You are on page 1of 44

Dynamic Control by Interconnection of

Port–Hamiltonian Systems

– p. 1/4
CbI: Controllers as Multiport Cyclo–passive Systems

Plant (Σ) and controller (Σc ), with states x ∈ Rn , ζ ∈ Rm , are cyclo–passive, that is,
∃H(·) : Rn → R, Hc : Rm → R, such that

Ḣ(·) = u> y(·) − d


Ḣc = u>
c yc − d c ,

where dRn → R+ , dc Rm → R+ are dissipations.

– p. 2/4
Adding the Energies via Power Preserving Interconnection

Interconnection subsystem (ΣI ) is power–preserving (lossless)

>
y(·) u + yc> uc = y(·)
>
v.

A (very) particular case is negative feedback: u = −y c + v, uc = y(·)


Interconnected system satisfies

Ḣ(·) + Ḣc = v > y(·) − d − dc

⇒ H(·) (x) + Hc (ζ) is the new energy and d + dc new dissipation.

Problem:
Although Hc (ζ) is free, not clear how to affect x?
Energy functions “coupled" via the generation of invariant spaces.
Another, poorly understood, alternative is to make H c (x, ζ).

– p. 3/4
Invariant Function Method: Basic Idea

Principle: Restrict the motion to a subspace of (x, ζ)


Define the set

ξ Ωκ , {(x, ζ)|ζ = F (x) + κ},

(κ determined by the controllers ICs).


Then, in Ω0 ,

Hd (x) , H(x) + Hc [F (x)]

It can be shaped selecting Hc (ζ).


x (o) x2 Mathematical problem Let

C(x, ζ) , F (x) − ζ

x (t) Finding F (·) that renders Ω invariant ⇔


x1
d
C|C=0 ≡ 0, (P DE)
dt

– p. 4/4
Preliminaries: Invariant Sets (Manifolds)

Definition Consider the dynamical systems ẋ = f (x), x ∈ Rn . The set M ⊂ Rn is


invariant (w.r.t. the flow of the system) iff

x(0) ∈ M ⇒ x(t) ∈ M, ∀t ≥ 0.

4
Fact Given D : Rn → Rn−m , D ∈ C 1 . The set M = {x ∈ Rn |D(x) = 0} is invariant iff

Ḋ|x∈M = 0.

4
Furthermore, the level sets (foliation) Mκ = {x ∈ Rn |D(x) = κ}, κ ∈ R are invariant iff

Ḋ = 0 (♥).

Remarks
For LTI systems, ẋ = Ax, the set spanned by the (real) eigenvectors of A is invariant.
(♥) is a PDE: ∇Df = 0.
The functions D(x) are called conserved quantities (dynamical invariants).

– p. 5/4
CbI for PH Systems with Natural Output y

Given a PH system,

 ẋ = F (x)∇H(x) + g(x)u
Σ(u,y) ⇒ Ḣ ≤ u> y
 y = g > (x)∇H(x),

where we defined F (x) := J (x) − R(x), J = −J > , R = R> ≥ 0.


PH controller (nonlinear integrators), ζ ∈ R m

 ζ̇ = uc
Σc : ⇒ Ḣc = u>
c yc
 yc = ∇ζ Hc (ζ),

Standard negative feedback interconnection


      
 u 0 −1 y v
ΣI :  =  +  ⇒ Ḣ + Ḣc ≤ v > y
 uc 1 0 yc 0

For ease of presentation, and with loss of generality, we have taken ζ ∈ Rm .

– p. 6/4
Energy–Casimir Method

Proposition Assume, ∃ C : Rn → Rm such that the level sets Ωκ , {(x, ζ)|ζ = C(x) + κ}
are invariant, for all κ ∈ R. Then, for all Φ : Rm → R, the function

4
W (x, ζ) = H(x) + Hc (ζ) + Φ(C(x) − ζ).

satisfies
Ẇ ≤ v > y

That is, the system is cyclo–passive w.r.t. W (x, ζ), which (given C) can be shaped selecting
Hc and Φ.

Proof Invariance of Ωκ is equivalent to


d
ζ̇ − C(x) = 0.
dt

Hence,
Φ̇ = ∇Φ(C˙ − ζ̇) = 0 ⇒ Ẇ = Ḣ + Ḣc .



– p. 7/4
Caveat: Achieving Asymptotic Stability

All level sets

ξ Ωκ = {(x, ζ)|ζ = C(x) + κ},

κ ∈ R are invariant. In principle, we must


set

ζ(0) = ζ? + C(x(0)) − C(x? )

to ensure that the trajectory starts (and


remains) in Ωκ? , with
x (o) x2
4
κ? = ζ? − C(x? ),

x (t) that contains the desired equilibrium.


x1 More practical solutions are to estimate
κ or adding damping to the controller:
(Castanos, et al., Automatica’09).

– p. 8/4
Casimir Functions for PH Systems

Recalling the total energy function

W (x, ζ) = H(x) + Hc (ζ) + Φ(C(x) − ζ),

where H is given, C will be computed and Hc , Φ selected to shape W .


We want C to be independent of H and Hc – these are called Casimir functions.
The dynamics of the interconnected system is given by
      
ẋ F −g ∇H g
 =  +  v.
ζ̇ g> 0 ∇Hc 0

We are looking for C such that C˙ − ζ̇ = 0 for all H and Hc . Thus, we get the PDEs
 
h i F −g
(∇C)> −Im   = 0.
g> 0

Note that (∇C)> g = 0 ensures C˙ − ζ̇ = 0 even with v 6= 0.

– p. 9/4
Conditions for CbI

Proposition Assume there exists a vector function C : Rn → Rm such that


   
F> g
  ∇C =   (CbI − P DE)
g> 0

Then, for all functions Φ : Rm → R, the following cyclo–passivity inequality is satisfied

Ẇ ≤ v > y,

where the shaped storage function W : Rn × Rm → R is defined as

4
W (x, ζ) = H(x) + Hc (ζ) + Φ(C(x) − ζ).

Dissipation Obstacle This design not applicable to systems with pervasive dissipation!

– p. 10/4
The Dissipation Obstacle

Proposition If (CbI–PDE) admits a solution then

R∇x Φ(C(x) − ζ) = 0,

for all Φ : Rm → R. Consequently, energy cannot be shaped for coordinates that are
affected by physical damping.
Proof (CbI–PDE) ⇔

F > ∇C = g, g > ∇C = 0 ⇒ (∇C)> F > ∇C = 0


⇒ R∇C = 0

Proof completed with ∇x Φ = ∇C∇Φ. 


Remarks
OK for mechanical systems where dissipation enters in the momenta equations—that
need not be shaped.
Note that J ∇C = −g. Hence, (CbI-PDE) is equivalent to

F ∇C = −g, g > ∇C = 0.

– p. 11/4
A Power Balance Perspective of the Dissipation Obstacle

We proved that (CbI–PDE) ⇒ F ∇C = −g, R∇C = 0. Assuming (wlog) F invertible,

RF −1 g = 0,

which is a necessary condition for the existence of Casimirs.


Denote (·)(x∗ ) := (·)∗ , where x∗ is an equilibrium to be stabilized and assume
(CBI–PDE) holds. From ẋ = F ∇H + gu we have that

0 = F∗ ∇H∗ + g∗ u∗ ⇒ ∇H∗ = −F∗−1 g∗ u∗


⇒ R∇H∗ = −RF∗−1 g∗ u∗ ⇒ R∇H∗ = 0 ⇒ ∇> H∗ R∗ ∇H∗ = 0.

From the power balance equation

Ḣ = −∇H > R∇H + u> y

we see that, if the system admits a Casimir then the dissipated power at the
equilibrium should be zero, i.e., u>
∗ y∗ = 0. That is, we should be able to stabilize the
system extracting a finite amount of energy from the source.

– p. 12/4
Dissipation of PH Systems and Casimirs

Fact For PH systems Casimirs are necessarily “independent of the damped coordinates"

Proof Assume C : Rn → Rm verifies

∇C > [J(x) − R(x)] = 0

Post-multiplication by ∇C yields

∇C > [J(x) − R(x)]∇C = 0

Now, by transposition
∇C > [J(x) + R(x)]∇C = 0

Adding them up
∇C > R(x)∇C = 0 ⇔ R(x)∇C = 0

– p. 13/4
Can General Dynamic Controllers Overcome the Obstacle?

Consider general PH controller



 ζ̇ = [Jc (ζ) − Rc (ζ)] ∂H
∂ζ
c (ζ) + g (ζ)u
c c
Σc :
 yc = gc> (ζ) ∂H
∂ζ
c
(ζ)

Jc (ζ) = −Jc> (ζ), Rc (ζ) = R>


c (ζ) ≥ 0, and standard feedback interconnection, i.e.,

u = −yc , uc = y.

Closed–loop is still pH:


    
∂H
ẋ J(x) − R(x) −g(x)gc> (ζ) ∂x
(x)
 =  
∂Hc
ζ̇ gc (ζ)g > (x) Jc (ζ) − Rc (ζ) ∂ζ
(ζ)

with total energy H(x) + Hc (ζ).

– p. 14/4
cont’

Casimir functions Look for C(x, ζ) = C(x) − ζ, such that

d
C=0
dt

for all H(x), Hc (ζ), hence


 
h i J(x) − R(x) −g(x)gc> (ζ)
∇C > −Im   = 0. (P DE)
gc (ζ)g > (x) Jc (ζ) − Rc (ζ)
| {z }
 
 ẋ 
 
ζ̇

– p. 15/4
Proposition

(PDE) admits a solution if and only if C(x) satisfies

 >
∂C ∂C
(x) (x)
J(x) = Jc (ζ)
∂x ∂x
∂C
R(x) (x) = 0 (DO)
∂x
Rc (ζ) = 0
 >
∂C
(x) J(x) = gc (ζ)g > (x)
∂x

Dynamics reduced to Ω = {(x, ζ)|ζ = C(x)}


Σd : ẋ = [J(x) − R(x)] ∂H
∂x
d
(x)
with
Hd (x) = H(x) + Hc [C(x) + κ].

Remark Result holds for R(x) ≥ 0. How about pumping energy? (Jeltsema,
WLHMNLC’12).

– p. 16/4
Example without Pervasive Dissipation
1 ϕL
L

L
u C1 1 qC
C1

R2

Energy function
H(qC , φL ) = HE (qC ) + HM (φL ).

Assume 0 = arg min HM .


Co-energy variables ∇H = col(vC , iL )
     
q̇C 0 1 0
PH model  =  ∇H +  u
φ̇L −1 −R2 1

Equilibria: ∇φL H(φ̄L ) = 0 ⇒ φ̄L = 0 ⇒ E = {(qC∗ , 0)}


No need to shape HM and no dissipation in the coordinate to be shaped (q C )
Power balance: Ḣ = −R2 (∇φL H)2 + u∇φL H ⇒ dissipation zero at equilibria.

– p. 17/4
cont’d
1 2 1 2
W.l.o.g. assume linear elements: H(qC , φL ) = q
2C C
+ φ .
2L L

(CBI–PDE): F ∇C = −g, g > ∇C = 0,

ˆ C )),
(g > ∇C = 0 ⇒ C = C(q (F ∇C = −q ⇒ C = qC + k).

We can add to H + Hc an arbitrary function Φ(qC − ζ),

1 2 1 2
W (qC , φL , ζ) = qC + φ + Hc (ζ) + Φ(qC − ζ),
2C 2L L

To assign a minimum at (qC∗ , 0, ζ∗ ), for some ζ∗ , check


   
1 1
q + Φ0
C C C
+ Φ00 0 −Φ00
   
∇W = 

1
φ
L L
,
 ∇2 W = 
 0 1
L
0 .

Hc0 − Φ0 −Φ00 0 Hc00 + Φ00

Fixing ζ∗ = 0, it is solvable with a linear Φ and a quadratic Hc , yielding a linear


controller.

– p. 18/4
Example with Pervasive Dissipation

1
ϕL
L

L
u C1 R2 1 q
C1 C

 
−1
R2
1
Same energy function, but the dissipation has changed: J − R =  
−1 0

Equilibria: (∇qC H, ∇φL H) = (u∗ , R2 u∗ ) 6= (0, 0) ⇒ Dissipation in a coordinate to be


shaped (qC )
Not stabilizable via CbI!
We have RF −1 g = col( −1
R
, 0). Therefore, the necessary condition for the existence of
2
Casimirs, RF −1 g = 0, is not satisfied
Power balance: Ḣ = − R1 (∇qC H)2 + u∇φL H ⇒ dissipation not zero at equilibria.
2

– p. 19/4
Control by State–Modulated Interconnection with Σ (u,y)

Proposition Assume the PDE


 
g⊥ F >
  ∇C = 0, , (CbI SM − P DE)
g>

admits a solution for some vector function C : R n → Rm . The PH system Σ(u,y) with the
PH controller Σc and the state–modulated power–preserving interconnection subsystem
      
 u 0 −α(x) y v
ΣSM
I :  =  + ,
 uc α> (x) 0 yc 0

where α : Rn → Rm×m is defined as

α = −(g > g)−1 g > F ∇C.

Then, for all functions Φ : Rm → R, Ẇ ≤ v > y, where

4
W (x, ζ) = H(x) + Hc (ζ) + Φ(C(x) − ζ).

– p. 20/4
CbI with Alternative Port Variables: yPS

Proposition For all full–rank matrices Fd solution of the PDE


  h  i>
∇ Fd−1 F ∇H = ∇ Fd−1 F ∇H (P O − P DE)

verifying
Fd + Fd> ≤ 0 (SY M ),

there exists a storage function HPS such that



 ẋ = F ∇H + gu
Σ(u,yPS ) ⇒ ḢPS ≤ u> yPS .
 yPS = −g > Fd−> (F ∇H + gu)

The system may be written as


ẋ = Fd> ∇HPS + gu.

– p. 21/4
CbI with Σ(u,yPS ) Overcomes the Dissipation Obstacle

Proposition Assume (PO–PDE) admits a solution Fd verifying (SYM) and such that

Fd ∇C = −g (CbIPS − P DE)

for some vector function C : Rn → Rm . Consider the PH system Σ(u,yPS ) coupled with the
PH controller Σc through the power–preserving interconnection subsystem
      
 u 0 −1 yPS v
ΣPS
I :  =  + .
 uc 1 0 yc 0

Then, for all functions Φ : Rm → R, the following cyclo–passivity inequality is satisfied

ẆPS ≤ v > yPS ,

where
4
WPS (x, ζ) = HPS (x) + Hc (ζ) + Φ(C(x) − ζ),
R
with HPS = (Fd−1 F ∇H)dx.

Remark The condition g > ∇C = 0 is absent ⇒ no dissipation obstacle.

– p. 22/4
An Example of Stabilization via CbIPS

Consider the nonlinear RC circuit with x the


Sfrag replacements
capacitor charge, ẋ = i and H 0 = v

ẋ The Thevenin–Norton transformation


PSfrag replacements
yields the equivalent circuit:
R + ẋ
u C H 0 (x)

+
H0
u
R R R H 0 (x)
Power balance equation –

1 1
Ḣ = − (H 0 )2 + H 0 u.
R R A more physically sensible passive map
is voltage to current, i.e., u → ẋ,
One PH model

 1 1
1 1
 ẋ = −R H0 + u
 ẋ = −R H0 + R
u Σ(u,y PS ) : R
Σ(u,y) : 1 1
1
 y PS = −R H0 + u = ẋ,
 y = R
H 0, R

y has no physical interpretation.

– p. 23/4
cont’d

1 0 1
There is no C that satisfies the Casimir conditions: − R C = , 1 C0
R R
= 0, so this
problem can not be solved using CbI.
1
For any x? , the corresponding u? = (H 0 )? and y? = R
(H 0 )? yield the power

1
u? y ? = (H 0 )2? .
R

Hence, u? y? = 0 can only be satisfied for the zero equilibrium.


1
RF −1 g = − R 6= 0 implies that y 6= y PS , so we will try the CbI PS technique.
1 0 1
The Casimirs for CbI PS are solutions of F C 0 = −g, which becomes R
C = R
, so
C(x) = x will be a Casimir.
Look for functions Hc , Φ such that the function

W PS (x, ζ) = H(x) + Hc (ζ) + Φ (x − ζ)

has an isolated minimum at a given equilibrium point (x ? , 0). For simplicity, take
ζ? = 0.

– p. 24/4
cont’d

Compute

 Φ0? = −u?
(∇W PS )? = 0 ⇔ .
 (H 0 )? = −u?
c

The Hessian ∇2 W PS > 0 if and only if

00 00
Φ > −H
00 00
00 −H Φ
Hc > 00 00 .
H +Φ

Fix a quadratic
1
Hc = (ζ − Cc u? )2 ,
2Cc
with Cc ∈ R+ .
β 2
For Φ we propose Φ(z) = 2
z + γz,

βx? + γ = −u? .

– p. 25/4
cont’d

Evaluating minima conditions) at the equilibrium turns into

00
β > −H?
00
1 H β
> − 00? ,
Cc H? + β

00
from where it is easy to see that if H? > 0, we can take β = 0 and the equilibrium
(x? , 0) will be stable, for all Cc > 0, with Lyapunov function W PS (x, ζ) − (W PS )? ,
where
1
W PS (x, ζ) = H(x) + (ζ − Cc u? )2 − u? (x − ζ).
2Cc

The controller is given by



1
 ζ̇ = R
(−H 0 + u? − C1 ζ + v)
Σc + ΣPS
I : c
 u = u? − C1 ζ + v.
c

– p. 26/4
cont’d
PSfrag replacements

Physical realization

ζ̇ ẋ
v
+
1
Cc
ζ Cc + R +
– u C H 0 (x)
u? – –

Remark It can be implemented without distinction of “inputs" and “outputs".

– p. 27/4
Control by State–Modulated Interconnection with Σ (u,yPS )

Proposition Assume (PO–PDE) admits a solution Fd verifying (SYM) and such that

g ⊥ Fd ∇C = 0, SM
(CbIPS − P DE)

for some vector function C : Rn → Rm , where g ⊥ ∈ R(n−m)×n is a full rank left annihilator
of g, that is, g ⊥ g = 0 and rank g ⊥ = n − m. The PH system Σ(u,yPS ) with the PH controller
Σc and the state–modulated power–preserving interconnection subsystem
      
 u 0 −α(x) yPS v
ΣSM
I :  =  + ,
 uc α> (x) 0 yc 0

where α : Rn → Rm×m is defined as

α = −(g > g)−1 g > Fd ∇C.

Then, for all functions Φ : Rm → R, ẆPS ≤ v > yPS , where

4
WPS (x, ζ) = HPS (x) + Hc (ζ) + Φ(C(x) − ζ).

– p. 28/4
Generating New Cyclo–Passive Port Variables

Proposition The pH system



 ẋ = F (x)∇H(x) + g(x)u
Σ(u,yN )
 yN = (g(x) + 2T (x))> ∇H(x) + (D(x) + S(x))u,

where F (x) := J (x) − R(x), J = −J > , R = R> ≥ 0, S = S > , D = −D > and


 
R T
  ≥ 0,
T> S

defines a cyclo–passive pair (u, yN ), with storage function H(x). That is,

Ḣ ≤ u> yN

Question What are the new PDE’s in CbI?

– p. 29/4
Relationship with yPS

Proposition Assume F is full rank and set Fd = F > , which clearly verifies (PO–PDE) and
(SYM). Select

T = RF −1 g, S = g −> F −> RF −1 g, D = −g −> F −> J F −1 g,

which verifies (DC). Then,


yN = yPS .

Proof Replacing Fd = F > in yPS yields

yPS = −g > F −> F ∇H − g > F −> gu.

On the other hand, with the definitions of T, S and D above we get

g + 2T = (In + 2RF −1 )g = (F + 2R)F −1 g = (J + R)F −1 g = −F −> F −1 g,

and
D + S = g > F −> (R − J )F −1 g = −g > F −> F F −1 g = −g > F −> g.

Replacing the two identities above in yN completes the proof.

– p. 30/4
CbI With y(·) = yN

Assumption 1 Assume F is full rank. Define gd : Rn → Rn×m

gd := F −> (g + 2T ),

and S : Rn → Rm×m as
1 >
S := (gd g + g > gd ).
2
Assume there exists T : Rn → Rn×m such that

rank (g + 2T ) = m,

and  
R T
  ≥ 0, (DC)
T> S

hold.

Remark We cannot take T = 0, because in that case S = g > F −1 g ≤ 0.

– p. 31/4
Proposition

Consider the PH system Σ(u,yN ) verifying Assumption 1 with

1 >
D := (gd g − g > gd ),
2

the PH controller Σc and the state–modulated power–preserving interconnection subsystem


      
 u 0 −α(x) yN v
ΣSM
I :  =  + ,
 uc α> (x) 0 yc 0

with α : Rn → Rm×m is defined as

α = (gd> gd )−1 gd> ∇C,

C : Rn → Rm solutions of the PDEs

(g + 2T )⊥ F > ∇C = 0 (N EW − P DE).

All functions C(x) − xc are Casimir functions of the interconnected system.

– p. 32/4
Proof

The dynamics of the interconnected system is given by


    
ẋ F −gα ∇H
 =  .
ẋc α> (g + 2T )> −α> (D + S)α ∇Hc

Now, C˙ − ẋc = 0 for all H and Hc , if and only if


   
F> (g + 2T )α
  ∇C =  .
α> g > α> (S − D)α

Now, (NEW–PDE) and α are equivalent to gd α = ∇C. Replacing in the PDEs above yields

F > gd α = (g + 2T )α
α> gd> gα = α> (S + D)α.

Recalling gd the first equation is verified. The second identity also holds, using the definition
of S and D.

– p. 33/4
Summary of CbI

(CbI) and (CbI SM )


     
F −g g⊥ F
  ∇C =  ,   ∇C = 0.
g> 0 g>

SM )
(Basic CbIPS ) and (Basic CbIPS

F ∇C = −g, g ⊥ F ∇C = 0.

SM )
(CbIPS ) and (CbIPS
Fd ∇C = −g, g ⊥ Fd ∇C = 0.

(CbIN ) and (CbINSM )

F > ∇C = g + 2T, (g + 2T )⊥ F > ∇C = 0,

plus dissipation assumption.

– p. 34/4
cont’d

Implication diagram (from the point of view of solvability of the PDEs)

Notation: A → B means that the set of solutions of the PDEs of B is strictly larger than the
PSfrag replacements
one of A, consequently the set of plants to which B is applicable is also strictly larger.

CbI Basic CbIps CbIps

sm sm sm
CbI Basic CbIps CbIps
Question Where is CbINSM ?

– p. 35/4
Comparison of CbI and State–Feedback PBC: Applicability
   
F −g
(CbI)   ∇C =  
 
g> 0
g⊥ F
  (EBC)   ∇Ha = 0
g⊥ F g>
(CbI SM )   ∇C = 0
g> (Basic IDA)
(Basic CbIPS ) F ∇C = −g
g ⊥ F ∇Ha = 0
(CbIPS ) Fd ∇C = −g plus (PO–PDE)
(F ∇H = Fd ∇HPS )
SM )
(Basic CbIPS (PS)
⊥ g ⊥ Fd ∇Ha = 0
g F ∇C = 0
plus (PO–PDE)
(IDA)
SM )
(CbIPS
g ⊥ Fd ∇C = 0 g ⊥ Fd ∇Ha = g ⊥ (F − Fd )∇H

plus (PO-PDE).

– p. 36/4
ementsImplication Diagram

CbI Basic CbIps CbIps

sm sm sm
CbI Basic CbIps CbIps

EB Basic IDA PS IDA

– p. 37/4
State–Feedback PBC and CbI: Connections

Dynamic CbI
Dynamic feedback control u = −yc + v = −∇ζ Hc (ζ) + v,
ζ controllers state with energy Hc (ζ) free,
Generate Casimir functions, C, that make Ω = {(x, ζ)|ζ = C(x)} invariant
⇒ For arbitrary Φ
Ḣ(x) + Ḣc (ζ) + Φ(C(x) − ζ) ≤ v > y

State–feedback PBCFind a static state feedback, û(x) that ensures the following.
CbI with regulated sources (EB–PBC):

Ḣa = −û> (x)y.

IDA–PBC:
f (x) + g(x)û(x) = Fd (x)∇Hd (x).

Questions
Is there a connection between the two methods?
What happens if we restrict CbI to the invariant subspace Ω?
Is there an advantage of dynamic extension from minimum assignment viewpoint?

– p. 38/4
Restricting a Dynamic CbI Controller Yields an EB–PBC

Proposition Assume (CbI–PDE) admit a solution. Then, for all H c : Rm → R, the PH


system Σ(u,y) in closed–loop with the static state–feedback control u = û EB (x) + v, where

ûEB (x) = −∇C Hc (C(x)),

satisfies the cyclo–passivity inequality

d
Ḣ + Hc (C(x)) ≤ v > y.
dt

Furthermore,
d
Hc (C(x)) = −y > ûEB .
dt
4
Proof Define Ha (x) = Hc (C(x))

Ḣa = (∇C Hc (C))> (∇C)> (F ∇H + gu)


= (∇C Hc (C))> g > ∇H (⇐ F > ∇C = g, g > ∇C = 0)
= −û>
EB y (⇐ ûEB = −∇C Hc (C), y = g > ∇H).

– p. 39/4
Restricting a CbIPS Controller Yields an IDA–PBC

Proposition Assume the conditions for CbIPS are satisfied. Then, for all Hc : Rm → R, the
state–feedback controller
ûIDA (x) = −∇C Hc (C(x)),

ensures that the IDA–PBC matching condition

F ∇H + g ûIDA = Fd ∇Hd

4
is satisfied with Hd = HPS + Ha and Ha (x) = Hc (C(x)).

Proof
Conditions for CbIPS :
(PO–PDE) ⇔ F ∇H = Fd ∇HPS ,
(CbIPS –PDE) ⇔ Fd ∇C = −g.
Replacing in the matching equation yields

Fd (∇HPS − (∇C)ûIDA ) = Fd ∇Hd ⇔ ∇Ha = −(∇C)ûIDA ,

which is satisfied with Ha and ûIDA above. 

– p. 40/4
Stabilization via CbI and State–feedback PBC

We have concentrated our attention on the ability of the various PBCs to modify the
energy function, without particular concern to stabilization.
Stability will be ensured if a (desired) strict minimum is assigned to the total energy
function

Proposition In the single input case, the use of a dynamic extension does not provide any
additional freedom for minimum assignment to the corresponding static state–feedback
solutions.

Proof Define
4
W (x, ζ) = H(x) + Hc (ζ) + Φ(C(x) − ζ)
4
Hd (x) = H(x) + Hc (C(x)).

We will prove that

∇W? = 0 and ∇2 W? > 0 ⇒ (∇Hd )? = 0 and (∇2 Hd )? > 0.

– p. 41/4
cont’d

Compute
 
∇H + Φ0 ∇C
∇W =  , ∇Hd = ∇H + Hc0 ∇C.
Hc0 − Φ0

Now,
∇W? := ∇W (x? , ζ? ) = 0 ⇒ Φ0? = (Hc0 )? ⇒ ∇x W? = (∇Hd )? .

On the other hand,

∇ 2 Hd = ∇2 H + Hc0 ∇2 C + Hc00 ∇C∇C >


   
∇2 H + Φ 0 ∇2 C 0 ∇C∇C > −∇C
∇2 W =  00
+Φ  
0 Hc 00 −∇C > 1

– p. 42/4
cont’d

Now,
   >  
I ∇C 2 I ∇C ∇2 H + Φ0 ∇2 C + Hc00 ∇C∇C > ∗
 ∇ W   = 
0 |∇C|2 0 |∇C|2 ∗ ∗

From Sylvester’s Law of Inertia we have

∇2 W? > 0 ⇒ (∇2 H + Φ0 ∇2 C + Hc00 ∇C∇C > )? > 0


⇔ (∇2 Hd )? > 0 (⇐ Φ0? = (Hc0 )? )



– p. 43/4
Current Research

Is there a CbI version of IDA?


Relax Assumption 1 of new CbI.
 
R T
New CbI: What is the class of F, g such that there is T verifying   ≥ 0 with
T> S

1 > −1
S= [g (F + F −> )g] + T > F −1 g + g > F −> T.
2

The order of the dynamic extension is m. There are some advantages for increasing
their number.
Beyond simple nonlinear integrators?
CbI does not help for minimum assignment, but certainly has an impact on
performance and simplicity.
Can CbI be used—as an alternative to the current artificial “perturbation" or “delayed"
framework—to formulate the problem of control over networks?
Key question: Impact of the network topology on the ability to shape the energy, i.e., to
generate the Casimir functions.

– p. 44/4

You might also like