Advanced Control

An Overview on Robust Control

'

Scope allow the student to assess the potential of different methods in

robust control without entering deep into theory. Sensitize for
the necessity of robust feedback control.

Keywords uncertainty representations, H∞ , µ synthesis, LMI

Prerequisites Nyquist criterion, gain and phase margin, LQG state space
control

Contact Raoul Herzog1 , Jürg Keller2

Version 1.0

Date May 6, 2009

1 raoul.herzog@heig–vd.ch
2 juerg.keller1@fhnw.ch
Advanced Control, An Overview on Robust Control MSE

2 Raoul Herzog, Jürg Keller May 6, 2009

Contents

1 Introduction to Robust Control . . . . . . . . . . . . . . . . . . . . . 3

1.1 Motivation of Robust Control . . . . . . . . . . . . . . . . . . 3
1.2 An attempt to define robust control . . . . . . . . . . . . . . . 4
1.3 Structure of this document . . . . . . . . . . . . . . . . . . . . 5
2 Review of Norms for Signals and Systems . . . . . . . . . . . . . . . . 5
3 The Nyquist Criterion and the Small Gain Theorem . . . . . . . . . . 8
3.1 Review of the Nyquist Criterion and Classical Stability Margins 8
3.2 The Small Gain Theorem . . . . . . . . . . . . . . . . . . . . 10
3.3 Applications of the Small Gain Theorem to Robust Control . 11
4 Description of Model Uncertainty . . . . . . . . . . . . . . . . . . . . 12
4.1 Unstructured Uncertainty . . . . . . . . . . . . . . . . . . . . 13
4.2 Structured Uncertainty . . . . . . . . . . . . . . . . . . . . . . 19
5 Formulation of the Standard H∞ Problem . . . . . . . . . . . . . . . 21
6 A Glimpse on the H∞ State Space Solution . . . . . . . . . . . . . . 24
7 Limitation of H∞ Methods . . . . . . . . . . . . . . . . . . . . . . . . 26
8 Outlook: µ Synthesis and LMI Methods . . . . . . . . . . . . . . . . 26
8.1 Structured Singular Values (SSV) and µ Synthesis . . . . . . . 26
8.2 Linear Matrix Inequalities (LMI) . . . . . . . . . . . . . . . . 27
9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1 First exercise: two cart problem . . . . . . . . . . . . . . . . . . . . . 30
2 Second exercise: drawback of classical stability margins . . . . . . . . 31

1 Introduction to Robust Control

1.1 Motivation of Robust Control
System modeling is generally one of the most important tasks in engineering, and
in particular in control engineering. Modeling is always a delicate task, since the
physical reality is often complicated, and all attempt to mathematically describe a
real physical process involves simplifications and assumptions, e.g.:

• dynamics are often consciously neglected1 to make the model tractable, e.g.
unmodeled sensor and actuator dynamics, higher order modes from large scale
structures modeled by finite elements (FEM),
1
It’s common to say: “My model is precise up to . . . Hz.”

May 6, 2009 Raoul Herzog, Jürg Keller 3

Advanced Control, An Overview on Robust Control MSE

• nonlinearities are often either hard to model or too complicated,

• parameters are often not exactly known, either because they are hard to mea-
sure precisely, or because of varying manufacturing conditions2 .

Furthermore, it is desirable to end up with simple plant models. Indeed, in modern

control, the controller is the output of an optimization problem, and the complex-
ity of the controller is directly linked with the complexity of the plant model: a
complicated high–order plant will automatically lead to a complicated high–order
controller, which is undesirable.
Robust control deals with system analysis and control design for such imperfectly
known process models. One of the main goals of feedback control is to maintain
overall stability and system performance despite uncertainties in the plant.
In general, robustness does not “come for free” from a controller designed via
optimal control and estimation theory (observer design): a controller designed for a
nominal process model generally works fine for the nominal plant model, but may
fail3 for even a “nearby” plant model.
An important point of all feedback control synthesis methods is the control en-
gineer’s awareness of inherent trade–offs : increasing the robustness will generally
make the controller “less aggressive”, and will thereby decrease system performance.
Robust control allows to specify more or less directly the plant uncertainty, and
allows to predict the possible trade–offs between robustness and closed–loop perfor-
mance.
Sometimes bachelor–level courses in control may give the impression that every-
thing is feasible with control, and that it’s only a matter of finding a “good” con-
troller. But this impression is completely false: the plant itself implies inherent lim-
itations4 , especially if the plant is unstable. The achievable performance/robustness
trade–offs strictly depend on the plant, and cannot be overcome by any sophisticated
control [1].

1.2 An attempt to define robust control

A definition of robust control could be stated as:
Robust control aims at designing a fixed (non–adaptive) controller such that some
defined level of performance5 of the controlled system is guaranteed, irrespective of
changes in plant dynamics within a predefined class.
2
For serial production of devices incorporating feedback control, individual tuning is not desir-
able. The robustness of the system should be sufficient to tolerate all uncertainties and tolerances
from the manufacturing processes. For mechatronical systems, the manufacturing uncertainties
are coming from the mechanical and electronical parts. Generally, there are no manufacturing
tolerances inside the digital controller because software is perfectly reproducable.
3
Instability or unacceptable performance degradations may occur.
4
e.g. by the Bode integral relation.
5
e.g. closed–loop stability, reference tracking performance, and disturbance rejection perfor-
mance.

4 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

Robust control offers a collection of powerful mathematical tools and efficient

software algorithms able to partly6 answer the following key questions:
Describing and characterizing plant uncertainty: What is a good way to de-
scribe plant variations or plant uncertainty? Some descriptions attempt to
faithfully describe the real situation (e.g. probability distributions on physical
parameters), but unfortunately there may exist no efficient method to solve the
problem. Other uncertainty descriptions are less direct, but more convenient
for the theory (e.g. analytic solutions exist).
Robustness analysis problems: As an example consider a nominal plant and a
given stabilizing controller. The uncertainty in the plant model is defined by
a class of perturbations. There is no “true” model, we have to deal with a
given set of possible plant models. Now, we would like to know whether or not
the closed–loop remains stable for the whole class of plants. This is a typical
robustness analysis problem.
Robustness synthesis problems: Find a controller which stabilizes a given class
of plants. Generally, synthesis problems are more difficult to solve than ana-
lysis problems.

1.3 Structure of this document

The document is structured as follows: In section 2, norms for signals and systems
are reviewed for the single input single output case (SISO). Section 3 recapitulates
the Nyquist criterion and the small gain theorem, an important working horse in
robust control. Section 4 addresses the question how to describe model uncertainty.
Sections 5 introduces the setup of H∞ control. It is not the goal of this document to
describe the underlying mathematical theory, which is quite demanding. Therefore,
section 6 only sketches the H∞ solution from a user point of view. Section 7 discusses
some limitations and drawbacks of standard H∞ methods. Finally, section 8 gives
an outlook to the actual state–of–the–art in robust control. Section 9 concludes
with some general remarks on robust control.

2 Review of Norms for Signals and Systems

Norms are used in many places of engineering to quantify the magnitude of an object
(e.g. the amplitude of a signal), or to quantify the proximity of two objects (e.g.
the proximity of two systems). In robust control, norms play a crucial role:
• the choice of metric used to quantify the amount of process uncertainty,
6
It should be noted that some simple problems in robust control (e.g. exact stability determi-
nation of a linear system in which several parameters vary over given ranges) have shown to be
NP–hard, hence as difficult as other famous problems for which no efficient solutions are known to
exist, or likely to be found.

May 6, 2009 Raoul Herzog, Jürg Keller 5

Advanced Control, An Overview on Robust Control MSE

• the choice of norm used in the optimization problem associated with the con-
troller synthesis. The linear quadratic gaussian LQG control is based on the
optimization of a ||·||2 norm, whereas H∞ control is based on the optimization
of a || · ||∞ norm.
The L2 (Euclidean) norm of a time–domain scalar signal u(t) is defined as:
sZ
∞
||u||2 = u2 (t) dt. (1)
−∞

If this integral is finite, then the signal u is square integrable, denoted as u ∈ L2 .

For vector–valued signals u(t),

u1 (t)
 

 u2 (t) 

u(t) =  .. . (2)
.
 
 
un (t)

the 2–norm is defined as

sZ sZ
∞ ∞
||u||2 = ||u(t)||22 dt = uT (t) u(t) dt, (3)
−∞ −∞

where the superscript T denotes transposition.

A Linear Time–Invariant (LTI) system G can be described either by a state space
realization

ẋ = Ax + Bu
y = Cx + Du

or by its corresponding transfer function

G(s) = C(sI − A)−1 B + D. (4)

Similar to signals, we would like to define different norms for systems, respectively for
transfer functions. Two systems G1 and G2 are “close” if the norm of the difference
of their transfer functions ||G1 − G2 || is “small”. Two systems might be “close” with
respect to one defined norm, but “far” with respect to another norm7 .
The H2 norm of a stable system G(s) is defined as the L2 norm of its impulse
response g(t). When dealing with stochastic signals, the H2 norm of a system
corresponds to the variance E[||y||2 ] of the system response y(t) when the system is
excited by a unit intensity white noise input u(t).
1
7
With the metric induced by the H∞ norm, the two systems P1 (s) = s+ǫ and P2 (s) = 1s are
infinitely “far”, no matter how small ǫ is. This may seem illogical because a reasonable controller
will stabilize both plants P1 and P2 simultaneously, and the resulting closed–loop systems will be
nearly identical. A remedy consists in using the so–called Vinnicombe metric, which also allows to
treat marginally stable and unstable systems.

6 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

The H∞ norm of a stable transfer function G(s) is defined as the maximum RMS
amplification over arbitrary square integrable input signals u 6= 0.

||y||2
||G||∞ = sup (5)
u∈L2 ||u||2

Note that the H∞ norm for a system is induced by the L2 norm for signals. The
physical interpretation of the H∞ norm corresponds simply to the maximum energy
amplification over all input signals. It can be shown that for single–input single–
output systems ||G||∞ equals the peak magnitude in the Bode diagram of the transfer
function G(j ω):
||G||∞ = sup |G(j ω)| (6)
ω

Bode Diagram
2
10
Magnitude (abs)

1
10

0
10

−1
10
0

−45
Phase (deg)

−90

−135

−180
−2 −1 0
10 10 10
Frequency (Hz)

1
Figure 1: Example: The || · ||∞ norm of G(s) = s2 +0.1s+1
is ||G||∞ ≈ 10.

In H∞ control the performance to be optimized is defined in terms of minimizing

the H∞ norm of closed–loop transfer functions, e.g. the sensitivity function S(s)
and the complementary sensitivity function T (s). This type of optimization prob-
lem is also called min–max problem: H∞ control seeks to minimize the worst–case
scenario, i.e. when the closed–loop function has its peak. In contrast, classical LQG
control minimizes the closed–loop behaviour for known input signals8 , whereas H∞
control works with unknown input signals and tries to optimize the worst–case sce-
nario. Therefore, the solution of H∞ control problems typically inhibits “flat9 ” Bode
8
e.g. the energy of the resulting closed–loop impulse response
9
allpass behaviour

May 6, 2009 Raoul Herzog, Jürg Keller 7

Advanced Control, An Overview on Robust Control MSE

magnitude plots (no more peak). When minimizing the absolute peak in the Bode
magnitude diagram, automatically other local peaks pop up. This effect is called
“waterbed–effect”. Finally, the value of all local peaks join the value of the absolute
peak, and the response becomes “flat”. The theoretical background of this comes
from the Bode integral theorem, see equation (12) and figure 6 page 12.

Waterbed effect
1
10

minimize
the peak !
magnitude will pop up
elsewhere !

0
10
magnitude of a closed−loop function

−1
10

−2
10

−3
10
−1 0
10 10
frequency Hz

Figure 2: Illustrating the waterbed effect.

Instead of “flat” Bode magnitude plots we can also impose a given shape. This
corresponds to a modern version of classical loop–shaping. In this context, important
design parameters are frequency dependent weighting functions W (s). They allow
to shape given closed–loop functions, e.g. the sensitivity S(s), by optimizing their
weighted norm ||W S||∞ .
The H∞ norm of a multivariable (MIMO) transfer matrix G(s) needs the im-
portant concept of singular values which is beyond the scope of this document.

3 The Nyquist Criterion and the Small Gain Theorem

The Nyquist criterion is a cornerstone of classical control, and is of fundamental
importance in robust control. The following chapter recapitulates the Nyquist cri-
terion, the small gain theorem, and shows its application to robust control.

3.1 Review of the Nyquist Criterion and Classical Stability Margins

The Nyquist criterion allows to check closed–loop stability based on the inspection
of the loop gain L(s) = C(s) · P (s), without computing the closed–loop poles, i.e.

8 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

the roots of 1 + L(s) = 0. The Nyquist criterion is based on Cauchy’s argument,

and says:
The closed–loop system with loop gain L(s) and a negative feedback polarity is
stable if and only if the Nyquist plot of L(jω) encircles NP anticlockwise times the
critical point scrit = −1 in the complex plane, where NP is the number of unstable
(right halfplane) poles of L(s).

loop gain

 L(s)

feedback with minus polarity

Figure 3: Elementary feedback system with loop gain L(s)

As a special case, if L(s) already is a stable open loop transfer function, NP is

zero, and the Nyquist plot of L(jω) must not encircle the critical point −1 in order
to ensure closed–loop stability. Figure 4 recapitulates the definition of the classical
gain margin Am and phase margin φm . For example a gain margin of Am = 2 means
that the closed-loop stays stable even if the loop gain doubles. The phase margin
φm indicates the amount of additional delay the feedback loop can tolerate before
becoming unstable. Robust control does not work with the classical stability margins
Am and φm for two reasons: 1) there exist no analytical optimization techniques for
these margins, and 2) there are cases where the classical margins indicate a good
robustness against individual gain and phase tolerances, whereas the feedback loop
is not at all robust against simultaneous variations of gain and phase, see exercice
2, page 31. For these reasons, robust control prefers as margin the critical distance
dcrit between the Nyquist plot of L(jω) and the critical point scrit = −1:

dcrit = min
ω
(dist(L(jω), scrit )) = min
ω
||L(jω) − scrit ||2 = min
ω
||1 + L(jω)||2 . (7)

It follows that the critical distance dcrit is the reciprocal of the sensitivity function
peak:
1 1
dcrit = , where S(jω) = . (8)
||S||∞ 1 + L(jω)
Maximizing the critical distance dcrit corresponds to minimizing the || · ||∞ norm of
the sensitivity function. Therefore, one of the natural objectives10 of robust control
consists of minimizing the || · ||∞ norm of the sensitivity function.
10
It is important to see that sensitivity peak minimization is not the only objective in real world
problems. In fact, minimizing only the || · ||∞ norm of the sensitivity function turns out to be an
ill-posed problem, because the gain of the resulting controller would be infinite.

May 6, 2009 Raoul Herzog, Jürg Keller 9

Advanced Control, An Overview on Robust Control MSE

Nyquist plot of L(s)

complex plane

0.5
imaginary part

−1 L(j ∞) L(j 0)
0
1/Am
dcrit

−0.5

φm

−1

L(j ω)

−1.5 −1 −0.5 0 0.5 1 1.5 2

real part

Figure 4: Definition of the critical distance dcrit and the classical stability margins

3.2 The Small Gain Theorem

The small gain theorem follows directly from the Nyquist criterion: if L(s) is stable
and ||L||∞ < 1, then the loop gain |L(jω)| is smaller than 1 for all frequencies ω,
and hence L(jω) does not encircle the critical point −1. It follows that the small
gain condition ||L||∞ < 1 is a sufficient but not necessary condition for closed–loop
stability. The small gain condition corresponds to an infinite phase margin φm = ∞.
If ||L||∞ < 1, the closed loop stays stable even if the feedback polarity is wrong!
The small gain theorem is not limited to linear feedback control: it can even
be generalized to nonlinear feedback control. In this case, L becomes a nonlinear
operator in time domain, and the || · ||∞ norm condition on L(s) must be replaced
by an operator norm condition. This may sound abstract, but a simple application
is the feedback of a linear dynamic system in cascade with a static nonlinearity. In
this case, the small gain theorem yields a sector bounded11 nonlinearity as a sufficient
condition for closed–loop stability.
It makes no sense to apply the small gain theorem directly for controller syn-
thesis: for good tracking performance a high loop gain is needed within the control
11
also called Popov criterion.

10 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

bandwidth12 . However, the small gain theorem has a great utility for analyzing
feedback loops with unstructured uncertainty.

3.3 Applications of the Small Gain Theorem to Robust Control

Suppose that P0 (s) denotes the nominal plant, and C(s) a stabilizing controller.
Now consider an additive perturbation which yields a “cloud” of plants P (s) =
P0 (s)+∆a (s), where ∆a (s) denotes an unknown stable transfer function representing
the modeling uncertainty of P0 . The question arises “how much” uncertainty the
closed loop may tolerate before becoming unstable. Figure 5 shows that the feedback

perturbed plant P
'a feedback
seen by '
additive uncertainty +

'a P0 +

-C
nominal plant +

P0 +

'a
controller
C -C
1+P0C

Figure 5: Application of the Small Gain Theorem to additive plant uncertainty.

−C −C
seen by ∆a is 1+P 0C
. The corresponding loop gain is ∆a · 1+P 0C
. Since both ∆a and
the nominal closed–loop are stable we can apply the small gain theorem for the
feedback loop seen by ∆a . A sufficient condition for the stability of the perturbed
system is therefore:
−C 
 

  ∆a · < 1. (9)

1 + P0 C ∞
Using the submultiplicative property of norms ||AB|| ≤ ||A|| · ||B||, the sufficient
stability condition becomes:

C 1
 
 
||∆a ||∞ ·   < 1 =⇒ ||∆a ||∞ <   . (10)
1 + P0 C ∞ C
 1+P

C

0 ∞

12
If the controller includes an integral action (e.g. the I part of PID), the loop gain is infinite at
0 Hz, and hence ||L||∞ = ∞.

May 6, 2009 Raoul Herzog, Jürg Keller 11

Advanced Control, An Overview on Robust Control MSE

Equation (10) is a conservative bound indicating the amount of tolerated additive

uncertainty which preserves stability.
In case of multiplicative unstructured uncertainty, the set of perturbed plants is
described as P (s) = P0 (s) · (1 + ∆m (s)). Using the small gain theorem, the following
stability bound can be found:
1 1
||∆m ||∞ <  1  = . (11)
 1+P C  ||S||∞
0 ∞

A high sensitivity peak value ||S||∞ leads to a small tolerable multiplicative pertur-
bation ∆m , which is intuitively clear since the distance dcrit of the nominal loop gain
L0 = P0 · C to the critical point −1 is small.
We will now discuss an extremely fundamental relationship, called Bode integral
theorem: Z ∞ X
ln |S(jω)| dω = Re(p). (12)
0 unstable poles p of L(s)

This law can be seen as a conservation law: the integrated value of the log of the
magnitude of sensitivity function S(jω) is conserved under the action of feedback.
If the open–loop L(s) is stable, then the integral becomes zero. At low frequencies,
in order to have good tracking performance, the sensitivity must be much smaller
than 1 (negative dB levels), i.e. ln |S(jω)| must become negative. The Bode in-
tegral theorem states that the average sensitivity improvement at low frequencies
is compensated by the average sensitivity deterioration at high frequencies. This is
illustrated by figure 6. If the plant is unstable, the situation is becoming worse since
the right hand side of equation (12) is now positive. This means that the average
sensitivity deterioration is always larger than the improvement. The more unstable
the plant is, the more positive the real part of the unstable poles, the more diffi-
cult the situation becomes. This applies to every controller, no matter how it was
designed. Unstable plants are inherently more difficult to control than stable plants
[1].

4 Description of Model Uncertainty

It is worthwhile to clarify what is meant by model uncertainty. In a control system
there are two categories of uncertainties: disturbance signals and perturbations in
the plant dynamics. Disturbances are external stochastic inputs which are not under
control. Examples of disturbance signals are sensor and actuator noise or changes
of the environment. Dynamic perturbations are model uncertainties caused by not
exactly known or slowly changing plant parameters, and unmodeled or approximated
system dynamics. In a house temperature control, disturbances and dynamic plant
perturbations would be :
• Disturbances:
changing outdoor temperature, wind speed, open windows and doors, heating
due to electrical equipment or human bodies, sun radiation

12 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

10

L og Ma gnitude
1.0

0.1
0.0 0.5 1.0 1.5 2.0
Frequency

Figure 6: Loopshaping constraints: Sensitivity reduction at low frequencies in-

evitably leads to sensitivity increase at higher frequencies. Picture taken from [1].

• Dynamic plant perturbations:

not exactly known isolation coefficients, unknown and changing heat capaci-
ties, uncertain efficiency of the heating device

The example shows that it is not always clear how to classify the disturbances.
Therefore, it is not surprising that in µ–synthesis the two categories of disturbances
are handled within the same formalism. For H∞ –controller design plant uncertainty
results from approximating the real system with a mathematical model of tractable
complexity. Additionally its physical parameters are not exactly known.
As explained in the motivation section, the aim of H∞ –based controller design is
to include uncertainty into controller design. To achieve this it is necessary to have
a mathematical description of model uncertainty. In this section, two different types
of uncertainty descriptions will be introduced. The first is unstructured uncertainty,
whereas the second is structured uncertainty.

4.1 Unstructured Uncertainty

A description of model uncertainty has to meet the following goals: it should be
simpler than a physical model of the neglected system dynamics, and it should be
tractable with a formalism which can easily be used in controller design. A simple
solution is found, if all dynamics, time–invariant perturbations that may occur in
different parts of the system are represented with a single perturbation transfer
function ∆(s). There are many possibilities to include ∆(s) into a control system,
but only the two most commonly used will be presented in the following. These are
shown in figures 7 and 8. The focus is on SISO systems. Figure 7 shows ∆a (s) as
an additive perturbation of the plant transfer function P (s):

P (s) = Po (s) + ∆a (s), (13)

where P (s) is the actual, perturbed plant, and P0 (s) is the nominal plant.

May 6, 2009 Raoul Herzog, Jürg Keller 13

Advanced Control, An Overview on Robust Control MSE

The transfer function ∆a (s) is used to describe a frequency dependent unstruc-

tured uncertainty as follows:

˜ a (s)
∆a (s) = W2 (s) · ∆ (14)

˜ a (s) is any stable transfer function with:

where the normalized perturbation ∆

˜ a ||∞ ≤ 1
||∆ (15)

This uncertainity description is closely related to a plant transfer function repre-

˜ a ||∞ = 1 defines a circle,
sentation in C, the Nyquist plot, as shown in figure 9a). ||∆
13
whose radius is scaled with W2 (s).

Figure 7: Additive uncertainty

Dynamic perturbations can also be described with multiplicative uncertainty.

The corresponding block diagram is shown in figure 8.

P (s) = P0 (s) · (1 + ∆m (s)) (16)

In multivariable systems, transfer functions are matrices. It is known that matrix
multiplication is not commutative, so for matrix–valued P0 and ∆m , P0 (s) · (I +
∆m (s)) is generally different from (I + ∆m (s)) · P0 (s). As a consequence, input and
output multiplicative perturbations have to be distinguished. Input multiplicative
perturbations are able to model actuator uncertainty, whereas output multiplicative
perturbations are used to describe sensor related uncertainties.

Figure 8: Multiplicative uncertainty

13
Index 2 for W is commonly used for uncertainty weighting functions.

14 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

The capabilities of additive or multiplicative uncertainty representation are il-

lustrated with an example. For comparison purposes the unmodeled high frequency
dynamic is specified. It will be investigated how this a priori known model error can
be represented with an additive or multiplicative uncertainty.
Example:
Many plants can roughly be approximated by a first order system, e.g. PT1. Such
a model usually neglects the high frequency dynamic of the system. As it is well
known, a control system with a PT1-plant can achieve any required closed loop
performance. Consequently, we are interested in an uncertainty description which
incorporates possible unmodeled phase loss into controller design and prevents a
controller design with unrealistic high bandwidth.
The nominal plant model is
5
P0 (s) = . (17)
s+1
The unmodeled dynamic is represented with a set of transfer functions with the
following elements:
1
P∆ (s) = , τ ∈ [0.02 . . . 0.05]. (18)
(τ s + 1)2

The plants to be controlled are:

P (s) = Po (s) · P∆ (s). (19)

For comparison purposes the unmodeled dynamic described above is represented

as an additive and multiplicative perturbation. Additive uncertainties are best rep-
resented in the complex plane C, i.e. in the Nyquist diagram of figure (9a), whereas
multiplicative uncertainties can easily be plotted in a Bode diagram of figure (9b).
The figures show the plant’s frequency response P (jω) for different values of τ . Next,
a single description of uncertainty has to be found for the specified range of values
for the uncertain parameter τ . To accomplish this, the additive and multiplicative
∆(jω) are determined for a set of values of τ according to the following formulas:

Additive unstructured perturbation:

∆a (s) = P (s) − P0 (s). (20)

Multiplicative unstructured perturbation:

P (s)
∆m (s) = − 1. (21)
P0 (s)

The magnitude of the perturbations ∆a (jω) and ∆m (jω) are shown in figure 10.
As defined in (14), frequency dependency is specified with the weighting transfer
function W2 (s). The solid line in the figures is an envelope of all error frequency
responses and ∆(s) = W2 (s) · ∆(s) is therefore an upper bound for the uncertainty.

May 6, 2009 Raoul Herzog, Jürg Keller 15

Advanced Control, An Overview on Robust Control MSE

0.5
nominal plant 10
5

perturbed plant set

0

0
10
−0.5

−1
−1 0 1 2 3
10 10 10 10 10
−1.5
0

−2
−100

−2.5
−200

−3 −300
−1 0 1 2 3 4 5 −1 0 1 2 3
10 10 10 10 10

(a) Additive uncertainty (b) Multiplicative uncertainty

Figure 9: Plant uncertainty

Additive weighting:
7.5s
W2a (s) = . (22)
(s + 1)(s + 15)
Multiplicative weighting:
0.997s(s + 170)
W2m (s) = . (23)
(s + 9)(s + 135)

0 1
10 10
approx. error bound approx. error bound
additive. errors multipl. errors

0
10
−1
10

−1
10

−2
10
−2
10

−3 −3
10 10
−1 0 1 2 3 −1 0 1 2 3
10 10 10 10 10 10 10 10 10 10

(a) Additive uncertainty (b) Multiplicative uncertainty

Figure 10: Uncertainty descriptions

In figure 11 the resulting uncertainty regions are plotted in the corresponding

plots.
Compare the resulting regions with the set of plants in figure 9! Obviously, the
uncertainty region is very large due to the fact, that the nominal plant is not in the
center of the plant set. Since the unnecessary uncertainty regions are mainly opposite
of the critical point, it might not have a large impact on controller design. After

16 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

5
0.5 10
perturbed plant set(w)

0
0
10
−0.5

−5
−1 10
−1 0 1 2 3
10 10 10 10 10

−1.5 100

0
−2
−100
−2.5
−200

−3 −300
−1 0 1 2 3 4 5 6 −1 0 1 2 3
10 10 10 10 10

(a) Additive bounds (b) Multiplicative bounds

Figure 11: Uncertainty bounds

ω = 10 rad/s the 0–gain point lies within the uncertainty set. As a consequence, a
plant phase cannot be determined anymore. In the Bode diagram of figure phase
bounds are set to −270 and 90 degrees, in order to get a nice bode plot.
In a similar way, static nonlinearities (sector criterion) or uncertain dead–time
can be described with unstructured uncertainties.
There arises the question, which of the representations of uncertainty should be
used. Since the optimal H∞ –controller has the order of the plant plus the orders of
all the weighting functions, it is preferable to choose a representation which leads
to a minimal order weighting. In the SISO case, additive uncertainty can be recast
into multiplicative uncertainty with simple algebraic operations:

P0 (s) ∆m (s) = ∆a (s) (24)

Since the additive representation includes the plant transfer function P0 (s) it is
expected, that an additive representation usually is of higher degree. To illustrate
this: uncertain dead–time can be fit into a multiplicative uncertainty description
which is independent of the the plant:

P (s) = Po (s) · e−sT (25)

∆m (s) = e−sT − 1 (26)

This function is shown in the amplitude plot of figure 12. It can be bounded with
a DT1 transfer function.
The chosen uncertainty representation determines which closed-loop transfer
function has to be considered in H∞ –controller design in order to guarantee ro-
bustness with respect to stability and performance.
Additive and multiplicative uncertainty are special cases of a more general frame-
work which will be explained below. First, notice that we have to distinguish two
different feedback loop: the feedback loop formed by the controller and the per-
turbed plant, and the feedback loop in which the uncertainty ∆ resides. A general

May 6, 2009 Raoul Herzog, Jürg Keller 17

Advanced Control, An Overview on Robust Control MSE

1
10

0
10

−1
10

−2
10

−3
10
−1 0 1 2 3
10 10 10 10 10

Figure 12: Dead–time uncertainty

solution for all different uncertainty descriptions can be obtained, if the control
system is described in the feedback structure shown in figure 13. The plant P is

h
g

Figure 13: Standard P–∆ configuration

partitioned according to the dimensions of ∆.

" #
P11 P12
P = . (27)
P21 P22

With simple algebraic calculations it follows:

z = [P22 + P21 ∆(I − P11 ∆)−1 P12 ] w (28)

if (I − P11 ∆)−1 exists.

Define:
F(P, ∆) := P22 + P21 ∆(I − P11 ∆)−1 P12 (29)

18 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

F(P, ∆) is called a linear fractional transformation LFT of P and ∆. In the single

input single output case the LFT corresponds to a bilinear transform. This represen-
tation is also used for more sophisticated uncertainty representation (next section)
and also to formulate the general H∞ –controller design problem.
Additive and multiplicative uncertainty representations are special cases of a
linear fractional transformation. For additive perturbations, the matrix P becomes:
" #
0 I
P = . (30)
I P0

For multiplicative perturbations, the matrix P becomes:

" #
0 P0
P = . (31)
I P0

The example before shows that the conservative error bounds on the plant can be
large, compared to the real plant perturbation. An unstructured uncertainty model
is also not suited for perturbations which only affects a part of an interconnected
system. This motivates a more general representation of uncertainty which will be
introduced in the next section.

4.2 Structured Uncertainty

Structured uncertainty means uncertainty (tolerances) of concrete physical parame-
ters. Some examples are listed below:
• electrical components like resistors or capacitors are always affected with tol-
erances, e.g. 2% for a standard resistor.

• In a magnetic bearing system, the nominal air gap is an important parameter

which is affected by manufacturing tolerances and by thermal growth when
the machine is running.

• The relative permeability µr of a magnetic material is usually not very precisely

known
The internal structure of a linear plant can be represented as a block diagram with
integrators, summators, and gains. The physical parameters are gains in the block
diagram. It can be shown that the plant transfer function is always an LFT (linear
fractional transform) with respect to every physical parameter p1 , . . . , pn .
As an example, consider the uncertain plant equation (17), (18) and (19) page 15.
We will treat the time constant τ in P∆ (s) as uncertain parameter lying in the range
of 0.02 . . . 0.05. Suppose that the nominal value τ0 corresponds to the mid–range
value τ0 = 0.035, and τ = τ0 + ∆τ , where ∆τ is a real parameter varying between
5 1
−0.015 . . . + 0.015. The uncertain system P (s) = P0 (s) · P∆ (s) = s+1 · (τ s+1) 2 can

be recasted with the following linear fractional transform in figure 14. Note that
the uncertainty block ∆ in figure 14 is diagonal with the same diagonal element ∆τ

May 6, 2009 Raoul Herzog, Jürg Keller 19

Advanced Control, An Overview on Robust Control MSE

diagonal uncertainty block

'
'W 
'W
h g

Paug
w z

Figure 14: Structured uncertainty with an unknown time constant τ appearing

twice.

repeated twice. The reason is that the uncertain time constant τ appears twice in
the second order term P∆ (s). The augmented plant Paug (s) will be of third order
with 3 inputs and 3 outputs.
Exercise : Calculate Paug (s) and show that the associated linear fractional transform
yields
5 1
F(Paug , ∆) = · (32)
s + 1 ((τ0 + ∆τ )s + 1)2
There are available software tools capable to directly define and handle systems
with structured or unstructured uncertainty. Using the Robust Control toolbox of
Matlab the example above can be entered as follows :

>> P0 = tf(5, [1, 1])

Transfer function:

5
-----
s + 1

>> tau = ureal(’tau’, 0.035, ’range’, [0.02, 0.05])

Uncertain Real Parameter: Name tau, NominalValue 0.035, Range [0.02 0.05]
>> Pdelta = tf(1, [tau, 1])*tf(1, [tau, 1])
USS: 2 States, 1 Output, 1 Input, Continuous System
tau: real, nominal = 0.035, range = [0.02 0.05], 2 occurrences
>> P = P0 * Pdelta
USS: 3 States, 1 Output, 1 Input, Continuous System
tau: real, nominal = 0.035, range = [0.02 0.05], 2 occurrences
>> bode(P)

20 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

Bode Diagram
50

Magnitude (dB)

−50

−100

−150
0

−90
Phase (deg)

−180

−270
−2 −1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)

Figure 15: Example : structured uncertainty with an unknown time constant τ .

which gives the family of Bode plot in figure 15.

In contrast to unstructured uncertainty, see figure 13, structured uncertainty
leads to a feedback configuration figure 15 with a static diagonal block ∆. In addi-
tion, the diagonal terms are real–valued, whereas in the unstructured case they are
mostly complex–valued norm–bounded transfer functions ∆(s).
It is obvious that structured uncertainty representations are often closer to phys-
ical specifications. However, the algorithms needed to tackle robust control analysis
or synthesis problems with structured uncertainty are much more complicated, see
outlook section page 26.
We disadvise from recasting highly structured parameter uncertainty as additive
or multiplicative unstructured uncertainty, because this implies a high degree of con-
servativeness. Indeed, the class of perturbations may become excessively large, and
the resulting controller, if there is any complying with the robustness specifications
is likely to show poor performance.

5 Formulation of the Standard H∞ Problem

In this section the controller design problem is formulated in the H∞ formalism.
The objective is to design a controller, which does not only have nominal stability
and performance, but has this properties with all the plants represented by some
uncertainty description. H∞ –controller design was specially developed, to solve this
problem in a systematic manner. As system which is stable with all the plant of
the uncertainty set is called robustly stable. It has a robust performance, if the
performance specifications are met for all the admissible plant behaviors.

May 6, 2009 Raoul Herzog, Jürg Keller 21

Advanced Control, An Overview on Robust Control MSE

The control system with controller and uncertainty model is shown in figure
16. In H∞ –controller design, the H∞ –norm of the mapping from w to z can be
minimized. As a consequence the control problem has to be formulated in this
formalism. The inputs w are typically reference or disturbance signals, whereas
the outputs z can be the control error or the controller output. The H∞ –controller

Figure 16: Control system with uncertainty

design problem cannot directly by solved in structure of figure 16. Since performance
specifications and stability conditions can both be expressed as H∞ –norm conditions
on some transfer functions, there are two ways to simplify the control system of
figure 16. Either the stability conditions are formulated in the same formalism
as performance specifications, resulting in a structure as in figure 17(a), or the
performance specification is formulated similarly to the uncertainty description ∆p
(see figure 17(b)). For unstructured uncertainties the structure of figure 17(a) is
easier to go, but for structured uncertainties, only the structure in figure 17(b) leads
to an elegant representation. The structured uncertainty problem is solved with
µ-synthesis, which is beyond the scope of this introduction. Consequently, the first
approach will be followed in the sequel.

(a) Robust Stability as norm condition (b) Performance as artificial plant uncer-
tainty

Figure 17: Possible system representations

As a result from the small gain section, it is clear that as system is robustly
stable, if
for additive uncertainty:
||W2 C(I + P C)−1 ||∞ ≤ 1 (33)

22 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

2
10

1
10

0
10
−2 −1 0 1 2
10 10 10 10 10

Figure 18: Performance Weighting Function

for multiplicative uncertainty:

||W2 P C(I + P C)−1 ||∞ ≤ 1 (34)

In a control system as shown in figure 19 the robust stability condition corre-

sponds to the H∞ –norm of the mapping from r to u for the additive case and form
r to y for the multiplicative case.
System performance is usually specified by conditions on disturbance attenua-
tion, i.e. on the H∞ –norm of the mapping from d to y, or for reference tracking on
the H∞ –norm of r → e. For both mappings an acceptable performance is achieved,
if the following condition on the sensitivity function S = (I + P C)−1 is met:

||γW1 (I + P C)−1 ||∞ ≤ 1 (35)

This condition is fulfilled, if S(jω) ≤ |W11(jω)| . Consequently, a typical weighting
function W1 (s) has a shape as shown in figure 18. An integrator in W1 (s) forces
the sensitivity function to be zero at ω = 0. The parameter γ < 1 is used as an
optimization parameter. The larger γ the better the disturbance attenuation.
Performance and robust stability can therefore be achieved, if the two conditions
are combined to the following condition (mixed sensitivity approach):
Additive uncertainty:
" #
γW1 (I + P C)−1
. (36)

W2,a C(I + P C)−1

∞

Multiplicative uncertainty:
" #
γW1 (I + P C)−1

W2,m P C(I + P C)
−1



. (37)
∞

The block diagram representation of the mixed sensitivity problem is shown in

19. Here it becomes obvious, that the robust stability condition can be viewed as a

May 6, 2009 Raoul Herzog, Jürg Keller 23

Advanced Control, An Overview on Robust Control MSE

Figure 19: Control System for mixed Sensitivity Optimization

mapping of signals.
A controller can be found by solving the following optimization problem:
Additive uncertainty:
" #
γW1 (s)(I + P C)−1
max
γ

W2,a (s)C(I + P C)−1
= 1, (38)
∞

(Similarly for the multiplicative uncertainty)

Figure 20: LFT for H∞ –optimization

With some matrix algebra, the equation (38) can be transformed into a lower
linear fractional transformation F(P, C) as shown in figure 20. Therefore, the fol-
lowing problem has to be solved:

Find a stabilizing controller C, for which:

||F(P, C)||∞ := ||P11 + P12 C(I − P22 C)−1 P21 ||∞ = 1 (39)
Nowadays, various solutions to this problem are known. The two most important
are state–space methods based on the solution of two Riccati equation and the
solution based on linear matrix inequalities (LMI). Special attention has to be payed
on additional assumptions, that are imposed for the algorithms. Some of them will
be treated in the next chapter.

6 A Glimpse on the H∞ State Space Solution

H∞ –controller design problems can be formulated in different ways, but finally, the
problem has to be brought to a representation as shown in figure 20. Its state–space

24 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

representation is:

ẋ = Ax + B1 w + B2 u (40)
z = C1 x + D11 w + D12 u (41)
y = C2 x + D21 w + D22 u (42)
(43)

The system matrices are usually represented as follows:

..
 
 A . B1 B2 
 ··· ··· ···
 

Gs =  .. . (44)
 C1 . D11 D12
 


. 
C2 .. D21 D22

A closed optimal solution for the above general system is not yet published. For
the H∞ problem a closed solution was first developped in [2] for the following special
case:
..
 
 A . B1 B2 

 · · · · · · " · ·# 
· 
.. 0 
 
Gs =  C . (45)
 1 . 0
I  

 .. h i 
C2 . 0 I 0

In a design tool the general state space description of equation (44) is transformed
into the representation of equation (45) by means of loop shifting [3]. The transfor-
mation of D21 and D12 into the form with the identity matrix, requires that the two
matrices have full rank. This condition is usually violated, when weighting func-
tions are strictly proper, i.e. have a zero D–matrix. To circumvent this problem,
the weighting D-matrix can usually be set to a small value, without influencing the
controller design.
For a solution to exist, (A, B2 ) must be stabilisable and (A, C2 ) must be de-
tectable.
Another important aspect is that the H∞ problem must have a solution at
ω = ∞. An unappropriate weighting at ω = ∞ can lead to unsatisfactory con-
troller designs. For a closed loop system with plant as in equation (44) the direct
feedthrough term (gain at ω = ∞) is:

Dcl = D11 + D12 (I − Dc D22 )−1 Dc D21 (46)

where Dc is the controller D-matrix.
The term D11 is modified with the controller Dc -matrix. This can only be done
to a limited extent, which is depending on the sizes of D12 and D21 . Bounds are
documented in [3].

May 6, 2009 Raoul Herzog, Jürg Keller 25

Advanced Control, An Overview on Robust Control MSE

7 Limitation of H∞ Methods
In this section some of the limitations of H∞ –controller design will be summarized.
• The H∞ –controller design methods offers powerful tools to solve various design
problems. Nevertheless H∞ performance measures are not always adequate
for the investigated design problem and although robustness and performance
measures are combined in the design procedure, it does not really guarantee,
that performance is robust with respect to plant uncertainty. Consequently
it was proposed to combine H∞ robustness measures with other performance
measures, for example measures similar to the LQR-design, the mixed H2 /H∞
design. Also LMI (see chapter ’outlook’) offers many possibilities for ’mixed’
design problems.
• H∞ –controller design leads to controllers of high order. A high order controller
needs more ressources for its implementation and is suspect to numerical prob-
lems. Therefore it is favorable to have low order controllers. The optimal
solution for a system as described by equation ((44)) has the same order as
the system itself (size of the A-matrix). The order of the system is the order
of the plant to be controlled and the sum of the orders of all the weighting
functions, see figure 19. As already mentioned in the section on uncertainty
description, the weighting functions have to be of minimal order. The prob-
lem of high order controllers can be diminished with order reduction methods.
There are several order reduction methods available. These can be applied
either on plant before H∞ –controller design or after design on the controller
itself. The designer should be cautious not to spoil the controller design with a
radical order reduction. There are also suboptimal H∞ –controller design algo-
rithm which allow a direct design of a reduced order controller, but these are
not currently available in the commercially available controller design tools,
although a direct low order controller design is doubtless the way to go.
• As it can be seen from figures 11 the uncertainty description is not tight in
the sense, that it incorporates a much larger set of plants than necessary. This
leads to a conservative controller design, which might result in very poor closed
loop performance. This problem is reduced, when a structured uncertainty
description is used in conjuction with µ–synthesis (see Chapter ’Outlook’).

8 Outlook: µ Synthesis and LMI Methods

8.1 Structured Singular Values (SSV) and µ Synthesis
The standard H∞ minimizes an H∞ norm between input signal w and output signal
z. Often these signals are vector–valued (MIMO), as it is already the case for z in
the mixed–sensitivity approach. In practice we are interested to individual transfer
functions between a scalar component wk and a scalar component zl . A large amount
of conservativeness is introduced in the design when optimizing an overall H∞ –norm.

26 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

The µ framework allows a more selective optimization related to structured un-

certainty. A detailed description is out of the scope of this document. Basically the
µ framework introduces a new norm based on structured singular values which is the
new quantity to be minimized. The solution procedure is iterative (D–K iteration)
and involves a sequence of minimizations, first over the controller K (holding the
scaling variable D associated with the µ), and then optimizing over the scaling D
(holding the controller K fixed). The D–K procedure is not guaranteed to converge
neither to the global minimum nor to a local minimum of the µ value, but often
works well in practice. It has been applied to a number of real–world applications
with success. These applications include vibration suppression for flexible struc-
tures, flight control, chemical process control, and acustic reverberation suppression
in enclosures.

8.2 Linear Matrix Inequalities (LMI)

Linear Matrix Inequalities (LMI) have emerged as powerful numerical design tools
in areas ranging from robust control (e.g. H∞ ) to system identification.
The canonical form of linear matrix inequality is any constraint of the form

A(x) = A0 + x1 A1 + . . . + xn An < 0 (47)

where x = [x1 , . . . , xn ] is a vector of unknowns called decision variables14 , and

A0 , . . . , An are given symmetric matrices. A(x) < 0 stands for negative definite,
i.e. all eigenvalues of A(x) are negative. There are three generic LMI problems :

Feasibility problem: Find a solution x to the LMI system A(x) < 0. The set of
all solutions is also called feasibility set.

Linear objective minimization problem: Minimize cT x subject to A(x) < 0.

It is easy to show that the feasibility set of equation (47) is convex. Convexity
has an important consequence : even though the optimization problem has no
analytical solution in general, it can be solved numerically with a guaranteed
convergence to the solution when one exists. This is in sharp contrast to
general nonlinear optimization algorithms which may not converge towards
the global optimum.

Generalized eigenvalue minimization problem: Minimize λ subject to A(x) <

λ B(x), B(x) > 0, and C(x) < 0.

Efficient15 interior–point algorithms [4] are available to solve these three generic
LMI problems. Many control problems and general design specifications can be
formulated in terms of LMI. The main strength of LMI formulations is the ability to
14
It is possible to convert from scalar decision variables to matrix valued variables.
15
However the complexity of LMI computations can grow quickly with the problem order. For
example, the number of operations required to solve a standard Riccati equation is o(n3 ) where n
is the state dimension. Solving an equivalent LMI Riccati inequality needs o(n6 ) operations!

May 6, 2009 Raoul Herzog, Jürg Keller 27

Advanced Control, An Overview on Robust Control MSE

combine various convex design constraints or objectives in a numerically tractable

manner [5] [6].
A non–exhaustive list of problems addressed by LMI techniques include the fol-
lowing:

1. robust pole placement

2. optimal LQG control

3. robust H∞ control

4. multi–objective synthesis

5. robust stability of systems with structered parameter uncertainty (µ–analysis)

It is fair to say the advent of LMI optimization has significantly influenced the
direction of research in robust control. A widely–accepted technique for numerically
solving robust control problems is to reduce them to LMI problems.

9 Conclusion
Robust control emerged around 1980, and the progress in theory and numerical
algorithms during the last 25 years has been enormous. Efficient commercial and
public domain software tools are available nowadays. It is possible to successfully
use these tools without understanding the deepest details of the underlying theory.
However, a minimum of theoretical knowledge is necessary, and as it is the case with
any complex scientific software tools16 , a lot of practical experience is needed before
being able to solve real–world17 problems.
In robust control, specific reasons for this are the following: Modelling is a
challenging engineering task, and finding an estimation the modelling uncertainty
needs some experience. Additionally, either control specifications are not entirely
known in the beginning, i.e. incomplete, or real–world specifications may be very
complex18 to such an extent that no direct synthesis procedure exist. No matter
which controller synthesis method is used it remains an iterative process which also
needs some experience. In robust control, the user often needs to play around with
frequency weighting functions in order to understand the performance/robustness
trade–offs of his specific problem.
One of the main achievements in robust control might be the consciousness about
the industrial importance of robustness. Generally speaking, system failure is not
accepted in our society, and robustness considerations will remain very important.

16
e.g. tools for finite element modelling, or any other complex scientific tool
17
= non–academic
18
e.g. mixed time domain and frequency domain specifications, low controller order, etc.

28 Raoul Herzog, Jürg Keller May 6, 2009

Bibliography

[1] G. Stein. Respect the unstable. IEEE Control Systems Magazine, August 2003.

[2] J. Doyle, K. Glover, P.P. Khargonekar, and B.A. Francis. State–space solutions
to standard H2 and H∞ control problems. IEEE Trans. Aut. Control, 34(8),
August 1989.

[3] M. Safonov and D. Limebeer. Simplifying H∞ theory via loop shifting. Proc.
27th Conference on Decision and Control, December 1988.

[4] Y. Nesterov and A. Nemirowski. Interior Point Polynomial Methods in Con-

vex Programming: Theory and Applications. Studies in Applied Mathematics,
SIAM books Philadelphia, 1994.

[5] S. Boyd, L. El Ghaoui, E. Feron, and V. Balakrishnan. Linear Matrix Inequal-

ities in Systems and Control Theory. Studies in Applied Mathematics, SIAM
books Philadelphia, vol. 15, 1994.

[6] P. Gahinet and P. Apkarian. A linear matrix inequality approach to H∞ . Int.

J. Robust and Nonlinear Control, 4, 1994.

[7] M. Green and D. Limebeer. Linear Robust Control. Pearson Education, Inc.,
2002.

[8] J. Doyle, B. Francis, and A. Tannenbaum. Feedback Control Theory. McMillan

Publishing Co., 1990.

[9] Robust Control Toolbox 3, User’s Guide, The Mathworks, 2008.

[10] LMITOOL: A Package for LMI Optimization in Scilab User’s Guide, 1995.

[11] D. Gu, P. Petkov, and M. Konstantinov. Robust Control Design with MATLAB.
Springer, 2005.

May 6, 2009 Raoul Herzog, Jürg Keller 29

Advanced Control, An Overview on Robust Control MSE

Exercises

1 First exercise: two cart problem

Consider the following mechanical plant: two carts are moving on a horizontal plane
with no friction. The control force w is acting on the left cart with mass m1 , and the
position of the right cart z = x2 is measured. The carts are connected by a linear
spring k. The nominal parameters are m1 = 1 kg, m2 = 1 kg, and k = 1 N/m.
All parameters are affected by independent tolerances of ±10% for the masses, and
±20% for the spring.

x1 z = x2
position
measurement
control
force
w
m1 m2
spring
k

Figure 21: Two cart ACC benchmark problem

diagonal uncertainty
block

'
h g
augmented plant

w P z

Figure 22: Structured uncertainty setup

30 Raoul Herzog, Jürg Keller May 6, 2009

MSE Advanced Control, An Overview on Robust Control

1. Find a state space realization of the augmented plant P, where the structured
parameter uncertainty is represented as a diagonal feedback block
 
δ1 0 0
∆ =  0 δ2 0 

.
0 0 δ3

The uncertainty δ1 should correspond to the uncertainty of m1 , and a variation

of −1 < δ1 < +1 should correspond to a variation of ±10% of m1 around its
nominal value. Similarly, δ2 , δ3 correspond to the uncertainties of m2 , k, and
both should also be normalized to a range of ±1.

2. Why is it inappropriate to formulate the uncertainties of this system as additive

or multiplicative uncertainties?

2 Second exercise: drawback of classical stability margins

Consider a feedback system with the following loop gain

0.38(s2 + 0.1s + 0.55)

L(s) = .
s(s + 1)(s2 + 0.06s + 0.5)

The feedback system exhibits very good gain and phase margins, but is not at all
robust with respect to simultaneous uncertainties of gain and phase.

1. Calculate the gain margin Am and the phase margin φm using Matlab.

2. Determine the peak of the sensitivity ||S||∞ .

3. Determine the distance between the Nyquist plot of loop gain L(jω) and the
critical point −1.

This example shows that it is much better to assess robustness using ||S||∞ instead
of using traditional gain and phase margins. A low value of ||S||∞ , e.g. implies good
gain and phase margins while the converse is not true. It follows from elementary
trigonometric relations that

||S||∞ 1
Am > , φm > 2 arcsin
||S||∞ − 1 ||S||∞

For example ||S||∞ = 2 gives a garanteed gain margin of Am > 2, corresponding to

a worst case gain margin of 6 dB, and a phase margin φm > 60◦ .

May 6, 2009 Raoul Herzog, Jürg Keller 31