20.07.2013 Views

Estimation optimale du gradient du semi-groupe de la chaleur sur le ...

Estimation optimale du gradient du semi-groupe de la chaleur sur le ...

Estimation optimale du gradient du semi-groupe de la chaleur sur le ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

S. Albeverio, A. Kosyak / Journal of Functional Analysis 236 (2006) 634–681 653<br />

(see (40)) and theirs generalizations (see (42)). We cannot calcu<strong>la</strong>te explicitly<br />

<br />

n = max<br />

pq<br />

Mξn (t) 2 ,<br />

Ξ pq<br />

t∈R p<br />

but we are ab<strong>le</strong> by Lemmas B.1 and B.2 to obtain the estimation Ξ pq<br />

n >Ψ pq<br />

n ,<br />

Ψ pq<br />

n :=<br />

(M 12...p−1p<br />

12...p−1q (C(n)<br />

p,q)) 2 exp(−1)<br />

M 12...p−1<br />

12...p−1 (C(n) p )M 12...p<br />

12...p (C(n) p ) + p k=2<br />

ˆλk(A<br />

p<br />

k (C(n) p )) 2<br />

(see (47) and (48)). The crucial for proving (17) is Lemma 16 <strong>de</strong>aling with some inequalities<br />

involving the generalized characteristic polynomials. This <strong>le</strong>mma is proved in Appendix C.<br />

We use the notations of Lemma 8 (see Remark 9):<br />

Let<br />

S L m<br />

pq μB =<br />

∞<br />

n=q+1<br />

c (n)<br />

pp b(n)<br />

qq =<br />

∞<br />

n=q+1<br />

c (n)<br />

pp A q q(C (n)<br />

m )<br />

<strong>de</strong>t C (n)<br />

m<br />

=<br />

∞<br />

n=q+1<br />

cppA q q(Cm)<br />

<strong>de</strong>t Cm<br />

λ = (λk) m k=1 ∈ Rm , ˆλ = (ˆλk) m k=1 , k−1<br />

ˆλ1 = 0, ˆλk = crr, 2 k m, (19)<br />

fq = e <br />

Ĉm =<br />

1rp

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!