• Keine Ergebnisse gefunden

Exercise 1. Prove:

N/A
N/A
Protected

Academic year: 2021

Aktie "Exercise 1. Prove:"

Copied!
2
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

V4E2 - Numerical Simulation

Sommersemester 2017 Prof. Dr. J. Garcke

G. Byrenheid

Exercise sheet 4. To be handed in on Thursday, 18.05.2017.

Exercise 1. Prove:

(i) Let η(·) be a nonnegative, absolutely continuous function on [0, T ], which satisfies for a.e. 0 ≤ t ≤ T the differential inequality

η 0 (t) ≤ ω(t)η(t) + ψ(t)

where ω(t) and ψ(t) are nonnegative, integrable functions on [0, T ]. Then η(t) ≤ e R

0t

ω(s)ds h

η(0) + Z t

0

ψ(s)ds i

for all 0 ≤ t ≤ T .

(ii) Let φ(·) be a nonnegative, integrable function on [0, T ] which satisfies for a.e. 0 ≤ t ≤ T the integral inequality

φ(t) ≤ C 2 + Z t

0

C 1 φ(s)ds

for constants C 1 , C 2 > 0. Then

φ(t) ≤ C 2 (1 + C 1 te C

1

t )

for a.e. 0 ≤ t ≤ T . Hints: (i): consider ds d (η(s)e R

0s

ω(r)dr ) (ii): Use (i) to prove (ii).

(6 Punkte) Exercise 2. Prove: Assuming

(i) f : R n × A → R n is continuous such that

|f (x, α)| ≤ C, |f (x, α) − f (y, α)| ≤ C|x − y|, ∀x, y ∈ R n , α ∈ A (ii) ψ : R n → R

|ψ(x)| ≤ C, |ψ(x) − ψ(y)| ≤ C|x − y|, ∀x, y ∈ R n (iii) ` : R n × A → R

|`(x, α)| ≤ C, |`(x, α) − `(y, α)| ≤ C|x − y|, ∀x, y ∈ R n , α ∈ A.

Then

V (x, t) := inf

α∈A J x,t (α) = inf

α∈A

Z T t

`(y x,t (s), α(s))dt + ψ(y x,t (T )) (1) is bounded and Lipschitz continuous, i.e.

|V (y, s)| ≤ C 0 , |V (y, s) − V (y 0 , s 0 )| ≤ C 0 (|s − s 0 | + |y − y 0 |), ∀x, y ∈ R d , 0 ≤ s, s 0 ≤ T (As usual: y x,t (s) is the unique solution of the state dynamics generated by x, t and f ) Hints:

1

(2)

• consider 2 trajectories z(s) := y x,t (s, α ε ), z 0 (s) := y x

0

,t

0

(s, α ε ) with different initial data and where α ε is an almost optimal control for the initial data (x, t),

• prove |z 0 (t) − z(t)| ≤ C|t − t 0 | + |x 0 − x|,

• prove |z 0 (s) − z(s)| ≤ e CT (C|t − t 0 | + |x 0 − x|) using Exercise 1),

• estimate J x,tε ) − J x

0

,t

0

ε ).

(6 Punkte) Exercise 3. Given the initial data y(t 0 ) = x 0 . We define the function

h(t) :=

Z t t

0

`(y x

0

,t

0

(s), α(s))ds + V (y x

0

,t

0

(t), t).

where V is the value function known from (1). Prove:

(i) h is nondecreasing for any control α,

(ii) h is constant if and only if the control α is optimal.

(4 Punkte)

2

Referenzen

ÄHNLICHE DOKUMENTE

Note how the computation of Q requires a complete knowledge of f , l, and that - if it can help to clarify - this is the Q used in the remark for proof of theorem 44 added later..

The definition of viscosity solution for this problem was given in the lecture..

By solving the second one you can earn extra points.

V4E2 - Numerical Simulation. Sommersemester

Prove monotonicity and consistency for

[r]

Da das ECAL nur eine Dicke von etwa 1λ hat, werden im Mittel nur etwa 1 − 1/e = 63% der Pionenenergie im ECAL deponiert, sodass die Energie- messung verf¨ alscht wird.. Dies

ermittelte Alter von 36 000 Jahren Resultat eines F¨ alschungsskandals und wurde sp¨ ater auf