Scribd
Upload
55 views6 pages

Existence Theorems: The Following Optimal Control Problem Has A Globally Optimal Solution

The document summarizes two theorems on existence of optimal solutions to optimal control problems and provides examples of singular optimal control problems and two specific optimal control problems: fuel-optimal horizontal motion of a rocket and Goh's fishing problem. The theorems state that optimal solutions exist under certain conditions on the problem formulation, such as convexity and time-invariance. Singular optimal control occurs when the Hamiltonian is affine in the control and the control drops out, requiring additional necessary conditions.

Uploaded by

Mohammed Abd El Djalil DJEHAF
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
55 views6 pages

Existence Theorems: The Following Optimal Control Problem Has A Globally Optimal Solution

The document summarizes two theorems on existence of optimal solutions to optimal control problems and provides examples of singular optimal control problems and two specific optimal control problems: fuel-optimal horizontal motion of a rocket and Goh's fishing problem. The theorems state that optimal solutions exist under certain conditions on the problem formulation, such as convexity and time-invariance. Singular optimal control occurs when the Hamiltonian is affine in the control and the control drops out, requiring additional necessary conditions.

Uploaded by

Mohammed Abd El Djalil DJEHAF
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Chapter 2.

Existence Theorems

Theorem 1:
The following optimal control problem has a globally optimal solution:

ẋ(t) = f (x(t)) + B(x(t))u(t)


u(t) ∈ Rm : unconstrained
f (x) and B(x) continuously differentiable
x(ta ) = xa
x(tb ) : free, tb fixed
 tb  
J(u) = K(x(tb )) + L1 (x(t)) + L2 (u(t)) dt
ta

K(x) and L1 (x) convex and bounded from below


L2 (u) strictly convex and growing without bounds for u → ∞ .
Chapter 2.7

Theorem 2:
The following optimal control problem has a globally optimal solution:

ẋ(t) = f (x(t), u(t))


f continuously differentiable
u(t) ∈ Ω
Ω closed, convex, bounded, and time-invariant
x(ta ) = xa
x(tb ) : free, tb fixed
 tb
J(u) = K(x(tb )) + L(x(t), u(t)) dt
ta

K(x) and L(x, u) continuously differentiable.


Chapter 2.6

Singular Optimal Control


Hamiltonian function affine in u, u ∈ [umin , umax ] ⊂ R:

H = g(x(t), λ(t), λ0 ) + h(x(t), λ(t), λ0 )u(t)

Singularity condition:

h(xo (t), λo (t), λo0 ) ≡ 0 for t ∈ [t1 , t2 ] ⊆ [ta , tb ]

Necessary conditions for h ≡ 0:

h ≡ 0 =⇒ ḣ ≡ 0 =⇒ ḧ ≡ 0 =⇒ h(3) ≡ 0 =⇒ . . .

Continue differentiating until u appears explicitely in h(2k) , k ≥ 1.


Chapter 2.6

Necessary conditions for an optimal control to evolve along a singular arc in the
time interval [t1 , t2 ]:

h(2k) (xo (t), uo (t), λo (t), λo0 ) ≡ 0 and


h(2k−1) (xo (t), λo (t), λo0 ) ≡ 0 and
..
.
ḣ(xo (t), λo (t), λo0 ) ≡ 0 and
h(xo (t), λo (t), λo0 ) ≡ 0 .

Furthermore:

H = g(xo (t), λo (t), λo0 ) ≡ 0 if tb is free.


Problem 4: Fuel-optimal horizontal motion of a rocket
x1 = position, x2 = velocity, x3 = mass, u = thrust force delivered by the
engine, u ∈ Ω = [0, Fmax ].

Find u : [0, tb ] → [0, Fmax ], such that


⎡ ⎤ ⎡ ⎤
ẋ1 (t) x2 (t)

⎣ ẋ2 (t) ⎦ = ⎣ x 1(t) u(t) − 12 Aρcw x22 (t) ⎦


3

ẋ3 (t) −αu(t)


⎡ ⎤ ⎡ ⎤
x1 (0) sa
⎣ x2 (0) ⎦ = ⎣ va ⎦
x3 (0) ma
⎡ ⎤ ⎡ ⎤
x1 (tb ) sb
⎣ x2 (tb ) ⎦ = ⎣ vb ⎦
x3 (tb ) frei
 tb
J(u) = u(t) dt minimal.
0

Remark 1: sa , va , ma , sb , vb , Fmax , and tb are fixed.

Remark 2: J(u) = −x3 (tb ) is an equivalent cost criterion.


Problem 6: Goh’s fishing problem
x = number of fish in an ocean, u = catching rate, 0 ≤ u(t) ≤ U .

Goal: Maximize the total catch over a fixed time interval [0, tb ].

Find u : [0, tb ] → [0, U ], such that


x2 (t)
ẋ(t) = a x(t) − − u(t)
b
x(0) = xa
x(t) ≥ 0 for all t ∈ [0, tb ]
 tb
J(u) = − u(t) dt minimal.
0

Remark: a > 0, b > 0; xa , tb , and U are fixed.

You might also like