21.06.2014 Views

H-Matrix approximation for the operator exponential with applications

H-Matrix approximation for the operator exponential with applications

H-Matrix approximation for the operator exponential with applications

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

110 I. P. Gavrilyuk et al.<br />

Using Lemma 2.3 and setting in (2.14) α = t a k<br />

, we get<br />

(6.22) ‖η N (F, h)‖ ≤<br />

Mc √ [<br />

π<br />

2 √ k exp(−2πd/h)<br />

√ + k exp[−(N +1)2 h 2 a k t]<br />

at(1 − exp(−2πd/h)) ath(N +1)<br />

Equalising<br />

√<br />

<strong>the</strong> exponents by setting −2πd/h = −(N +1) 2 h 2 a/k, weget<br />

h = 3 2πdk<br />

a<br />

(N +1)−2/3 . Substituting this value into (6.22) leads to <strong>the</strong><br />

estimate<br />

‖η N (F, h)‖ ≤<br />

Mc √ [<br />

(6.23) π<br />

2 √ ke −s(N+1)2/3<br />

√<br />

at(1 − e −s(N+1) 2/3 ) +<br />

]<br />

.<br />

]<br />

ke −ts(N+1)2/3<br />

t(N +1) 1/3 3√ ,<br />

2πdka 2<br />

which completes our proof.<br />

1<br />

Note that our estimate implies ‖η N (F, h)‖ = O( ) as t → 0,<br />

t(N+1) 1/3<br />

1<br />

but numerical tests even indicate an error order O( ) as t → 0.<br />

(N+1) 1/3<br />

Acknowledgements. The authors would like to thank Prof. C. Lubich (University Tübingen)<br />

<strong>for</strong> valuable comments und useful suggestions. We appreciate Lars Grasedyck (University<br />

of Kiel) <strong>for</strong> providing <strong>the</strong> numerical computations.<br />

References<br />

1. H.Bateman, A.Erdelyi: Higher transcendental functions, Vol. 1, Mc Graw-Hill Book<br />

Company, Inc. (1953)<br />

2. R. Dautray, J.-L. Lions: Ma<strong>the</strong>matical analysis and numerical methods <strong>for</strong> science and<br />

technology, Vol. 5, Evolutions problems I, Springer (1992)<br />

3. Z. Gajić, M.T.J. Qureshi: Lyapunov matrix equation in system stability and control,<br />

Academic Press, San Diego (1995)<br />

4. I.P. Gavrilyuk: Strongly P-positive <strong>operator</strong>s and explicit representation of <strong>the</strong> solutions<br />

of initial value problems <strong>for</strong> second order differential equations in Banachspace. Journ.<br />

of Math. Analysis and Appl. 236 (1999), 327–349<br />

5. I.P. Gavrilyuk, V.L. Makarov: Exponentially convergent parallel discretization methods<br />

<strong>for</strong> <strong>the</strong> first order evolution equations, Preprint NTZ 12/2000, Universität Leipzig<br />

6. I.P. Gavrilyuk, V.L. Makarov: Explicit and approximate solutions of second order elliptic<br />

differential equations in Hilbert- and Banachspaces, Numer. Funct. Anal. Optimization<br />

20 (1999), 695–717<br />

7. I.P. Gavrilyuk, V.L. Makarov: Exact and approximate solutions of some <strong>operator</strong> equations<br />

based on <strong>the</strong> Cayley trans<strong>for</strong>m, Linear Algebra and its Applications 282 (1998),<br />

97–121<br />

8. I.P. Gavrilyuk, V.L. Makarov: Representation and <strong>approximation</strong> of <strong>the</strong> solution of an<br />

initial value problem <strong>for</strong> a first order differential eqation in Banachspace, Z. Anal.<br />

Anwend. 15 (1996), 495–527

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

Saved successfully!

Ooh no, something went wrong!