Nonlinear Control

Lecture # 25
State Feedback Stabilization

Nonlinear Control Lecture # 25 State Feedback Stabilization

Backstepping
η̇ = fa (η) + ga (η)ξ
ξ˙ = fb (η, ξ) + gb (η, ξ)u, gb 6= 0, η ∈ Rn , ξ, u ∈ R

Stabilize the origin using state feedback

View ξ as “virtual” control input to the system

η̇ = fa (η) + ga (η)ξ

Suppose there is ξ = φ(η) that stabilizes the origin of

η̇ = fa (η) + ga (η)φ(η)

∂Va
[fa (η) + ga (η)φ(η)] ≤ −W (η)
∂η

Nonlinear Control Lecture # 25 State Feedback Stabilization

 z = ξ − φ(η)

η̇ = [fa (η) + ga (η)φ(η)] + ga (η)z

ż = F (η, ξ) + gb (η, ξ)u

V (η, ξ) = Va (η) + 21 z 2 = Va (η) + 21 [ξ − φ(η)]2

∂Va ∂Va
V̇ = [fa (η) + ga (η)φ(η)] + ga (η)z
∂η ∂η
+zF (η, ξ) + zgb (η, ξ)u
 
∂Va
≤ −W (η) + z ga (η) + F (η, ξ) + gb (η, ξ)u
∂η

Nonlinear Control Lecture # 25 State Feedback Stabilization

  
∂Va
V̇ ≤ −W (η) + z ga (η) + F (η, ξ) + gb (η, ξ)u
∂η
 
1 ∂Va
u=− ga (η) + F (η, ξ) + kz , k > 0
gb (η, ξ) ∂η

V̇ ≤ −W (η) − kz 2

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.9

ẋ1 = x21 − x31 + x2 , ẋ2 = u

ẋ1 = x21 − x31 + x2

x2 = φ(x1 ) = −x21 − x1 ⇒ ẋ1 = −x1 − x31

Va (x1 ) = 21 x21 ⇒ V̇a = −x21 − x41 , ∀ x1 ∈ R

z2 = x2 − φ(x1 ) = x2 + x1 + x21

ẋ1 = −x1 − x31 + z2

ż2 = u + (1 + 2x1 )(−x1 − x31 + z2 )

Nonlinear Control Lecture # 25 State Feedback Stabilization

 V (x) = 21 x21 + 12 z22

V̇ = x1 (−x1 − x31 + z2 )
+ z2 [u + (1 + 2x1 )(−x1 − x31 + z2 )]

V̇ = −x21 − x41
+ z2 [x1 + (1 + 2x1 )(−x1 − x31 + z2 ) + u]

u = −x1 − (1 + 2x1 )(−x1 − x31 + z2 ) − z2

V̇ = −x21 − x41 − z22

The origin is globally asymptotically stable

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.10

ẋ1 = x21 − x31 + x2 , ẋ2 = x3 , ẋ3 = u

ẋ1 = x21 − x31 + x2 , ẋ2 = x3

def
x3 = −x1 − (1 + 2x1 )(−x1 − x31 + z2 ) − z2 = φ(x1 , x2 )

Va (x) = 21 x21 + 12 z22 , V̇a = −x21 − x41 − z22

z3 = x3 − φ(x1 , x2 )

ẋ1 = x21 − x31 + x2 , ẋ2 = φ(x1 , x2 ) + z3

∂φ 2 ∂φ
ż3 = u − (x1 − x31 + x2 ) − (φ + z3 )
∂x1 ∂x2

Nonlinear Control Lecture # 25 State Feedback Stabilization

 V = Va + 12 z32

∂Va 2 ∂Va
V̇ = (x1 − x31 + x2 ) + (z3 + φ)
∂x1 ∂x2
 
∂φ 2 3 ∂φ
+ z3 u − (x − x1 + x2 ) − (z3 + φ)
∂x1 1 ∂x2

V̇ = −x21 − x41 − (x2 + x1 + x21 )2

 
∂Va ∂φ 2 ∂φ
+z3 − (x1 − x31 + x2 ) − (z3 + φ) + u
∂x2 ∂x1 ∂x2

∂Va ∂φ 2 ∂φ
u=− + (x1 − x31 + x2 ) + (z3 + φ) − z3
∂x2 ∂x1 ∂x2

The origin is globally asymptotically stable

Nonlinear Control Lecture # 25 State Feedback Stabilization

Strict-Feedback Form

ẋ = f0 (x) + g0 (x)z1
ż1 = f1 (x, z1 ) + g1 (x, z1 )z2
ż2 = f2 (x, z1 , z2 ) + g2 (x, z1 , z2 )z3
..
.
żk−1 = fk−1 (x, z1 , . . . , zk−1 ) + gk−1(x, z1 , . . . , zk−1 )zk
żk = fk (x, z1 , . . . , zk ) + gk (x, z1 , . . . , zk )u

gi (x, z1 , . . . , zi ) 6= 0 for 1 ≤ i ≤ k

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.12

ẋ = −x + x2 z, ż = u

ẋ = −x + x2 z

z = 0 ⇒ ẋ = −x, Va = 12 x2 ⇒ V̇a = −x2

V = 21 (x2 + z 2 )

V̇ = x(−x + x2 z) + zu = −x2 + z(x3 + u)

u = −x3 − kz, k > 0, ⇒ V̇ = −x2 − kz 2

Global stabilization

Compare with semiglobal stabilization in Example 9.7

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.13

ẋ = x2 − xz, ż = u

ẋ = x2 − xz
1
z = x + x2 ⇒ ẋ = −x3 , V0 (x) = x2 ⇒ V̇ = −x4
2
1
V = V0 + (z − x − x2 )2
2
V̇ = −x4 + (z − x − x2 )[−x2 + u − (1 + 2x)(x2 − xz)]
u = (1 + 2x)(x2 − xz) + x2 − k(z − x − x2 ), k > 0

V̇ = −x4 − k(z − x − x2 )2 Global stabilization

Nonlinear Control Lecture # 25 State Feedback Stabilization

Passivity-Based Control
ẋ = f (x, u), y = h(x), f (0, 0) = 0

∂V
uT y ≥ V̇ = f (x, u)
∂x

Theorem 9.1
If the system is
(1) passive with a radially unbounded positive definite
storage function and
(2) zero-state observable,
then the origin can be globally stabilized by

u = −φ(y), φ(0) = 0, y T φ(y) > 0 ∀ y 6= 0

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Proof

∂V
V̇ = f (x, −φ(y)) ≤ −y T φ(y) ≤ 0
∂x

V̇ (x(t)) ≡ 0 ⇒ y(t) ≡ 0 ⇒ u(t) ≡ 0 ⇒ x(t) ≡ 0

Apply the invariance principle
A given system may be made passive by
(1) Choice of output,
(2) Feedback,
or both

Nonlinear Control Lecture # 25 State Feedback Stabilization

Choice of Output
∂V
ẋ = f (x) + G(x)u, f (x) ≤ 0, ∀x
∂x
No output is defined. Choose the output as
 T
def ∂V
y = h(x) = G(x)
∂x

∂V ∂V
V̇ =
f (x) + G(x)u ≤ y T u
∂x ∂x
Check zero-state observability

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.14

ẋ1 = x2 , ẋ2 = −x31 + u

V (x) = 41 x41 + 12 x22

With u = 0 V̇ = x31 x2 − x2 x31 = 0

∂V ∂V
Take y = G= = x2
∂x ∂x2
Is it zero-state observable?

with u = 0, y(t) ≡ 0 ⇒ x(t) ≡ 0

u = −kx2 or u = −(2k/π) tan−1 (x2 ) (k > 0)

Nonlinear Control Lecture # 25 State Feedback Stabilization

Feedback Passivation
Definition
The system

ẋ = f (x) + G(x)u, y = h(x) (∗)

is equivalent to a passive system if ∃ u = α(x) + β(x)v such

that

ẋ = f (x) + G(x)α(x) + G(x)β(x)v, y = h(x)

is passive

Theorem [20]
The system (*) is locally equivalent to a passive system (with
a positive definite storage function) if it has relative degree
one at x = 0 and the zero dynamics have a stable equilibrium
point at the origin with a positive definite Lyapunov function

Nonlinear Control Lecture # 25 State Feedback Stabilization

 Example 9.15 (m-link Robot Manipulator)

M(q)q̈ + C(q, q̇)q̇ + D q̇ + g(q) = u

M = M T > 0, (Ṁ − 2C)T = −(Ṁ − 2C), D = D T ≥ 0

Stabilize the system at q = qr
e = q − qr , ė = q̇

M(q)ë + C(q, q̇)ė + D ė + g(q) = u

(e = 0, ė = 0) is not an open-loop equilibrium point

u = g(q) − Kp e + v, (Kp = KpT > 0)

M(q)ë + C(q, q̇)ė + D ė + Kp e = v

Nonlinear Control Lecture # 25 State Feedback Stabilization

 M(q)ë + C(q, q̇)ė + D ė + Kp e = v

V = 12 ėT M(q)ė + 21 eT Kp e

V̇ = 21 ėT (Ṁ − 2C)ė − ėT D ė − ėT Kp e + ėT v + eT Kp ė ≤ ėT v

y = ė
Is it zero-state observable? Set v = 0
ė(t) ≡ 0 ⇒ ë(t) ≡ 0 ⇒ Kp e(t) ≡ 0 ⇒ e(t) ≡ 0

v = −φ(ė), [φ(0) = 0, ėT φ(ė) > 0, ∀ė 6= 0]

u = g(q) − Kp e − φ(ė)

Special case: u = g(q) − Kp e − Kd ė, Kd = KdT > 0

Nonlinear Control Lecture # 25 State Feedback Stabilization