.
 Model gain and phase uncertainty. Use the umargin control design block to
analyze the effect of gain and uncertainty on system performance and stability .

Functions
Disk-based stability margins of
diskmargin
feedback loops

Worst-case disk-based stability

wcdiskmargin
margins of uncertain feedback loops

Visualize disk-based stability margins

diskmarginplot
(Since R2020a)

Visualize worst-case disk-based

wcdiskmarginplot
stability margins (Since R2020a)

Customize disk-based stability-margin

diskmarginoptions
plots (Since R2020a)

Convert gain and phase variation

getDGM
into disk-based gain variation (Since R2020a)

Disk-based phase variation corresponding to

getDPM
disk-based gain variation (Since R2020a)

Get disk-based margins from disk

dm2gm
size and skew (Since R2020a)

Convert disk-based gain margin to

gm2dm
disk size and skew (Since R2020a

18
4.4. Modelling Gain and Phase Variations
Both umargin and diskmargin represent gain and phase variation as a multiplicative complex
factor F(s), replacing the nominal open-loop response L(s) with L(s)*F(s). The factor F takes
values in a disk that includes the nominal value F = 1. This multiplicative factor models both
gain and phase variations. For instance, the following plot shows one such disk in the
complex plane.

Figure 6 Disk Visualization on SIMULINK

Both umargin and diskmargin model gain and phase uncertainty with a family of disks
described by two parameters, α and σ. For SISO systems, the disk is parameterized by:

In this model,
 δ is the normalized uncertainty (an arbitrary complex value in the unit disk |δ| < 1).
 α sets the amount of gain and phase variation modeled by F. For fixed σ, the
parameter α controls the size of the disk. For α = 0, the multiplicative factor is 1,
corresponding to the nominal L.
 σ, called the skew, biases the modeled uncertainty toward gain increase or gain
decrease.

19
Each α,σ pair corresponds to a disk that models a particular gain-variation range DGM =
[gmin,gmax], given by the points where the disk intercepts the real (x) axis. The
corresponding phase variation DPM is determined by the angle between the real axis and a
line through the origin and tangent to the disk. Thus you can describe a modeled set of gain
and phase variations entirely by either the two values α,σ or the two values DGM =
[gmin,gmax]. σ = 0 models a balanced gain variation with [gmin,gmax] such that gmin =
1/gmax. When σ < 0, then F represents a larger gain decrease than increase (gmin < 1/gmax).
Conversely σ > 0 represents a larger gain increase than decrease. For instance, consider the
disks parameterized by α = 0.5 and three different skews, σ = –2, 0, and 2.

Command

diskmarginplot(0.5,[-2 0 2],'disk')

Figure 7 Variation of Skew in disk

20
Each α,σ pair corresponds to a disk that models a different gain-variation range DGM =
[gmin,gmax]. Examine the gain variations that correspond to each of these three disks.
Get
 Ranges = dm2gm(0.5,[-2 0 2])
Ranges = 3×2
0.3333 1.4000
0.6000 1.6667
0.7143 3.0000
Get
diskmarginplot(Ranges)

Figure 8 Range of Gain and phase variation

The balanced σ = 0 range is symmetric around the nominal value, allowing the gain to
increase or decrease by a factor of about 1.67. The negative σ value corresponds to more gain
decrease than increase, while positive σ gives more increase than decrease.
The umargin control design block uses this model to represent gain and phase uncertainty in a
feedback loop, setting α and σ from the gain and phase margin values you specify when you
create the block. The diskmargin command uses this model to compute disk-based gain and
phase margins as discussed in the next section.

21

4.5. Disk Margins for SISO Loops

For a loop transfer L and a given skew σ, the diskmargin command finds the largest disk
size α at which the closed-loop system feedback(L*F,1) is stable for all values of F. This
value of α is called the disk margin. The disk-based gain margin DGM and disk-based phase
margin DPM are the range of gain and phase variations represented by the corresponding
disk.
For instance, compute the disk margin and associated disk-based gain and phase margins for
a SISO transfer function, using the default σ = 0.

sigma = 0;
L = tf(25,[1 10 10 10]);
DM = diskmargin(L,sigma);
alpha = DM.DiskMargin

alpha =
0.4581
DGM = DM.GainMargin
DGM = 1×2

0.6273 1.5942

Get
DPM = DM.PhaseMargin
DPM = 1×2

-25.8017 25.8017

For this system, the balanced (σ = 0) disk margin α is about 0.46. The corresponding disk-based gain
margin DGM shows that the system remains stable for relative variations in gain between about 0.63
and 1.6, or for phase variations of about ±26 degrees. This result establishes stability for all values
of F of in the disk:

22
Get
Disk margin plot (DGM,'disk')
Figure 9 Output Disk for SISO loop

Uncertainty Disk in the Nyquist Plane

figure
h = nyquistplot(L);
hold on
diskmarginplot([DGMdec;DGM;DGMinc],'nyquist')
h.Responses(2).Name = '\sigma = -2';
h.Responses(3).Name = '\sigma = 0';
h.Responses(4).Name = '\sigma = 2';
hold off

23
Figure 10 Uncertainty Disk in the Nyquist Plane

24
CHAPTER-5
RESULTS

25
 Variation of Skew

26
27
CHAPTER-6
WORK PLAN

28
 CHAPTER-7
REFERENCES

[1] Blight, James D., R. Lane Dailey, and Dagfinn Gangsaas. “Practical Control
Law Design for Aircraft Using Multivariable Techniques.” International
Journal of Control 59, no. 1 (January 1994): 93–
137. https://doi.org/10.1080/00207179408923071.
[2] Seiler, Peter, Andrew Packard, and Pascal Gahinet. “An Introduction to Disk
Margins [Lecture Notes].” IEEE Control Systems Magazine 40, no. 5 (October
2020): 78–95.
[3] S.Ricardo, Sanchez-Pena and Mario Sznaier, "Robust systems theory and
applications", John Wiley & Sons, 1998.
[4] Zhou K.and J.Doyle, "Essential of the robust Control", Printice Hall, New
Jercy, 1998.
[5] G.Zames,"Feedback and Optimal sensitivity:. Model reference
Transformations, multiplicative semi norms, and approximate Inverses",IEEE
Transaction on Automatic Control, vol. AC-26,pp.301-320,1981.
[6] G.Zames and B.A.Francis, "Feedback, Minimax Sensitivity and Optimal
Robustness", IEEE Transaction on Automatic Control AC-28,pp.585-600,1983.

29