Sampling-based Approach for Design Optimization in the Presence of Interval Variables

Size: px
Start display at page:

Download "Sampling-based Approach for Design Optimization in the Presence of Interval Variables"

Transcription

1 0 th Word Congress on Structura and Mutdscpnary Optmzaton May 9-4, 03, Orando, orda, USA Sampng-based Approach for Desgn Optmzaton n the Presence of nterva Varabes Davd Yoo and kn Lee Unversty of Connectcut, Storrs, Connectcut, USA, davd.yoo@engr.uconn.edu Unversty of Connectcut, Storrs, Connectcut, USA, ee@engr.uconn.edu. Abstract Ths paper proposes a methodoogy for sampng-based desgn optmzaton n the presence of nterva varabes. Assumng that an accurate surrogate mode s avaabe, the proposed method frst searches the combnaton of nterva varabes for constrants when ony nterva varabes are present or for probabstc constrants when both nterva and random varabes are present. Due to the fact that the combnaton of nterva varabes for probabty of faure does not aways concde wth that for a performance functon, the proposed method drecty uses the probabty of faure to obtan the combnaton of nterva varabes when both nterva and random varabes are present. To cacuate senstvtes of the constrants and probabstc constrants wth respect to nterva varabes by the sampng-based method, behavor of nterva varabes at the case s defned by the Drac deta functon. Then, Monte Caro smuaton s apped to cacuate the constrants and probabstc constrants wth the combnaton of nterva varabes, and ther senstvtes. The mportant mert of the proposed method s that t does not requre gradents of performance functons and transformaton from -space to U-space for reabty anayss after the combnaton of nterva varabes s obtaned, thus there s no appromaton or restrcton n cacuatng senstvtes of constrants or probabstc constrants. Numerca resuts ndcate that the proposed method can search the case probabty of faure wth both effcency and accuracy and that t can perform desgn optmzaton wth mture of random and nterva varabes by utzng the case probabty of faure search.. Keywords: nterva Varabes, Sampng-Based Method, Drac Deta functon, Monte Caro smuaton, Surrogate mode 3. ntroducton eabty anayss and reabty-based desgn optmzaton (BDO) have been deveoped to take uncertanty nto consderaton, and have been successfuy adapted to many engneerng appcatons such as crashworthness of vehce and structura-acoustc system desgn [-0]. The uncertanty s generay categorzed nto aeatory and epstemc uncertantes, where aeatory uncertanty s consdered as rreducbe whereas epstemc uncertanty s reducbe by coectng more data. n case when suffcent amount of data for statstca nformaton s unavaabe, possbty-based (or fuzzy set) methods have utzed membershp functon to mode nsuffcenty coected data [], and adusted standard devaton and correaton coeffcent nvovng confdence ntervas have been utzed to offset an naccurate modeng of data [,3]. When degree of nsuffcency of data s even greater as ony ower and upper bounds of data are avaabe, the methods sted above are not appcabe anymore, thus the dfferent approach s requred. To dea wth data of whch ony ower and upper bounds are avaabe, a method of mut-pont appromaton that evauates the weghtng functon and oca appromatons separatey has been frst deveoped for nterva anayss [4]. Then, the most probabe pont (MPP) based frst-order reabty method (OM) has been utzed for desgn optmzaton wth mture of random and nterva varabes [5]. As bounds of probabty of faure or reabty est n the presence of nterva varabes, desgn optmzaton for the and best cases has been aso deveoped [6], and senstvty anayss consderng bounds of nterva varabes and probabty of faure has been deveoped accordngy [7]. By usng the MPP-based OM, a desgn optmum s very effcenty searched; however t s generay ess accurate for hghy nonnear performance functons and hgh-dmensona nput varabes [8-]. To mprove the accuracy on ths occason, the second order reabty method (SOM) can be apped after the MPP search; however, ts effcency s sacrfced due to the fact that computaton of the Hessan matr s requred by the SOM [-5]. The MPP-based dmenson reducton method (DM) can be aso used for appromatey assessng the reabty of a system, whch s used as a probabstc constrant n BDO [6-8]. n absence of accurate senstvtes of performance functons, the MPP-based reabty anayss or BDO, whch utzes senstvtes of performance functons to fnd the MPP, cannot be drecty used, nstead the sampng-based reabty anayss or BDO can be used [9-3]. Assumng an accurate surrogate mode s gven [3-36], Monte Caro smuaton (MCS) [37] can be apped to fnd a desgn optmum wth affordabe computatona burden. Ths study ntroduces nterva anayss and desgn optmzaton utzng the sampng-based method n the

2 presence of ony nterva varabes and n the presence of both random and nterva varabes. Due to the presence of nterva varabes, obtanng the combnaton of nterva varabes for both constrants and probabstc constrants s nvoved [5]. When both random and nterva varabes are present, the combnaton of nterva varabes for probabty of faure s drecty searched usng the probabty of faure and ts senstvty snce the desgn pont where the case probabty of faure occurs does not aways concde wth that for the case performance functon; t s hghy key as many studes have assumed, however not aways. To evauate senstvtes of probabty of faure wth respect to nterva varabes, the Drac deta functon s utzed to defne behavor of the nterva varabes at the case [38-4]. Assumng an accurate surrogate mode s gven, one mert of the proposed method ests not ony durng the case probabty of faure search but aso durng reabty anayss after the case probabty of faure search. The case probabty of faure search, whch w be epaned n Secton 4., utzes a vector of nterva varabes nstead of ndvdua components of the vector, and t thus promses effcency. Aso, t resoves the probem that the case probabty of faure does not aways occur where the case performance occurs. Durng the reabty anayss after the case probabty of faure search, another mert of the proposed method s that t does not make further appromatons snce t does not requre gradents of the performance functon and transformaton of desgn varabes from -space to U-space, thus there s no appromaton or restrcton n cacuatng the senstvtes of constrants or probabstc constrants [30]. The paper s organzed as foow. Secton 4 brefy revews the sampng-based BDO. Secton 5 epans desgn optmzaton wth nterva varabes ony, ncudng the agorthm to obtan the combnaton of nterva varabes for a performance functon, mathematca dervaton for senstvtes of each constrant wth respect to nterva varabes by defnng behavor of the nterva varabes at the case usng the Drac deta functon, and ther computaton by the MCS. Secton 6 epans the sampng-based desgn optmzaton wth both random and nterva varabes ncudng detas to obtan the combnaton of nterva varabes for probabty of faure and computaton of probabstc constrants and ther senstvtes by the MCS. Secton 7 ustrates search for the combnaton of nterva varabes for probabty of faure and desgn optmzaton wth random and nterva varabes usng numerca eampes. Secton 8 summarzes and concudes the paper wth dscusson of future research. 4. evew of Sampng-based BDO 4. ormuaton of BDO The mathematca formuaton of BDO s epressed as mnmze cost( d) tar subect to P G 0 P,,..., NC L U ndv d d d d,, and N () where d ( ) s the desgn vector, whch s the mean vaue of the N-dmensona random vector T,,... N ; P s the target probabty of faure for the th constrant; and NC, ndv, and N are the tar number of probabstc constrants, desgn varabes, and random varabes, respectvey [30]. To carry out BDO usng Eq. (), the probabstc constrants and ther senstvtes must be evauated. evews on the reabty and ts senstvty anayses are epaned n Sectons. and.3, respectvey. 4. Probabty of aure The probabty of faure wth random varabes, denoted by P, s defned usng a mut-dmensona ntegra ; P P f d E () N where s a matr of dstrbuton parameters, whch ncudes mean ( ) and/or standard devaton ( ) of P represents a probabty measure; Ω s defned as a faure set; f ; ; and functon (PD) of ndcator functon and defned as E represents the epectaton operator [30,4]. ; s a ont probabty densty n Eq. () s caed an

3 ,. (3) 0, otherwse 4.3 Senstvty of Probabty of aure Wth the four reguarty condtons satsfed, whch are aso epaned n deta n ef. [30], takng the parta dervatve of Eq. () wth respect to yeds P f ; d (4) N and the dfferenta and ntegra operators can be nterchanged due to the 4 th reguarty condton n ef. [30] and the Lebesgue domnated convergence theorem [43,44] gvng P E n f ;. (5) The parta dervatve of the og functon of the ont PD n Eq. (5) wth respect to s known as the frst-order score functon [30] for and s denoted as () s n f ; ;. (6) To derve the senstvty of the probabty of faure n Eq. (), t s requred to know the frst-order score functon n Eq. (6), whch s obtaned usng the foowng equaton for ndependent random varabes where ; () s n f ; n f ; ; (7) f s the margna PD correspondng to the th random varabe foowng equaton for correated random varabes () s cu v n f ;, and obtaned usng the n f ; n, ; ; (8) where c s a copua densty functon, u ; and v ; are margna CDs for and respectvey, and s the correaton coeffcent between and [30]. The nformaton of margna PDs, CDs, and commony used copua densty functons s sted n deta n ef. [30]. 4.4 Smpfcaton of the nonnear characterstcs of vehce behavor The MCS can be apped to cacuate the probabstc constrants n Eq. () and ther senstvtes. Denotng a surrogate mode for the th constrant functon wth random varabes as ˆ G Eq. () can be cacuated as K k,, the probabstc constrants n K ( k ) tar P P G 0 ˆ P (9) where K s the MCS sampe sze, ( k ) s the k th reazaton of, and the faure set ˆ for the surrogate mode 3

4 s defned as ˆ G ˆ : 0 [30]. Senstvtes of the probabstc constrants n Eq. () are cacuated usng the score functon as P ; (0) K ( k) () ( k) ˆ s K k () ( k where s ) ; s obtaned usng Eqs. (7) and (8) for ndependent and correated random varabes, respectvey. 5. Desgn Optmzaton wth nterva Varabes 5. ormuaton of Desgn Optmzaton wth nterva Varabes The mathematca formuaton of desgn optmzaton wth nterva varabes ony s epressed as mnmze cost( d), subect to G 0,,..., NC () L U ndv d d d d,, and N where d s the desgn vector, whch s the md-pont of the N-dmensona nterva vector T,,,..., N where N s the number of nterva varabes. n Eq. () s the case nterva varabes for the th constrant, whch s obtaned by sovng the optmzaton probem to where s the nterva ength of s not avaabe, mamze G subect to,..., for N (). t shoud be noted that as any statstca nformaton of an nterva varabe, must be consdered for the desgn optmzaton. To carry out the desgn optmzaton wth nterva varabes usng Eq. (), constrants wth the case nterva varabes, namey the case constrants or the case performance, and ther senstvtes must be evauated. Each of the case constrant s obtaned by the case performance search that soves Eq. () and w be epaned n Secton 3., and senstvty anayss of each of the case constrant and ts cacuaton are epaned n Secton 3.3. t s assumed n ths study that gradents of performance functons are not avaabe; however t can be drecty used f avaabe. 5. Agorthm Searchng for Worst Case Performance The agorthm epaned n ths secton searches the case performance, and the agorthm was orgnay deveoped by Lu et a. n ef. for the mama possbty search (MPS) for possbty-based desgn optmzaton. An mportant mert of the proposed agorthm s that t utzes a vector of nterva varabes and a vector of senstvtes of a performance functon wth respect to a nterva varabes, thus ts effcency s not affected by the dmenson of the nterva varabes. The agorthm for the case performance search s summarzed as foowng, whch s aso shown n the fowchart n g.. Step. Normaze nterva varabes usng Z Z (3) or such that Z

5 Step. Set the teraton counter k = 0 wth the convergence parameter = 0-3 (0). Set =. Let Z 0. Cacuate the performance G (0) (0) Z and the senstvty G Z (0) (0) (0) to obtan G Z. Let the drecton vector be G ( k) ( k) Step 3. Search the net pont as 0.5sgn ( ) Step 4. Cacuate the performance G k ( k ) Z and ts senstvty G d Z.. t s epaned n Secton 3.3 how Z d where 0.5 s obtaned from Step. Let k = k +. Z. Let a conugate drecton vector ( k ) k ) k d G Z d ) where G ( k) / G ( k). k ) sgn sgn k G Z Z ), t s the case and go to Step. Step 5. ( f ) ( ) ( ) G Z GZ, et = k and go to Step 3. Otherwse, go to Step 6 wth Z, G ( ) ( ) G Z. Z Z f Z and f behavor of the performance functon s not monotonc wthn an nterva doman, n other words, f any component of the case nterva vector does not occur at the verte of ts nterva doman, nterpoaton agorthm must be addtonay apped to obtan more accurate case performance []. ( Step 6. Let = 0 and a drecton vector be ) ( ) G Step 7. d Z. ( ) Cacuate the new pont Z k on the boundary of the doman from the start pont search drecton () d. Let k = k +. ( ) Step 8. Cacuate the performance G k ( ) Z and ts senstvty G Z k ( k ) sgn G Z / Z sgn Z, for Z 0.5 ( k ) GZ / Z, for Z 0.5 then t s the case and go to Step. Otherwse, go to Step 9. ( ) Step 9. Use G ( ) Z, G k ( ) Z, G ( ) Z, and G k. f ( ) Z aong the Z to construct the thrd order poynoma f() t ( ) on the straght ne between Z ( ) and Z k where t s the parameter for the ne. Cacuate the mamum pont t * ( ) for ths poynoma. Let Z k be the pont on the ne correspondng to t *. Let k = k +. ( k ) G ( k ) G Z. Check the convergence crtera Step 0. Cacuate the performance Z and ts senstvty Step. usng the equaton n Step 8. f converged, t s the case and go to Step. Otherwse, et the new ( ) ( k ) ( d G Z d ) where β s gven conugate drecton vector be ( k) ( k) by G G De-normaze Z / Z. Let = k, = +, and go to Step 7., Z by Eq. (3) n Step to obtan,. The proposed agorthm requres evauaton of senstvtes of a performance functon wth respect to nterva varabes. When gradents of the performance functon are not avaabe, senstvtes of each performance functon wth respect to nterva varabes can be cacuated by the sampng-based method, and dervaton of the senstvtes of the performance functon wth respect to nterva varabes and ts cacuaton are epaned n Secton

6 nterpoaton nta nterva Varabes Step 5. Store nterva Varabes, Performance, and Senstvty Step. Normaze nterva Varabes Yes Step 6. Cacuate Drecton Vector Step. Cacuate Performance and Senstvty Step 3. Search the Net Pont by Drecton Vector No Check Convergence n Step 5. Step 7. Search the Net Pont by Drecton Vector on the Boundary on the Doman Check Convergence n Step 8. Step 4. Cacuate Performance and Senstvty Check Convergence n Step 4. Yes No Yes Step. De-Normaze nterva Varabes Step 9. Search the Net Pont on the Lne between the Pont n Step 5 and the Pont n Step 7. Check Convergence n Step 0. Yes No No Worst Case nterva Varabes gure. owchart for Worst Case Performance Search 5.3 Senstvty Anayss of Worst Case Performance uncton and ts Cacuaton The behavor of any pont wthn the nterva of an nterva varabe can be epressed usng the Drac deta functon () [45] as 0, 0, 0, (4) and shftng Eq. (4) by the case of denoted as, yeds whch s constraned to satsfy the dentty,, 0,,, (5) Aso, the property of the Drac deta functon [45] yeds Usng Eqs. (4)~(7) and assumng, d. (6) G, d G,. (7) G s a contnuousy dfferentabe functon of any rea number, senstvty 6

7 of the case performance functon wth respect to the th case nterva varabe n genera dmenson becomes G,,, G,, d G d, N N (8) N,, where. Based on the defnton of the Drac deta functon, behavor of a snge nterva varabe at ts case,, can be treated as a Gaussan norma dstrbuton wth of and approachng to 0, whch mpes,, 0.5 / = m e m f. (9) 0 0 Equaton (9) s verfed n ths secton frst. Consder senstvty of an one-dmensona performance functon, G wth respect to the case of nterva varabe, whch by usng Eq. (8) becomes G d. (0), G, G n Eq. (0) usng the Tayor seres epanson at ( m), where m,,, can be epressed as ( m), G m m m () G a m!,, m0 m0 a G / m!. Usng Eqs. (9) and () and the score functon epaned n Secton.3, the rght hand sde of Eq. (0) s evauated as,,, m G d a, m m f d () m0 Usng the epectaton operator, the Eq. () s further smpfed as 0 m m m,,, G d E m a E a, a 0 0 m0 (3) where, E p, p p 0 f p s odd and E p!! f p s even accordng to the property of centra moments of a norma dstrbuton. Usng Eq. (), the eft hand sde of Eq. (0) s evauated as,, G am G m0,, m,, m amm a. (4) m, The dentca resuts n Eqs. (3) and (4) demonstrate the vadty of treatng behavor of, Gaussan norma dstrbuton wth of and approachng to 0. at, as a 7

8 nay, usng Eq. (9), Eq. (8) s further deveoped as,,, G d E G,, m 0 N G (5) and senstvty of the case constrant wth respect to desgn pont where,, N n Eq. () becomes G G (6), N,,, EG m 0,, /, f /,, / 0, f /. (7) Addtonay, t s noted that the Drac deta functon can be aso appcabe to defne behavor of determnstc varabes when senstvtes of performance functons wth respect to determnstc varabes are not avaabe. Thus, n the presence of determnstc varabes, the proposed sampng-based method can be apped to evauate senstvtes of performance functons even when gradents of the performance functons are not obtanabe. Gˆ. The MCS can be apped to Denote a surrogate mode for the constrant functon wth nterva varabes as cacuate senstvty of a performance functon wth respect to the th case nterva varabe durng the case performance search n Secton 3. usng Eq. (5) as where s for,, ˆ, G G ˆ (8), () can be cacuated usng Eqs. (6) and (7) as The desred vaue of K ( k ),, ( k ) G,, K k, comng from n Eq. (6), and senstvtes of the case constrants n Eq., ˆ, G G ˆ. (9) N K ( k ),,, ( k ),,, G K k,,, used n Eqs. (8) and (9) for the sampng-based method s determned through the foowng smuaton anayss. Durng the smuaton anayss, rato of to consdered snce depends on or nstead of ust s, and senstvty of a performance functon G wth respect to s cacuated by the MCS whe changes from 0. to 0.00 n descendng order. The resut s then compared to the true senstvty, whch s anaytcay obtaned as. rom the resut shown n gure, t s demonstrated that for the desred vaue of. rom the neggbe amount of error ess than 0.3% n gure, vadty of cacuatng senstvtes of a performance functon wth respect to nterva varabes usng the sampng-based method wth a very sma standard devaton can be aso shown. 8

9 3.5 (% error) gure. Error of Senstvty as changes from 0.00 to Desgn Optmzaton wth andom and nterva Varabes 6. ormuaton of Desgn Optmzaton wth Mture of andom and nterva Varabes The mathematca formuaton of desgn optmzaton wth mture of random and nterva varabes s epressed as mnmze cost( d) where,..., N, N,..., NN, tar subect to P P G, 0,,..., NC P L U ndv N d d d d d s the desgn vector;,,, and (30) N, n Eq. (30) s the case nterva varabes for the th probabstc constrant, whch s obtaned by sovng the optmzaton probem to mamze PG 0 subect to for,..., N. (3) To carry out the desgn optmzaton wth nterva varabes usng Eq. (30), probabstc constrants wth the case nterva varabes, namey the case probabstc constrants or the case probabty of faure, and ther senstvtes must be evauated. Each of the case probabstc constrant s obtaned by the case probabty of faure search that soves Eq. (3) and w be epaned n Secton 4., and senstvty anayss of each of the case probabstc constrant and ts cacuaton are epaned n Secton 4.3. t shoud be noted that the case probabty of faure does not aways occur at the pont where the case performance occurs, whch s demonstrated wth an eampe n Secton 4.. Thus, by appyng an agorthm for the case performance search n Secton 3. by drecty utzng probabty of faure and ts senstvty n repacement of performance vaue and ts senstvty, the probem ponted out n the prevous sentence can be resoved. 6. Worst Case Probabty of aure The case probabty of faure wth random and nterva varabes, denoted by P, s defned usng Eq. (30) and a mut-dmensona ntegra as NN,, P, f d d E., (3) The case probabty of faure n Eq. (3) s obtaned usng the agorthm for the case performance search epaned n Secton 3. by utzng probabty of faure and ts senstvty n repacement of the 9

10 - -5 performance functon and ts senstvty. Dervaton of the senstvty of the probabty of faure wth respect to the case nterva varabes and ts cacuaton are epaned n Secton 4.3. Usuay, the case probabty of faure occurs at the case performance, so conventonay the case probabty of faure s cacuated by evauatng the probabty of faure at the case performance [5-7]. However, ths s not aways the case and the foowng eampe demonstrates t. Consder a D hghy nonnear poynoma functon, G (33) 3 4 ( ) (W 6) (W 6) 0.6 (W 6) Z Tabe. Property of nput Varabes Varabes Types Dstrbuton Parameters nterva N/A andom Norma.5 W where Z As shown n Tabe, respectvey. The md-pont and nterva ength of and are nterva and random varabes, are 6.5 and 3, respectvey. The mean and standard devaton of are.5 and, respectvey. Then, s dvded nto 00 sub-ntervas, for each of whch, the performance functons and probabty of faures are evauated. or the evauaton of the probabty of faure, MCS sampe are used for each sub-nterva The case performance The case probabty of faure 5-4 G = = gure 3. Worst Case Performance and Worst Case Probabty of aure As shown n g. 3, the case probabty of faure does not occur where the case performance occurs. The case probabty of faure occurs at / 8 where performance and probabty of faure are and 0.48, respectvey, whe the case performance occurs at 5, where the performance and probabty of faure are.547 and 0., respectvey. Thus, ths study suggests usng the agorthm for the case probabty faure search drecty nstead of obtanng the case probabty of faure by cacuatng the probabty of faure where the case performance occurs. The MCS can be apped to cacuate probabty of faure durng the case probabty of faure search. Denotng the surrogate mode for constrant functons wth random and nterva varabes as Ĝ,, the probabty of faure durng the case probabty of faure search can be cacuated usng Eq. (3) as P P G P (34) K ( k 0 ) tar ˆ,,, K, k 0

11 where the faure set ˆ for the surrogate mode s defned as ˆ, G ˆ : 0,. 6.3 Senstvty Anayss of Worst Case Probabty of aure and ts Cacuaton Takng parta dervatve of Eq. (3) wth respect to the th case nterva varabe yeds N P,, E,. (35), Then, takng parta dervatve of Eq. (3) wth respect to the md-pont of the th nterva varabe usng Eq. (35) yeds N, N, P P,, E,, (36),, where s obtaned from Eq. (7). Takng parta dervatve of Eq. (3) wth respect to the mean of the th random varabe yeds P NN n f ;,, f ; d d. (37) Usng Eq. (7), Eq. (37) s further smpfed as n f ; n f ; ;. (38) P,, f d E,, NN The MCS can be apped to cacuate senstvty of the probabty of faure wth respect to the th case nterva varabe durng the case probabty of faure search n Secton 4. based on Eq. (35) as P,. (39) K ( k ), ( k ),, ˆ K, Senstvtes of the case probabstc constrants n Eq. (30) wth respect to the th nterva varabe at the md-pont as P (40) N K ( k ),, ( k ),,,, ˆ, K k,,, based on Eq. (36). Senstvtes of the case probabstc constrants n Eq. (30) wth respect to the th random varabe at the mean pont are cacuated as P K ( k), () ( k) ˆ s K k, ; (4) based on Eq. (38). 7. Numerca Eampes

12 Numerca studes are carred out n ths secton to verfy the agorthm that searches the case probabty of faure n Secton 4 for both ow-dmensona and hgh-dmensona cases. Aso, desgn optmzaton wth mture of nterva and random varabes that utzes the case probabty of faure search s carred out. 7. Worst Case Probabty of aure Search for Two-Dmensona nputs n ths numerca eampe, the agorthm that searches the case probabty of faure s apped to a two-dmensona case, and one of nput varabes s an nterva and the other s a random. Consder a nonnear performance functon gven as G (4) Tabe. Property of nput Varabes Varabes Types Dstrbuton Parameters nterva NA 0.5 andom Norma. As shown n Tabe, s an nterva varabe wth ts md-pont at 0.5 and ts nterva ength of, and s a normay dstrbuted random varabe wth ts mean at. and ts standard devaton of G = 0 G = G = P =.00 The case probabty of faure gure 4. Search Hstory of Worst Case Probabty of aure Tabe 3. Search Hstory of Worst Case Probabty of aure teraton P P / Wth the gven property of these nput varabes and the performance functon n Eq. (4), the case probabty of faure s obtaned usng the case probabty of faure search epaned n Secton 4.. The resuts are shown both n Tabe 3 and g. 4. The case nterva varabes at the 4 th teraton n Tabe 3 s obtaned by an nterpoaton of two case nterva varabes canddates at the nd and the 3 rd teraton durng Step 9 of the case probabty of faure search. The obtaned resut s compared wth the resut obtaned by 7 dvdng the nterva doman nto 00 sub-ntervas and performng the MCS wth 50 sampes for a sub-ntervas, whch s shown n Tabe 4.

13 Methods Tabe 4. Comparson of esuts Obtaned by Dfferent Methods P Number of MCS Proposed Agorthm Performng MCS for a 00 sub-ntervas n terms of effcency, the proposed agorthm requres (4teratons) (MCS/teraton) = 4MCSs, and performng the MCS for a 00 sub-dvded ntervas requres (00sub-ntervas) (MCS/sub-nterva) = 00MCSs. Thus, the proposed agorthm s 5 tmes more effcent than the crude MCS whe mantanng accuracy n ths eampe. 7. Worst Case Probabty of aure Search for Hgh-Dmensona nputs gure 5. Schematc Dagram of Cantever Tube n ths numerca eampe, the agorthm that searches the case probabty of faure s apped to a hgh-dmensona case where of nput varabes are nterva and 9 of them are random varabes. Consder the cantever tube shown n g. 5 subected to eterna forces,, and P, and torson T [6]. The performance functon s defned as the dfference between the yed strength S y and the mamum stress, namey, where G g S (43) ma y ma s the mamum von Mses stress on the top surface of the tube at the orgn, whch s gven by y (44) ma 3 z where the norma stress s obtaned P sn sn L cos L cos d 4 4 d d t d d t 4 64 (45) and the shear stress z s obtaned as z Td d d t , (46) respectvey. The property of random and nterva varabes are gven n Tabes 5 and 6, respectvey. As shown n Tabes 5 and 6, nne random varabes ~ havng varous dstrbutons and two nterva varabes 9 0 and havng the dentca nterva ength at dfferent md-ponts are used as nput varabes. Wth the property of nput varabes and the performance functon n Eq. (43), the case probabty of faure 7 s obtaned usng the case probabty of faure search. The MCS wth 5 0 sampes s tred for every 4 teraton, and the toerance of 0 3 nstead of 0 s set for ths eampe snce the senstvty of probabty of faure wth respect to both nterva varabes s ess than 0 throughout the nterva doman. By usng the 3

14 Tabe 5. Property of andom Varabes Varabes Parameter Parameter Dstrbuton 5 mm (mean) 0. mm (std* ) Norma ( ) d 4 mm (mean) 4 mm (mean) Norma ( L ) 9.75 mm (b ** ) 0.5 mm (ub***) Unform 3 ( L ) mm (b) 60.5 mm (ub) Unform 4 ( ) 3.0 kn (mean) 0.3 kn (std) Norma 5 ( ) 3.0 kn (mean) 0.3 kn (std) Norma 6 P.0 kn (mean). kn (std) Gumbe ( ) 7 ( ) 8 T 90.0 Nm (mean) 9.0 Nm (std) Norma ( ) 9 S y 33.7 MPa (mean).0 MPa (std) Norma *: std-standard devaton **: b ower bound of a unform dstrbuton ***: ub upper bound of a unform dstrbuton Tabe 6. Property of nterva Varabes Varabes ( ) 0 ( ) Parameters 0 5, 0 0 0, 0 proposed agorthm, the case probabty of faure s obtaned n 8 teratons ncudng the one wth the nterpoaton and the dscard one. n Tabe 7, snce the probabty of faure at the 4 th teraton s smaer than that at the 3 rd teraton, t s dscarded durng the Step 5 of the case probabty of faure search n Secton 4.. Search hstory s shown n both Tabe 7 and g. 6. The case probabty of faure s obtaned as and the case nterva varabes are obtaned as [3.993, 7.887]. The obtaned resut s then compared wth the 7 resut obtaned by dvdng both nterva domans nto 00 sub-ntervas and performng MCS wth 50 sampes for a combnatons of sub-ntervas. teraton ( ) 0 Tabe 7. Search Hstory of Worst Case Probabty of aure ( ) P P / 0 P / E E E E E E-04 4 * E E-03 3' E E E E-04 4' E-04.36E E-05.6E-07 *: Dscarded durng Step 5 of Worst Case Probabty of aure Search ' : Utzed for nterpoaton The resut of the comparson s shown n Tabe 8. n terms of effcency, the proposed agorthm requres (8teratons) (MCS/teraton) = 8MCSs, and performng MCS for a combnatons of 00 sub-ntervas requres (00 00combnatons) (MCS/combnaton) = 0000MCSs. Thus, the proposed agorthm s 50 tmes more effcent whe mantanng accuracy n ths eampe. As suggested by the current and the prevous eampes, the more nterva varabes there are, the ess effcent performng the crude MCS eponentay becomes. n genera dmenson, performng the MCS for a combnatons of 00 sub-ntervas of every nterva varabe requres (00 N 4

15 8 G = ' G = 0 4* 3 G = ' P =0.0 0 G 3 = 0.03 =0.0 The case probabty of faure gure 6. Search Hstory of Worst Case Probabty of aure combnatons) (MCS/combnaton) = (0) N MCSs. On the other hand, the proposed agorthm requres smar number of MCSs regardess of dmenson of nterva varabes snce t utzes a vector of nterva varabes and ts senstvty vector nstead of ther ndvdua components. Tabe 8. Comparson of esuts Obtaned by Dfferent Methods Methods Worst Case nterva Varabe Probabty of aure Number of MCSs Proposed Agorthm [ ] Performng MCS for a combnatons of 00 sub-ntervas [ ] Desgn Optmzaton wth Mture of andom and nterva nputs Ths numerca eampe shows the desgn optmzaton wth mture of random and nterva varabes, utzng the case probabty of faure search. Consder a D mathematca desgn optmzaton probem, whch s formuated to mnmze where three constrants are gven by d C d d, tar subect to P P G 0.75%, ~ 3 d, P L U d d d d,,, and (47) G G G (48) The propertes of two nput varabes, one nterva and one random varabe, are shown n Tabe 9. As shown n Eq. tar (47), the target probabty of faure P s set to.75% for a constrants. g. 7 shows the optmum desgn of the sampng-based desgn optmzaton wth nterva and random varabes. As can be seen n g. 7, the determnstc desgn optmum (d dopt ) was frst searched to enhance effcency of the 5

16 Tabe 9. Property of nput Varabes L nput Varabes Varabe Types d O d U d Parameters nterva andom desgn optmzaton procedure. n g. 7, the dotted bo ustrated around the desgn optmum (d opt ) shows the tar ont range of and. Wth P of.75%, aowed tota range of dstrbuton of becomes 4.6, and wth. of, sze of the dotted bo becomes..6. Wth the dotted bo around d opt t s easy dentfed that d opt s the desred optmum as vertces of the bo are rght on two actve constrants, G G. P and P occur on the eft and the rght bounds of 0 are 0.03 and 0.08, respectvey, whch are very cose to P tar, respectvey where 0 P and and P. Desgn search hstory and number of teratons taken to obtan the case probabty of faure at each desgn are shown n Tabe 0. One MCS for each teraton s used to obtan P and whe appyng the case probabty of faure search epaned P n Secton 4.. At the nd teraton durng the desgn search, P does not behave monotonc wthn the doman of the nterva varabe, thus the nterpoaton agorthm s apped to fnd more accurate case probabty of faure, whch s why 5 teratons are taken to obtan P. Overa teratons taken to obtan P s around for each desgn search snce concuded P behaves monotonc most tmes wthn the doman of the nterva varabe. Thus t s G 3 () = 0 4 G () = 0 d opt d dopt G () = gure 7. Optmum Desgn of Sampng-Based Desgn Optmzaton wth andom and nterva Varabes that the computatona burden to obtan P n ths eampe s affordabe. Tabe 0. Desgn Search Hstory and Number of teratons for Worst Case Probabty of aure Desgn Pont # teratons # teratons # teratons teraton (d, d ) for P for P for P (3.39,.0639) (4.384, ) 5 3 (3.84,.9765) 4 (.8478, 3.096) 5 (3.377, 3.037) 6 (3.3407, 3.005) 7 (3.374, 3.) 8 (3.3838, 3.74) 3 6

17 8. Concusons Sampng-based desgn optmzatons wth ony nterva varabes and wth both nterva and random varabes are deveoped n ths study. or the desgn optmzaton wth nterva varabes ony, each of the case constrant s evauated by the deveoped case performance search where nterva and senstvty vector are utzed, thus effcency s promsed regardess of the dmenson of the nterva varabes. t s assumed that gradents of performance functons are not avaabe n ths study. Therefore, senstvtes of a performance functon wth respect to nterva varabes are derved by defnng behavor of the nterva varabes at the case by the Drac deta functon to cacuate t by the sampng-based method. Through the smuaton anayss, desred vaue of standard devaton for the nterva varabes at the case s determned, and the error of the resut turns out to be neggbe at the desred vaue. Usng the obtaned vaue, the senstvtes of each of the case constrants both at the case and desgn ponts are cacuated by the MCS. or the desgn optmzaton wth random and nterva varabes, the case probabstc constrants are evauated by the case probabty of faure search. Snce probabty of faure does not aways occur where the case performance occurs as demonstrated n ths study, the case probabty of faure s obtaned by drecty usng the probabty of faure and ts senstvty. Smary to desgn optmzaton wth nterva varabes ony, the senstvtes of the probabty of faure both at the case and desgn ponts are derved, whch are then cacuated by the MCS. Numerca eampes show the case probabty of faure s obtaned effcenty for both ow and hgh-dmensona nputs regardess of the dmenson of the nterva varabes and the desgn optmzaton wth random and nterva varabes s successfuy carred out wth effcency utzng the case probabty of faure search. 9. eferences [] Youn, B.D., Cho, K.K., Yang,.J., and Gu, L., eabty-based Desgn Optmzaton for Crashworthness of Vehce Sde mpact, Structura and Mutdscpnary Optmzaton, Vo. 6, No. 3-4, pp. 7-83, 004. [] Youn, B.D., Cho, k.k., and Y, K., Performance Moment ntegraton (PM) Method for Quaty Assessment n eabty-based obust Optmzaton, Mechancs Based Desgn of Structures and Machnes, Vo. 33, No., pp. 85-3, 005. [3] Youn, B.D., Cho, K.K., and Tang, J., Structura Durabty Desgn Optmzaton and ts eabty Assessment, nternatona Journa of Product Deveopment, Vo., Nos. 3/4, pp , 005. [4] Jung, B.C., Lee, D., Youn, B.D., Lee, S., A Statstca Characterzaton Method for Dampng Matera Propertes and ts Appcaton to Structura-Acoustc System Desgn, Journa of Mechanca Scence and Technoogy, Vo. 5, No. 8, pp , 0. [5] Daskewcz, M.J., German, B.J., Takahash, T.T., Donovan, S., and Shaanan, A., Effects of dscpnary uncertanty on mut-obectve optmzaton n arcraft conceptua desgn, Structura and Mutdscpnary Optmzaton, Vo. 44, No. 6, pp , 0. [6] Acar, E. and Soank, K., System eabty-based Vehce Desgn for Crashworthness and Effects of Varous Uncertanty educton Measures, Structura and Mutdscpnary Optmzaton, Vo. 39, No. 3, pp. 3-35, 009. [7] Snha, K., eabty-based Mutobectve Optmzaton for Automotve Crashworthness and Occupant Safety, Structura and Mutdscpnary Optmzaton, Vo. 33, No. 3, pp , 007. [8] rangopo, D.M., and Maute, K., Lfe-Cyce eabty-based Optmzaton of Cv and Aerospace Structures, Computers & Structures, Vo. 8, No. 7, pp , 003. [9] Lan, Y., Km, N.H., eabty-based Desgn Optmzaton of a Transonc Compressor, AAA ourna, Vo. 44, No., pp , 006. [0] Dong, J., Cho, K.K., Vahopouos, N., Wang, A., and Zhang, W., Desgn Senstvty Anayss and Optmzaton of Hgh requency adaton Probems Usng Energy nte Eement and Energy Boundary Eement Methods, AAA Journa, Vo. 45, No. 6, pp , 007. [] Du, L., Cho, K.K., and Youn, B.D., An nverse Possbty Anayss Method for Possbty-Based Desgn Optmzaton, The Amercan nsttute of Aeronautcs and Astronauts, Vo. 44, No., pp , 006. [] Noh, Y., Cho, K.K., Lee,., Gorsch, D., and Lamb, D., eabty-based Desgn Optmzaton wth Confdence Leve for Non-Gaussan Dstrbutons Usng Bootstrap Method, Journa of Mechanca Desgn, Vo. 33, No. 9, pp ~, 0. [3] Noh, Y., Cho, K.K., and Lee,., eabty-based Desgn Optmzaton wth Confdence Leve Under nput Mode Uncertanty due to Lmted Test Data, Structura and Mutdscpnary Optmzaton, Vo. 43, No. 4, pp , 0. [4] Penmetsa,.C. and Grandh.V., Effcent Estmaton of Structura eabty for Probems wth Uncertan ntervas, Computers & Structures, Vo. 80, No., pp. 03-, 00. 7

18 [5] Du,., Agus, S., and Huang, B., eabty-based Desgn wth Mture of andom and nterva varabes, Journa of Mechanca Desgn, Vo. 7, No. 6, pp , 005. [6] Du,., nterva eabty Anayss, Proceedngs of ASME 007 nternatona Desgn Technca Conferences and Computers and nformaton n Engneerng Conference, Las Vegas, Nevada, 007. [7] Guo, J. and Du,., eabty Senstvty Anayss wth andom and nterva Varabes, nternatona Journa for Numerca Methods n Engneerng, Vo. 78, No.3, pp , 009. [8] Hader, A., and Mahadevan, S., Probabty, eabty and Statstca Methods n Engneerng Desgn, John Wey & Sons, New York, NY, 000. [9] Hasofer, A.M. and Lnd, N.C., An Eact and nvarant rst Order eabty ormat, ASCE Journa of the Engneerng Mechancs Dvson, Vo. 00, No., pp. -, 974. [0] Tu, J. and Cho, K.K., A New Study on eabty-based Desgn Optmzaton, Journa of Mechanca Desgn, Vo., No. 4, pp , 999. [] Tu, J., Cho, K.K., and Park, Y.H., Desgn Potenta Method for eabty-based System Parameter Desgn Usng Adaptve Probabstc Constrant Evauaton, AAA Journa, Vo. 39, No. 4, pp , 00. [] Bretung, K., Asymptotc Appromatons for Mutnorma ntegras, ASCE Journa of Engneerng Mechancs, Vo. 0, No. 3, pp , 984. [3] Hohenbcher, M. and ackwtz,., mprovement of Second-Order eabty Estmates by mportance Sampng, ASCE Journa of Engneerng Mechancs, Vo. 4, No., pp , 988. [4] Adhkar, S., eabty Anayss Usng Paraboc aure Surface Appromaton, ASCE Journa of Engneerng Mechancs, Vo. 30, No., pp , 004. [5] Zhang, J., and Du,., A Second-Order eabty Method Wth rst-order Effcency, Journa of Mechanca Desgn, Vo. 3, No. 0, paper #:0006, 00. [6] ahman, S. and We, D., A Unvarate Appromaton at Most Probabe Pont for Hger-Order eabty Anayss, nternatona Journa of Sods and Structures, Vo. 43, No. 9, pp , 006. [7] Lee,., Cho, K.K., Du, L., and Gorsch, D., nverse Anayss Method Usng MPP-Based Dmenson educton for eabty-based Desgn Optmzaton of Nonnear and Mut-Dmensona Systems, Computer Methods n Apped Mechancs and Engneerng, Vo. 98, No., pp. 4-7, 008. [8] ong,., Greene, S., Chen, W., ong, Y., and Yang, S., A New Sparse Grd Based Method for Uncertanty Propagaton, Structura and Mutdscpnary Optmzaton, Vo. 4, No. 3, pp , 00. [9] Lee,., Cho, K.K., and Zhao, L., Sampng-Based BDO Usng the Dynamc Krgng (D-Krgng) Method and Stochastc Senstvty Anayss, Journa of Structura and Mutdscpnary Optmzaton, Vo. 44, No. 3, pp , 0. [30] Lee,., Cho, K. K., and Zhao, L., Sampng-Based Stochastc Senstvty Anayss Usng Score unctons for BDO Probems Wth Correated andom Varabes, Journa of Mechanca Desgn, Vo. 33, No., pp , 00. [3] Gu, L., Yang,. J., Tho, C.H., Makowskt, M., aruquet, O., and L, Y., Optmzaton and obustness for Crashworthness of Sde mpact, nternatona Journa of Vehce Desgn, Vo. 6, No. 4, pp , 00. [3] Youn, B.D., Zhmn,., and Wang, P., Egenvector Dmenson educton (ED) Method for Senstvty-ree Uncertanty Quantfcaton, Structura and Mutdscpnary Optmzaton, Vo. 37, No., pp. 3-8, 008. [33] We, D.L., Cu, Z.S., and Chen, J., Uncertanty Quantfcaton Usng Poynoma Chaos Epanson wth Ponts of Monoma Cubature ues, Computers & Structures, Vo. 86, No. 3-34, pp.0-08, 008. [34] Chowdhury,., ao, B.N., and Prasad, A.M., Hgh Dmensona Mode epresentaton or Structura eabty Anayss, Communcaton n Numerca Methods n Engneerng, Vo. 5, No. 4, pp , 009. [35] Hu, C. and Youn, B.D., Adaptve-Sparse Poynoma Chaos Epanson for eabty Anayss and Desgn of Compe Engneerng Systems, Structura and Mutdscpnary Optmzaton, Vo. 43, No. 3, pp , 0. [36] Hu, C., and Youn, B.D., An Asymmetrc Dmenson-Adptve Tensor-Product Method for eabty Anayss, Structura Safety, Vo. 33, No. 3, pp. 8-3, 0. [37] ubnsten,.y., Smuaton and Monte Caro Method, John Wey & Sons, New York, NY, 98. [38] Khur, A.., Appcatons of Drac s Deta uncton n Statstcs, nternatona Journa of Mathematca Educaton n Scence and Technoogy, Vo. 35, No., pp , 004. [39] Hoskns,.., Generazed unctons, John Wey & Sons, New York, NY, 979. [40] Kanwa,.P., Generazed unctons Theory and Technque, Boston, Boston, MA, 998. [4] Sachev, A.. and Woyczynsk, W.A., Dstrbutons n the Physca and Engneerng Scences, Boston, Boston, MA, 997. [4] McDonad, M. and Mahadevan. S., Desgn Optmzaton wth System eabty Constrants, Journa of 8

19 Mechanca Desgn, Vo. 30, No., pp , 008. [43] ahman, S., Stochastc Senstvty Anayss by Dmensona Decomposton and Score unctons, Probabstc Engneerng Mechancs, Vo. 4, pp , 009. [44] ubnsten,.y. and Shapro, A., Dscrete Event Systems Senstvty Anayss and Stochastc Optmzaton by the Score uncton Method, John Wey & Sons, New York, NY, 993. [45] Browder, A., Mathematca Anayss: An ntroducton, Sprnger-Verag, New York, NY,

A DIMENSION-REDUCTION METHOD FOR STOCHASTIC ANALYSIS SECOND-MOMENT ANALYSIS

A DIMENSION-REDUCTION METHOD FOR STOCHASTIC ANALYSIS SECOND-MOMENT ANALYSIS A DIMESIO-REDUCTIO METHOD FOR STOCHASTIC AALYSIS SECOD-MOMET AALYSIS S. Rahman Department of Mechanca Engneerng and Center for Computer-Aded Desgn The Unversty of Iowa Iowa Cty, IA 52245 June 2003 OUTLIE

More information

MARKOV CHAIN AND HIDDEN MARKOV MODEL

MARKOV CHAIN AND HIDDEN MARKOV MODEL MARKOV CHAIN AND HIDDEN MARKOV MODEL JIAN ZHANG JIANZHAN@STAT.PURDUE.EDU Markov chan and hdden Markov mode are probaby the smpest modes whch can be used to mode sequenta data,.e. data sampes whch are not

More information

Research on Complex Networks Control Based on Fuzzy Integral Sliding Theory

Research on Complex Networks Control Based on Fuzzy Integral Sliding Theory Advanced Scence and Technoogy Letters Vo.83 (ISA 205), pp.60-65 http://dx.do.org/0.4257/ast.205.83.2 Research on Compex etworks Contro Based on Fuzzy Integra Sdng Theory Dongsheng Yang, Bngqng L, 2, He

More information

A finite difference method for heat equation in the unbounded domain

A finite difference method for heat equation in the unbounded domain Internatona Conerence on Advanced ectronc Scence and Technoogy (AST 6) A nte derence method or heat equaton n the unbounded doman a Quan Zheng and Xn Zhao Coege o Scence North Chna nversty o Technoogy

More information

Sampling-Based Stochastic Sensitivity Analysis Using Score Functions for RBDO Problems with Correlated Random Variables

Sampling-Based Stochastic Sensitivity Analysis Using Score Functions for RBDO Problems with Correlated Random Variables Proceedngs of the ASME 00 Internatonal Desgn Engneerng Techncal Conferences & Computers and Informaton n Engneerng Conference IDETC/CIE 00 August 5 8, 00, Montreal, Canada DETC00-859 Samplng-Based Stochastc

More information

Associative Memories

Associative Memories Assocatve Memores We consder now modes for unsupervsed earnng probems, caed auto-assocaton probems. Assocaton s the task of mappng patterns to patterns. In an assocatve memory the stmuus of an ncompete

More information

Neural network-based athletics performance prediction optimization model applied research

Neural network-based athletics performance prediction optimization model applied research Avaabe onne www.jocpr.com Journa of Chemca and Pharmaceutca Research, 04, 6(6):8-5 Research Artce ISSN : 0975-784 CODEN(USA) : JCPRC5 Neura networ-based athetcs performance predcton optmzaton mode apped

More information

REAL-TIME IMPACT FORCE IDENTIFICATION OF CFRP LAMINATED PLATES USING SOUND WAVES

REAL-TIME IMPACT FORCE IDENTIFICATION OF CFRP LAMINATED PLATES USING SOUND WAVES 8 TH INTERNATIONAL CONERENCE ON COMPOSITE MATERIALS REAL-TIME IMPACT ORCE IDENTIICATION O CRP LAMINATED PLATES USING SOUND WAVES S. Atobe *, H. Kobayash, N. Hu 3 and H. ukunaga Department of Aerospace

More information

COXREG. Estimation (1)

COXREG. Estimation (1) COXREG Cox (972) frst suggested the modes n whch factors reated to fetme have a mutpcatve effect on the hazard functon. These modes are caed proportona hazards (PH) modes. Under the proportona hazards

More information

Supplementary Material: Learning Structured Weight Uncertainty in Bayesian Neural Networks

Supplementary Material: Learning Structured Weight Uncertainty in Bayesian Neural Networks Shengyang Sun, Changyou Chen, Lawrence Carn Suppementary Matera: Learnng Structured Weght Uncertanty n Bayesan Neura Networks Shengyang Sun Changyou Chen Lawrence Carn Tsnghua Unversty Duke Unversty Duke

More information

Probabilistic Sensitivity Analysis for Novel Second-Order Reliability Method (SORM) Using Generalized Chi-Squared Distribution

Probabilistic Sensitivity Analysis for Novel Second-Order Reliability Method (SORM) Using Generalized Chi-Squared Distribution th World Congress on Structural and Multdscplnary Optmzaton May 9 -, 3, Orlando, lorda, USA Probablstc Senstvty Analyss for ovel Second-Order Relablty Method (SORM) Usng Generalzed Ch-Squared Dstrbuton

More information

Nested case-control and case-cohort studies

Nested case-control and case-cohort studies Outne: Nested case-contro and case-cohort studes Ørnuf Borgan Department of Mathematcs Unversty of Oso NORBIS course Unversty of Oso 4-8 December 217 1 Radaton and breast cancer data Nested case contro

More information

Characterizing Probability-based Uniform Sampling for Surrogate Modeling

Characterizing Probability-based Uniform Sampling for Surrogate Modeling th Word Congress on Structura and Mutdscpnary Optmzaton May 9-4, 3, Orando, Forda, USA Characterzng Probabty-based Unform Sampng for Surrogate Modeng Junqang Zhang, Souma Chowdhury, Ache Messac 3 Syracuse

More information

Dynamic Analysis Of An Off-Road Vehicle Frame

Dynamic Analysis Of An Off-Road Vehicle Frame Proceedngs of the 8th WSEAS Int. Conf. on NON-LINEAR ANALYSIS, NON-LINEAR SYSTEMS AND CHAOS Dnamc Anass Of An Off-Road Vehce Frame ŞTEFAN TABACU, NICOLAE DORU STĂNESCU, ION TABACU Automotve Department,

More information

REPORT DOCUMENTATION PAGE

REPORT DOCUMENTATION PAGE REPORT DOCUMENTATION PAGE orm Approved OMB NO. 0704-088 The publc reportng burden for ths collecton of nformaton s estmated to average hour per response, ncludng the tme for revewng nstructons, searchng

More information

[WAVES] 1. Waves and wave forces. Definition of waves

[WAVES] 1. Waves and wave forces. Definition of waves 1. Waves and forces Defnton of s In the smuatons on ong-crested s are consdered. The drecton of these s (μ) s defned as sketched beow n the goba co-ordnate sstem: North West East South The eevaton can

More information

3. Stress-strain relationships of a composite layer

3. Stress-strain relationships of a composite layer OM PO I O U P U N I V I Y O F W N ompostes ourse 8-9 Unversty of wente ng. &ech... tress-stran reatonshps of a composte ayer - Laurent Warnet & emo Aerman.. tress-stran reatonshps of a composte ayer Introducton

More information

The line method combined with spectral chebyshev for space-time fractional diffusion equation

The line method combined with spectral chebyshev for space-time fractional diffusion equation Apped and Computatona Mathematcs 014; 3(6): 330-336 Pubshed onne December 31, 014 (http://www.scencepubshnggroup.com/j/acm) do: 10.1164/j.acm.0140306.17 ISS: 3-5605 (Prnt); ISS: 3-5613 (Onne) The ne method

More information

G : Statistical Mechanics

G : Statistical Mechanics G25.2651: Statstca Mechancs Notes for Lecture 11 I. PRINCIPLES OF QUANTUM STATISTICAL MECHANICS The probem of quantum statstca mechancs s the quantum mechanca treatment of an N-partce system. Suppose the

More information

LECTURE 21 Mohr s Method for Calculation of General Displacements. 1 The Reciprocal Theorem

LECTURE 21 Mohr s Method for Calculation of General Displacements. 1 The Reciprocal Theorem V. DEMENKO MECHANICS OF MATERIALS 05 LECTURE Mohr s Method for Cacuaton of Genera Dspacements The Recproca Theorem The recproca theorem s one of the genera theorems of strength of materas. It foows drect

More information

Reliability Sensitivity Algorithm Based on Stratified Importance Sampling Method for Multiple Failure Modes Systems

Reliability Sensitivity Algorithm Based on Stratified Importance Sampling Method for Multiple Failure Modes Systems Chnese Journa o Aeronautcs 3(010) 660-669 Chnese Journa o Aeronautcs www.esever.com/ocate/ca Reabty Senstvty Agorthm Based on Strated Importance Sampng Method or Mutpe aure Modes Systems Zhang eng a, u

More information

Predicting Model of Traffic Volume Based on Grey-Markov

Predicting Model of Traffic Volume Based on Grey-Markov Vo. No. Modern Apped Scence Predctng Mode of Traffc Voume Based on Grey-Marov Ynpeng Zhang Zhengzhou Muncpa Engneerng Desgn & Research Insttute Zhengzhou 5005 Chna Abstract Grey-marov forecastng mode of

More information

Numerical Investigation of Power Tunability in Two-Section QD Superluminescent Diodes

Numerical Investigation of Power Tunability in Two-Section QD Superluminescent Diodes Numerca Investgaton of Power Tunabty n Two-Secton QD Superumnescent Dodes Matta Rossett Paoo Bardea Ivo Montrosset POLITECNICO DI TORINO DELEN Summary 1. A smpfed mode for QD Super Lumnescent Dodes (SLD)

More information

Delay tomography for large scale networks

Delay tomography for large scale networks Deay tomography for arge scae networks MENG-FU SHIH ALFRED O. HERO III Communcatons and Sgna Processng Laboratory Eectrca Engneerng and Computer Scence Department Unversty of Mchgan, 30 Bea. Ave., Ann

More information

Strain Energy in Linear Elastic Solids

Strain Energy in Linear Elastic Solids Duke Unverst Department of Cv and Envronmenta Engneerng CEE 41L. Matr Structura Anass Fa, Henr P. Gavn Stran Energ n Lnear Eastc Sods Consder a force, F, apped gradua to a structure. Let D be the resutng

More information

MACHINE APPLIED MACHINE LEARNING LEARNING. Gaussian Mixture Regression

MACHINE APPLIED MACHINE LEARNING LEARNING. Gaussian Mixture Regression 11 MACHINE APPLIED MACHINE LEARNING LEARNING MACHINE LEARNING Gaussan Mture Regresson 22 MACHINE APPLIED MACHINE LEARNING LEARNING Bref summary of last week s lecture 33 MACHINE APPLIED MACHINE LEARNING

More information

A MIN-MAX REGRET ROBUST OPTIMIZATION APPROACH FOR LARGE SCALE FULL FACTORIAL SCENARIO DESIGN OF DATA UNCERTAINTY

A MIN-MAX REGRET ROBUST OPTIMIZATION APPROACH FOR LARGE SCALE FULL FACTORIAL SCENARIO DESIGN OF DATA UNCERTAINTY A MIN-MAX REGRET ROBST OPTIMIZATION APPROACH FOR ARGE SCAE F FACTORIA SCENARIO DESIGN OF DATA NCERTAINTY Travat Assavapokee Department of Industra Engneerng, nversty of Houston, Houston, Texas 7704-4008,

More information

Greyworld White Balancing with Low Computation Cost for On- Board Video Capturing

Greyworld White Balancing with Low Computation Cost for On- Board Video Capturing reyword Whte aancng wth Low Computaton Cost for On- oard Vdeo Capturng Peng Wu Yuxn Zoe) Lu Hewett-Packard Laboratores Hewett-Packard Co. Pao Ato CA 94304 USA Abstract Whte baancng s a process commony

More information

On the Power Function of the Likelihood Ratio Test for MANOVA

On the Power Function of the Likelihood Ratio Test for MANOVA Journa of Mutvarate Anayss 8, 416 41 (00) do:10.1006/jmva.001.036 On the Power Functon of the Lkehood Rato Test for MANOVA Dua Kumar Bhaumk Unversty of South Aabama and Unversty of Inos at Chcago and Sanat

More information

A General Column Generation Algorithm Applied to System Reliability Optimization Problems

A General Column Generation Algorithm Applied to System Reliability Optimization Problems A Genera Coumn Generaton Agorthm Apped to System Reabty Optmzaton Probems Lea Za, Davd W. Cot, Department of Industra and Systems Engneerng, Rutgers Unversty, Pscataway, J 08854, USA Abstract A genera

More information

Boundary Value Problems. Lecture Objectives. Ch. 27

Boundary Value Problems. Lecture Objectives. Ch. 27 Boundar Vaue Probes Ch. 7 Lecture Obectves o understand the dfference between an nta vaue and boundar vaue ODE o be abe to understand when and how to app the shootng ethod and FD ethod. o understand what

More information

Distributed Moving Horizon State Estimation of Nonlinear Systems. Jing Zhang

Distributed Moving Horizon State Estimation of Nonlinear Systems. Jing Zhang Dstrbuted Movng Horzon State Estmaton of Nonnear Systems by Jng Zhang A thess submtted n parta fufment of the requrements for the degree of Master of Scence n Chemca Engneerng Department of Chemca and

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

Optimization of JK Flip Flop Layout with Minimal Average Power of Consumption based on ACOR, Fuzzy-ACOR, GA, and Fuzzy-GA

Optimization of JK Flip Flop Layout with Minimal Average Power of Consumption based on ACOR, Fuzzy-ACOR, GA, and Fuzzy-GA Journa of mathematcs and computer Scence 4 (05) - 5 Optmzaton of JK Fp Fop Layout wth Mnma Average Power of Consumpton based on ACOR, Fuzzy-ACOR, GA, and Fuzzy-GA Farshd Kevanan *,, A Yekta *,, Nasser

More information

Multispectral Remote Sensing Image Classification Algorithm Based on Rough Set Theory

Multispectral Remote Sensing Image Classification Algorithm Based on Rough Set Theory Proceedngs of the 2009 IEEE Internatona Conference on Systems Man and Cybernetcs San Antono TX USA - October 2009 Mutspectra Remote Sensng Image Cassfcaton Agorthm Based on Rough Set Theory Yng Wang Xaoyun

More information

Image Classification Using EM And JE algorithms

Image Classification Using EM And JE algorithms Machne earnng project report Fa, 2 Xaojn Sh, jennfer@soe Image Cassfcaton Usng EM And JE agorthms Xaojn Sh Department of Computer Engneerng, Unversty of Caforna, Santa Cruz, CA, 9564 jennfer@soe.ucsc.edu

More information

Numerical integration in more dimensions part 2. Remo Minero

Numerical integration in more dimensions part 2. Remo Minero Numerca ntegraton n more dmensons part Remo Mnero Outne The roe of a mappng functon n mutdmensona ntegraton Gauss approach n more dmensons and quadrature rues Crtca anass of acceptabt of a gven quadrature

More information

A Derivative-Free Algorithm for Bound Constrained Optimization

A Derivative-Free Algorithm for Bound Constrained Optimization Computatona Optmzaton and Appcatons, 21, 119 142, 2002 c 2002 Kuwer Academc Pubshers. Manufactured n The Netherands. A Dervatve-Free Agorthm for Bound Constraned Optmzaton STEFANO LUCIDI ucd@ds.unroma.t

More information

Equivalent Standard Deviation to Convert High-reliability Model to Low-reliability Model for Efficiency of Samplingbased

Equivalent Standard Deviation to Convert High-reliability Model to Low-reliability Model for Efficiency of Samplingbased roceedngs of the ASME 0 Internatonal Desgn Engneerng echncal Conferences & Computers and Informaton n Engneerng Conference IDEC/CIE 0 August 8 3, 0, Washngton, D.C., USA DEC0-47537 Equvalent Standard Devaton

More information

The Application of BP Neural Network principal component analysis in the Forecasting the Road Traffic Accident

The Application of BP Neural Network principal component analysis in the Forecasting the Road Traffic Accident ICTCT Extra Workshop, Bejng Proceedngs The Appcaton of BP Neura Network prncpa component anayss n Forecastng Road Traffc Accdent He Mng, GuoXucheng &LuGuangmng Transportaton Coege of Souast Unversty 07

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

Lecture Notes on Linear Regression

Lecture Notes on Linear Regression Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume

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

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

Thermodynamics II. Department of Chemical Engineering. Prof. Kim, Jong Hak

Thermodynamics II. Department of Chemical Engineering. Prof. Kim, Jong Hak Thermodynamcs II Department o Chemca ngneerng ro. Km, Jong Hak .5 Fugacty & Fugacty Coecent : ure Speces µ > provdes undamenta crteron or phase equbrum not easy to appy to sove probem Lmtaton o gn (.9

More information

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M CIS56: achne Learnng Lecture 3 (Sept 6, 003) Preparaton help: Xaoyng Huang Lnear Regresson Lnear regresson can be represented by a functonal form: f(; θ) = θ 0 0 +θ + + θ = θ = 0 ote: 0 s a dummy attrbute

More information

ON AUTOMATIC CONTINUITY OF DERIVATIONS FOR BANACH ALGEBRAS WITH INVOLUTION

ON AUTOMATIC CONTINUITY OF DERIVATIONS FOR BANACH ALGEBRAS WITH INVOLUTION European Journa of Mathematcs and Computer Scence Vo. No. 1, 2017 ON AUTOMATC CONTNUTY OF DERVATONS FOR BANACH ALGEBRAS WTH NVOLUTON Mohamed BELAM & Youssef T DL MATC Laboratory Hassan Unversty MORO CCO

More information

Using T.O.M to Estimate Parameter of distributions that have not Single Exponential Family

Using T.O.M to Estimate Parameter of distributions that have not Single Exponential Family IOSR Journal of Mathematcs IOSR-JM) ISSN: 2278-5728. Volume 3, Issue 3 Sep-Oct. 202), PP 44-48 www.osrjournals.org Usng T.O.M to Estmate Parameter of dstrbutons that have not Sngle Exponental Famly Jubran

More information

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number

More information

Markov Chain Monte Carlo Lecture 6

Markov Chain Monte Carlo Lecture 6 where (x 1,..., x N ) X N, N s called the populaton sze, f(x) f (x) for at least one {1, 2,..., N}, and those dfferent from f(x) are called the tral dstrbutons n terms of mportance samplng. Dfferent ways

More information

Cyclic Codes BCH Codes

Cyclic Codes BCH Codes Cycc Codes BCH Codes Gaos Feds GF m A Gaos fed of m eements can be obtaned usng the symbos 0,, á, and the eements beng 0,, á, á, á 3 m,... so that fed F* s cosed under mutpcaton wth m eements. The operator

More information

Sensitivity Analysis Using Neural Network for Estimating Aircraft Stability and Control Derivatives

Sensitivity Analysis Using Neural Network for Estimating Aircraft Stability and Control Derivatives Internatona Conference on Integent and Advanced Systems 27 Senstvty Anayss Usng Neura Networ for Estmatng Arcraft Stabty and Contro Dervatves Roht Garhwa a, Abhshe Hader b and Dr. Manoranan Snha c Department

More information

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm Desgn and Optmzaton of Fuzzy Controller for Inverse Pendulum System Usng Genetc Algorthm H. Mehraban A. Ashoor Unversty of Tehran Unversty of Tehran h.mehraban@ece.ut.ac.r a.ashoor@ece.ut.ac.r Abstract:

More information

Xin Li Department of Information Systems, College of Business, City University of Hong Kong, Hong Kong, CHINA

Xin Li Department of Information Systems, College of Business, City University of Hong Kong, Hong Kong, CHINA RESEARCH ARTICLE MOELING FIXE OS BETTING FOR FUTURE EVENT PREICTION Weyun Chen eartment of Educatona Informaton Technoogy, Facuty of Educaton, East Chna Norma Unversty, Shangha, CHINA {weyun.chen@qq.com}

More information

Optimum Selection Combining for M-QAM on Fading Channels

Optimum Selection Combining for M-QAM on Fading Channels Optmum Seecton Combnng for M-QAM on Fadng Channes M. Surendra Raju, Ramesh Annavajjaa and A. Chockangam Insca Semconductors Inda Pvt. Ltd, Bangaore-56000, Inda Department of ECE, Unversty of Caforna, San

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

Example: Suppose we want to build a classifier that recognizes WebPages of graduate students.

Example: Suppose we want to build a classifier that recognizes WebPages of graduate students. Exampe: Suppose we want to bud a cassfer that recognzes WebPages of graduate students. How can we fnd tranng data? We can browse the web and coect a sampe of WebPages of graduate students of varous unverstes.

More information

Possibility-Based Design Optimization Method for Design Problems with both Statistical and Fuzzy Input Data

Possibility-Based Design Optimization Method for Design Problems with both Statistical and Fuzzy Input Data 6 th World Congresses of Structural and Multdscplnary Optmzaton Ro de Janero, 30 May - 03 June 2005, Brazl Possblty-Based Desgn Optmzaton Method for Desgn Problems wth both Statstcal and Fuzzy Input Data

More information

Analysis of Bipartite Graph Codes on the Binary Erasure Channel

Analysis of Bipartite Graph Codes on the Binary Erasure Channel Anayss of Bpartte Graph Codes on the Bnary Erasure Channe Arya Mazumdar Department of ECE Unversty of Maryand, Coege Par ema: arya@umdedu Abstract We derve densty evouton equatons for codes on bpartte

More information

Uncertainty Specification and Propagation for Loss Estimation Using FOSM Methods

Uncertainty Specification and Propagation for Loss Estimation Using FOSM Methods Uncertanty Specfcaton and Propagaton for Loss Estmaton Usng FOSM Methods J.W. Baer and C.A. Corne Dept. of Cv and Envronmenta Engneerng, Stanford Unversty, Stanford, CA 94305-400 Keywords: Sesmc, oss estmaton,

More information

QUARTERLY OF APPLIED MATHEMATICS

QUARTERLY OF APPLIED MATHEMATICS QUARTERLY OF APPLIED MATHEMATICS Voume XLI October 983 Number 3 DIAKOPTICS OR TEARING-A MATHEMATICAL APPROACH* By P. W. AITCHISON Unversty of Mantoba Abstract. The method of dakoptcs or tearng was ntroduced

More information

Note 2. Ling fong Li. 1 Klein Gordon Equation Probablity interpretation Solutions to Klein-Gordon Equation... 2

Note 2. Ling fong Li. 1 Klein Gordon Equation Probablity interpretation Solutions to Klein-Gordon Equation... 2 Note 2 Lng fong L Contents Ken Gordon Equaton. Probabty nterpretaton......................................2 Soutons to Ken-Gordon Equaton............................... 2 2 Drac Equaton 3 2. Probabty nterpretaton.....................................

More information

COMBINING SPATIAL COMPONENTS IN SEISMIC DESIGN

COMBINING SPATIAL COMPONENTS IN SEISMIC DESIGN Transactons, SMRT- COMBINING SPATIAL COMPONENTS IN SEISMIC DESIGN Mchae O Leary, PhD, PE and Kevn Huberty, PE, SE Nucear Power Technooges Dvson, Sargent & Lundy, Chcago, IL 6060 ABSTRACT Accordng to Reguatory

More information

A Unified Elementary Approach to the Dyson, Morris, Aomoto, and Forrester Constant Term Identities

A Unified Elementary Approach to the Dyson, Morris, Aomoto, and Forrester Constant Term Identities A Unfed Eementary Approach to the Dyson, Morrs, Aomoto, and Forrester Constant Term Identtes Ira M Gesse 1, Lun Lv, Guoce Xn 3, Yue Zhou 4 1 Department of Mathematcs Brandes Unversty, Watham, MA 0454-9110,

More information

ANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U)

ANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U) Econ 413 Exam 13 H ANSWERS Settet er nndelt 9 deloppgaver, A,B,C, som alle anbefales å telle lkt for å gøre det ltt lettere å stå. Svar er gtt . Unfortunately, there s a prntng error n the hnt of

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

Correspondence. Performance Evaluation for MAP State Estimate Fusion I. INTRODUCTION

Correspondence. Performance Evaluation for MAP State Estimate Fusion I. INTRODUCTION Correspondence Performance Evauaton for MAP State Estmate Fuson Ths paper presents a quanttatve performance evauaton method for the maxmum a posteror (MAP) state estmate fuson agorthm. Under dea condtons

More information

Particular Solutions of Chebyshev Polynomials for Polyharmonic and Poly-Helmholtz Equations

Particular Solutions of Chebyshev Polynomials for Polyharmonic and Poly-Helmholtz Equations Copyrght c 2008 Tech Scence Press CMES, vo.27, no.3, pp.151-162, 2008 Partcuar Soutons of Chebyshev Poynomas for Poyharmonc and Poy-Hemhotz Equatons Cha-Cheng Tsa 1 Abstract: In ths paper we deveop anaytca

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

Approximate merging of a pair of BeÂzier curves

Approximate merging of a pair of BeÂzier curves COMPUTER-AIDED DESIGN Computer-Aded Desgn 33 (1) 15±136 www.esever.com/ocate/cad Approxmate mergng of a par of BeÂzer curves Sh-Mn Hu a,b, *, Rou-Feng Tong c, Tao Ju a,b, Ja-Guang Sun a,b a Natona CAD

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

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

2.29 Numerical Fluid Mechanics

2.29 Numerical Fluid Mechanics REVIEW Lecture 10: Sprng 2015 Lecture 11 Classfcaton of Partal Dfferental Equatons PDEs) and eamples wth fnte dfference dscretzatons Parabolc PDEs Ellptc PDEs Hyperbolc PDEs Error Types and Dscretzaton

More information

Gaussian Processes and Polynomial Chaos Expansion for Regression Problem: Linkage via the RKHS and Comparison via the KL Divergence

Gaussian Processes and Polynomial Chaos Expansion for Regression Problem: Linkage via the RKHS and Comparison via the KL Divergence entropy Artce Gaussan Processes and Poynoma Chaos Expanson for Regresson Probem: Lnkage va the RKHS and Comparson va the KL Dvergence Lang Yan * ID, Xaojun Duan, Bowen Lu and Jn Xu Coege of Lbera Arts

More information

Journal of Multivariate Analysis

Journal of Multivariate Analysis Journa of Mutvarate Anayss 3 (04) 74 96 Contents sts avaabe at ScenceDrect Journa of Mutvarate Anayss journa homepage: www.esever.com/ocate/jmva Hgh-dmensona sparse MANOVA T. Tony Ca a, Yn Xa b, a Department

More information

Application of support vector machine in health monitoring of plate structures

Application of support vector machine in health monitoring of plate structures Appcaton of support vector machne n heath montorng of pate structures *Satsh Satpa 1), Yogesh Khandare ), Sauvk Banerjee 3) and Anrban Guha 4) 1), ), 4) Department of Mechanca Engneerng, Indan Insttute

More information

arxiv: v1 [physics.comp-ph] 17 Dec 2018

arxiv: v1 [physics.comp-ph] 17 Dec 2018 Pressures nsde a nano-porous medum. The case of a snge phase fud arxv:1812.06656v1 [physcs.comp-ph] 17 Dec 2018 Oav Gateand, Dck Bedeaux, and Sgne Kjestrup PoreLab, Department of Chemstry, Norwegan Unversty

More information

Networked Cooperative Distributed Model Predictive Control Based on State Observer

Networked Cooperative Distributed Model Predictive Control Based on State Observer Apped Mathematcs, 6, 7, 48-64 ubshed Onne June 6 n ScRes. http://www.scrp.org/journa/am http://dx.do.org/.436/am.6.73 Networed Cooperatve Dstrbuted Mode redctve Contro Based on State Observer Ba Su, Yanan

More information

Kernel Methods and SVMs Extension

Kernel Methods and SVMs Extension Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general

More information

Ebrahim Sharifi Tashnizi Tafresh University Dep. of Industrial and Mechanical Engineering Tafresh, Iran

Ebrahim Sharifi Tashnizi Tafresh University Dep. of Industrial and Mechanical Engineering Tafresh, Iran Ebrahm Sharf ashnz et a. Ebrahm Sharf ashnz Dr.sharf@taut.ac.r afresh Unversty Dep. of Industra and Mechanca Engneerng afresh, Iran Azadeh Akhavan aher A.AkhavanaherBoroen@student.tudeft.n U Deft Unversty

More information

Quantum Runge-Lenz Vector and the Hydrogen Atom, the hidden SO(4) symmetry

Quantum Runge-Lenz Vector and the Hydrogen Atom, the hidden SO(4) symmetry Quantum Runge-Lenz ector and the Hydrogen Atom, the hdden SO(4) symmetry Pasca Szrftgser and Edgardo S. Cheb-Terrab () Laboratore PhLAM, UMR CNRS 85, Unversté Le, F-59655, France () Mapesoft Let's consder

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

IV. Performance Optimization

IV. Performance Optimization IV. Performance Optmzaton A. Steepest descent algorthm defnton how to set up bounds on learnng rate mnmzaton n a lne (varyng learnng rate) momentum learnng examples B. Newton s method defnton Gauss-Newton

More information

Time-Varying Systems and Computations Lecture 6

Time-Varying Systems and Computations Lecture 6 Tme-Varyng Systems and Computatons Lecture 6 Klaus Depold 14. Januar 2014 The Kalman Flter The Kalman estmaton flter attempts to estmate the actual state of an unknown dscrete dynamcal system, gven nosy

More information

Chapter 6. Rotations and Tensors

Chapter 6. Rotations and Tensors Vector Spaces n Physcs 8/6/5 Chapter 6. Rotatons and ensors here s a speca knd of near transformaton whch s used to transforms coordnates from one set of axes to another set of axes (wth the same orgn).

More information

DISTRIBUTED PROCESSING OVER ADAPTIVE NETWORKS. Cassio G. Lopes and Ali H. Sayed

DISTRIBUTED PROCESSING OVER ADAPTIVE NETWORKS. Cassio G. Lopes and Ali H. Sayed DISTRIBUTED PROCESSIG OVER ADAPTIVE ETWORKS Casso G Lopes and A H Sayed Department of Eectrca Engneerng Unversty of Caforna Los Angees, CA, 995 Ema: {casso, sayed@eeucaedu ABSTRACT Dstrbuted adaptve agorthms

More information

Heuristic Algorithm for Finding Sensitivity Analysis in Interval Solid Transportation Problems

Heuristic Algorithm for Finding Sensitivity Analysis in Interval Solid Transportation Problems Internatonal Journal of Innovatve Research n Advanced Engneerng (IJIRAE) ISSN: 349-63 Volume Issue 6 (July 04) http://rae.com Heurstc Algorm for Fndng Senstvty Analyss n Interval Sold Transportaton Problems

More information

Development of whole CORe Thermal Hydraulic analysis code CORTH Pan JunJie, Tang QiFen, Chai XiaoMing, Lu Wei, Liu Dong

Development of whole CORe Thermal Hydraulic analysis code CORTH Pan JunJie, Tang QiFen, Chai XiaoMing, Lu Wei, Liu Dong Deveopment of whoe CORe Therma Hydrauc anayss code CORTH Pan JunJe, Tang QFen, Cha XaoMng, Lu We, Lu Dong cence and technoogy on reactor system desgn technoogy, Nucear Power Insttute of Chna, Chengdu,

More information

Which Separator? Spring 1

Which Separator? Spring 1 Whch Separator? 6.034 - Sprng 1 Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng 3 Margn of a pont " # y (w $ + b) proportonal

More information

we have E Y x t ( ( xl)) 1 ( xl), e a in I( Λ ) are as follows:

we have E Y x t ( ( xl)) 1 ( xl), e a in I( Λ ) are as follows: APPENDICES Aendx : the roof of Equaton (6 For j m n we have Smary from Equaton ( note that j '( ( ( j E Y x t ( ( x ( x a V ( ( x a ( ( x ( x b V ( ( x b V x e d ( abx ( ( x e a a bx ( x xe b a bx By usng

More information

Downlink Power Allocation for CoMP-NOMA in Multi-Cell Networks

Downlink Power Allocation for CoMP-NOMA in Multi-Cell Networks Downn Power Aocaton for CoMP-NOMA n Mut-Ce Networs Md Shpon A, Eram Hossan, Arafat A-Dwe, and Dong In Km arxv:80.0498v [eess.sp] 6 Dec 207 Abstract Ths wor consders the probem of dynamc power aocaton n

More information

Adaptive Reduction of Design Variables Using Global Sensitivity in Reliability-Based Optimization

Adaptive Reduction of Design Variables Using Global Sensitivity in Reliability-Based Optimization Adaptve Reducton of Desgn Varables Usng Global Senstvty n Relablty-Based Optmzaton Nam H. Km * and Haoyu Wang Dept. of Mechancal & Aerospace Engneerng, Unversty of Florda, Ganesvlle, Florda, 326 Nestor

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

Laboratory 1c: Method of Least Squares

Laboratory 1c: Method of Least Squares Lab 1c, Least Squares Laboratory 1c: Method of Least Squares Introducton Consder the graph of expermental data n Fgure 1. In ths experment x s the ndependent varable and y the dependent varable. Clearly

More information

The Geometry of Logit and Probit

The Geometry of Logit and Probit The Geometry of Logt and Probt Ths short note s meant as a supplement to Chapters and 3 of Spatal Models of Parlamentary Votng and the notaton and reference to fgures n the text below s to those two chapters.

More information

Analytical Uncertainty Propagation via Metamodels in Simulation-Based Design under Uncertainty

Analytical Uncertainty Propagation via Metamodels in Simulation-Based Design under Uncertainty Anaytca Uncertanty Propagaton va etamodes n Smuaton-Based Desgn under Uncertanty We Chen * Integrated DEsgn Automaton Laoratory (IDEAL), orthwestern Unversty, Evanston, IL 68-3, USA Ruchen Jn and Agus

More information

Composite Hypotheses testing

Composite Hypotheses testing Composte ypotheses testng In many hypothess testng problems there are many possble dstrbutons that can occur under each of the hypotheses. The output of the source s a set of parameters (ponts n a parameter

More information

RESEARCH ARTICLE. Solving Polynomial Systems Using a Fast Adaptive Back Propagation-type Neural Network Algorithm

RESEARCH ARTICLE. Solving Polynomial Systems Using a Fast Adaptive Back Propagation-type Neural Network Algorithm Juy 8, 6 8:57 Internatona Journa of Computer Mathematcs poynomas Internatona Journa of Computer Mathematcs Vo., No., Month, 9 RESEARCH ARTICLE Sovng Poynoma Systems Usng a Fast Adaptve Back Propagaton-type

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

Decentralized Adaptive Control for a Class of Large-Scale Nonlinear Systems with Unknown Interactions

Decentralized Adaptive Control for a Class of Large-Scale Nonlinear Systems with Unknown Interactions Decentrazed Adaptve Contro for a Cass of Large-Scae onnear Systems wth Unknown Interactons Bahram Karm 1, Fatemeh Jahangr, Mohammad B. Menhaj 3, Iman Saboor 4 1. Center of Advanced Computatona Integence,

More information