Preconditioning techniques in Chebyshev collocation method for elliptic equations

Size: px
Start display at page:

Download "Preconditioning techniques in Chebyshev collocation method for elliptic equations"

Transcription

1 Precondtonng technques n Chebyshev collocaton method for ellptc equatons Zh-We Fang Je Shen Ha-We Sun (n memory of late Professor Benyu Guo Abstract When one approxmates ellptc equatons by the spectral collocaton method on the Chebyshev-Gauss-Lobatto (CGL grd, the resultng coeffcent matrx s dense and llcondtoned. It s known that a good precondtoner, n the sense that the precondtoned system becomes well condtoned, can be constructed wth fnte dfference on the CGL grd. However, there s a lack of an effcent solver for ths precondtoner n mult-dmenson. A modfed precondtoner based on the approxmate nverse technque s constructed n ths paper. The computatonal cost of each teraton n solvng the precondtoned system s O(l y log, where, y are the grd szes n each drecton and l s a small nteger. umercal examples are gven to demonstrate the effcency of the proposed precondtoner. Key words: Chebyshev collocaton method, ellptc equaton, fnte-dfference precondtoner, approxmate nverse Mathematcs Subject Classfcaton: 35J25, 65F10, 6522, Introducton We consder a two-dmensonal separable ellptc equaton ( a(x u ( b(y u + c(xd(yu(x, y = f(x, y x x y y n Ω = ( 1, 1 2 (1.1 Department of Mathematcs, Unversty of Macau, Macao (fzw913@yeah.net. Department of Mathematcs, Purdue Unversty, West Lafayette, I , USA (shen7@purdue.edu. Ths author was supported n part by SF DMS and AFOSR FA Department of Mathematcs, Unversty of Macau, Macao (hsun@umac.mo. Ths author was supported n part by research grants MYRG FST from Unversty of Macau and 054/2015/A2 from FDCT of Macao. 1

2 wth homogeneous Drchlet boundary condtons u = 0 on Ω, where the coeffcent functons a(x, b(y, c(x, d(y and f(x, y are contnuous, and 0 < α a(x, b(y β n Ω for some postve constants α and β, and c(xd(y 0. A very effcent accurate method for obtanng approxmate soluton of the above boundaryvalue problem s the Chebyshev collocaton method [3, 4, 7, 9, 11, 12], whch uses the Lagrange nodal bass functons based on the Chebyshev collocaton ponts. However, due to the global nature of the Lagrange bass polynomals, the assocated lnear systems are dense and llcondtoned. Thus t becomes prohbtve to use an drect nverson method or an teratve method wthout precondtonng n the mult-dmensonal case, so t s mperatve to use an teratve method wth a good precondtoner. Fnte element/fnte dfference precondtoners have been wdely used snce the orgnal work by Orszag [9]. Haldenwang et al. [5] proved that the fnte dfference method based on the Chebyshev collocaton ponts n the one-dmensonal case leads to a good precondtoner. The propertes of the fnte element/fnte dfference precondtoners n the two-dmensonal case were rgorously establshed by Km and Parter [6, 7]. Thus, the applcaton of the Krylov subspace methods [1], such as generalzed mnmal resdual method (GMRES, leads to an teratve solvers convergng to the algebrac soluton wthn a constant number of steps that depends on the requred accuracy, but not on the number of unknowns. However, such a precondtoned method requres to solvng the precondtoner system,.e., solvng the fnte element/fnte dfference system on the spectral collocaton ponts. How to effcently apply the precondtoners s a challengng problem snce the grd formed by spectral collocaton ponts, contanng long-thn elements, s not shape-regular. We note that Shen et al. [13] developed a qute fnte element multgrd precondtoner for the second-order ellptc equatons. In ths paper, we seek to develop an approxmate precondtoner by explorng the algebrac propertes of the fnte dfference precondtoner. It s obvous that the two-dmensonal fnte dfference precondtoner s a nonsymmetrc block trdagonal matrx. Approxmatng ths matrx to construct a new effcent precondtoner s a natural dea. In [8], g and Pan proposed an approxmate nverse method to modfy crculant-plus-dagonal precondtoners for solvng Toepltz-plus-dagonal systems. Ther dea s to use crculant matrces to approxmate the nverson of Toepltz matrces and then combne the rows of these matrces together. As the resultng precondtoner s already of the nverted form, only matrx-vector multplcatons are requred n the precondtonng step. Recently, Pan et al. [10] also proposed approxmate nverse precondtoners for dagonal-tmes-toepltz matrces. The man purpose of ths paper s to propose and develop approxmate nverse precond- 2

3 toners for two-dmensonal ellptc operators, based on the modfcaton of the fnte-dfference operator dscretzed on the CGL grd. Frst, we use a scalng strategy to approxmate the fntedfference operator. Then we construct an approxmate nverse precondtoner to approxmate the nverse of scaled Laplacan-plus-dagonal matrces and combne them together row-by-row. In order to reduce the nfluence of the varous coeffcents, an nterpolaton method wth the egenvalues of Laplacan s utlzed. Specal nterpolaton nodes are chosen to mprove the accuracy of approxmaton. By usng of the dscrete sne transform (DST, the resultng precondtoner can be effcently mplemented wth O(l y log operatons, where the small nteger l s ndependent of and y. umercal examples are gven to demonstrate the effectveness of the proposed precondtoner. The paper s organzed as follows. In Secton 2, we ntroduce the Chebyshev collocaton method for the ellptc operator and the assocated fnte-dfference operator. In Secton 3, we construct the proposed precondtoners. umercal examples are gven to demonstrate the performance of the proposed precondtoner n Secton 4. In the fnal secton, concludng remarks are gven. 2 The Chebyshev-collocaton and the fnte-dfference operator In ths secton we recall the Chebyshev-collocaton method for the ellptc operator and the assocated fnte-dfference operator. Let P be the space of polynomals of degree less than or equal to. Let whch are the CGL ponts. x j = cos ( jπ, j = 0, 1,...,, 2.1 The one-dmensonal case Consder the one-dmensonal ellptc problems (a(xu (x + c(xu(x = f(x, x ( 1, 1; u(±1 = 0. (2.1 The Chebyshev-collocaton method for (2.1 s to fnd u X := {v P : v(±1 = 0} such that (au x=xk + c(x k u (x k = f(x k, k = 1, 2,..., 1. (2.2 Let {p j (x} j=0 be the Lagrange bass polynomals assocated wth {x j} j=0. Then, we can express u (x = u (x j p j (x. Denotng the Chebyshev dfferentaton matrx by G = j=0 3

4 (p j (x k k,j=0,1,...,, we have and u (x k = u (x k = u (x j p j(x k = j=0 j=0 u (x j p j (x k = (G kj u (x j, j=0 (G 2 kj u (x j. We lst below the formulas for the entres of G for arbtrary (cf. [2, 14]: Lemma 1 For each 1, let the rows and columns of the ( + 1 ( + 1 Chebyshev spectral dfferentaton matrx G be ndexed from 0 to. The entres of ths matrx are (G 00 = j=0, (G = , 6 (G jj = x j 2(1 x 2 j = 1,..., 1, j, (G kj = γ k ( 1 k+j, k j, k, j = 0,...,, γ j (x k x j where γ 0 = γ = 2 and γ k = 1 for k = 1,..., 1. Explct formulae for the entres of G 2 s also avalable n [3, 11]. By usng the dfferentaton matrx G, t costs O( 2 to compute the dervatve of u at all CGL ponts. However, ths process can be accelerated to O( log by expressng u usng the fast cosne transform [4, 12]. n Chebyshev polynomals and Set D a = dag(a(x 0, a(x 1,..., a(x, Dc = dag(c(x 1, c(x 2,..., c(x 1, A = {( G D a G k,j } 1 k,j 1, and A 1 = A + D c. The Chebyshev-collocaton scheme (2.2 reduces to the followng lnear system A 1 ū = f, where ū = (u (x 1,..., u (x 1 and f = (f(x 1,..., f(x 1. The matrx A s full and ll-condtoned. As proposed n [9], a good precondtoner for A 1 s to use a fnte-dfference operator on the CGL grd. Denote h j = x j x j 1 (j = 1,...,, h j = (x j+1 x j 1 /2 (j = 1,..., 1, and a k+1/2 = a((x k + x k+1 /2 (k = 0, 1,..., 1. We consder the followng fnte dfference approxmaton: (au x=x a 1/2 u(x 1 h h ( a 1/2 h h + a +1/2 u(x + a +1/2 u(x +1, = 1, 2,..., 1. h h +1 h h +1 4

5 Then, the precondtoner for A 1 based on the above approxmaton can be wrtten as follows: where (F j := B 1 := F + D c, a 1/2 h h, j = 1, a 1/2 h h + a +1/2 h h +1, j =, a +1/2 h h +1, j = + 1. (2.3 ote that B 1 s a nonsymmetrc trdagonal matrx. Invertng B 1 or solvng a lnear system wth B 1 as the coeffcent matrx requres about O( operatons. 2.2 The two-dmensonal case For the two-dmensonal ellptc equatons (1.1, the collocaton ponts are the tensor product of unvarate CGL nodes. Assume that and y are the number of the CGL ponts n each drecton respectvely. Then, the Chebyshev-collocaton method wll lead to a lnear system wth the matrx A 2 := I x A b y + A a I y + D c x D d y, where denotes the Kronecker product, A a s the matrx A defned n the last sub-secton, A b y s smlar to A a wth b(y replacng a(x, Dc x s the dagonal matrx D c defned n the last sub-secton, Dd y s smlar to D c where d(y s used nstead of c(x, and both I and I y are dentty matrces. The fnte-dfference operator assocated wth the two-dmensonal ellptc operator s defned as follows: B 2 := I x F b y + F a x I y + D c x D d y, (2.4 where F a x s defned by the formula (2.3, F b y s defned analogously wth a(x replaced by b(y. We remark that B 2 s a non-symmetrc block trdagonal wth trdagonal blocks matrx. Therefore, unlke B 1 n the one-dmensonal case, t s not an easy task to nvert B 2. In the followng secton, we shall construct a precondtoner based on the approxmate nverse strategy for B 2. 3 Constructon of the precondtoner For the nterest of smplcty, we frst dscuss the basc technques n the one-dmensonal case, and then these technques are utlzed to approxmate B 2. 5

6 3.1 Constructon n the one-dmensonal case We shall construct an effectve precondtoner n the one-dmensonal case through a sequence of approxmatons. Takng the structure of the matrx F nto consderaton, we frstly propose a scaled matrx as an approxmaton. Defne ( 1 a 1/2 t = + a +1/2, = 1, 2,..., 1, 2 h h +1 and T = dag(t 1, t 2,..., t 1. (3.1 The frst approxmaton s as follows: where and Then, we construct the frst precondtoner Denote F H T L T, (3.2 ( H = dag, 1 h1 1 h2 1,...,, (3.3 h L =.... R ( 1 ( 1. ( P 1 = H T L T + D c = H T (L + D c H 1 T 2 T. M = L + D c H 1 T 2. (3.5 Then P 1 = H T MT. We note that the matrces H, T are dagonal and can be easly to handle. Therefore, we only consder how to nvert the matrx M n (3.5. Defne K = L + h c t 2 I, = 1, 2,..., 1, (3.6 where c = c(x. Let e be the -th column of the dentty matrx. Accordng to the fact that e M = e K, we construct our precondtoner based on the followng approxmaton [8, 10] e M 1 e K 1. 6

7 Ths means that the -th row of the nverse of M s approxmated by the -th row of the nverse of K. Therefore, we propose our second precondtoner P 2 whose nverse s defned by ( 1 P 1 2 = T 1 =1 e e K 1 H 1 T 1. (3.7 We see from above that, to construct P 1 2, we need to compute the nverse of K ( = 1, 2,..., 1. Snce the matrx L can be dagonalzed n O( log operatons by the DST, the product K 1 v for any vector v can be computed n O( log operatons. Let S be the ( 1 ( 1 DST matrx. ote that S s symmetrc, orthogonal and ts (, j-th entry s gven by ( 2 πj sn, 1, j 1. Thus, the nverse of K can be computed by K 1 = S ( Λ + h 1 c t 2 I S, where Λ s a dagonal matrx whose entres are 2 2 cos( jπ, j = 1, 2,..., 1, the egenvalues of L. Hence, mplementng a precondtoner based on P 2 requres O( DST per teraton, whch s stll too expensve. In order to reduce the computatonal cost, we propose to use the nterpolaton method to construct a more effcent precondtoner. We choose a small number l(l of values {θ j } l {ξ = π } 1 =1, whch covers (most of the range of values of {ξ } 1 =1. Defne q (θ = 1 λ Λ (θ + w, θ (0, π, where λ Λ (θ = 2 2 cos θ and w = h c. Let t 2 p (θ = φ 1 (θq (θ 1 + φ 2 (θq (θ φ l (θq (θ l (3.8 be the pecewse lnear nterpolaton for q (θ based on the l ponts {θ j, q (θ j } l. We apply nterpolaton formula (3.8 to approxmate K 1 K 1 : S Φ j q (θ j S, = 1, 2,..., 1, (3.9 where Φ j = dag (φ j (ξ 1, φ j (ξ 2,..., φ j (ξ 1 are the nterpolaton coeffcent matrces. 7

8 Fnally,combnng the above consderaton, we defne our fnal precondtoner P 3 by 1 P3 1 =T 1 e e S Φ j q (θ j S H 1 =1 1 =T 1 =T 1 =T 1 =1 ( 1 =1 e e q (θ j S Φ j S H 1 T 1 e e q (θ j W j S Φ j S H 1 T 1, T 1 S Φ j S H 1 T 1 where W j = dag (q 1 (θ j, q 2 (θ j,..., q 1 (θ j are dagonal matrces. ow applyng P 1 3 to any vector requres about O(l log operatons whch s acceptable for a small number l. Snce the orgnal functon q (θ has weak sngulartes near θ = 0, the nterpolaton nodes should be slghtly dense near ξ Constructon n the two-dmensonal case In the followng, we apply smlar technques to construct a sequence of approxmate precondtoner for B 2 defned n (2.4. Frst, usng the approxmaton (3.2 n x-drecton, we defne our frst precondtoner by ˆP 1 =I x F b y + H x T x L x T x I y + D c x D d y ( =(H x T x I y H 1 T 2 F b y + L x I y + H 1 T 2 Dc x x D d y (T x I y (H x T x I y ˆM(T x I y, (3.10 where H x, T x and L x are defned as formulas (3.3, (3.1 and (3.4 respectvely, and ˆM denotes the mddle term n the above decomposton. ote that the frst and the last terms n the decomposton of ˆP1 n (3.10 are dagonal. precondtoner for the mddle term ˆM. Defne ˆK = h t 2 Therefore, we only need to construct a I x F b y + L x I y + h c t 2 I x D d y, = 1, 2,..., 1, where t, h and c are as n (3.6. Let e be the -th column of the dentty matrx I x. ote the fact that (e I y ˆM = (e I y ˆK. 8

9 As n one-dmensonal case, we use the followng approxmaton (e I y ˆM 1 (e I y ˆK 1, whch means that the (( 1( y th to ( y 1-th rows of the nverse of ˆM are approxmated by the (( 1( y 1+1-th to ( y 1-th rows of the nverse of ˆK. Therefore, smlarly to P 2 n (3.7 for the one-dmensonal case, we propose the precondtoner ˆP 2 whose nverse s defned by ˆP 1 2 = (T 1 I y We note that ˆK can be factored as ( x 1 (e e I y =1 ˆK 1 (T 1 H 1 I y. ˆK =(S x I y ( h t 2 (S x I y C (S x I y, I x F b y + Λ x I y + h c t 2 I x D d y (S x I y (3.11 where Λ x = dag(λ 1, λ 2,..., λ x 1 s a dagonal matrx whose dagonals are the egenvalues of L x, and C, whch s a block dagonal wth trdagonal blocks matrx, denotes the mddle factor n the factorzaton of ˆK. We remark that the nverse of the frst and the last term n (3.11, multplyng any vector, can be mplemented by the DST n O( y log operatons, whle the mddle term C can be nverted n O( y operatons;.e., ˆK 1 = (S x I y C 1 (S x I y. 1 evertheless, t s too expensve to calculate ˆP 2 snce we need to compute about nverses of ˆK ( = 1, 2,..., 1. In order to reduce the computatonal workload, as n the onedmensonal case, we propose to explot the nterpolaton method to construct the practcal precondtoner. Denote C = dag(c 1, C 2,..., C (x 1 where C k s an ( y 1 ( y 1 trdagonal matrx as follows, C k = h t 2 F b y + h c D t 2 d y + λ Λ ( ξ k I y, n whch λ Λ (θ = 2 2 cos θ and ξ k = kπ. We nvestgate ts nverse usng the nterpolaton method. Let ( h Q (θ t 2 F b y + h 1 c D d t 2 y + λ Λ (θi y, θ (0, π, be an ( y 1 ( y 1 matrx functon. Then we have C 1 k = Q ( ξ k. 9

10 We choose a small number l(l of values {θ j } l { ξ k = kπ } x 1 k=1. Let P (θ = φ 1 (θ Q (θ 1 + φ 2 (θ Q (θ φ l (θ Q (θ l (3.12 be the pecewse lnear nterpolaton for Q (θ based on the l matrces {θ j, Q (θ j } l. Thus, by nterpolaton formula (3.12 to approxmate C 1, we have C 1 k P ( ξ k = φ j ( ξ k Q (θ j, = 1, 2,..., 1, and C 1 ( dag φ j ( ξ 1, φ j ( ξ 2,..., φ j ( ξ x 1 Q (θ j = Φ j Q (θ j, where the dagonal matrx Φ j s as n (3.9. Fnally, we defne the practcal precondtoner ˆP 3 whose nverse s defned as follows, [ x 1 ( ] ˆP 3 1 =(T 1 I y (e e S I y Φ j Q (θ j (S x I y (T 1 H 1 I y =(T 1 I y =1 [ x 1 =1 ( x 1 =(T 1 I y =1 ( e e Q (Sx ] (θ j Φ j S x I y (T 1 H 1 I y e e Q (θ j (Sx Φ j S x I y =(T 1 I y W j (S x Φ j S x I y ( T 1 H 1 I y, (T 1 H 1 I y (3.13 where W j = dag( Q 1 (θ j, Q 2 (θ j,..., Q x 1(θ j, j = 1, 2,..., l, are block dagonal matrces n whch each block s the nverse of a trdagonal matrx. Therefore, Wj can be nverted wth ˆP 1 only O( y operatons. Applyng 3 to multply any vector requres O(l y log operatons whch s acceptable for a small number l. It s expected that as l, the number of nterpolaton nodes, ncreases, the number of teratons requred for convergence decreases. However, the cost of formng and applyng the precondtoner grows proportonally to l. Hence there s a trade-off to determne a sutable number of nterpolaton ponts. 4 umercal experments In ths secton, we carry out numercal experments to study the performance of the proposed precondtoner ˆP 3 n (3.13. We employ the precondtoned GMRES method to solve the 10

11 collocaton system. In all numercal experments, the stoppng crteron s r k 2 r 0 2 < 10 10, where r k s the resdual vector after k teratons and r 0 s the ntal resdual vector. All numercal experments are mplemented usng Matlab on a Dell Optplex 3020 wth the confguraton: Intel(R Core(TM CPU GHz and 8.00 GB of memory. We consder the two-dmensonal ellptc equaton (1.1 wth the source term f(x, y = 1. ote that no approxmaton s used n the y-drecton n dervng the precondtoner ˆP 3 whch ndcates that the coeffcent functon b(y wll not affect the convergence rate of the precondtoner. Therefore we set b(y = 1 n all the experments. On the other hand, the coeffcent functon c(xd(y s to add nformaton to the man dagonal of the system matrx. c(xd(y = 0 s chosen to demonstrate the bad condtonal cases. In consequence of the weak sngulartes of the orgnal functon, the nterpolaton nodes are selected as θ j = π ( 1 j 1 l 1, j = 1, 2,..., l, where x denotes the celng of x. We remark the every θ j should be dfferent, and we set θ r = θ r 1 + π/ once θ r θ r 1. For the purpose of comparsons, we take the average of each dagonal of the fnte dfference precondtoner, resultng n a block trdagonal Toepltz wth trdagonal Toepltz blocks structured precondtoner for the collocaton system whch s denoted by Ave-GMRES. The numercal results are lsted n Table 1-4, where ˆP 3 (l = 8, 10, 12, 14 denotes the GMRES method wth the precondtoner ˆP 3 wth l beng the number of nterpolaton nodes, Iter denotes the number of teratons requred to solve (1.1, CPU denotes the CPU tme n seconds for solvng the dscretzed system, and - means that the methods do not converge wthn 6000 teratons. Table 1: umercal results for a(x = b(y = 1. = y ˆP3 (l = 8 ˆP3 (l = 10 ˆP3 (l = 12 ˆP3 (l = 14 Ave-GMRES Iter CPU Iter CPU Iter CPU Iter CPU Iter CPU In Table 1, the numercal results are reported wth constant coeffcent a(x = 1. We see that the precondtoned GMRES methods exhbt excellent performance both n terms of teraton steps and CPU tme, and the teraton number only ncreases slghtly as the number 11

12 of grd ponts ncreases. The number of teratons decreases as expected whle the number of nterpolaton nodes ncreases. Table 2: umercal results for a(x = e x, b(y = 1. = y ˆP3 (l = 8 ˆP3 (l = 10 ˆP3 (l = 12 ˆP3 (l = 14 Ave-GMRES Iter CPU Iter CPU Iter CPU Iter CPU Iter CPU Table 2 and 3 lst the numercal results for non-constant coeffcents. The results wth coeffcent a(x = e x s tested n Table 2. We observe that l = 8 provde better results n terms of the teraton number and CPU tme. In Table 3, we lst the results wth a(x beng functons wth large varatons. We observe that the precondtoners s not senstve for problems wth coeffcents havng large varatons. Table 3: umercal results for a(x n large varaton case and b(y = 1 wth l = 8. a(x e 6x 10 3 cos(x cos(8πx + 10 = y Iter CPU Iter CPU Iter CPU In the last example, we examne the effectveness of the precondtoner to non-separable ellptc equatons: x ( a(x, y u ( b(x, y u = f(x, y n Ω = ( 1, 1 2. x y y For problems wth non-separable coeffcents, we buld the precondtoner by usng averages of the coeffcents as ā(x = a(x, ydy,

13 and b(y n the smlar way. Then our precondtoners could be appled to the relatve separable coeffcent systems. umercal results are gven for a(x, y = b(x, y = e (x+y +1 n Table 4. We observe that the teraton numbers are larger than the separable case, but are stll acceptable. Table 4: umercal results for a(x, y = b(x, y = e (x+y + 1. = y ˆP3 (l = 5 ˆP3 (l = 6 ˆP3 (l = 7 Iter CPU Iter CPU Iter CPU We remark that the current algorthm s based on the approxmaton n x drecton. However, problem (1.1 s symmetrc wth respect to x and y. The roles of x and y can be swtched to get an alternatve algorthm. When y <, the alternatve one would be less computatonal expensve. 5 Concludng remarks The man contrbuton of ths paper s to develop a precondtoner based on the approxmate nverses n Chebyshev collocaton method for two-dmensonal ellptc equatons. The complexty of the matrx-vector multplcaton of ˆP 3 s of O(l y log. It s shown numercally 1 that the precondtoned GMRES method for solvng these precondtoned collocaton systems converges very quckly. We only consdered two-dmensonal case n ths paper. But snce one drecton s approxmated n the two-dmensonal case, our strategy for constructng precondtoners can be easly extended to three-dmensonal cases. Indeed, n the three-dmensonal case, we can frst apply the approxmaton (3.2 to x and y drectons of the fnte dfference operator, resultng n a scaled tensor product. Usng the row-by-row approxmaton, we can obtan the second precondtoner, whch s of nverted form. Fnally, by defnng a two-dmensonal tensor functon, the nterpolaton method can be utlzed to construct a practcal precondtoner whch requres only O(l y z log y operatons. evertheless, even the numercal results show the effcency and fast convergence of the proposed method, the convergence of our algorthm has not been studed theoretcally, but wll be tackled n our future work. 13

14 References [1] M. Benz, Precondtonng technques for large lnear systems: A survey, J. Comput. Phys., 182(2002, pp [2] D. Gottleb, M. Y. hussan, and S. A. Orszag, Introducton: Theory and Applcatons of Spectral Methods, n R. G. Vogt, D. Gottleb, and M. Y. Hussan, eds., Spectral Methods for Partal Dfferental Equatons, SIAM, Phladelpha, [3] D. Gottleb and L. Lustman, The Dufort-Frankel Chebyshev method for parabolc ntal boundary value problems, Comput. Fluds, 11(1983, pp [4] D. Gottleb and S. A. Orszag, umercal Analyss of Spectral Methods: Theory and Applcatons, SIAM, [5] P. Haldenwang, G. Labrosse, S. Abboud, and M. DeVlle, Chebyshev 3-D spectral and 2-D pseudospectral solvers for the Helmholtz equaton, J. Comput. Phys, 55(1984, pp [6] S. Km and S. Parter, Precondtonng Chebyshev spectral collocaton method for ellptc partal dfferental equatons, SIAM J. umer. Anal., 33(1996, pp [7] S. Km and S. Parter, Precondtonng Chebyshev spectral collocaton by fnte-dfferences operators, SIAM J. umer. Anal., 34(1997, pp [8] M. g and J. Pan, Approxmate nverse crculant-plus-dagonal precondtoners for Toepltzplus-dagonal matrces, SIAM J. Sc. Comput., 32(2010, pp [9] S. A. Orszag, Spectral methods for problems n complex geometres, J. Comput. Phys., 37(1980, pp [10] J. Pan, R. Ke, M. g, and H. Sun, Precondtonng technques for dagonal-tmes-toepltz matrces n fractonal dffuson equatons, SIAM J. Sc. Comput., 36(2014, pp. A2698 A2719. [11] R. Peyret, Introducton to Spectral Methods, Von Karman Insttute, Rhode-St.-Genèse, Belgum, [12] J. Shen, T. Tang, and L. Wang, Spectral Methods: Algorthms, Analyss and Applcatons, Sprnger Seres n Computatonal Mathematcs, Sprnger, [13] J. Shen, F. Wang, and J. Xu, A fnte element multgrd precondtoner for Chebyshevcollocaton methods, Appl. umer. Math., 33(2000, pp

15 [14] J. A. C. Wedeman and S. C. Reddy, A MATLAB Dfferentaton Matrx Sute, ACM Trans. Math. Software,

PRECONDITIONING TECHNIQUES IN CHEBYSHEV COLLOCATION METHOD FOR ELLIPTIC EQUATIONS

PRECONDITIONING TECHNIQUES IN CHEBYSHEV COLLOCATION METHOD FOR ELLIPTIC EQUATIONS ITERATIOAL JOURAL OF UMERICAL AALYSIS AD MODELIG Volume 15 umber 1-2 Pages 277 287 c 2018 Insttute for Scentfc Computng and Informaton PRECODITIOIG TECHIQUES I CHEBYSHEV COLLOCATIO METHOD FOR ELLIPTIC

More information

Numerical Heat and Mass Transfer

Numerical Heat and Mass Transfer Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and

More information

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons

More information

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS IJRRAS 8 (3 September 011 www.arpapress.com/volumes/vol8issue3/ijrras_8_3_08.pdf NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS H.O. Bakodah Dept. of Mathematc

More information

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16 STAT 39: MATHEMATICAL COMPUTATIONS I FALL 218 LECTURE 16 1 why teratve methods f we have a lnear system Ax = b where A s very, very large but s ether sparse or structured (eg, banded, Toepltz, banded plus

More information

Convexity preserving interpolation by splines of arbitrary degree

Convexity preserving interpolation by splines of arbitrary degree Computer Scence Journal of Moldova, vol.18, no.1(52), 2010 Convexty preservng nterpolaton by splnes of arbtrary degree Igor Verlan Abstract In the present paper an algorthm of C 2 nterpolaton of dscrete

More information

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could

More information

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k)

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k) ISSN 1749-3889 (prnt), 1749-3897 (onlne) Internatonal Journal of Nonlnear Scence Vol.17(2014) No.2,pp.188-192 Modfed Block Jacob-Davdson Method for Solvng Large Sparse Egenproblems Hongy Mao, College of

More information

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS)

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS) Some Comments on Acceleratng Convergence of Iteratve Sequences Usng Drect Inverson of the Iteratve Subspace (DIIS) C. Davd Sherrll School of Chemstry and Bochemstry Georga Insttute of Technology May 1998

More information

Appendix B. The Finite Difference Scheme

Appendix B. The Finite Difference Scheme 140 APPENDIXES Appendx B. The Fnte Dfference Scheme In ths appendx we present numercal technques whch are used to approxmate solutons of system 3.1 3.3. A comprehensve treatment of theoretcal and mplementaton

More information

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence.

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence. Vector Norms Chapter 7 Iteratve Technques n Matrx Algebra Per-Olof Persson persson@berkeley.edu Department of Mathematcs Unversty of Calforna, Berkeley Math 128B Numercal Analyss Defnton A vector norm

More information

Deriving the X-Z Identity from Auxiliary Space Method

Deriving the X-Z Identity from Auxiliary Space Method Dervng the X-Z Identty from Auxlary Space Method Long Chen Department of Mathematcs, Unversty of Calforna at Irvne, Irvne, CA 92697 chenlong@math.uc.edu 1 Iteratve Methods In ths paper we dscuss teratve

More information

Salmon: Lectures on partial differential equations. Consider the general linear, second-order PDE in the form. ,x 2

Salmon: Lectures on partial differential equations. Consider the general linear, second-order PDE in the form. ,x 2 Salmon: Lectures on partal dfferental equatons 5. Classfcaton of second-order equatons There are general methods for classfyng hgher-order partal dfferental equatons. One s very general (applyng even to

More information

Lecture 12: Discrete Laplacian

Lecture 12: Discrete Laplacian Lecture 12: Dscrete Laplacan Scrbe: Tanye Lu Our goal s to come up wth a dscrete verson of Laplacan operator for trangulated surfaces, so that we can use t n practce to solve related problems We are mostly

More information

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 13

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 13 CME 30: NUMERICAL LINEAR ALGEBRA FALL 005/06 LECTURE 13 GENE H GOLUB 1 Iteratve Methods Very large problems (naturally sparse, from applcatons): teratve methods Structured matrces (even sometmes dense,

More information

The Finite Element Method: A Short Introduction

The Finite Element Method: A Short Introduction Te Fnte Element Metod: A Sort ntroducton Wat s FEM? Te Fnte Element Metod (FEM) ntroduced by engneers n late 50 s and 60 s s a numercal tecnque for solvng problems wc are descrbed by Ordnary Dfferental

More information

Inexact Newton Methods for Inverse Eigenvalue Problems

Inexact Newton Methods for Inverse Eigenvalue Problems Inexact Newton Methods for Inverse Egenvalue Problems Zheng-jan Ba Abstract In ths paper, we survey some of the latest development n usng nexact Newton-lke methods for solvng nverse egenvalue problems.

More information

Errors for Linear Systems

Errors for Linear Systems Errors for Lnear Systems When we solve a lnear system Ax b we often do not know A and b exactly, but have only approxmatons  and ˆb avalable. Then the best thng we can do s to solve ˆx ˆb exactly whch

More information

THE STURM-LIOUVILLE EIGENVALUE PROBLEM - A NUMERICAL SOLUTION USING THE CONTROL VOLUME METHOD

THE STURM-LIOUVILLE EIGENVALUE PROBLEM - A NUMERICAL SOLUTION USING THE CONTROL VOLUME METHOD Journal of Appled Mathematcs and Computatonal Mechancs 06, 5(), 7-36 www.amcm.pcz.pl p-iss 99-9965 DOI: 0.75/jamcm.06..4 e-iss 353-0588 THE STURM-LIOUVILLE EIGEVALUE PROBLEM - A UMERICAL SOLUTIO USIG THE

More information

Difference Equations

Difference Equations Dfference Equatons c Jan Vrbk 1 Bascs Suppose a sequence of numbers, say a 0,a 1,a,a 3,... s defned by a certan general relatonshp between, say, three consecutve values of the sequence, e.g. a + +3a +1

More information

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method Journal of Electromagnetc Analyss and Applcatons, 04, 6, 0-08 Publshed Onlne September 04 n ScRes. http://www.scrp.org/journal/jemaa http://dx.do.org/0.46/jemaa.04.6000 The Exact Formulaton of the Inverse

More information

A New Refinement of Jacobi Method for Solution of Linear System Equations AX=b

A New Refinement of Jacobi Method for Solution of Linear System Equations AX=b Int J Contemp Math Scences, Vol 3, 28, no 17, 819-827 A New Refnement of Jacob Method for Soluton of Lnear System Equatons AX=b F Naem Dafchah Department of Mathematcs, Faculty of Scences Unversty of Gulan,

More information

A Hybrid Variational Iteration Method for Blasius Equation

A Hybrid Variational Iteration Method for Blasius Equation Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method

More information

Report on Image warping

Report on Image warping Report on Image warpng Xuan Ne, Dec. 20, 2004 Ths document summarzed the algorthms of our mage warpng soluton for further study, and there s a detaled descrpton about the mplementaton of these algorthms.

More information

Application of B-Spline to Numerical Solution of a System of Singularly Perturbed Problems

Application of B-Spline to Numerical Solution of a System of Singularly Perturbed Problems Mathematca Aeterna, Vol. 1, 011, no. 06, 405 415 Applcaton of B-Splne to Numercal Soluton of a System of Sngularly Perturbed Problems Yogesh Gupta Department of Mathematcs Unted College of Engneerng &

More information

A FORMULA FOR COMPUTING INTEGER POWERS FOR ONE TYPE OF TRIDIAGONAL MATRIX

A FORMULA FOR COMPUTING INTEGER POWERS FOR ONE TYPE OF TRIDIAGONAL MATRIX Hacettepe Journal of Mathematcs and Statstcs Volume 393 0 35 33 FORMUL FOR COMPUTING INTEGER POWERS FOR ONE TYPE OF TRIDIGONL MTRIX H Kıyak I Gürses F Yılmaz and D Bozkurt Receved :08 :009 : ccepted 5

More information

1 GSW Iterative Techniques for y = Ax

1 GSW Iterative Techniques for y = Ax 1 for y = A I m gong to cheat here. here are a lot of teratve technques that can be used to solve the general case of a set of smultaneous equatons (wrtten n the matr form as y = A), but ths chapter sn

More information

Singular Value Decomposition: Theory and Applications

Singular Value Decomposition: Theory and Applications Sngular Value Decomposton: Theory and Applcatons Danel Khashab Sprng 2015 Last Update: March 2, 2015 1 Introducton A = UDV where columns of U and V are orthonormal and matrx D s dagonal wth postve real

More information

2.29 Numerical Fluid Mechanics Fall 2011 Lecture 12

2.29 Numerical Fluid Mechanics Fall 2011 Lecture 12 REVIEW Lecture 11: 2.29 Numercal Flud Mechancs Fall 2011 Lecture 12 End of (Lnear) Algebrac Systems Gradent Methods Krylov Subspace Methods Precondtonng of Ax=b FINITE DIFFERENCES Classfcaton of Partal

More information

= = = (a) Use the MATLAB command rref to solve the system. (b) Let A be the coefficient matrix and B be the right-hand side of the system.

= = = (a) Use the MATLAB command rref to solve the system. (b) Let A be the coefficient matrix and B be the right-hand side of the system. Chapter Matlab Exercses Chapter Matlab Exercses. Consder the lnear system of Example n Secton.. x x x y z y y z (a) Use the MATLAB command rref to solve the system. (b) Let A be the coeffcent matrx and

More information

EFFICIENT DOMAIN DECOMPOSITION METHOD FOR ACOUSTIC SCATTERING IN MULTI-LAYERED MEDIA

EFFICIENT DOMAIN DECOMPOSITION METHOD FOR ACOUSTIC SCATTERING IN MULTI-LAYERED MEDIA European Conference on Computatonal Flud Dynamcs ECCOMAS CFD 2006 P. Wesselng, E. Oñate and J. Péraux (Eds) c TU Delft, The Netherlands, 2006 EFFICIENT DOMAIN DECOMPOSITION METHOD FOR ACOUSTIC SCATTERING

More information

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:

More information

APPENDIX A Some Linear Algebra

APPENDIX A Some Linear Algebra APPENDIX A Some Lnear Algebra The collecton of m, n matrces A.1 Matrces a 1,1,..., a 1,n A = a m,1,..., a m,n wth real elements a,j s denoted by R m,n. If n = 1 then A s called a column vector. Smlarly,

More information

On a direct solver for linear least squares problems

On a direct solver for linear least squares problems ISSN 2066-6594 Ann. Acad. Rom. Sc. Ser. Math. Appl. Vol. 8, No. 2/2016 On a drect solver for lnear least squares problems Constantn Popa Abstract The Null Space (NS) algorthm s a drect solver for lnear

More information

Formulas for the Determinant

Formulas for the Determinant page 224 224 CHAPTER 3 Determnants e t te t e 2t 38 A = e t 2te t e 2t e t te t 2e 2t 39 If 123 A = 345, 456 compute the matrx product A adj(a) What can you conclude about det(a)? For Problems 40 43, use

More information

Robust Norm Equivalencies and Preconditioning

Robust Norm Equivalencies and Preconditioning Robust Norm Equvalences and Precondtonng Karl Scherer Insttut für Angewandte Mathematk, Unversty of Bonn, Wegelerstr. 6, 53115 Bonn, Germany Summary. In ths contrbuton we report on work done n contnuaton

More information

The Order Relation and Trace Inequalities for. Hermitian Operators

The Order Relation and Trace Inequalities for. Hermitian Operators Internatonal Mathematcal Forum, Vol 3, 08, no, 507-57 HIKARI Ltd, wwwm-hkarcom https://doorg/0988/mf088055 The Order Relaton and Trace Inequaltes for Hermtan Operators Y Huang School of Informaton Scence

More information

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity LINEAR REGRESSION ANALYSIS MODULE IX Lecture - 30 Multcollnearty Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur 2 Remedes for multcollnearty Varous technques have

More information

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng

More information

ρ some λ THE INVERSE POWER METHOD (or INVERSE ITERATION) , for , or (more usually) to

ρ some λ THE INVERSE POWER METHOD (or INVERSE ITERATION) , for , or (more usually) to THE INVERSE POWER METHOD (or INVERSE ITERATION) -- applcaton of the Power method to A some fxed constant ρ (whch s called a shft), x λ ρ If the egenpars of A are { ( λ, x ) } ( ), or (more usually) to,

More information

NUMERICAL DIFFERENTIATION

NUMERICAL DIFFERENTIATION NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the

More information

A property of the elementary symmetric functions

A property of the elementary symmetric functions Calcolo manuscrpt No. (wll be nserted by the edtor) A property of the elementary symmetrc functons A. Esnberg, G. Fedele Dp. Elettronca Informatca e Sstemstca, Unverstà degl Stud della Calabra, 87036,

More information

arxiv: v1 [math.ho] 18 May 2008

arxiv: v1 [math.ho] 18 May 2008 Recurrence Formulas for Fbonacc Sums Adlson J. V. Brandão, João L. Martns 2 arxv:0805.2707v [math.ho] 8 May 2008 Abstract. In ths artcle we present a new recurrence formula for a fnte sum nvolvng the Fbonacc

More information

Lecture 21: Numerical methods for pricing American type derivatives

Lecture 21: Numerical methods for pricing American type derivatives Lecture 21: Numercal methods for prcng Amercan type dervatves Xaoguang Wang STAT 598W Aprl 10th, 2014 (STAT 598W) Lecture 21 1 / 26 Outlne 1 Fnte Dfference Method Explct Method Penalty Method (STAT 598W)

More information

New Method for Solving Poisson Equation. on Irregular Domains

New Method for Solving Poisson Equation. on Irregular Domains Appled Mathematcal Scences Vol. 6 01 no. 8 369 380 New Method for Solvng Posson Equaton on Irregular Domans J. Izadan and N. Karamooz Department of Mathematcs Facult of Scences Mashhad BranchIslamc Azad

More information

arxiv: v1 [math.co] 12 Sep 2014

arxiv: v1 [math.co] 12 Sep 2014 arxv:1409.3707v1 [math.co] 12 Sep 2014 On the bnomal sums of Horadam sequence Nazmye Ylmaz and Necat Taskara Department of Mathematcs, Scence Faculty, Selcuk Unversty, 42075, Campus, Konya, Turkey March

More information

Relaxation Methods for Iterative Solution to Linear Systems of Equations

Relaxation Methods for Iterative Solution to Linear Systems of Equations Relaxaton Methods for Iteratve Soluton to Lnear Systems of Equatons Gerald Recktenwald Portland State Unversty Mechancal Engneerng Department gerry@pdx.edu Overvew Techncal topcs Basc Concepts Statonary

More information

Norms, Condition Numbers, Eigenvalues and Eigenvectors

Norms, Condition Numbers, Eigenvalues and Eigenvectors Norms, Condton Numbers, Egenvalues and Egenvectors 1 Norms A norm s a measure of the sze of a matrx or a vector For vectors the common norms are: N a 2 = ( x 2 1/2 the Eucldean Norm (1a b 1 = =1 N x (1b

More information

Module 3: Element Properties Lecture 1: Natural Coordinates

Module 3: Element Properties Lecture 1: Natural Coordinates Module 3: Element Propertes Lecture : Natural Coordnates Natural coordnate system s bascally a local coordnate system whch allows the specfcaton of a pont wthn the element by a set of dmensonless numbers

More information

Lecture 10: May 6, 2013

Lecture 10: May 6, 2013 TTIC/CMSC 31150 Mathematcal Toolkt Sprng 013 Madhur Tulsan Lecture 10: May 6, 013 Scrbe: Wenje Luo In today s lecture, we manly talked about random walk on graphs and ntroduce the concept of graph expander,

More information

Workshop: Approximating energies and wave functions Quantum aspects of physical chemistry

Workshop: Approximating energies and wave functions Quantum aspects of physical chemistry Workshop: Approxmatng energes and wave functons Quantum aspects of physcal chemstry http://quantum.bu.edu/pltl/6/6.pdf Last updated Thursday, November 7, 25 7:9:5-5: Copyrght 25 Dan Dll (dan@bu.edu) Department

More information

Perron Vectors of an Irreducible Nonnegative Interval Matrix

Perron Vectors of an Irreducible Nonnegative Interval Matrix Perron Vectors of an Irreducble Nonnegatve Interval Matrx Jr Rohn August 4 2005 Abstract As s well known an rreducble nonnegatve matrx possesses a unquely determned Perron vector. As the man result of

More information

FTCS Solution to the Heat Equation

FTCS Solution to the Heat Equation FTCS Soluton to the Heat Equaton ME 448/548 Notes Gerald Recktenwald Portland State Unversty Department of Mechancal Engneerng gerry@pdx.edu ME 448/548: FTCS Soluton to the Heat Equaton Overvew 1. Use

More information

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016 U.C. Berkeley CS94: Spectral Methods and Expanders Handout 8 Luca Trevsan February 7, 06 Lecture 8: Spectral Algorthms Wrap-up In whch we talk about even more generalzatons of Cheeger s nequaltes, and

More information

A new Approach for Solving Linear Ordinary Differential Equations

A new Approach for Solving Linear Ordinary Differential Equations , ISSN 974-57X (Onlne), ISSN 974-5718 (Prnt), Vol. ; Issue No. 1; Year 14, Copyrght 13-14 by CESER PUBLICATIONS A new Approach for Solvng Lnear Ordnary Dfferental Equatons Fawz Abdelwahd Department of

More information

Procedia Computer Science

Procedia Computer Science Avalable onlne at www.scencedrect.com Proceda Proceda Computer Computer Scence Scence 1 (01) 00 (009) 589 597 000 000 Proceda Computer Scence www.elsever.com/locate/proceda Internatonal Conference on Computatonal

More information

General viscosity iterative method for a sequence of quasi-nonexpansive mappings

General viscosity iterative method for a sequence of quasi-nonexpansive mappings Avalable onlne at www.tjnsa.com J. Nonlnear Sc. Appl. 9 (2016), 5672 5682 Research Artcle General vscosty teratve method for a sequence of quas-nonexpansve mappngs Cuje Zhang, Ynan Wang College of Scence,

More information

An efficient algorithm for multivariate Maclaurin Newton transformation

An efficient algorithm for multivariate Maclaurin Newton transformation Annales UMCS Informatca AI VIII, 2 2008) 5 14 DOI: 10.2478/v10065-008-0020-6 An effcent algorthm for multvarate Maclaurn Newton transformaton Joanna Kapusta Insttute of Mathematcs and Computer Scence,

More information

Lecture 13 APPROXIMATION OF SECOMD ORDER DERIVATIVES

Lecture 13 APPROXIMATION OF SECOMD ORDER DERIVATIVES COMPUTATIONAL FLUID DYNAMICS: FDM: Appromaton of Second Order Dervatves Lecture APPROXIMATION OF SECOMD ORDER DERIVATIVES. APPROXIMATION OF SECOND ORDER DERIVATIVES Second order dervatves appear n dffusve

More information

Digital Signal Processing

Digital Signal Processing Dgtal Sgnal Processng Dscrete-tme System Analyss Manar Mohasen Offce: F8 Emal: manar.subh@ut.ac.r School of IT Engneerng Revew of Precedent Class Contnuous Sgnal The value of the sgnal s avalable over

More information

ELASTIC WAVE PROPAGATION IN A CONTINUOUS MEDIUM

ELASTIC WAVE PROPAGATION IN A CONTINUOUS MEDIUM ELASTIC WAVE PROPAGATION IN A CONTINUOUS MEDIUM An elastc wave s a deformaton of the body that travels throughout the body n all drectons. We can examne the deformaton over a perod of tme by fxng our look

More information

Least squares cubic splines without B-splines S.K. Lucas

Least squares cubic splines without B-splines S.K. Lucas Least squares cubc splnes wthout B-splnes S.K. Lucas School of Mathematcs and Statstcs, Unversty of South Australa, Mawson Lakes SA 595 e-mal: stephen.lucas@unsa.edu.au Submtted to the Gazette of the Australan

More information

ME 501A Seminar in Engineering Analysis Page 1

ME 501A Seminar in Engineering Analysis Page 1 umercal Solutons of oundary-value Problems n Os ovember 7, 7 umercal Solutons of oundary- Value Problems n Os Larry aretto Mechancal ngneerng 5 Semnar n ngneerng nalyss ovember 7, 7 Outlne Revew stff equaton

More information

4DVAR, according to the name, is a four-dimensional variational method.

4DVAR, according to the name, is a four-dimensional variational method. 4D-Varatonal Data Assmlaton (4D-Var) 4DVAR, accordng to the name, s a four-dmensonal varatonal method. 4D-Var s actually a drect generalzaton of 3D-Var to handle observatons that are dstrbuted n tme. The

More information

1 Matrix representations of canonical matrices

1 Matrix representations of canonical matrices 1 Matrx representatons of canoncal matrces 2-d rotaton around the orgn: ( ) cos θ sn θ R 0 = sn θ cos θ 3-d rotaton around the x-axs: R x = 1 0 0 0 cos θ sn θ 0 sn θ cos θ 3-d rotaton around the y-axs:

More information

The Minimum Universal Cost Flow in an Infeasible Flow Network

The Minimum Universal Cost Flow in an Infeasible Flow Network Journal of Scences, Islamc Republc of Iran 17(2): 175-180 (2006) Unversty of Tehran, ISSN 1016-1104 http://jscencesutacr The Mnmum Unversal Cost Flow n an Infeasble Flow Network H Saleh Fathabad * M Bagheran

More information

Global Sensitivity. Tuesday 20 th February, 2018

Global Sensitivity. Tuesday 20 th February, 2018 Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values

More information

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

More information

One-sided finite-difference approximations suitable for use with Richardson extrapolation

One-sided finite-difference approximations suitable for use with Richardson extrapolation Journal of Computatonal Physcs 219 (2006) 13 20 Short note One-sded fnte-dfference approxmatons sutable for use wth Rchardson extrapolaton Kumar Rahul, S.N. Bhattacharyya * Department of Mechancal Engneerng,

More information

Structure and Drive Paul A. Jensen Copyright July 20, 2003

Structure and Drive Paul A. Jensen Copyright July 20, 2003 Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.

More information

Generalized Linear Methods

Generalized Linear Methods Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set

More information

Developing an Improved Shift-and-Invert Arnoldi Method

Developing an Improved Shift-and-Invert Arnoldi Method Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 93-9466 Vol. 5, Issue (June 00) pp. 67-80 (Prevously, Vol. 5, No. ) Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) Developng an

More information

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin Proceedngs of the 007 Wnter Smulaton Conference S G Henderson, B Bller, M-H Hseh, J Shortle, J D Tew, and R R Barton, eds LOW BIAS INTEGRATED PATH ESTIMATORS James M Calvn Department of Computer Scence

More information

Queueing Networks II Network Performance

Queueing Networks II Network Performance Queueng Networks II Network Performance Davd Tpper Assocate Professor Graduate Telecommuncatons and Networkng Program Unversty of Pttsburgh Sldes 6 Networks of Queues Many communcaton systems must be modeled

More information

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography CSc 6974 and ECSE 6966 Math. Tech. for Vson, Graphcs and Robotcs Lecture 21, Aprl 17, 2006 Estmatng A Plane Homography Overvew We contnue wth a dscusson of the major ssues, usng estmaton of plane projectve

More information

Finite Element Modelling of truss/cable structures

Finite Element Modelling of truss/cable structures Pet Schreurs Endhoven Unversty of echnology Department of Mechancal Engneerng Materals echnology November 3, 214 Fnte Element Modellng of truss/cable structures 1 Fnte Element Analyss of prestressed structures

More information

Speeding up Computation of Scalar Multiplication in Elliptic Curve Cryptosystem

Speeding up Computation of Scalar Multiplication in Elliptic Curve Cryptosystem H.K. Pathak et. al. / (IJCSE) Internatonal Journal on Computer Scence and Engneerng Speedng up Computaton of Scalar Multplcaton n Ellptc Curve Cryptosystem H. K. Pathak Manju Sangh S.o.S n Computer scence

More information

Grover s Algorithm + Quantum Zeno Effect + Vaidman

Grover s Algorithm + Quantum Zeno Effect + Vaidman Grover s Algorthm + Quantum Zeno Effect + Vadman CS 294-2 Bomb 10/12/04 Fall 2004 Lecture 11 Grover s algorthm Recall that Grover s algorthm for searchng over a space of sze wors as follows: consder the

More information

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS SECTION 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS 493 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS All the vector spaces you have studed thus far n the text are real vector spaces because the scalars

More information

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 ) Kangweon-Kyungk Math. Jour. 4 1996), No. 1, pp. 7 16 AN ITERATIVE ROW-ACTION METHOD FOR MULTICOMMODITY TRANSPORTATION PROBLEMS Yong Joon Ryang Abstract. The optmzaton problems wth quadratc constrants often

More information

DUE: WEDS FEB 21ST 2018

DUE: WEDS FEB 21ST 2018 HOMEWORK # 1: FINITE DIFFERENCES IN ONE DIMENSION DUE: WEDS FEB 21ST 2018 1. Theory Beam bendng s a classcal engneerng analyss. The tradtonal soluton technque makes smplfyng assumptons such as a constant

More information

EEE 241: Linear Systems

EEE 241: Linear Systems EEE : Lnear Systems Summary #: Backpropagaton BACKPROPAGATION The perceptron rule as well as the Wdrow Hoff learnng were desgned to tran sngle layer networks. They suffer from the same dsadvantage: they

More information

arxiv: v1 [quant-ph] 6 Sep 2007

arxiv: v1 [quant-ph] 6 Sep 2007 An Explct Constructon of Quantum Expanders Avraham Ben-Aroya Oded Schwartz Amnon Ta-Shma arxv:0709.0911v1 [quant-ph] 6 Sep 2007 Abstract Quantum expanders are a natural generalzaton of classcal expanders.

More information

COMPLEX NUMBERS AND QUADRATIC EQUATIONS

COMPLEX NUMBERS AND QUADRATIC EQUATIONS COMPLEX NUMBERS AND QUADRATIC EQUATIONS INTRODUCTION We know that x 0 for all x R e the square of a real number (whether postve, negatve or ero) s non-negatve Hence the equatons x, x, x + 7 0 etc are not

More information

Linear Approximation with Regularization and Moving Least Squares

Linear Approximation with Regularization and Moving Least Squares Lnear Approxmaton wth Regularzaton and Movng Least Squares Igor Grešovn May 007 Revson 4.6 (Revson : March 004). 5 4 3 0.5 3 3.5 4 Contents: Lnear Fttng...4. Weghted Least Squares n Functon Approxmaton...

More information

High resolution entropy stable scheme for shallow water equations

High resolution entropy stable scheme for shallow water equations Internatonal Symposum on Computers & Informatcs (ISCI 05) Hgh resoluton entropy stable scheme for shallow water equatons Xaohan Cheng,a, Yufeng Ne,b, Department of Appled Mathematcs, Northwestern Polytechncal

More information

Solving Fractional Nonlinear Fredholm Integro-differential Equations via Hybrid of Rationalized Haar Functions

Solving Fractional Nonlinear Fredholm Integro-differential Equations via Hybrid of Rationalized Haar Functions ISSN 746-7659 England UK Journal of Informaton and Computng Scence Vol. 9 No. 3 4 pp. 69-8 Solvng Fractonal Nonlnear Fredholm Integro-dfferental Equatons va Hybrd of Ratonalzed Haar Functons Yadollah Ordokhan

More information

Linear Algebra and its Applications

Linear Algebra and its Applications Lnear Algebra and ts Applcatons 4 (00) 5 56 Contents lsts avalable at ScenceDrect Lnear Algebra and ts Applcatons journal homepage: wwwelsevercom/locate/laa Notes on Hlbert and Cauchy matrces Mroslav Fedler

More information

Binomial transforms of the modified k-fibonacci-like sequence

Binomial transforms of the modified k-fibonacci-like sequence Internatonal Journal of Mathematcs and Computer Scence, 14(2019, no. 1, 47 59 M CS Bnomal transforms of the modfed k-fbonacc-lke sequence Youngwoo Kwon Department of mathematcs Korea Unversty Seoul, Republc

More information

2.3 Nilpotent endomorphisms

2.3 Nilpotent endomorphisms s a block dagonal matrx, wth A Mat dm U (C) In fact, we can assume that B = B 1 B k, wth B an ordered bass of U, and that A = [f U ] B, where f U : U U s the restrcton of f to U 40 23 Nlpotent endomorphsms

More information

Haar Wavelet Collocation Method for the Numerical Solution of Nonlinear Volterra-Fredholm-Hammerstein Integral Equations

Haar Wavelet Collocation Method for the Numerical Solution of Nonlinear Volterra-Fredholm-Hammerstein Integral Equations Global Journal of Pure and Appled Mathematcs. ISS 0973-768 Volume 3, umber 2 (207), pp. 463-474 Research Inda Publcatons http://www.rpublcaton.com Haar Wavelet Collocaton Method for the umercal Soluton

More information

Polynomial Regression Models

Polynomial Regression Models LINEAR REGRESSION ANALYSIS MODULE XII Lecture - 6 Polynomal Regresson Models Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur Test of sgnfcance To test the sgnfcance

More information

Problem Set 9 Solutions

Problem Set 9 Solutions Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem

More information

MATH 5630: Discrete Time-Space Model Hung Phan, UMass Lowell March 1, 2018

MATH 5630: Discrete Time-Space Model Hung Phan, UMass Lowell March 1, 2018 MATH 5630: Dscrete Tme-Space Model Hung Phan, UMass Lowell March, 08 Newton s Law of Coolng Consder the coolng of a well strred coffee so that the temperature does not depend on space Newton s law of collng

More information

Finding The Rightmost Eigenvalues of Large Sparse Non-Symmetric Parameterized Eigenvalue Problem

Finding The Rightmost Eigenvalues of Large Sparse Non-Symmetric Parameterized Eigenvalue Problem Fndng he Rghtmost Egenvalues of Large Sparse Non-Symmetrc Parameterzed Egenvalue Problem Mnghao Wu AMSC Program mwu@math.umd.edu Advsor: Professor Howard Elman Department of Computer Scences elman@cs.umd.edu

More information

Construction of Serendipity Shape Functions by Geometrical Probability

Construction of Serendipity Shape Functions by Geometrical Probability J. Basc. Appl. Sc. Res., ()56-56, 0 0, TextRoad Publcaton ISS 00-0 Journal of Basc and Appled Scentfc Research www.textroad.com Constructon of Serendpty Shape Functons by Geometrcal Probablty Kamal Al-Dawoud

More information

MMA and GCMMA two methods for nonlinear optimization

MMA and GCMMA two methods for nonlinear optimization MMA and GCMMA two methods for nonlnear optmzaton Krster Svanberg Optmzaton and Systems Theory, KTH, Stockholm, Sweden. krlle@math.kth.se Ths note descrbes the algorthms used n the author s 2007 mplementatons

More information

A MODIFIED METHOD FOR SOLVING SYSTEM OF NONLINEAR EQUATIONS

A MODIFIED METHOD FOR SOLVING SYSTEM OF NONLINEAR EQUATIONS Journal of Mathematcs and Statstcs 9 (1): 4-8, 1 ISSN 1549-644 1 Scence Publcatons do:1.844/jmssp.1.4.8 Publshed Onlne 9 (1) 1 (http://www.thescpub.com/jmss.toc) A MODIFIED METHOD FOR SOLVING SYSTEM OF

More information

Section 8.3 Polar Form of Complex Numbers

Section 8.3 Polar Form of Complex Numbers 80 Chapter 8 Secton 8 Polar Form of Complex Numbers From prevous classes, you may have encountered magnary numbers the square roots of negatve numbers and, more generally, complex numbers whch are the

More information

The lower and upper bounds on Perron root of nonnegative irreducible matrices

The lower and upper bounds on Perron root of nonnegative irreducible matrices Journal of Computatonal Appled Mathematcs 217 (2008) 259 267 wwwelsevercom/locate/cam The lower upper bounds on Perron root of nonnegatve rreducble matrces Guang-Xn Huang a,, Feng Yn b,keguo a a College

More information