Wavelet based matrix compression for boundary integral equations on complex geometries

HTML
DOWNLOAD

Μέγεθος: px

Εμφάνιση ξεκινά από τη σελίδα:

Download "Wavelet based matrix compression for boundary integral equations on complex geometries"

Ωκεανός Μπουκουβαλαίοι
8 χρόνια πριν
Προβολές:

1 1 Wavelet based matrix compression for boundary integral equations on complex geometries Ulf Kähler Chemnitz University of Technology Workshop on Fast Boundary Element Methods in Industrial Applications (Hirschegg)

2 2 Overview Motivation - presentation of problem

3 2 Overview Motivation - presentation of problem Wavelet basis - stiffness matrix

4 2 Overview Motivation - presentation of problem Wavelet basis - stiffness matrix Wavelet construction

5 2 Overview Motivation - presentation of problem Wavelet basis - stiffness matrix Wavelet construction Computation of the stiffness matrix

6 2 Overview Motivation - presentation of problem Wavelet basis - stiffness matrix Wavelet construction Computation of the stiffness matrix Numerical results

7 3 Preliminaries Aρ = f on Γ = Ω R 2 A : H q (Γ) H q (Γ) (Aρ)(x) = k(x, y)ρ(y) Γ y Γ

8 4 single layer potential: q = 1 2, A = K K = 1 log y x ρ(y) Γ y 2π double layer potential: q = 0, A = K K = 1 2π Γ Γ n(y), y x y x 2 ρ(y) Γ y

9 5 Galerkin scheme Variational formulation: find ρ H q (Γ) : (Aρ, v) L 2 (Γ) = (f, v) L 2 (Γ) v H q (Γ) V N = span{φ 1,..., φ N } H q (Γ) A Φ ρ Φ = f Φ

10 6 Geometry Γ N - polygonial approximations of the surface Γ finest level is fixed diam(ω) < 1

11 7 ansatzfunctions: φ i (x) = 1 Γ i Γ, for x Γ i, 0, else

12 7 ansatzfunctions: φ i (x) = 1 Γ i Γ, for x Γ i, 0, else φ i (x(s)) = 13 s Γ i 1 Γ s 13 Γ Γ+ 1 i 1 3 Γ i Γ for x(s) Γ i 1, Γ i Γ for x(s) Γ i, 0 else

13 8 Motivation - presentation of problem classical single scale method

14 8 Motivation - presentation of problem classical single scale method advantages: reduction of the space dimension good convergence rates

15 8 Motivation - presentation of problem classical single scale method advantages: reduction of the space dimension good convergence rates disadvantages: densly populated stiffness matrix expensive quadrature rules

16 8 Motivation - presentation of problem classical single scale method advantages: reduction of the space dimension good convergence rates disadvantages: densly populated stiffness matrix expensive quadrature rules

17 Objectives 9

18 9 Objectives sparse stiffness matrix

19 9 Objectives sparse stiffness matrix low complexity of solving

20 Wavelet basis - stiffness matrix 10

21 10 Wavelet basis - stiffness matrix hierarchical wavelets:

22 10 Wavelet basis - stiffness matrix hierarchical wavelets: hierarchical structure

23 10 Wavelet basis - stiffness matrix hierarchical wavelets: hierarchical structure ψ i := N k=1 ω i,kφ k, ω i1 ω i2 = δ i1 i 2

24 10 Wavelet basis - stiffness matrix hierarchical wavelets: hierarchical structure ψ i := N k=1 ω i,kφ k, ω i1 ω i2 = δ i1 i 2 (ψ i, x α ) = 0, α < d ( x α = x α 1 1 xα 2 2 )

25 10 Wavelet basis - stiffness matrix hierarchical wavelets: hierarchical structure ψ i := N k=1 ω i,kφ k, ω i1 ω i2 = δ i1 i 2 (ψ i, x α ) = 0, α < d ( x α = x α 1 1 xα 2 2 ) Dahmen-Prößdorf-Schneider/ von Petersdorff-Schwab: A ψ is a quasi sparse matrix with O(N log(n)) entries.

26 11 idea: = Γ k(x, y)ψ k (x)ψ k (y) Γ x Γ y Γ (D α+β k)(x 0, y 0 ) (α,β) N 2 0 N 2 0 Γ ψ k(x)(x x 0 ) α Γ x α! Γ ψ k (y)(y y 0) β Γ y β!

27 11 idea: = Γ k(x, y)ψ k (x)ψ k (y) Γ x Γ y Γ (D α+β k)(x 0, y 0 ) (α,β) N 2 0 N 2 0 Γ ψ k(x)(x x 0 ) α Γ x α! D α+β k(x 0, y 0 ) ( 1 C x 0 y 0 Γ ψ k (y)(y y 0) β Γ y β! ) α+β+1 2q

28 Wavelet construction 12

29 12 Wavelet construction hierarchical structure coarsening cluster tree

30 12 Wavelet construction hierarchical structure coarsening cluster tree [ ] [ ] Φ ν,j 1 V ν,j 1 Ψ ν,j 1 0 = V ν,j 1 Φ ν,j 0

31 13 Let M ν,j 1 be the moment matrix of the cluster ν from level j 1 [ ] M ν,j 1 = x α Φ ν,j dx Γ α < d

32 13 Let M ν,j 1 be the moment matrix of the cluster ν from level j 1 [ ] M ν,j 1 = x α Φ ν,j dx Γ α < d SV D M ν,j 1 = UΣV = U [S, 0] [ V ν,j 1 0 V ν,j 1 0 ]

33 13 Let M ν,j 1 be the moment matrix of the cluster ν from level j 1 [ ] M ν,j 1 = x α Φ ν,j dx Γ α < d SV D M ν,j 1 = UΣV = U [S, 0] [ V ν,j 1 0 V ν,j 1 0 ] constant/linear ansatzfunctions Ψ is orthonormal/ Riesz-basis.

34 14 complexity of computing cluster tree and wavelets: O(N)

35 15 Computation of the stiffness matrix transformation matrix Ω Ψ,Φ := (ω i,k ) N i,k=1

36 15 Computation of the stiffness matrix transformation matrix Ω Ψ,Φ := (ω i,k ) N i,k=1 A Ψ ρ Ψ = Ω Ψ,Φ A Φ Ω Ψ,Φ Ω Ψ,Φ ρ Φ = Ω Ψ,Φ (f, φ i ) N i=1 = f Ψ

37 15 Computation of the stiffness matrix transformation matrix Ω Ψ,Φ := (ω i,k ) N i,k=1 A Ψ ρ Ψ = Ω Ψ,Φ A Φ Ω Ψ,Φ Ω Ψ,Φ ρ Φ = Ω Ψ,Φ (f, φ i ) N i=1 = f Ψ calculation of f Ψ in O(N) possible

38 16 Ω Ψ,Φ = ψ ν ψ ν1 1 0 ψ ν0 1 0 ψ ν ψ ν ψ ν ψ ν

39 16 Ω Ψ,Φ = ψ ν ψ ν1 1 0 ψ ν0 1 0 ψ ν ψ ν ψ ν ψ ν ψ ν2 1 ψ ν1 1 ψ ν0 1 ψ ν2 2.. ψ ν2 3 ψ ν1 2. ψ ν2 4..

40 16 Ω Ψ,Φ = ψ ν ψ ν1 1 0 ψ ν0 1 0 ψ ν ψ ν ψ ν ψ ν O(log(N)) columns ψ ν2 1 ψ ν1 1 ψ ν0 1 ψ ν2 2.. ψ ν2 3 ψ ν1 2. ψ ν2 4..

41 17 Using multipole method the multipole method

42 17 Using multipole method the multipole method iterative solving of A Φ ρ Φ = f Φ

43 17 Using multipole method the multipole method iterative solving of A Φ ρ Φ = f Φ fast matrix-vector product

44 17 Using multipole method the multipole method iterative solving of A Φ ρ Φ = f Φ fast matrix-vector product expansion of kernel

45 17 Using multipole method the multipole method iterative solving of A Φ ρ Φ = f Φ fast matrix-vector product expansion of kernel k(x, y) = (D α+β k)(x 0, y 0 ) (α,β) N 2 0 N 2 0 (x x 0 ) α (y y 0 ) β α! β!

46 low rank approximation of parts of A Φ 18

47 18 low rank approximation of parts of A Φ (A Φ i,j) i I,j J XkY

48 18 low rank approximation of parts of A Φ subdivision of A Φ (A Φ i,j) i I,j J XkY

49 18 low rank approximation of parts of A Φ subdivision of A Φ hierarchical matrix (A Φ i,j) i I,j J XkY

50 18 low rank approximation of parts of A Φ (A Φ i,j) i I,j J XkY subdivision of A Φ hierarchical matrix cluster-cluster interactions possible

51 18 low rank approximation of parts of A Φ (A Φ i,j) i I,j J XkY subdivision of A Φ hierarchical matrix cluster-cluster interactions possible complexity of a matrix-vector product: O(N log 2 (N))

52 18 low rank approximation of parts of A Φ (A Φ i,j) i I,j J XkY subdivision of A Φ hierarchical matrix cluster-cluster interactions possible complexity of a matrix-vector product: O(N log 2 (N))

53 application on wavelets 19

54 19 application on wavelets using fast matrix-vector products on ψ ν2 1 ψ ν1 1 ψ ν0 1 ψ ν2 2.. ψ ν2 3 ψ ν1 2. ψ ν2 4..

55 19 application on wavelets using fast matrix-vector products on ψ ν2 1 ψ ν1 1 ψ ν0 1 ψ ν2 2.. ψ ν2 3 ψ ν1 2. ψ ν2 4.. complexity of computing A Ψ : O(N log 3 (N))

56 19 application on wavelets using fast matrix-vector products on ψ ν2 1 ψ ν1 1 ψ ν0 1 ψ ν2 2.. ψ ν2 3 ψ ν1 2. ψ ν2 4.. complexity of computing A Ψ : O(N log 3 (N))

57 H.Harbecht: Error estimates for entries of A Ψ 20

58 Numerical results 21

59 Future research 22

60 22 Future research application to the 3D-case

61 22 Future research application to the 3D-case improved combination of multipole and wavelets

Σχετικά έγγραφα

Partial Differential Equations in Biology The boundary element method. March 26, 2013

Partial Differential Equations in Biology The boundary element method. March 26, 2013 The boundary element method March 26, 203 Introduction and notation The problem: u = f in D R d u = ϕ in Γ D u n = g on Γ N, where D = Γ D Γ N, Γ D Γ N = (possibly, Γ D = [Neumann problem] or Γ N = [Dirichlet