LMI Methods in Optimal and Robust Control

Matthew M. Peet
Arizona State University

Lecture 15: Nonlinear Systems and Lyapunov Functions

Overview

Our next goal is to extend LMI’s and optimization to nonlinear systems analysis.

Today we will discuss

1. Nonlinear Systems Theory
1.1 Existence and Uniqueness
1.2 Contractions and Iterations
1.3 Gronwall-Bellman Inequality
2. Stability Theory
2.1 Lyapunov Stability
2.2 Lyapunov’s Direct Method
2.3 A Collection of Converse Lyapunov Results
The purpose of this lecture is to show that Lyapunov stability can be solved
Exactly via optimization of polynomials.

M. Peet Lecture 15: 1 / 44

Ordinary Nonlinear Differential Equations
Computing Stability and Domain of Attraction

Consider: A System of Nonlinear Ordinary Differential Equations

ẋ(t) = f (x(t))

Problem: Stability
Given a specific polynomial f : Rn → Rn ,
find the largest X ⊂ Rn
such that for any x(0) ∈ X,
limt→∞ x(t) = 0.

M. Peet Lecture 15: 2 / 44

 Ordinary Nonlinear Differential Equations
Lecture 15 Computing Stability and Domain of Attraction

Consider: A System of Nonlinear Ordinary Differential Equations

2018-04-24
ẋ(t) = f (x(t))

Ordinary Nonlinear Differential Equations

Problem: Stability
Given a specific polynomial f : Rn → Rn ,
find the largest X ⊂ Rn
such that for any x(0) ∈ X,
limt→∞ x(t) = 0.

Linearity refers to the map from inputs to outputs vs. linearity in the RHS of
the representation.
Nonlinear Dynamical Systems
Long-Range Weather Forecasting and the Lorentz Attractor

A model of atmospheric convection analyzed by E.N. Lorenz, Journal of

Atmospheric Sciences, 1963.
ẋ = σ(y − x) ẏ = rx − y − xz ż = xy − bz

M. Peet Lecture 15: 3 / 44

 Nonlinear Dynamical Systems
Lecture 15 Long-Range Weather Forecasting and the Lorentz Attractor

2018-04-24

Nonlinear Dynamical Systems

A model of atmospheric convection analyzed by E.N. Lorenz, Journal of
Atmospheric Sciences, 1963.
ẋ = σ(y − x) ẏ = rx − y − xz ż = xy − bz

Nonlinear Systems may have

• Multiple Equilibria
• Regions of Attraction
• Limit Cycles
• Chaos
• Invariant Manifolds
• Non-exponential stability
• Finite-Escape Time
• Implicit (vs Excplicit) Algebraic Constraints
Stability and Periodic Orbits
The Poincaré-Bendixson Theorem and van der Pol Oscillator

An oscillating circuit model:

1
2
ẏ = −x − (x − 1)y
0

y
ẋ = y
−1

−2

Domain−of−attraction
−3
−3 −2 −1 0 1 2 3
x

Figure: The van der Pol oscillator in reverse

Theorem 1 (Poincaré-Bendixson).
Invariant sets in R2 always contain a limit cycle or fixed point.

M. Peet Lecture 15: 4 / 44

Stability of Ordinary Differential Equations

Consider

ẋ(t) = f (x(t))
with x(0) ∈ Rn .

Theorem 2 (Lyapunov Stability).

Suppose there exists a continuous V and α, β, γ > 0 where

βkxk2 ≤ V (x) ≤ αkxk2

−∇V (x)T f (x) ≥ γkxk2

for all x ∈ X. Then any sub-level set of V in X is a Domain of Attraction.

A Sublevel Set: Has the form Vδ = {x : V (x) ≤ δ}.

M. Peet Lecture 15: 5 / 44
Do The Equations Have a Solution?
The Cauchy Problem

The first question people ask is the Cauchy problem:

For Autonomous (Uncontrolled) Systems:

Definition 3 (Cauchy Problem).

The Cauchy problem is to find a unique, continuous x : [0, tf ] → Rn for some
tf such that ẋ is defined and ẋ(t) = f (t, x(t)) for all t ∈ [0, tf ].

If f is continuous, the solution must be continuously differentiable.

Controlled Systems:
• For a controlled system, we have ẋ(t) = f (x(t), u(t)) and assume u(t) is
given.
I This precludes feedback
• In this lecture, we focus on the autonomous system.
I Including t complicates the analysis.

I However, results are almost all the same.

M. Peet Lecture 15: 6 / 44

 ◮ exponential stability
Ordinary Differential Equations quadratic stability
time-varying systems
◮

◮
Existence of Solutions ◮ invariant sets
◮ center manifold theorem
There exist many systems for which no solution exists or for which a solution
blems
only of solutions
exists over a finite time interval. Finite Escape Time

Even for something as simple as Finite escape time of dx/dt = x2

uation 5

ẋ(t) = x(t)2 x(0) = x0 4.5

x(0) = x0 4

has the solution 3.5

3
1 x(t) x0
, 0≤t< = 2.5

x(t)
0t x0 1 − x0 t 2

1.5
which clearly has escape time
1 1
f =
x0 1 0.5
te = 0
x0 0 1 2 3 4 5
Time t

Figure: Simulation of ẋ = x2 for several

x(0)
ess Problems Previous problem is like the water-tank problem in backwar
M. Peet time Lecture 15: 7 / 44
Ordinary Differential Equations
Non-Uniqueness

A classical example of a system without a unique solution is

ẋ(t) = x(t)1/3 x(0) = 0
For the given initial condition, it is easy to verify that
 3/2
2t
x(t) = 0 and x(t) =
3
both satisfy the differential equation.
1.6 1.6

1.4 1.4

1.2 1.2

1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Figure: Matlab simulation of Figure: Matlab simulation of

ẋ(t) = x(t)1/3 with x(0) = 0 ẋ(t) = x(t)1/3 with x(0) = .000001
M. Peet Lecture 15: 8 / 44
Ordinary Differential Equations Uniqueness Problems
Non-Uniqueness
√
Example: The equation ẋ = x, x(0) = 0 has man
An Example of a system with several solutions is given by

p (t − C)2 /4 t > C
ẋ(t) = x(t) x(t)x(0)
== 0
0 t≤C

For the given initial condition, it is easy 2

to verify that for any C,

1.5
(
(t−C)2
4 t>C 1

x(t) =
0 t≤C

x(t)
0.5

satisfies the differential equation.

−0.5

−1
0 1 2 3 4 5
Time t

√
Figure: Several solutions of ẋ = x
Compare with water tank:
M. Peet Lecture 15: 9 / 44
Continuity of a Function
Customary Notions of Continuity

Nonlinear Stability requires some additional Math Definitions.

Definition 4 (Continuity at a Point).

For normed spaces X, Y , a function f : X → Y is continuous at the point
x0 ∈ X if for any  > 0, there exists a δ > 0 such that kx − x0 k < δ (U )
implies kf (x) − f (x0 )k <  (V ).

M. Peet Lecture 15: 10 / 44

Ordinary Differential Equations
Customary Notions of Continuity

Definition 5 (Continuity on a Set of Points (B)).

For normed spaces X, Y , a function f : A ⊂ X → Y is continuous on B if it is
continuous at any point x0 ∈ B. A function is simply continuous if B = A.

Dropping some of the notation,

Definition 6 (Uniform Continuity on a Set of Points (B)).

f : A ⊂ X → Y is uniformly continuous on B if for any  > 0, there exists a
δ > 0 such that for x, y ∈ B, kx − yk < δ implies kf (x) − f (y)k < . A
function is simply uniformly continuous if B = A.

Example: f (x) = x3 is uniformly continuous on B = [0, 1], but not B = R

f 0 (x) = 3x2 < 3 for x ∈ [0, 1]

hence |f (x) − f (y)| ≤ 3|x − y|. So given  > 0, choose δ < 13 .

M. Peet Lecture 15: 11 / 44

Lipschitz Continuity
A Quantitative Notion of Continuity

Definition 7 (Lipschitz Continuity).

The function f is Lipschitz continuous on X if there exists some L > 0 such
that
kf (x) − f (y)k ≤ Lkx − yk for any x, y ∈ X.
The constant L is referred to as the Lipschitz constant for f on X.

Definition 8 (Local Lipschitz Continuity).

The function f is Locally Lipschitz continuous on X if for every x ∈ X, there
exists a neighborhood, B of x such that f is Lipschitz continuous on B.

Definition 9.
The function f is Globally Lipschitz if it is Lipschitz on its entire domain.
Example: f (x) = x3 is Locally Lipschitz on [−1, 1] with L = 3.
• But f (x) = x3 is NOT Globally Lipschitz on R
• L is typically just a bound on the derivative.

M. Peet Lecture 15: 12 / 44

A Theorem on Existence of Solutions
Existence and Uniqueness

Let B(x0 , r) be the unit ball, centered at x0 of radius r.

Theorem 10 (A Typical Existence Theorem).

Suppose x0 ∈ Rn , f : Rn → Rn and there exist L, r such that for any
x, y ∈ B(x0 , r),
kf (x) − f (y)k ≤ Lkx − yk
and kf (x)k ≤ c for x ∈ B(x0 , r). Let tf < min{ L1 , rc }. Then there exists a
unique differentiable x : [0, tf ] 7→ Rn , such that x(0) = x0 , x(t) ∈ B(x0 , r) and
ẋ(t) = f (x(t)).

M. Peet Lecture 15: 13 / 44

 Examples on Existence of Solutions Lecture 1++, 2006

• Nonlinear Phenomena and Stability theory

Theorem 11 (A Typical Existence Theorem).

Nonlinear phenomena [Khalil Ch 3.1] ◮
◮ existence and uniqueness
◮ finite escape time
n Nonlinear
n Control
n Theory 2006 ◮ peaking
Suppose x0 ∈ R , f : R → R and there exist L, r suchLinear
that for any
system theory revisited ◮
Second order systems [Khalil Ch 2.4, 2.6] ◮
x, y ∈ B(x0 , r), periodic solutions / limit cycles ◮

kf (x) − f (y)k ≤ Lkx − yk Stability theory [Khalil Ch. 4]

Lyapunov Theory revisited
◮
◮
◮ exponential stability

and kf (x)k ≤ c for x ∈ B(x0 , r). Let tf < min{ L1 , rc }. Then

quadratic stability
there
time-varying exists a
systems
◮

invariant sets ◮

unique differentiable solution on interval [0, tf ]. center manifold theorem ◮

Existence problems of solutions Finite Escape Time

Recall:
2 Finite escape time of dx/dt = x2
ẋ(t) = x(t)
Example: The differential equation x(0) = x0 5

4.5
dx
= x2 , x(0) = x0
dt 4
has the solution
has the solution x0 3.5

x(t) = x 3

x(t) = 1 0
−
, x0 0≤tt < x1 2.5

x(t)
1−x t 0 0 2

Lets take r = 1, xFinite

0 =escape Then L =1 supx∈[0,2] |f 0 (x)| = 4.
1. time 1.5

1
t = f
c = supx∈[0,2] |f (x)| = 4. Then we xhave a solution for 0 0.5

0
0 1 2 3 4 5

tf < min{ L1 , rc } = min{ 41 , 14 } = 14 and where |x(t)| < 2 for Time t

t ∈ [0, tf ].
1
We can verify that the solution x(t)Problems
Uniqueness = 1−t < 34 for t < tfPrevious
. problem is like the water-tank problem in backw
time
√
Example: The equation ẋ = x, x(0) = 0 has many solutions:
M. Peet Lecture 15: (Substitute τ = −t in differential equation). 14 / 44
Examples on Existence of Solutions
Non-Uniqueness

Theorem 12 (A Typical Existence Theorem).

Suppose x0 ∈ Rn , f : Rn → Rn and there exist L, r such that for any
x, y ∈ B(x0 , r),
kf (x) − f (y)k ≤ Lkx − yk
and kf (x)k ≤ c for x ∈ B(x0 , r). Let tf < min{ L1 , rc }. Then there exists a
unique differentiable solution on interval [0, tf ].

Recall the system without a unique solution is

ẋ(t) = x(t)1/3 x(0) = 0

2
The problem here is that f 0 (x) = 13 x− 3 = 1
2 .
3x 3

1
L = sup |f 0 (x)| = sup | 2 |=∞
x∈[0,2] x∈[0,2] 3x 3
1
Since 0 = ∞. So there is no Lipschitz Bound.
M. Peet Lecture 15: 15 / 44
Typical Existence of Solutions Proof Approach
Step 1: Contraction Mapping Theorem

Theorem 13 (Contraction Mapping Principle).

Let (X, k·k) be a complete normed space and let P : X → X. Suppose there
exists a ρ < 1 such that

kP x − P yk ≤ ρkx − yk for all x, y ∈ X.

Then there is a unique x∗ ∈ X such that P x∗ = x∗ . Furthermore for y ∈ X,

define the sequence {xi } as x1 = y and xi = P xi−1 for i > 2. Then
limi→∞ xi = x∗ .
Some Observations:
• Proof of CM: Show that P k y is a Cauchy sequence for any y ∈ X.
• For existence of solutions, X is the space of solutions. The contraction is
Z t
(P x)(t) = x0 + f (x(s))ds,
0

and kxk := supt∈[0,tf ] kx(t)k.

M. Peet Lecture 15: 16 / 44
Ordinary Differential Equations
Contraction Mapping Theorem

Key Point: P x = x if and only if x is a solution!

Z t
(P x)(t) = x0 + f (x(s))ds,
0

Theorem 14 (Fundamental Theorem of Calculus).

Suppose x and f : Rn → Rn are continuous. Then the following are equivalent.
1. x is differentiable at any t ∈ [0, tf ] and

ẋ(t) = f (x(t)) at all t ∈ [0, tf ]

x(0) = x0

2. Z t
x(t) = x0 + f (x(s))ds for all t ∈ [0, tf ]
0

M. Peet Lecture 15: 17 / 44

Ordinary Differential Equations
Existence and Uniqueness

First recall what we are trying to prove:

Theorem 15 (A Typical Existence Theorem).

Suppose x0 ∈ Rn , f : Rn → Rn and there exist L, r such that for any
x, y ∈ B(x0 , r),
kf (x) − f (y)k ≤ Lkx − yk
and kf (x)k ≤ c for x ∈ B(x0 , r). Let tf < min{ L1 , rc }. Then there exists a
unique differentiable solution on interval [0, tf ].

We will show that P is a contraction where

Z t
(P x)(t) = x0 + f (x(s))ds t ∈ [0, Tf ]
0

if Tf is sufficiently small. Then P x = x has a solution!

M. Peet Lecture 15: 18 / 44

Proof of Existence Theorem
Proof.
For given x0 , define the complete normed space
B = {x ∈ C[0, tf ] : x(0) = x0 , x(t) ∈ B(x0 , r)} with norm supt∈[0,tf ] kx(t)k
(Proof of Completeness omitted). Recall we define P as
Z t
P x(t) = x0 + f (x(s))ds
0
Step 1: We first show that if x ∈ B, then P x ∈ B. Suppose x ∈ B. To see that
P x is continuous in t, we note that
Z t2 Z t2
kP x(t2 ) − P x(t1 )k = k f (x(s))dsk ≤ kf (x(s))kds ≤ c(t2 − t1 )
t1 t1

Which is the definition of uniform continuity. Clearly P x(0) = x0 .

Finally, we need to show (P x)(t) ∈ B(x0 , r). Since kf (x)k ≤ c when
x ∈ B(x0 , r), we have
Z t Z tf
sup kP x(t) − x0 k = sup k f (x(s))dsk ≤ kf (x(s))kds ≤ tf c < r
t∈[0,tf ] t∈[0,tf ] 0 0

where the final inequality is from the constraint tf < min{ L1 , rc }.

M. Peet Lecture 15: 19 / 44
Proof of Existence Theorem

Proof.
We have shown that P : B → B.
Step 2: Now show that P is a contraction. For any two x, y ∈ B, we have
Z t
kP x − P yk = sup k f (x(s)) − f (y(s))dsk
t∈[0,tf ] 0
Z t Z tf
≤ sup ( kf (x(s)) − f (y(s))kds) = kf (x(s)) − f (y(s))kds
t∈[0,tf ] 0 0
Z tf
≤L kx(s) − y(s)kds ≤ Ltf kx − yk < kx − yk
0

Where the last inequality comes from the constraint tf < min{ L1 , rc }. Thus P is
a contraction which means there exists a solution x ∈ B such that
ẋ = f (x(t))

M. Peet Lecture 15: 20 / 44

Picard Iteration
Constructing the Solution

This proof provides a way of actually constructing the solution.

Picard-Lindelöf Iteration:
• From the proof, unique solution of P x∗ = x∗ is a solution of ẋ∗ = f (x∗ ),
where Z t
(P x)(t) = x0 + f (x(s))ds
0
• From the contraction mapping theorem, the solution P x∗ = x∗ can be
found as
x∗ = lim P k z for ANY z ∈ B
k→∞

Note that this existence theorem only guarantees existence on the interval
  h ri
1
t ∈ 0, or t ∈ 0,
L c
Where
• L is a Lipschitz constant for f in the ball of radius r centered at x0
• c is a bound for f in the ball of radius r centered at x0 .
Q: What does a Lyapunov function prove if there is no guaranteed solution?
A: Nothing. (Particularly problematic for PDEs)
M. Peet Lecture 15: 21 / 44
Illustration of Picard Iteration
This is a plot of Picard iterations for the solution map of ẋ = −x3 .
z(t, x) = 0; P z(t, x) = x; P 2 z(t, x) = x − tx3 ;
P 3 z(t, x) = x − tx3 + 3t2 x5 − 3t3 x7 + t4 x9

1.5

1.0

0.5

0.2 0.4 0.6 0.8 1.0

Figure: The solution for x0 = 1

Convergence is only guaranteed on interval t ∈ [0, .333].

M. Peet Lecture 15: 22 / 44
Extension of Existence Theorem
For Lyapunov Functions to work, we need to extend the solution Indefinitely!
Step 1: Prove that solutions exist for a fixed r (yielding L, c, and tf ).
Step 2: Use a Lyapunov function to show that there exists a δ < r, such that
any solution which begins in B(0, δ) will stay in B(0, r).
Step 3: Since kx(0)k < δ < r, a solution exists on [0, tf ]. Further, kx(tf )k ≤ r.
Step 4: Since kx(tf )k < r, a solution exists on [0, 2tf ]. Further, Step 2 proves
kx(2tf )k ≤ r.
Repeat Step 4 Indefinitely
For a given set W , define Wr := {x : kx − yk ≤ r, y ∈ W }.
Theorem 16 (Formal Solution Extension Theorem).
Suppose that there exists a domain D and L > 0 such that
kf (x) − f (y)k ≤ Lkx − yk for all x, y ∈ D ⊂ Rn and t > 0. Suppose there
exists a compact set W and r > 0 such that Wr ⊂ D. Furthermore suppose
that it has been proven that for x0 ∈ W , any solution to
ẋ(t) = f (x), x(0) = x0
must lie entirely in W . Then, for x0 ∈ W , there exists a unique solution x with
x(0) = x0 such that x lies entirely in W .
M. Peet Lecture 15: 23 / 44
Illustration of Extended Picard Iteration
We take the previous approximation to the solution map and extend it.
xHtL
1.0

0.9

0.8

0.7

0.6

0.5

t
0.2 0.4 0.6 0.8 1.0
Figure: The Solution map φ and the functions Gki for k = 1, 2, 3, 4, 5 and i = 1, 2, 3
for the system ẋ(t) = −x(t)3 . The interval of convergence of the Picard Iteration is
T = 31 .

M. Peet Lecture 15: 24 / 44

Stability Definitions
Whenever you are trying to prove stability, Please define your notion of stability!

Denote the set of bounded continuous functions by

C¯ := {x ∈ C : kx(t)k ≤ r, r ≥ 0} with norm
kxk = supt kx(t)k.

We define g : D → C¯ to be the solution map: g(x0 , t) if

∂
g(x0 , t) = f (g(x0 , t)) and g(x0 , 0) = x0 x0 ∈ D
∂t
Definition 17.
The system is locally Lyapunov stable on D where D contains an open
neighborhood of the origin if it defines a unique map g : D → C¯ which is
continuous at the origin (x0 = 0).

The system is locally Lyapunov stable on D if for any  > 0, there exists a δ()
such that for kx(0)k ≤ δ(), x(0) ⊂ D we have kx(t)k ≤  for all t ≥ 0

M. Peet Lecture 15: 25 / 44

Stability Definitions
Definition 18.
The system is globally Lyapunov stable if it defines a unique map g : Rn → C¯
which is continuous at the origin.
We define the subspace of bounded continuous functions which tend to the
origin by G := {x ∈ C¯ : limt→∞ x(t) = 0} with norm kxk = supt kx(t)k.
Definition 19.
The system is locally asymptotically stable on D where D contains an open
neighborhood of the origin if it defines a map g : D → G which is continuous at
the origin.

M. Peet Lecture 15: 26 / 44

Stability Definitions

Definition 20.
The system is globally asymptotically stable if it defines a map g : Rn → G
which is continuous at the origin.

Definition 21.
The system is locally exponentially stable on D if it defines a map g : D → G
where
kg(x, t)k ≤ Ke−γt kxk
for some positive constants K, γ > 0 and any x ∈ D.

Definition 22.
The system is globally exponentially stable if it defines a map g : Rn → G
where
kg(x, t)k ≤ Ke−γt kxk
for some positive constants K, γ > 0 and any x ∈ Rn .

M. Peet Lecture 15: 27 / 44

Lyapunov Theorem
Consider the system:
ẋ = f (x), f (0) = 0

Theorem 23.
Let V : D → R be a continuously differentiable function and D compact such
that

V (0) = 0
V (x) > 0 for x ∈ D, x 6= 0
T
∇V (x) f (x) ≤ 0 for x ∈ D.

• Then ẋ = f (x) is well-posed and locally Lyapunov stable on the largest

sublevel set Vγ = {x : V (x) ≤ γ} of V contained in D.
• Furthermore, if ∇V (x)T f (x) < 0 for x ∈ D, x 6= 0, then ẋ = f (x) is
locally asymptotically stable on the largest sublevel set
Vγ = {x : V (x) ≤ γ} contained in D.

M. Peet Lecture 15: 28 / 44

Lyapunov Theorem
Sublevel Set: For a given Lyapunov function V and positive constant γ, we
denote the set Vγ = {x : V (x) ≤ γ}.
Proof.
Existence: Denote the largest bounded sublevel set of V contained in the
interior of D by Vγ ∗ . Because V̇ (x(t)) = ∇V (x(t))T f (x(t)) ≤ 0, if x(0) ∈ Vγ ∗ ,
then x(t) ∈ Vγ ∗ for all t ≥ 0. Therefore since f is locally Lipschitz continuous
on the compact Vγ ∗ , by the extension theorem, there is a unique solution for
any initial condition x(0) ∈ Vγ ∗ .
Lyapunov Stability: Given any 0 > 0, choose  < 0 with B() ⊂ Vγ ∗ , choose
γi such that Vγi ⊂ B(). Now, choose δ > 0 such that B(δ) ⊂ Vγi . Then
B(δ) ⊂ Vγi ⊂ B() and hence if x(0) ∈ B(δ), we have
x(0) ∈ Vγi ⊂ B() ⊂ B(0 ).
Asymptotic Stability:
• V monotone decreasing implies limt→ V (x(t)) = 0.
• V (x) = 0 implies x = 0.
• Proof omitted.

M. Peet Lecture 15: 29 / 44

Lyapunov Theorem
Exponential Stability

Theorem 24.
Suppose there exists a continuously differentiable function V and constants
c1 , c2 , c3 > and radius r > 0 such that the following holds for all x ∈ B(r).

c1 kxkp ≤ V (x) ≤ c2 kxkp

∇V (x)T f (x) ≤ −c3 kxkp

Then ẋ = f (x) is exponentially stable on any ball contained in the largest

sublevel set contained in B(r).

Exponential Stability allows a quantitative prediction of system behavior.

M. Peet Lecture 15: 30 / 44

The Gronwall-Bellman Inequality
Exponential Stability

Lemma 25 (Gronwall-Bellman).
Let λ be continuous and µ be continuous and nonnegative. Let y be continuous
and satisfy for t ≤ b,
Z t
y(t) ≤ λ(t) + µ(s)y(s)ds.
a

Then Z Z 
t t
y(t) ≤ λ(t) + λ(s)µ(s) exp µ(τ )dτ ds
a s

If λ and µ are constants, then

y(t) ≤ λeµt .

For λ(t) = y0 , the condition is equivalent to

ẏ(t) ≤ µ(t)y(t), y(0) = y(t).
M. Peet Lecture 15: 31 / 44
Lyapunov Theorem
Exponential Stability

Proof.
We begin by noting that we already satisfy the conditions for existence,
uniqueness and asymptotic stability and that x(t) ∈ B(r).
Now, observe that
c3
V̇ (x(t)) ≤ −c3 kx(t)k2 ≤ − V (x(t))
c2
−c3
Which implies by the Gronwall-Bellman inequality (µ = c2 , λ = V (x(0)))
that c3
V (x(t)) ≤ V (x(0))e− c2 t .
Hence
1 1 c3 c2 c3
kx(t)k2 ≤ V (x(t)) ≤ e− c2 t V (x(0)) ≤ e− c2 t kx(0)k2 .
c1 c1 c1

M. Peet Lecture 15: 32 / 44

Lyapunov Theorem
Invariance

Sometimes, we want to prove convergence to a set. Recall

Vγ = {x , V (x) ≤ γ}

Definition 26.
A set, X, is Positively Invariant if x(0) ∈ X implies x(t) ∈ X for all t ≥ 0.

Theorem 27.
Suppose that there exists some continuously differentiable function V such that
V (x) > 0 for x ∈ D, x 6= 0
T
∇V (x) f (x) ≤ 0 for x ∈ D.

for all x ∈ D. Then for any γ such that the level set
X = {x : V (x) = γ} ⊂ D, we have that Vγ is positively invariant.
Furthermore, if ∇V (x)T f (x) ≤ 0 for x ∈ D, then for any δ such that
X ⊂ Vδ ⊂ D, we have that any trajectory starting in Vδ will approach the
sublevel set Vγ .
M. Peet Lecture 15: 33 / 44
Problem Statement 1: Global Lypunov Stability

Given:
• Vector field, f (x)

P P
Find: function V , scalars α, β such that i αi = .01, i βi = .01 and
X
V (x) ≥ αi (xT x)i for all x
i
X
V (x) ≤ βi (xT x)i for all x
i
∇V (x)T f (x) ≤ 0 for all x

Conclusion:
• Lyapunov stability for any x(0) ∈ Rn .
P
• Can replace V (x) ≤ i βi (xT x)i with V (0) = 0 if it is well-behaved.

M. Peet Lecture 15: 34 / 44

Examples of Lyapunov Functions

Mass-Spring: Pendulum:

c k g
ẍ = − ẋ − x ẋ2 = − sin x1 ẋ1 = x2
m m l
1 1 1
V (x) = mẋ2 + kx2 V (x) = (1 − cos x1 )gl + l2 x22
2 2 2

V̇ (x) = ẋ(−cẋ − kx) + kxẋ V̇ (x) = glx2 sin x1 − glx2 sin x1

2
= −cẋ − k ẋx + kxẋ =0

= −cẋ2 ≤ 0
M. Peet Lecture 15: 35 / 44
Problem Statement 2: Global Exponential Stability
Given:
• Vector field, f (x), exponent, p
P P
Find: function V , scalars α, β, δ such that i αi = .01, i βi = .01 and
X
V (x) ≥ αi x2p
i for all x
i
X
V (x) ≤ βi x2p
i for all x
i
∇V (x)T f (x) ≤ −δV (x) for all x

Conclusion:
• Exponential stability for any x(0) ∈ Rn .

Convergence Rate:
r
βmax δ
kx(t)k ≤ 2p
kx(0)ke− 2p t
αmin
M. Peet Lecture 15: 36 / 44
Problem Statement 2: Global Exponential Stability
Example

Consider: Attitude Dynamics of a rotating Spacecraft:

J1 ω̇1 = (J2 − J3 )ω2 ω3
J2 ω̇2 = (J3 − J1 )ω3 ω1
J3 ω̇3 = (J1 − J2 )ω1 ω2
What about:
V (x) = ω12 + ω22 + ω32 ?
 T  J2 −J3 
ω1 J1 ω2 ω3
 
∇V (x)T f (x) = ω2   J3J−J 2
1
ω3 ω1 
ω3 J1 −J2
J3 ω1 ω2
 
J2 − J3 J3 − J1 J1 − J2
= + + ω1 ω2 ω3
J1 J2 J3
 2 
J2 J3 − J32 J2 + J32 J1 − J12 J3 + J2 J12 − J22 J1
= ω1 ω2 ω3
J1 J2 J3
OK, maybe not. Try ui = −ki ωi .
M. Peet Lecture 15: 37 / 44
Problem Statement 3: Local Exponential Stability
Given:
• Vector field, f (x), exponent, p
• Ball of radius r, Br := {x ∈ Rn : xT x ≤ r 2 }
P P
Find: function V , scalars α, β, δ such that i αi = .01, i βi = .01 and
X
V (x) ≥ αi x2p
i for all x : xT x ≤ r2
i
X
V (x) ≤ βi x2p
i for all x : xT x ≤ r2
i
∇V (x)T f (x) ≤ −δV (x) for all x : xT x ≤ r2

Conclusion: A Domain of Attraction! of the origin

• Exponential stability for x(0) ∈ Vγ := {x : V (x) ≤ γ} if Vγ ⊂ Br .
Sub-Problem: Given, V , r,
max γ such that
γ

V (x) ≤ γ for all x ∈ {xT x ≤ r}

M. Peet Lecture 15: 38 / 44
Domain of Attraction
The van der Pol Oscillator
An oscillating circuit: (in reverse time)
ẋ = −y
3
ẏ = x + (x2 − 1)y
2

Choose:
V (x) = x2 + y 2 , r = 1 1

y
Derivative
 T   −1
T 2x y
∇V (x) f (x) =
2y −x − (x2 − 1)y −2

Domain−of−attraction
= −xy + xy + (x2 − 1)y 2 −3
−3 −2 −1 0 1 2 3
x

≤0 for x2 ≤ 1 Figure: The van der Pol oscillator

in reverse
Level Set:
Vγ=1 = {(x, y) : x2 + y 2 ≤ 1} = B1

So B1 = Vγ=1 is a Domain of Attraction!

M. Peet Lecture 15: 39 / 44

Problem Statement 4: Invariant Regions/Manifolds
Given:
• Vector field, f (x), exponent, p
• Ball of radius r, Br
P P
Find: function V , scalars α, β, δ such that i αi = .01, i βi = .01 and
X
V (x) ≥ αi x2p
i for all x : xT x ≥ r2
i
X
V (x) ≤ βi x2p
i for all x : xT x ≥ r2
i
∇V (x)T f (x) ≤ −δV (x) for all x : xT x ≥ r2

Conclusion:
• There exist a T such that x(t) ∈ {x : V (x) ≤ γ} for all t ≥ T if Br ⊂ Vγ .
Sub-Problem: Given, V , r,
min γ such that
γ

xT x ≥ r 2 for all x ∈ {V (x) ≥ γ}

M. Peet Lecture 15: 40 / 44
Nonlinear Dynamical Systems
Long-Range Weather Forecasting and the Lorentz Attractor

A model of atmospheric convection analyzed by E.N. Lorenz, Journal of

Atmospheric Sciences, 1963.
ẋ = σ(y − x) ẏ = rx − y − xz ż = xy − bz
M. Peet Lecture 15: 41 / 44
Problem Statement 5: Controller Synthesis (Local)
Suppose
ẋ(t) = f (x) + g(x)u(t) u(t) = k(x)
Given:
• Vector fields, f (x), g(x), exponent, p

P
Find: function functions, k, V , scalars α ≥ 0, β ≥ 0 such that i αi = .01,
P
i βi = .01 and
X
V (x) ≥ αi x2p
i for all x : xT x ≤ r2
i
X
V (x) ≤ βi x2p
i for all x : xT x ≤ r2
i
∇V (x)T f (x) + ∇V (x)T g(x)k(x) ≤ 0 for all x : xT x ≤ r2 (BILINEAR)

Conclusion:
• Controller u(u) = k(x(t)) stabilizes the system for x(0) ∈ {x : V (x) ≤ γ}
if Vγ ⊂ Br .
M. Peet Lecture 15: 42 / 44
Problem Statement 6: Output Feedback Controller
Synthesis (Global Exponential )
Suppose
ẋ(t) = f (x) + g(x)u(t) u(t) = k(y(t))
y(t) = h(x(t))
Given:
• Vector fields, f (x), g(x), h(x) exponent, p

Find:
P functionPfunctions, k, V , scalars α ≥ 0, β ≥ 0, δ > 0 such that
i α i = .01, i βi = .01 and X
V (x) ≥ αi x2p
i for all x
i
X
V (x) ≤ βi x2p
i for all x
i
∇V (x)T f (x) + ∇V (x)T g(x)k(h(x)) ≤ −δV (x) for all x
Conclusion:
• Controller u(t) = k(y(t)) exponentially stabilizes the system for any
x(0) ∈ Rn .
M. Peet Lecture 15: 43 / 44
How to Solve these Problems?
General Framework for solving these problems

Convex Optimization of Functions: Variables V ∈ C[Rn ] and γ ∈ R

max γ
V ,γ

subject to
V (x) − xT x ≥ 0 ∀x ∈ X
T T
∇V (x) f (x) + γx x ≤ 0 ∀x ∈ X

Going Forward
• Assume all functions are polynomials or rationals.
• Assume X := {x : gi (x) ≥ 0} (Semialgebraic)

M. Peet Lecture 15: 44 / 44