We analyze the problem faced by companies that rely on TL (Truckload) and LTL (Less

Size: px
Start display at page:

Download "We analyze the problem faced by companies that rely on TL (Truckload) and LTL (Less"

Transcription

1 Effecve Zero-Invenory-Orderng Polces for he Sngle-Warehouse Mulrealer Problem wh Pecewse Lnear Cos Srucures LapMu Ann Chan Ana Murel Zuo-Jun Max Shen Davd Smch-Lev Chung-Paw Teo School of Managemen, Unversy of Torono, Torono, Onaro, Canada Deparmen of Mechancal and Indusral Engneerng, Unversy of Massachuses, Amhers, Massachuses Deparmen of Indusral and Sysems Engneerng, Unversy of Florda, Ganesvlle, Florda The Engneerng Sysems Dvson and he Deparmen of Cvl and Envronmenal Engneerng, Massachuses Insue of Technology, Cambrdge, Massachuses Deparmen of Decson Scences, Naonal Unversy of Sngapore, Sngapore We analyze he problem faced by companes ha rely on TL (Truckload) and LTL (Less han Truckload) carrers for he dsrbuon of producs across her supply chan. Our goal s o desgn smple nvenory polces and ransporaon sraeges o sasfy mevaryng demands over a fne horzon, whle mnmzng sysemwde cos by akng advanage of quany dscouns n he ransporaon cos srucures. For hs purpose, we sudy he cos effecveness of resrcng he nvenory polces o he class of zero-nvenory-orderng (ZIO) polces n a sngle-warehouse mulrealer scenaro n whch he warehouse serves as a cross-dock facly. In parcular, we demonsrae ha here exss a ZIO nvenory polcy whose oal nvenory and ransporaon cos s no more han 4/ (5.6/4.6 f ransporaon coss are saonary) mes he opmal cos. However, fndng he bes ZIO polcy s an NPhard problem as well. Thus, we propose wo algorhms o fnd an effecve ZIO polcy: An exac algorhm whose runnng me s polynomal for any fxed number of realers, and a lnear-programmng-based heursc whose effecveness s demonsraed n a seres of compuaonal expermens. Fnally, we exend he wors-case resuls developed n hs paper o sysems n whch he warehouse does hold nvenory. (Approxmaon Algorhms; Zero-Invenory-Orderng Polces; Invenory; Transporaon; Truckload (TL); Less han ruckload (LTL)). Inroducon In recen years, many companes have realzed ha mporan cos savngs can be acheved by negrang nvenory conrol and ransporaon polces hroughou her supply chans. Thus, he problem faced by hese companes s o fnd an opmal replenshmen plan,.e., an nvenory and ransporaon sraegy, so as o mnmze oal nvenory and ransporaon coss over a fne plannng horzon. The dffculy n desgnng a coordnaed sraegy, however, s compounded by he fac ha ypcally hese companes rely on exernal hrd-pary logscs provders for he Managemen Scence 2002 INFORMS Vol. 48, No., November 2002 pp /02/48/446$ elecronc ISSN

2 Zero-Invenory-Orderng Polces ransporaon of goods from supplers hrough warehouses o realers. These problems are dfferen from radonal nework flow problems, as he ransporaon cos srucure, also referred o as orderng cos, offered by he carrers s usually pecewse lnear bu no necessarly convex. Ths cos srucure, represenng quany dscouns, volume-based prce ncenves, and oher forms of economes of scale, has a major mpac on he replenshmen sraegy. I usually reflecs eher ncremenal or all-un dscoun effecs, leadng o he followng ypes of cos funcons. Fgure Cos Modfed All-Un Dscoun Cos Srucure α α α 2 Incremenal Dscoun Cos Srucures. Ths class can be fully characerzed by pecewse lnear concave cos funcons. Of course, a specal case of hs cos funcon s he fxed-charge funcon, where a fxed cos, ndependen of he shpmen sze, s ncurred whenever here s a shpmen. All-Un Dscoun Cos Srucures. Such a cos funcon mples ha f he facly orders Q uns, he ransporaon cos funcon s 0 f Q =0, c f 0<Q<M, G Q = Q f M Q<M 2, Q f M Q<M + =2 n, n Q f M n Q, where > 2 > 0, and M = c. Thus, c s a mnmum charge for shppng a small volume,.e., c s he oal cos when he number of uns shpped s no more han M. Ineresngly, n pracce, when he shpper s plannng o shp Q uns, M Q<M +, he cos s calculaed as F Q = mn G Q G M + = mn Q + M + Tha s, f he order quany s greaer han a ceran value, he shppers pay as f hey were shppng M + uns. Ths s called n he ndusry shppng Q bu declarng M +. Ths commonly used pracce mples ha he rue ransporaon cos funcon, F, has he srucure descrbed by he heavy sold lne n Fgure. As he dashed lnes ndcae, he assocaed sold lnes orgnae a pon 0 0. We refer o such cos funcons as modfed all-un dscoun cos funcons. Noce ha hese cos funcons sasfy he followng properes: M0 M M M2 M2 M Quany () hey are nondecreasng funcons of he amoun shpped, and () he cos per un s nonncreasng n he amoun shpped. Indeed, hese wo properes are necessary and suffcen o derve our resuls. These cos funcons are common n ndusry pracce. Mos LTL carrers use an ndusry sandard ransporaon-rang engne called CZAR (Souhern Moor Carrer s Complee Zp Audng and Rang engne). Ths engne allows he shpper o fnd he ransporaon cos of every shpmen whch s a funcon of he source, desnaon, produc class, and dscoun. The carrer and he shpper conracually agree on he produc class (ypcally class 00) and on he level of dscoun, whch mples ha he shpper wll pay only a gven fracon, say 90%, of he cos generaed by he rang engne. Now, gven hs npu, he ransporaon cos as a funcon of he amoun shpped enjoys he all-un dscoun srucure descrbed above. In hs paper, we sudy a class of mulperod dsrbuon problems wh ransporaon cos srucures ha model boh he ncremenal and all-un dscoun cos funcons. Specfcally, we consder a classcal nvenory dsrbuon model n whch a sngle warehouse receves nvenory from a sngle suppler and replenshes he nvenory of n realers. Each realer provdes he warehouse wh forecas demand for he nex T me perods. We assume ha shorages and backloggng are no allowed eher a he warehouse or a he realers. Managemen Scence/Vol. 48, No., November

3 Zero-Invenory-Orderng Polces Furhermore, we assume ha he warehouse uses a common logsc sraegy, referred o as cross-dockng, n whch he warehouse acs merely as a coordnaor of he supply process and as a ransshpmen pon for ncomng orders from he suppler, bu does no hold any socks. In hese suaons, (large) shpmens from he suppler o he warehouse are ofen delvered by TL carrers whose coss can be approxmaed by pecewse lnear concave funcons (for example, a fxed-charge cos funcon). Economes of scale n producon can also be modeled n hs way. Henceforh, we wll assume ha he orderng cos funcon from he suppler o he warehouse s of he ncremenal dscoun ype. By conras, snce shpmen szes from he warehouse o a realer are relavely small, hese shpmens are ypcally delvered by LTL carrers whose coss follow he modfed all-un dscoun cos srucure. The objecve s o fnd an opmal shpmen plan ha explos he quany dscoun effec and, a he same me, conrols he nvenory holdng cos a he realers end. Observe ha he sngle-warehouse mulrealer problem descrbed here can also be used o model he jon replenshmen problem; see Joneja (990). In hs problem, a sngle facly replenshes a se of ems over a fne horzon. Whenever he facly places an order for a subse of he ems, wo ypes of coss are ncurred: a jon se-upcos and an em-dependen se-upcos. The objecve n he jon replenshmen problem s o decde when and how many uns o order for each em so as o mnmze nvenoryholdng and orderng coss over he plannng horzon. Because he jon replenshmen problem s NP-hard (see Arkn e al. 989), he sngle-warehouse mulrealer problem s also NP-hard even f all ransporaon cos funcons are fxed-charge cos funcons. Evdenly, he fxed-charge cos funcon s a specal case of he all-un dscoun cos funcon consdered n hs paper. Ths mples ha he problem analyzed n hs paper s NP-hard n general. An neresng queson s wheher s NP-hard for a sngle or fxed number of realers. Ths queson was answered by Chan e al. (2002), who show ha a specal case of our problem, n whch a sngle realer s replenshed by a sngle warehouse wh zero ransporaon cos for shpmens o he warehouse and modfed allun dscoun ransporaon coss for shpmens o he realer, s NP-hard. Thus, he sngle-warehouse mulrealer problem descrbed above s NP-hard even for a fxed number of realers. Our focus n hs paper s on a class of polces referred o as zero-nvenory-orderng (ZIO) polces, n whch orders are placed by he realers only when her nvenory levels dropo zero. I s easy o see ha n he case of concave ransporaon cos funcons here exss an opmal ZIO polcy for he snglewarehouse mulrealer problem. Unforunaely, he resul of Arkn e al. (989) mples ha fndng he bes ZIO s n self NP-hard. Of course, our model s more general and, as shown by Chan e al. (2002) for he sngle-warehouse sngle-realer case, he opmal polcy may no be a ZIO polcy. Ths paper bulds upon resuls developed n Chan e al. (999), whch consders more general producon-dsrbuon problems wh pecewse lnear concave coss, and n Chan e al. (2002), whch sudes a smpler dynamc lo-szng problem wh a sngle realer and modfed all-un dscouns. In we exend he resuls n Chan e al. (2002) o he onewarehouse mulrealer case and show ha () here s a ZIO polcy whose assocaed cos s no more han 4/ mes he cos of he opmum replenshmen plan, and () f he orderng cos funcon does no vary over me, hen he cos of he opmal ZIO polcy s no more han 5 6/4 6 mes he opmal cos. Ths leads o he developmen of effcen algorhms. Secon 4 descrbes an exac algorhm ha fnds he bes ZIO polcy and whose runnng me s polynomal for a fxed number of realers. The second algorhm, descrbed n 5, s based on formulang he problem of fndng he bes ZIO polcy as an neger program. Ths neger program has a smlar srucure o he one developed n Chan e al. (999) for he concave case. In fac, we fnd ha he properes of he lnear programmng relaxaon derved n ha paper can be appled o our new neger programmng formulaon as well. The lnear programmng relaxaon of hs model s consequenly solved and s soluon and srucural properes are used o generae a ZIO polcy. An emprcal sudy shows (see 6) ha he heursc algorhm s compuaonally effcen and generaes soluons very close o he opmal ZIO polcy. 448 Managemen Scence/Vol. 48, No., November 2002

4 Zero-Invenory-Orderng Polces Fnally, n 7 we pon ou exensons of he worscase resuls o more general cos srucures and o he case of a radonal nvenory dsrbuon sysem n whch he warehouse may hold nvenory. 2. Noaon and Man Resuls Le n be he number of realers served by he warehouse and T be he lengh of he plannng horzon under consderaon. For each = 2 T,wele K 0 Q0 be he pecewse lnear concave ransporaon cos funcon assocaed wh shppng a quany Q 0 from he suppler o he warehouse a me. Smlarly, for each = 2 n and = 2 T, we denoe by K Q he modfed all-un dscoun ransporaon cos funcon assocaed wh shppng a quany Q from he warehouse o realer a me. Fnally, for each = 2 n and = 2 T, le I denoe he nvenory a realer a he end of perod, h he cos of holdng an em a realer a he end of perod, and d he demand of realer a me. Our objecve s o fnd he sze and mng of shpmens so as o mnmze oal ransporaon and nvenory coss whle sasfyng all demands whou shorages. In wha follows, we wll refer o hs problem as he sngle-warehouse mulrealer problem (SWMR): Problem SWMR s.. Q 0 = n Mn Q = T = [ K 0 Q0 + n K Q + h I = = 2 T Q + I = d + I = n = 2 T I 0 = 0 = 2 n Q 0 I = 0 n = 2 T 0 = n = 2 T () where whou loss of generaly we assume ha nal nvenory s 0; ha s, I0 = 0, for all = 2 n. Le Z be he cos of he opmal soluon o he snglewarehouse mulrealer problem and for any heursc H, le Z H be he cos of he soluon generaed by heursc H. We frs show ha unless P = NP, s no possble o developan algorhm ha runs n polynomal me and generaes, for any nsance of he problem, a soluon whch s whn a facor of O log n from opmaly. Theorem 2.. Suppose here exss a >0 and a polynomal me heursc, H, for he sngle-warehouse mulrealer problem such ha for all nsances hen P = NP Z H Z log n Proof. The proof s based on showng ha he se-coverng problem can be reduced o he snglewarehouse mulrealer problem. I s well known (see Fege 998 or Arora and Sudan 997) ha here s no polynomal me algorhm for he se-coverng problem wh wors-case bound beer han log n, for >0, unless P = NP. Consder an nsance of he se-coverng problem: mn m = x Ax, where A = a s a n m 0- marx. I can be reduced o he sngle-warehouse mulrealer problem wh n realers and m + perods as follows. Le K x = { M x f a = 0 0 f a = K m+ x = M x K 0 x = x { 0 d = for all and = 2 m for all for = 2 m m+ f = 2 m f = m + h = 0 for all for all where M s some large number no less han m, and x = when x>0, and 0 oherwse. The hgh se-upcos a me m + forces realers o order n earler perods. In addon, an order for realer s placed a me only f a = because here s a large fxed cos assocaed wh shpmens Managemen Scence/Vol. 48, No., November

5 Zero-Invenory-Orderng Polces n perods n whch a = 0. Thus, fndng he bes nvenory orderng polcy n hs suaon s equvalen o fndng he mnmum number of orderng perods, whch s deermned by cluserng realers ha wll be served ogeher a a ceran me. Thus, n he remander of hs paper we focus on he analyss of a class of polces referred o as zeronvenory-orderng (ZIO) polces. In hs class, orders are placed only a mes when on-hand nvenory has been fully depleed; ha s, Q I = 0. Le ZZIO be he cos assocaed wh he opmal ZIO polcy. Of course, he opmal polcy for he sngle-warehouse mulrealer problem may no be a ZIO polcy. However, n we show he followng wo heorems. Theorem 2.2. For every nsance of he snglewarehouse mulrealer problem, Z ZIO 4 Z and hs bound s gh. In pracce, he orderng cos funcon does no vary from perod o perod;.e., for all, K 0 = K0 and K = K, = 2 n. In hs case, we show ha he wors-case rao of he cos of he bes ZIO polcy o he opmal cos s no more han 5 6/ Tha s, Theorem 2.. For every nsance of he snglewarehouse mulrealer problem n whch he ransporaon cos funcons are saonary, Z ZIO Z The opmal ZIO polcy can be found n polynomal me for any fxed number of realers, usng he algorhm presened n 4. No surprsngly, however, he compuaonal complexy of hs mehod grows exponenally as he number of realers ncreases. To overcome hs problem, n 5 we propose a lnearprogrammng-based heursc ha runs n polynomal me. Ths algorhm s shown o be very effcen n our compuaonal sudy.. The Effecveness of Zero-Invenory-Orderng Polces In hs secon we analyze he effecveness of ZIO polces and prove Theorems 2.2 and 2.. We sar by showng some srucural properes of feasble soluons o he sngle-warehouse mulrealer problem. Le S be a feasble replenshmen plan for he sysem. Le m be he number of shpmens from he suppler o he warehouse and T 0 = = 2 m be he me epochs n whch hese shpmens occur. Smlarly, le m be he number of orders placed by realer, = 2 n, and T = = 2 m T 0 be he me epochs n whch goods are delvered from he warehouse o realer. Le Q denoe he sze of he shpmen o he realer a me perod and Q 0 = n = Q be he quany requred a he warehouse. Obvously, Q = 0 for all T, = 0 2 n. Observe ha all hese parameers are assocaed wh he nvenory polcy S under consderaon. We om he reference o S n our noaon because s clear wha polcy hey are referrng o a any pon. However, we add he superscrp * o he parameers correspondng o an opmal polcy, S. Consder he nvenory level a realer. Whou loss of generaly, we can assume ha n any replenshmen plan, he shpmen ha arrved a me l+ wll only be used afer he depleon of he shpmen ha arrved a me l. Tha s, whou loss of generaly we assume ha orders are used o sasfy demand n a frs-n-frs-ou bass. Thus, le s l, = 2 n and l = 2 m, s l l, be he earles (or frs) me a shpmen ha arrves n me l s used o sasfy cusomer demand. Observe ha n he opmal sraegy, a shpmen ha arrves a me l may only be needed a me s l (s l > l ). The early shpmen may be due o he need o explo he characerscs of he ransporaon cos funcon o he warehouse or he realer. Tha s, he early delvery may ake advanage of changes n he cos funcon as a funcon of me, and/or he effec of he ransporaon dscoun on he order quany. Consder realer and le H k denoe he cos of holdng an em a realer from me k o he begnnng of perod k+. Tha s, H k h + h + + +h + In addon, le K 0 Q 0 K 0 Q 0 Q + K Q A k Q k 450 Managemen Scence/Vol. 48, No., November 2002

6 Zero-Invenory-Orderng Polces Observaon.. A k represens he orderng cos per un assocaed wh shppng he quany Q o realer a perod k, and s also an upper bound on he un cos assocaed wh orderng addonal uns n ha perod for realer. Ths s due o he concavy of K 0 k and he fac ha cos per un resulng from he LTL charges s nonncreasng wh volume. Lemma.2. Gven any feasble polcy S, here exss a feasble polcy wh lower or equal cos n whch a posve fracon of any order s used o sasfy demand for perods pror o he arrval of he subsequen order. Tha s, k+ > s k k for all = 2 n and k = 2 m. Proof. Suppose ha he curren polcy S does no sasfy hs propery;.e., here exss an ndex k such ha k+ s k. Because boh he orders a perods k and k+ cover demand occurrng on or afer perod k+, he wo orders can be combned and sen n he same perod, eher k or k+. The oal coss assocaed wh orderng he combned quany and holdng he uns n nvenory unl perod k+ s no more han Q + Q mn A + k + H k A k+ Q A k + H k + Q +A k+ whch s he cos assocaed wh orderng hose quanes and holdng he uns n nvenory unl perod k+ n he curren polcy S. Snce all oher coss reman he same when combnng he orders, he above argumen shows ha we can always oban a polcy wh lower or equal cos sasfyng he propery. Gven a polcy S sasfyng he condon n Lemma.2, he order placed by realer n any perod j can be wren as Q j = j Q + j j Q j where j Q 0 < j j denoes he poron of he jh shpmen ha s used o sasfy demands from some me s j < j+ unl he arrval of he j + h shpmen. Ths s used n provng Theorems 2.2 and 2., whch are he subjecs of he subsequen secons... Tme-Varyng Orderng Cos Funcons In wha follows we show ha gven an opmal polcy of he one-warehouse mulrealer problem we can consruc a ZIO polcy whose cos s no more han 4/ mes he cos of he orgnal soluon. For hs purpose, we frs show ha we can focus on a subse of feasble polces. Lemma.. Gven any feasble polcy S, here exss a feasble polcy wh lower or equal cos sasfyng he followng properes.. A fracon of any order s used o sasfy demand for perods pror o he arrval of he subsequen order. Tha s, k+ >s k k for all = 2 n and k = 2 m. 2. If he order a k covers demands occurrng afer he nex order has arrved a perod k+, hen he cos per un assocaed wh orderng and holdng hose uns n nvenory n he earler perod s hgher. Tha s, A k +H k > A k+. Proof. The frs propery s ha n Lemma.2. If he second propery does no hold, hen followng he same reasonng as n Lemma.2, he wo orders can be consoldaed and delvered n perod k whou ncreasng oal coss. For = 2 n and j = 2 m, le K 0 n C j = j x= Qx K 0 n j j x=+ Qx + K Q j j j In wha follows, we allocae an orderng cos of C j o each un of he jh order of realer. Usng hs noaon and addng a superscrp o denoe values a opmaly, he oal cos assocaed wh an opmal polcy S, sasfyng he condons n Lemma., can be wren as: Z = n = [ C Q + C j m j=2 Q j + C j + H j j Q j j Q j + H (2) where H = oal holdng cos of polcy S mnus he cos of carryng, for each = 2 n and j = m, he poron of demand delvered n perod j bu used only afer j+ from perod j o j+. Tha s, H = (oal holdng cos of polcy S n m = j= H j. Managemen Scence/Vol. 48, No., November

7 Zero-Invenory-Orderng Polces We now ransform he opmal polcy S no a ZIO polcy S whose cos s whn 4/ of he opmal cos (2). For ha purpose, we apply he followng procedure. Transformaon Procedure. Sep 0. Le S = S. Sep. Fnd he smalles ndex of a realer, say realer, ha does no sasfy he ZIO polcy and he smalles ndex, k, such ha k < ; ha s, k s he earles perod n whch realer places an order before nvenory has been fully depleed. Sep 2. Eher, Combne : Move k Q from he order a perod k o ha a perod k,or Combne 2: Move k Q from he order a perod k o ha a perod k and k Q from he order a perod k o ha a perod k+, whchever resuls n a lower cos. Sep. If necessary, combne orders whou ncreasng oal cos unl he curren polcy sasfes he condons n Lemma.. Sep 4. Repea Seps o unl all realers sasfy he ZIO propery. Noe ha hs requres a mos nt eraons because he soluon generaed afer each eraon ensures ha one more orderng perod a a realer sasfes he ZIO propery. See Fgures 2,, and 4 for llusraon. Observe ha before he execuon of Sep2 for a ceran ndex k and realer, he oal order quan- Fgure Polcy Obaned f Combne Is Performed ( αk ) Q k d d2 d d4 5 d6 k Q k + d d2 d d4 d 5 d6 d Q k + k Noe. The op depcs he combne operaon: Add he srped par of he order a perod o ha a perod k k. The boom fgure represens he new polcy obaned, wh demand sasfed by each order n a dfferen shade of gray. y a any perod k n he curren polcy S sasfes Q 0 n x= Q x. Ths s rue because all he orders of realers + hrough n, and hose of a perods greaer han k, have no been modfed and he order of realer a perod k may have been ncreased n he prevous eraon. Noe ha f combne 2 s performed for a prevous ndex l, he order a perod l s reduced bu l+ wll no be consdered n Sep of he subsequen sages as he nvenory (ordered a l ) would have been fully depleed exacly a perod l+. Furhermore, even hough he order a perod l+ may have been ncreased, he poron ha wll be used afer perod Fgure 2 Inal Polcy Fgure 4 Polcy Obaned f Combne 2 Is Performed Q Q Q k k k + d d2 d d4 d5 d6 (α k )Q k ( αk )Q k k + d d2 d d4 d5 d6 k + sk s k sk+ Noe. Each shaded block represens he demand (here denoed as d d 2 ) faced by realer a subsequen me perods, from perod k onwards. The sze of he block depcs he magnude of he correspondng demand. The curren polcy S s supermposed n he pcure by ponng ou he perods, k k k+, a whch orders are placed and he parcular demands ha each of he ordered quanes (Q Q Q ) wll cover. k k k+ d d2 d d4 d5 d6 k Noe. The op depcs he combne operaon: Add he vercally srped par of he order a perod k o ha a perod k and he horzonally srped par of he order a perod k o ha a perod k+. The boom fgure represens he new polcy obaned, wh demand sasfed by each order n a dfferen shade of gray. 452 Managemen Scence/Vol. 48, No., November 2002

8 Zero-Invenory-Orderng Polces l+2 (.e., l+ Q remans unchanged under he l+) modfed polcy. Noe ha a each eraon of he ransformaon procedure no new orderng perods are added. Thus, we can consder whou loss of generaly ha he se of orderng perods k remans he same over he eraons;.e., k = k for k = 2 m and = 2 n, even f he order quany n some of hose perods becomes 0 afer Seps 2 and are execued. The followng wo lemmas demonsrae ha a each eraon of Sep2 for a parcular ndex k and realer, he ncrease n cos accrued s no more han onehrd of he correspondng kh erm assocaed wh he realer under consderaon n he expresson of he opmal cos (Equaon (2)), [ C k + H k k Q + C k k Q Ths proves ha he cos of he ZIO polcy generaed by he ransformaon procedure s no more han 4 Z, snce he ndex k s srcly ncreasng n he number of eraons performed and he sum of hose erms for all k and all s no larger han he opmal value, Z. Thus, he cos of he opmal ZIO polcy sasfes he 4 Z bound, as saed n Theorem 2.2. and Lemma.4. Le A x = K0 Q 0 K 0 Q 0 Q + K Q x x x x x x x Q x x+ H x = h = x The ncrease n cos a he combne sep for any realer and ndex k s no more han. A k k Q f combne s execued, 2. A k + H k A k k Q f combne 2 s execued. Proof.. The cos ncrease s no more han A k k Q f combne s execued. Ths s rue because A k s an upper bound on he cos per un assocaed wh exra uns sen a perod k ; see Observaon.. 2. The cos ncrease s no more han A k +H k A k k Q f combne 2 s execued. We prove hs n hree seps. () By reducng he quany ordered a perod k by Q orderng coss n ha perod are reduced by, exacly A k Q by defnon of A, k. () Movng k Q from perod k o k ncreases orderng coss and nvenory coss a perod k by a mos A k + H k k Q. () Movng k Q from perod k o k+ resuls n an ncrease n cos a perod k+ of a mos A k+ k Q A k +H k k Q where he, nequaly s rue because he soluon a any eraon sasfes he second propery n Lemma., and a reducon n nvenory coss of H k k Q. Lemma.5. A each eraon of he combnng sep for a ceran ndex k and realer, he ncrease n cos s no more han [ C k + H k k Q + C k k Q Proof. The proof of hs resul uses he followng key resul n Chan e al. (2002, see proof of Lemma 4.): mn A B + A + B for all A B 0 () The ncrease n cos n he combne sepassocaed wh he kh order of realer s no more han { } mn A k k Q A k + H k A k k Q [ A k + H k k Q + A k k Q where he nequaly s a drec consequence of (). Now, because K 0 s concave and a each eraon x Q 0 n x y= Qy for x k, we have ha x ( ) ( ) n n K 0 Q 0 K 0 Q 0 Q K 0 K 0 x x x x x x x Q y x y= Therefore, for all x k, A x K0 Q 0 K 0 Q 0 Q + K Q x x x x x x x Q x Q y x y=+ K0 n x j= Qj K x 0 n x j=+ Qj + K x x Q x C Q x x Managemen Scence/Vol. 48, No., November

9 Zero-Invenory-Orderng Polces Now observe ha Cx C x for all x k because x = x, Q x = Q x for x>k, and Q k Q k. Thus, he ncrease n cos s bounded by [ C k + H k k Q + C k k Q = [ C k + H k k Q + C k k Q The las equaly s due o he fac ha he curren soluon S, before he eraon s performed, sasfes s x = s x, x = x, x = x, Q x k Q k = k Q k. = Q x for x k, and I remans o show ha he bound s gh. Chan e al. (2002) show ha here exs nsances of he onewarehouse mulrealer problem wh a sngle realer for whch he rao Z ZIO /Z s arbrarly close o 4/. Thus, he 4/ bound canno be mproved n he case of mulple realers. Observe ha he only properes of he modfed all-un dscoun funcon used n he proof of Theorem 2.2 are ha s nondecreasng n he quany shpped and ha he cos per un s nonncreasng n ha quany. Hence, he heorem holds rue for any LTL ransporaon funcon sasfyng hose properes. In a smlar way, holdng coss can be generalzed o be any funcon of he quany held ha sasfes hose wo properes..2. Saonary Orderng Cos Funcons In hs case, we need o consder soluons sasfyng condons slghly dfferen han hose n he prevous secon. For hs purpose, le and K 0 Q = lm 0 + K 0 Q K 0 Q K Q B k = K0 Q 0 + k Q k for each realer, = 2 n and orderng ndex k = 2 m. Noe ha we have dropped he me subndex n he ransporaon cos funcons because hey are consan over me. Tha s, K 0 = K0 and K = K for all and. Observaon.6. K 0 Q 0 k s an upper bound on he per-un cos ha addonal uns o be delvered o he warehouse a me k would ncur and, a he same me, a lower bound on he cos per un ncurred by he curren Q 0 uns beng sen n ha perod. Ths s due o he concavy of he warehouse orderng cos funcon, K 0. Observaon.7. Smlarly, B k s an upper bound on he per-un ransporaon cos ha addonal uns delvered o he realer a me k would ncur, and, a he same me, a lower bound on he cos per un ncurred by he curren Q 0 uns beng sen n ha perod. Ths s explaned by he prevous observaon and he fac ha he cos per un resulng from he LTL charges does no ncrease as he shpmen becomes larger. We are now ready o nroduce he counerpar of Lemma. for he saonary case. Lemma.8. Gven any feasble polcy S, here exss a feasble polcy wh lower or equal cos sasfyng he followng properes.. A posve fracon of any order s used o sasfy demand for perods prevous o he arrval of he subsequen order. Tha s, k+ >s k k for all = 2 n and k = 2 m. 2. If he order a k covers demands occurrng afer he nex order a perod k+ has arrved, hen B k + H k >B k+. Proof. The proof s smlar o ha of Lemma.. Please refer o he echncal appendx for deals ( mansc.pubs.nforms.org/ecompanon.hml ). As n he prevous secon, we ransform an opmal polcy whch sasfes he properes n Lemma.8 no a ZIO polcy and show ha he ncrease n cos due o he ransformaon s no more han /4 6 mes he cos of he opmal soluon. However, hs case requres furher deal o show ha he gher bound holds. In parcular, one more way of combnng orders needs o be nroduced: Combne : Move Q uns from perod k o perod k, and Sep2 of he ransformaon procedure s replaced by he Sep 2 descrbed below. Le K Q K H k + K0 Q 0 + k H k + B k Q k K Q K 2 K 0 Q 0 + k B k Q k 454 Managemen Scence/Vol. 48, No., November 2002

10 Zero-Invenory-Orderng Polces H H k + K0 Q 0 K 0 Q 0 D k Q + k Q k Q D k Q D Proof. Please refer o he echncal appendx for deals. Lemma.0. If he curren soluon S sasfes (4), hen connues o hold afer Sep 2 has been execued. Proof. Ths s proven separaely for each of he hree possble combne seps. Please refer o he echncal appendx for deals. Sep 2. Combnng sep k a realer. If 4.6 K 2 H K + K 2, execue combne. Else f 4.6 K K 2 K + K 2 + K 2, execue combne 2. Else execue combne. Thus, n he remander of hs secon we consder a ransformaon procedure dencal o ha n. excep ha Sep 2 s replaced by Sep 2 and he soluon consdered a each eraon sasfes he condons n Lemma.8 raher han hose n Lemma.. For any nvenory orderng polcy S, le Z S be he sysemwde cos assocaed wh polcy S and Z S k be he oal cos assocaed wh sasfyng he demands of realer a perods k hrough T and all he demands (from perod o T ) of realers + hrough n. Le k be he ndex of he earles orderng perod a realer n whch an order s placed before all nvenory from he prevous order has been depleed. Observe ha o prove Theorem 2., suffces o show ha a any eraon of he ransformaon procedure he curren on-hand soluon S sasfes Z S Z S k Z Z S k (4) Obvously, hs condon holds for S = S. The followng lemmas show ha f he curren on-hand soluon sasfes (4), hen also holds afer execung Sep 2 and Sep. Ths mples ha he curren soluon a any eraon of he ransformaon procedure sasfes (4). Fnally, observe ha he soluon S generaed a he las eraon sasfes Z S k = 0, snce s a ZIO polcy, and hus Theorem 2. follows drecly from (4). Lemma.9. If he curren soluon S sasfes (4), hen connues o hold afer Sep has been execued. 4. Opmal Zero-Invenory- Orderng Polcy In hs secon we show ha when he number of realers s fxed, we can fnd he bes ZIO polcy by formulang an assocaed shores-pah problem, n me whch s polynomal n T and exponenal n he number of realers n. As we have seen, hs ZIO polcy has a cos whn a facor of 4/ from he opmal cos. Le = 2 T + be he se of dfferen me perods, where T + s used for noaonal convenence. Le N = 2 n be he se of realers. Consruc an acyclc graph G = V A, where V = ū = u u n u = n = } {{} n mes A = { u u n v v n v u for all ; here s a leas one componen such ha u <v ; for every wh u <v we have u = mn j= 2 n u j u (.e., all he componens ha changed had he same value, u } Gven an arc ū v, where ū = u u n and v = v v n, le k be he number of componens ha are dfferen n ū and v, and I = 2 k be he se of ndces of hose componens; ha s, for l = 2 k, l s such ha u l <v l. Observe ha k and by consrucon u = u 2 = =u k = u. The arc ū v represens orderng a perod u o sasfy demands of each realer l, l = 2 k, from perod u hrough v l. Thus, he cos assocaed wh hs arc s he cos of orderng hose uns a perod u and Managemen Scence/Vol. 48, No., November

11 Zero-Invenory-Orderng Polces holdng hem n nvenory unl her consumpon. Specfcally, he cos of hs arc s K 0 u d uv + d 2 uv2 + +d k uvk + c uv + c 2 uv2 + +c k uvk where duv s he oal demand faced by realer from perod u o v Ku 0 d uv + d 2 uv2 + +d k uvk s he cos of shppng d uv + d 2 uv2 + +d k uvk uns from he suppler o he warehouse a me u; and, cuv s he oal shppng and holdng cos for realer f we order a perod u o cover he demands n perods u u + v ;.e., v 2 c uv = K u d uv + h j d j+v I s easy o see ha he shores pah from o T + T+ T+ n G = V A corresponds o fndng he bes ZIO polcy. Usng a Fbonacc heapmplemenaon of Djksra s algorhm (Cormen e al. 999), he shores pah can be found n me O V log V + A, where V =O T + n, A =O T 2n, and n s he number of realers. The algorhm grows o be compuaonally expensve as he number of realers ncreases. Thus, he nex seps o developa heursc ha fnds a good ZIO polcy n polynomal me. Ths s he objecve of he nex secon. 5. Lnear-Programmng-Based Algorhm In hs secon we nroduce a lnear-programmngbased heursc ha generaes close-o-opmal ZIO polces and, hus, effecve soluons o he snglewarehouse mulrealer problem. We sar by formulang he problem of fndng an opmal ZIO polcy as an neger program. The algorhm s based on solvng he lnear programmng relaxaon of he resulng model and ransformng he fraconal soluon obaned no an neger soluon. Our mehod s smlar o he mehod presened n Chan e al. (999) for mulcommody nework flows wh pecewse lnear concave coss. The pecewse lnear concave coss assocaed wh shpmens from he suppler o he warehouse are j=u modeled as follows. Consder any me perod and he assocaed warehouse orderng cos funcon, K 0. Le R be he number of dfferen slopes n he cos funcon, whch we assume, whou loss of generaly, s he same for all me perods o avod cumbersome noaon. Le M r, M r, for r = R, denoe he lower and upper lms, respecvely, on he nerval correspondng o he rh slope of he pecewse lnear cos funcon. Noe ha M 0 = 0 and M R can be se o he oal quany ha may be shpped a perod, n l = d l. We assocae wh each of hese nervals, say r, a varable cos per un, denoed by r, equal o he slope of he correspondng lne segmen, and a fxed cos, f r, defned as he y-nercep of he lnear prolongaon of ha segmen. See Fgure 5 for a graphcal represenaon. Observe ha he cos ncurred by shppng a quany on a ceran range s he sum of s assocaed fxed cos plus he cos of shppng all uns a s correspondng lnear cos. Tha s, f we le Q 0 denoe he warehouse order a me, we can express he assocaed ransporaon cos, K 0 Q0,asK0 Q0 = f r + r Q0 where r s such ha Q0 r M M r Fnally, we defne he followng varables. For each = 2 T and r = 2 R, le Fgure 5 Cos f f 2 f M 0 U r = { f Q 0 M r M r, 0 oherwse. Pecewse Lnear Concave Funcon slope α slope α 2 slope α M M 2 M Quany 456 Managemen Scence/Vol. 48, No., November 2002

12 Zero-Invenory-Orderng Polces For each realer = 2 n, and perods k T, le Z = quany ordered by realer a me o sasfy demand a perod k and { Z = f Q 0 r M M r, 0 oherwse, Z r for each r = 2 R. In wha follows we refer o he U varables as nerval varables and o he Z varables as quany varables. To model orderng and nvenory coss a he realer level, we consder a dummy perod T + and defne, as n 4, for each realer = 2 n and perods <k T +, c = oal cos of orderng a perod o sasfy demand for perods hrough k and holdng he uns n nvenory unl her consumpon. Tha s, c = K ( k ) d j j= k 2 + j= h j ( k d l l=j+ Observe ha a ZIO polcy for realer can be nerpreed as a pah from o T + on a nework wh nodes 2 T + and arcs k, for <k T +, wh assocaed cos c. In wha follows, we refer o hs nework as he h realer s nework or G. Thus, o calculae orderng and nvenory coss a realer we formulae a shores-pah model on G usng he varables f an order s placed by realer a me X = o sasfy demands for perods hrough k, 0 oherwse, and flow conservaon consrans. We refer o X = X as he vecor of pah flows. The bes ZIO polcy can be found by solvng he followng neger program: Problem P Mn T R = r= + n [ f r U r + r T n T + c X = = k=+ T = k= Z r ) s R Z r d k U r r = 2 R = 2 n and k T (5) T Z r = + d k X l r= l=k+ k T = 2 n (6) X lj f l = X jl = f l = T + (7) j j>l j j<l 0 f <l T = 2 n Z r 0 r = 2 R = 2 n and k T U r 0 r = 2 R and = 2 T X 0 = 2 n and k T + The frs se of consrans, (5), specfes ha f some quany s ordered a me by any realer and shpped on nerval r of he ransporaon cos funcon, hen he assocaed nerval varable, U r, mus be. These, ogeher wh he negraly of he U varables, are he only consrans needed o model he pecewse lnear concave coss; see Chan e al. (999). Obvously, consrans (5) could be aggregaed for all k. However, hs would consderably weaken he lnear programmng relaxaon of Problem P. Equaon (6) guaranees ha f a posve amoun s shpped o realer a me o sasfy demand a he realer a perod k, hen he realer mus order a perod o cover demands for perods hrough some l k. Observe ha hese consrans lnk he supplerwarehouse model wh he realer model. Fnally, he flow conservaon consrans (7) correspond o fndng, for each realer, a pah from o T + onhe realer s nework, G. Unforunaely, solvng hs neger program s compuaonally nracable for all bu small-sze problems. To overcome hs dffculy, we focus on analyzng he behavor of s lnear programmng relaxaon and ake advanage of s srucural properes o developan effecve heursc ha consrucs Managemen Scence/Vol. 48, No., November

13 Zero-Invenory-Orderng Polces an neger soluon o Problem P from s opmal fraconal soluon. For ha purpose, we sar by fxng he fraconal pah flows and sudyng he behavor of he resulng lnear program. Le X = X be he vecor of pah flows n a feasble soluon o he lnear programmng relaxaon of Problem P. Wha s he cos ha he lnear program assocaes wh hs soluon? Wha are he values of he correspondng nerval and quany varables, U r and Z r? Observe ha gven he vecor of fraconal pah flows X, he amoun o be shpped from he suppler o he warehouse a any me perod s known and herefore he lnear programmng relaxaon of Problem P can be decomposed no mulple subproblems, one for every me perod. For each me perod, he oal shppng cos, as well as he correspondng varables Z r and U r, can be obaned by solvng he followng problem, whch we refer o as he fxed-flow subproblem a perod or Problem FFX : [ R Problem FF X Mn n T f r U r + r Z r s.. R r= = k= Z r d k U r k = T and r = R (8) T Z r = + d k X s r= s=k+ Z r 0 k = T and = n (9) k = T = n and r = R U r 0 r = R The followng lemma, presened n Chan e al. (999), explcly characerzes he soluon o he lnear program for any gven (fxed) vecor of pah flows X. Lemma 5.. For any gven perod of me and fxed vecor of pah flows X, le he proporon of he demand of realer a me k whch s shpped a perod be T + s=k+ X s for k and = 2 n Rank he proporons,,n nondecreasng order of her values. Followng ha order, assocae a sngle ndex l o each par k so ha l and 2 L where L = n T +. Smlarly, le d l d k for each l = 2 L and s correspondng par k. Then, he opmal soluon o he fxed-flow subproblem a me s L l= K 0 ( L d s ) l l where 0 = 0 (0) s=l We are now ready o descrbe a polynomal me heursc ha fnds an effecve ZIO polcy based on he soluon o he lnear programmng relaxaon of Problem P. Lemma 5. wll be exensvely used by he algorhm o compue he ncrease n coss n he soluon o he lnear program when he vecor X s modfed n he search for an neger soluon. Lnear-Programmng-Based Algorhm Sep. Solve he lnear programmng relaxaon of Problem P. Le X = X be he opmal soluon. Inalze =. Sep 2. For each arc k, <k T + n nework G compue a margnal cos, c, as follows. The margnal cos s he oal ncrease n cos n he soluon o he lnear program ncurred when augmenng he flow on ha arc from he fraconal X o. Tha s, c = W + X k c k where W s he ncrease n ransporaon cos o he warehouse resulng from modfyng flow n he lnear program from X o. Ths cos ncrease can be easly calculaed usng Lemma 5.. Sep. Deermne he orderng epochs of realer by fndng he mnmum cos pah from o T + on nework G wh edge coss equal o he margnal coss. Sep 4. Updae he amoun and coss of warehouse orders a each perod o accoun for realer s orderng sraegy. Coss are updaed usng Lemma 5.. Sep 5. Le = + and repea Seps (2) (5) unl = n Managemen Scence/Vol. 48, No., November 2002

14 Zero-Invenory-Orderng Polces Ths algorhm runs n polynomal me. I requres solvng a lnear program wh O T 2 Rn varables and O T 2 Rn consrans, and hen performng Seps 2 hrough 5 o oban an neger soluon. Observe ha when applyng Lemma 5. a any sep of he algorhm, he rankng of he proporons assocaed wh he flow on an edge from suppler o warehouse can be updaed easly n O nt because proporons are only changed o values of 0 or. The ransporaon cos on edge can hen be compued n O nt + R. Ths s done a mos O nt 2 mes over all eraons of Seps 2 and 4. Consequenly, he complexy of performng Seps 2 hrough 5 s no more han O n 2 T + nt 2 R. 6. Compuaonal Resuls In hs secon we es he performance of he lnear-programmng-based algorhm n erms of boh compuaonal me and relave devaon from he opmal ZIO polcy. For hs purpose, he lnearprogrammng-based algorhm was coded n C++ and a varey of es problems, descrbed below, were solved on a Sun Ulra SPARCsaon; CPLEX 5.0 was used o solve he lnear and neger programmng problems. We apply he algorhm o wo ypes of problems. The frs s he sngle-warehouse mulrealer problem wh realer orderng cos represened by he modfed all-un dscoun cos funcon. The second s he sngle-warehouse mulrealer problem wh concave orderng cos funcons for he realers. Of course, n boh ypes of problems, warehouse orderng cos s a pecewse lnear concave funcon. Observe ha here exss an opmal ZIO polcy for he second ype of problems. Type Insances. We consder four problem classes correspondng o 5, 25, 50, and 00 realers. The plannng horzon s 2 perods and demands for each realer are generaed from a normal dsrbuon wh mean 00 and sandard devaon 20. Holdng coss are randomly generaed n he nerval [0.,0.6. Suppler-warehouse ransporaon coss are descrbed by pecewse lnear concave funcons wh fve breakpons beween 0 and he maxmum amoun ha could possbly be ordered from he suppler o sasfy realer demands. We consder he breakpons fxed and randomly vary fxed coss and slopes (whn a ceran sensble range) over me. Smlarly, he warehouse-realer ransporaon cos funcon for a parcular realer s a modfed all-un dscoun funcon wh eher 6, 7, or 8 prce breaks and agan fxed coss and slopes are randomly vared over me. Table shows, for each problem class, he average compuaon me of he lnear-programmng-based algorhm over 200 nsances generaed. For hese moderae-sze nsances esed, he opmal ZIO can be calculaed by solvng he neger program, Problem P, and used o evaluae he performance of he heursc soluon. The assocaed average compuaon mes are gven n he fourh column of Table. The cos of he opmal ZIO polcy obaned s compared o he heursc soluon n he las wo columns: The frs column repors he average rao for cases n whch he soluon o he lnear programmng relaxaon of Problem P was no neger. The second repors he average over all problems esed. Type 2 Insances. Here we sudy he performance of he lnear-programmng-based algorhm for nsances n whch all he ransporaon coss are pecewse lnear and concave. We agan consder dfferen problem classes, wh normally and ndependenly, dencally dsrbued realer demands wh mean 00 and sandard devaon 20, and generae 0 nsances for each class. Holdng coss are se o 0.2 per un per perod. The pecewse lnear concave ransporaon coss consdered have hree prce breaks (.e., four segmens wh dfferen slope) n he range from 0 o he maxmum possble demand ha could be sasfed usng ha lnk. Assocaed fxed coss and varable coss are randomly generaed over me. Table 2 descrbes he sx problem classes esed and repors he average compuaon me and he average rao of heursc o opmal soluons over he fve nsances esed for each class. We observe ha n all he nsances esed, he soluon o he lnear programmng relaxaon concdes wh he opmal neger soluon. Managemen Scence/Vol. 48, No., November

15 Zero-Invenory-Orderng Polces Table Compuaonal Resuls for Modfed All-Un Dscoun Warehouse-Realer Coss Number CPU Tme Frequency of Z H /Z ZIO Z H /Z ZIO Problem of CPU me wh IP fraconal Fraconal All class realers (seconds) (seconds) soluon cases cases Class 5 0 4/ Class / Class / Class / Table 2 Compuaonal Resuls for Concave Transporaon Coss on All Lnks Problem Number of Number of CPU Tme Z H /Z ZIO class perods realers (seconds) = Z H /Z Class Class Class Class Class Class Conclusons and Exensons We should pon ou ha he wors-case resuls descrbed n hs paper hold under farly general sengs. Indeed, a careful nspecon of he proofs of Theorems 2.2 and 2. reveals ha hese resuls hold for any general warehouse-realer orderng funcons and realer holdng coss sasfyng he followng properes: () hey are nondecreasng funcons of he amouns shpped and held, respecvely, and () he assocaed cos per un s nonncreasng n hose amouns. Moreover, boh condons are necessary o oban a fne wors-case performance bound on he bes ZIO polcy. Tha s, f a leas one of hese condons does no hold, hen he bes ZIO polcy has an unbounded wors-case performance (see Chan e al. 2002). Fnally, he bounds on he performance of ZIO polces developed n hs paper can be easly exended o a more general dsrbuon problem wh cenral socks, n whch he warehouse s allowed o carry nvenory. For a proof, please refer o he echncal appendx o hs paper avalable on he Managemen Scence elecronc companon page a hp:\\mansc. pubs.nforms.org/ecompanon./hml. Acknowledgmens The auhors hank Mchael Deampel and Mchael Zemer from Schneder Logscs for nroducng hem o hs problem. Research suppored n par by ONR Conracs N and N , and by NSF Conracs DDM , DMI , and DMI References Arkn E., D. Joneja, R. Roundy Compuaonal complexy of uncapacaed mul-echelon producon plannng problems. Oper. Res. Le Arora, S., M. Sudan Improved low-degree esng and s applcaons. Proc. 29h Annual ACM Sympos. Theory Compu. El Paso, TX, Chan, L. M. A., A. Murel, D. Smch-Lev. 999a. Producon/dsrbuon plannng problems wh pece-wse lnear and concave cos srucures. Workng paper, Norhwesern Unversy, Evanson, IL.,, Z. J. Shen, D. Smch-Lev. 999b. On he effecveness of zero-nvenory-orderng polces for he economc lo szng problem wh a class of pece-wse lnear cos srucures. Oper. Res. Forhcomng. Cormen, T. H., C. E. Leserson, R. L. Rves Inroducon o Algorhms. MIT Press, Cambrdge, MA. Fege, U A hreshold of ln n for approxmang se cover. J. ACM 45(4) Joneja, D The jon replenshmen problem: New heurscs and wors case performance bounds. Oper. Res. 8(4) Acceped by Thomas M. Leblng; receved Aprl, 200. Ths paper was wh he auhors 2 monhs for revson. 460 Managemen Scence/Vol. 48, No., November 2002

Capacity Planning. Operations Planning

Capacity Planning. Operations Planning Operaons Plannng Capacy Plannng Sales and Operaons Plannng Forecasng Capacy plannng Invenory opmzaon How much capacy assgned o each producon un? Realsc capacy esmaes Sraegc level Moderaely long me horzon

More information

Pedro M. Castro Iiro Harjunkoski Ignacio E. Grossmann. Lisbon, Portugal Ladenburg, Germany Pittsburgh, USA

Pedro M. Castro Iiro Harjunkoski Ignacio E. Grossmann. Lisbon, Portugal Ladenburg, Germany Pittsburgh, USA Pedro M. Casro Iro Harjunkosk Ignaco E. Grossmann Lsbon Porugal Ladenburg Germany Psburgh USA 1 Process operaons are ofen subjec o energy consrans Heang and coolng ules elecrcal power Avalably Prce Challengng

More information

How To Calculate Backup From A Backup From An Oal To A Daa

How To Calculate Backup From A Backup From An Oal To A Daa 6 IJCSNS Inernaonal Journal of Compuer Scence and Nework Secury, VOL.4 No.7, July 04 Mahemacal Model of Daa Backup and Recovery Karel Burda The Faculy of Elecrcal Engneerng and Communcaon Brno Unversy

More information

Cooperative Distributed Scheduling for Storage Devices in Microgrids using Dynamic KKT Multipliers and Consensus Networks

Cooperative Distributed Scheduling for Storage Devices in Microgrids using Dynamic KKT Multipliers and Consensus Networks Cooperave Dsrbued Schedulng for Sorage Devces n Mcrogrds usng Dynamc KK Mulplers and Consensus Newors Navd Rahbar-Asr Yuan Zhang Mo-Yuen Chow Deparmen of Elecrcal and Compuer Engneerng Norh Carolna Sae

More information

HEAT CONDUCTION PROBLEM IN A TWO-LAYERED HOLLOW CYLINDER BY USING THE GREEN S FUNCTION METHOD

HEAT CONDUCTION PROBLEM IN A TWO-LAYERED HOLLOW CYLINDER BY USING THE GREEN S FUNCTION METHOD Journal of Appled Mahemacs and Compuaonal Mechancs 3, (), 45-5 HEAT CONDUCTION PROBLEM IN A TWO-LAYERED HOLLOW CYLINDER BY USING THE GREEN S FUNCTION METHOD Sansław Kukla, Urszula Sedlecka Insue of Mahemacs,

More information

MORE ON TVM, "SIX FUNCTIONS OF A DOLLAR", FINANCIAL MECHANICS. Copyright 2004, S. Malpezzi

MORE ON TVM, SIX FUNCTIONS OF A DOLLAR, FINANCIAL MECHANICS. Copyright 2004, S. Malpezzi MORE ON VM, "SIX FUNCIONS OF A DOLLAR", FINANCIAL MECHANICS Copyrgh 2004, S. Malpezz I wan everyone o be very clear on boh he "rees" (our basc fnancal funcons) and he "fores" (he dea of he cash flow model).

More information

SPC-based Inventory Control Policy to Improve Supply Chain Dynamics

SPC-based Inventory Control Policy to Improve Supply Chain Dynamics Francesco Cosanno e al. / Inernaonal Journal of Engneerng and Technology (IJET) SPC-based Invenory Conrol Polcy o Improve Supply Chan ynamcs Francesco Cosanno #, Gulo Gravo #, Ahmed Shaban #3,*, Massmo

More information

RESOLUTION OF THE LINEAR FRACTIONAL GOAL PROGRAMMING PROBLEM

RESOLUTION OF THE LINEAR FRACTIONAL GOAL PROGRAMMING PROBLEM Revsa Elecrónca de Comuncacones y Trabajos de ASEPUMA. Rec@ Volumen Págnas 7 a 40. RESOLUTION OF THE LINEAR FRACTIONAL GOAL PROGRAMMING PROBLEM RAFAEL CABALLERO rafael.caballero@uma.es Unversdad de Málaga

More information

MULTI-WORKDAY ERGONOMIC WORKFORCE SCHEDULING WITH DAYS OFF

MULTI-WORKDAY ERGONOMIC WORKFORCE SCHEDULING WITH DAYS OFF Proceedngs of he 4h Inernaonal Conference on Engneerng, Projec, and Producon Managemen (EPPM 203) MULTI-WORKDAY ERGONOMIC WORKFORCE SCHEDULING WITH DAYS OFF Tar Raanamanee and Suebsak Nanhavanj School

More information

Spline. Computer Graphics. B-splines. B-Splines (for basis splines) Generating a curve. Basis Functions. Lecture 14 Curves and Surfaces II

Spline. Computer Graphics. B-splines. B-Splines (for basis splines) Generating a curve. Basis Functions. Lecture 14 Curves and Surfaces II Lecure 4 Curves and Surfaces II Splne A long flexble srps of meal used by drafspersons o lay ou he surfaces of arplanes, cars and shps Ducks weghs aached o he splnes were used o pull he splne n dfferen

More information

Lecture 40 Induction. Review Inductors Self-induction RL circuits Energy stored in a Magnetic Field

Lecture 40 Induction. Review Inductors Self-induction RL circuits Energy stored in a Magnetic Field ecure 4 nducon evew nducors Self-nducon crcus nergy sored n a Magnec Feld 1 evew nducon end nergy Transfers mf Bv Mechancal energy ransform n elecrc and hen n hermal energy P Fv B v evew eformulaon of

More information

The Multi-shift Vehicle Routing Problem with Overtime

The Multi-shift Vehicle Routing Problem with Overtime The Mul-shf Vehcle Roung Problem wh Overme Yngao Ren, Maged Dessouy, and Fernando Ordóñez Danel J. Epsen Deparmen of Indusral and Sysems Engneerng Unversy of Souhern Calforna 3715 McClnoc Ave, Los Angeles,

More information

Y2K* Stephanie Schmitt-Grohé. Rutgers Uni ersity, 75 Hamilton Street, New Brunswick, New Jersey 08901 E-mail: grohe@econ.rutgers.edu.

Y2K* Stephanie Schmitt-Grohé. Rutgers Uni ersity, 75 Hamilton Street, New Brunswick, New Jersey 08901 E-mail: grohe@econ.rutgers.edu. Revew of Economc Dynamcs 2, 850856 Ž 1999. Arcle ID redy.1999.0065, avalable onlne a hp:www.dealbrary.com on Y2K* Sephane Schm-Grohé Rugers Unersy, 75 Hamlon Sree, New Brunswc, New Jersey 08901 E-mal:

More information

GUIDANCE STATEMENT ON CALCULATION METHODOLOGY

GUIDANCE STATEMENT ON CALCULATION METHODOLOGY GUIDANCE STATEMENT ON CALCULATION METHODOLOGY Adopon Dae: 9/28/0 Effecve Dae: //20 Reroacve Applcaon: No Requred www.gpssandards.org 204 CFA Insue Gudance Saemen on Calculaon Mehodology GIPS GUIDANCE STATEMENT

More information

The Virtual Machine Resource Allocation based on Service Features in Cloud Computing Environment

The Virtual Machine Resource Allocation based on Service Features in Cloud Computing Environment Send Orders for Reprns o reprns@benhamscence.ae The Open Cybernecs & Sysemcs Journal, 2015, 9, 639-647 639 Open Access The Vrual Machne Resource Allocaon based on Servce Feaures n Cloud Compung Envronmen

More information

INTERNATIONAL JOURNAL OF STRATEGIC MANAGEMENT

INTERNATIONAL JOURNAL OF STRATEGIC MANAGEMENT IJSM, Volume, Number, 0 ISSN: 555-4 INTERNATIONAL JOURNAL OF STRATEGIC MANAGEMENT SPONSORED BY: Angelo Sae Unversy San Angelo, Texas, USA www.angelo.edu Managng Edors: Professor Alan S. Khade, Ph.D. Calforna

More information

Network Effects on Standard Software Markets: A Simulation Model to examine Pricing Strategies

Network Effects on Standard Software Markets: A Simulation Model to examine Pricing Strategies Nework Effecs on Sandard Sofware Markes Page Nework Effecs on Sandard Sofware Markes: A Smulaon Model o examne Prcng Sraeges Peer Buxmann Absrac Ths paper examnes sraeges of sandard sofware vendors, n

More information

A Heuristic Solution Method to a Stochastic Vehicle Routing Problem

A Heuristic Solution Method to a Stochastic Vehicle Routing Problem A Heursc Soluon Mehod o a Sochasc Vehcle Roung Problem Lars M. Hvaum Unversy of Bergen, Bergen, Norway. larsmh@.ub.no Arne Løkkeangen Molde Unversy College, 6411 Molde, Norway. Arne.Lokkeangen@hmolde.no

More information

HEURISTIC ALGORITHM FOR SINGLE RESOURCE CONSTRAINED PROJECT SCHEDULING PROBLEM BASED ON THE DYNAMIC PROGRAMMING

HEURISTIC ALGORITHM FOR SINGLE RESOURCE CONSTRAINED PROJECT SCHEDULING PROBLEM BASED ON THE DYNAMIC PROGRAMMING Yugoslav Journal o Operaons Research Volume 19 (2009) Number 2, 281-298 DOI:10.2298/YUJOR0902281S HEURISTIC ALGORITHM FOR SINGLE RESOURCE CONSTRAINED PROJECT SCHEDULING PROBLEM BASED ON THE DYNAMIC PROGRAMMING

More information

Linear Extension Cube Attack on Stream Ciphers Abstract: Keywords: 1. Introduction

Linear Extension Cube Attack on Stream Ciphers Abstract: Keywords: 1. Introduction Lnear Exenson Cube Aack on Sream Cphers Lren Dng Yongjuan Wang Zhufeng L (Language Engneerng Deparmen, Luo yang Unversy for Foregn Language, Luo yang cy, He nan Provnce, 47003, P. R. Chna) Absrac: Basng

More information

Genetic Algorithm with Range Selection Mechanism for Dynamic Multiservice Load Balancing in Cloud-Based Multimedia System

Genetic Algorithm with Range Selection Mechanism for Dynamic Multiservice Load Balancing in Cloud-Based Multimedia System ISSN : 2347-8446 (Onlne) Inernaonal Journal of Advanced Research n Genec Algorhm wh Range Selecon Mechansm for Dynamc Mulservce Load Balancng n Cloud-Based Mulmeda Sysem I Mchael Sadgun Rao Kona, II K.Purushoama

More information

SHIPPING ECONOMIC ANALYSIS FOR ULTRA LARGE CONTAINERSHIP

SHIPPING ECONOMIC ANALYSIS FOR ULTRA LARGE CONTAINERSHIP Journal of he Easern Asa Socey for Transporaon Sudes, Vol. 6, pp. 936-951, 2005 SHIPPING ECONOMIC ANALYSIS FOR ULTRA LARGE CONTAINERSHIP Chaug-Ing HSU Professor Deparen of Transporaon Technology and Manageen

More information

CLoud computing has recently emerged as a new

CLoud computing has recently emerged as a new 1 A Framework of Prce Bddng Confguraons for Resource Usage n Cloud Compung Kenl L, Member, IEEE, Chubo Lu, Keqn L, Fellow, IEEE, and Alber Y. Zomaya, Fellow, IEEE Absrac In hs paper, we focus on prce bddng

More information

THE IMPACT OF QUICK RESPONSE IN INVENTORY-BASED COMPETITION

THE IMPACT OF QUICK RESPONSE IN INVENTORY-BASED COMPETITION Workng Paper WP no 722 November, 2007 THE IMPACT OF QUICK RESPONSE IN INVENTORY-BASED COMPETITION Felpe Caro Vícor Marínez de Albénz 2 Professor, UCLA Anderson School of Managemen 2 Professor, Operaons

More information

Methodology of the CBOE S&P 500 PutWrite Index (PUT SM ) (with supplemental information regarding the CBOE S&P 500 PutWrite T-W Index (PWT SM ))

Methodology of the CBOE S&P 500 PutWrite Index (PUT SM ) (with supplemental information regarding the CBOE S&P 500 PutWrite T-W Index (PWT SM )) ehodology of he CBOE S&P 500 PuWre Index (PUT S ) (wh supplemenal nformaon regardng he CBOE S&P 500 PuWre T-W Index (PWT S )) The CBOE S&P 500 PuWre Index (cker symbol PUT ) racks he value of a passve

More information

DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS. Exponential Smoothing for Inventory Control: Means and Variances of Lead-Time Demand

DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS. Exponential Smoothing for Inventory Control: Means and Variances of Lead-Time Demand ISSN 440-77X ISBN 0 736 094 X AUSTRALIA DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS Exponenal Smoohng for Invenory Conrol: Means and Varances of Lead-Tme Demand Ralph D. Snyder, Anne B. Koehler,

More information

12/7/2011. Procedures to be Covered. Time Series Analysis Using Statgraphics Centurion. Time Series Analysis. Example #1 U.S.

12/7/2011. Procedures to be Covered. Time Series Analysis Using Statgraphics Centurion. Time Series Analysis. Example #1 U.S. Tme Seres Analyss Usng Sagraphcs Cenuron Nel W. Polhemus, CTO, SaPon Technologes, Inc. Procedures o be Covered Descrpve Mehods (me sequence plos, auocorrelaon funcons, perodograms) Smoohng Seasonal Decomposon

More information

arxiv:1407.5820v1 [cs.sy] 22 Jul 2014

arxiv:1407.5820v1 [cs.sy] 22 Jul 2014 Approxmae Regularzaon Pah for Nuclear Norm Based H Model Reducon Nclas Blomberg, Crsan R. Rojas, and Bo Wahlberg arxv:47.58v [cs.sy] Jul 4 Absrac Ths paper concerns model reducon of dynamcal sysems usng

More information

Insurance. By Mark Dorfman, Alexander Kling, and Jochen Russ. Abstract

Insurance. By Mark Dorfman, Alexander Kling, and Jochen Russ. Abstract he Impac Of Deflaon On Insurance Companes Offerng Parcpang fe Insurance y Mar Dorfman, lexander Klng, and Jochen Russ bsrac We presen a smple model n whch he mpac of a deflaonary economy on lfe nsurers

More information

A robust optimisation approach to project scheduling and resource allocation. Elodie Adida* and Pradnya Joshi

A robust optimisation approach to project scheduling and resource allocation. Elodie Adida* and Pradnya Joshi In. J. Servces Operaons and Informacs, Vol. 4, No. 2, 2009 169 A robus opmsaon approach o projec schedulng and resource allocaon Elode Adda* and Pradnya Josh Deparmen of Mechancal and Indusral Engneerng,

More information

PerfCenter: A Methodology and Tool for Performance Analysis of Application Hosting Centers

PerfCenter: A Methodology and Tool for Performance Analysis of Application Hosting Centers PerfCener: A Mehodology and Tool for Performance Analyss of Applcaon Hosng Ceners Rukma P. Verlekar, Varsha Ape, Prakhar Goyal, Bhavsh Aggarwal Dep. of Compuer Scence and Engneerng Indan Insue of Technology

More information

An Anti-spam Filter Combination Framework for Text-and-Image Emails through Incremental Learning

An Anti-spam Filter Combination Framework for Text-and-Image Emails through Incremental Learning An An-spam Fler Combnaon Framework for Tex-and-Image Emals hrough Incremenal Learnng 1 Byungk Byun, 1 Chn-Hu Lee, 2 Seve Webb, 2 Danesh Iran, and 2 Calon Pu 1 School of Elecrcal & Compuer Engr. Georga

More information

Market-Clearing Electricity Prices and Energy Uplift

Market-Clearing Electricity Prices and Energy Uplift Marke-Clearng Elecrcy Prces and Energy Uplf Paul R. Grbk, Wllam W. Hogan, and Susan L. Pope December 31, 2007 Elecrcy marke models requre energy prces for balancng, spo and shor-erm forward ransacons.

More information

How Much Life Insurance is Enough?

How Much Life Insurance is Enough? How Much Lfe Insurance s Enough? Uly-Based pproach By LJ Rossouw BSTRCT The paper ams o nvesgae how much lfe nsurance proecon cover a uly maxmsng ndvdual should buy. Ths queson s relevan n he nsurance

More information

Boosting for Learning Multiple Classes with Imbalanced Class Distribution

Boosting for Learning Multiple Classes with Imbalanced Class Distribution Boosng for Learnng Mulple Classes wh Imbalanced Class Dsrbuon Yanmn Sun Deparmen of Elecrcal and Compuer Engneerng Unversy of Waerloo Waerloo, Onaro, Canada y8sun@engmal.uwaerloo.ca Mohamed S. Kamel Deparmen

More information

The Rules of the Settlement Guarantee Fund. 1. These Rules, hereinafter referred to as "the Rules", define the procedures for the formation

The Rules of the Settlement Guarantee Fund. 1. These Rules, hereinafter referred to as the Rules, define the procedures for the formation Vald as of May 31, 2010 The Rules of he Selemen Guaranee Fund 1 1. These Rules, herenafer referred o as "he Rules", defne he procedures for he formaon and use of he Selemen Guaranee Fund, as defned n Arcle

More information

A Background Layer Model for Object Tracking through Occlusion

A Background Layer Model for Object Tracking through Occlusion A Background Layer Model for Obec Trackng hrough Occluson Yue Zhou and Ha Tao Deparmen of Compuer Engneerng Unversy of Calforna, Sana Cruz, CA 95064 {zhou,ao}@soe.ucsc.edu Absrac Moon layer esmaon has

More information

Index Mathematics Methodology

Index Mathematics Methodology Index Mahemacs Mehodology S&P Dow Jones Indces: Index Mehodology Ocober 2015 Table of Conens Inroducon 4 Dfferen Varees of Indces 4 The Index Dvsor 5 Capalzaon Weghed Indces 6 Defnon 6 Adjusmens o Share

More information

Finance and Economics Discussion Series Divisions of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C.

Finance and Economics Discussion Series Divisions of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C. Fnance and Economcs Dscusson Seres Dvsons of Research & Sascs and Moneary Affars Federal Reserve Board, Washngon, D.C. Prcng Counerpary Rs a he Trade Level and CVA Allocaons Mchael Pyhn and Dan Rosen 200-0

More information

Estimating intrinsic currency values

Estimating intrinsic currency values Cung edge Foregn exchange Esmang nrnsc currency values Forex marke praconers consanly alk abou he srenghenng or weakenng of ndvdual currences. In hs arcle, Jan Chen and Paul Dous presen a new mehodology

More information

Monopolistic Competition and Macroeconomic Dynamics

Monopolistic Competition and Macroeconomic Dynamics Monopolsc Compeon and Macroeconomc Dynamcs Pasquale Commendaore, Unversà d Napol Federco II Ingrd Kubn, Venna Unversy of Economcs and Busness Admnsraon Absrac Modern macroeconomc models wh a Keynesan flavor

More information

Time Series. A thesis. Submitted to the. Edith Cowan University. Perth, Western Australia. David Sheung Chi Fung. In Fulfillment of the Requirements

Time Series. A thesis. Submitted to the. Edith Cowan University. Perth, Western Australia. David Sheung Chi Fung. In Fulfillment of the Requirements Mehods for he Esmaon of Mssng Values n Tme Seres A hess Submed o he Faculy of Communcaons, ealh and Scence Edh Cowan Unversy Perh, Wesern Ausrala By Davd Sheung Ch Fung In Fulfllmen of he Requremens For

More information

The Feedback from Stock Prices to Credit Spreads

The Feedback from Stock Prices to Credit Spreads Appled Fnance Projec Ka Fa Law (Keh) The Feedback from Sock Prces o Cred Spreads Maser n Fnancal Engneerng Program BA 3N Appled Fnance Projec Ka Fa Law (Keh) Appled Fnance Projec Ka Fa Law (Keh). Inroducon

More information

II. IMPACTS OF WIND POWER ON GRID OPERATIONS

II. IMPACTS OF WIND POWER ON GRID OPERATIONS IEEE Energy2030 Alana, Georga, USA 17-18 November 2008 Couplng Wnd Generaors wh eferrable Loads A. Papavaslou, and S. S. Oren UC Berkeley, eparmen of Indusral Engneerng and Operaons esearch, 4141 Echeverry

More information

Selected Financial Formulae. Basic Time Value Formulae PV A FV A. FV Ad

Selected Financial Formulae. Basic Time Value Formulae PV A FV A. FV Ad Basc Tme Value e Fuure Value of a Sngle Sum PV( + Presen Value of a Sngle Sum PV ------------------ ( + Solve for for a Sngle Sum ln ------ PV -------------------- ln( + Solve for for a Sngle Sum ------

More information

Levy-Grant-Schemes in Vocational Education

Levy-Grant-Schemes in Vocational Education Levy-Gran-Schemes n Vocaonal Educaon Sefan Bornemann Munch Graduae School of Economcs Inernaonal Educaonal Economcs Conference Taru, Augus 26h, 2005 Sefan Bornemann / MGSE Srucure Movaon and Objecve Leraure

More information

Ground rules. Guide to the calculation methods of the FTSE Actuaries UK Gilts Index Series v1.9

Ground rules. Guide to the calculation methods of the FTSE Actuaries UK Gilts Index Series v1.9 Ground rules Gude o he calculaon mehods of he FTSE Acuares UK Gls Index Seres v1.9 fserussell.com Ocober 2015 Conens 1.0 Inroducon... 4 1.1 Scope... 4 1.2 FTSE Russell... 5 1.3 Overvew of he calculaons...

More information

Power-of-Two Policies for Single- Warehouse Multi-Retailer Inventory Systems with Order Frequency Discounts

Power-of-Two Policies for Single- Warehouse Multi-Retailer Inventory Systems with Order Frequency Discounts Power-of-wo Polces for Sngle- Warehouse Mult-Retaler Inventory Systems wth Order Frequency Dscounts José A. Ventura Pennsylvana State Unversty (USA) Yale. Herer echnon Israel Insttute of echnology (Israel)

More information

Both human traders and algorithmic

Both human traders and algorithmic Shuhao Chen s a Ph.D. canddae n sascs a Rugers Unversy n Pscaaway, NJ. bhmchen@sa.rugers.edu Rong Chen s a professor of Rugers Unversy n Pscaaway, NJ and Peng Unversy, n Bejng, Chna. rongchen@sa.rugers.edu

More information

Prot sharing: a stochastic control approach.

Prot sharing: a stochastic control approach. Pro sharng: a sochasc conrol approach. Donaen Hanau Aprl 2, 2009 ESC Rennes. 35065 Rennes, France. Absrac A majory of lfe nsurance conracs encompass a guaraneed neres rae and a parcpaon o earnngs of he

More information

Cost- and Energy-Aware Load Distribution Across Data Centers

Cost- and Energy-Aware Load Distribution Across Data Centers - and Energy-Aware Load Dsrbuon Across Daa Ceners Ken Le, Rcardo Banchn, Margare Maronos, and Thu D. Nguyen Rugers Unversy Prnceon Unversy Inroducon Today, many large organzaons operae mulple daa ceners.

More information

An Architecture to Support Distributed Data Mining Services in E-Commerce Environments

An Architecture to Support Distributed Data Mining Services in E-Commerce Environments An Archecure o Suppor Dsrbued Daa Mnng Servces n E-Commerce Envronmens S. Krshnaswamy 1, A. Zaslavsky 1, S.W. Loke 2 School of Compuer Scence & Sofware Engneerng, Monash Unversy 1 900 Dandenong Road, Caulfeld

More information

Proceedings of the 2008 Winter Simulation Conference S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds.

Proceedings of the 2008 Winter Simulation Conference S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds. Proceedngs of he 008 Wner Smulaon Conference S. J. Mason, R. R. Hll, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds. DEMAND FORECAST OF SEMICONDUCTOR PRODUCTS BASED ON TECHNOLOGY DIFFUSION Chen-Fu Chen,

More information

Analyzing Energy Use with Decomposition Methods

Analyzing Energy Use with Decomposition Methods nalyzng nergy Use wh Decomposon Mehods eve HNN nergy Technology Polcy Dvson eve.henen@ea.org nergy Tranng Week Pars 1 h prl 213 OCD/ 213 Dscusson nergy consumpon and energy effcency? How can energy consumpon

More information

Distribution Channel Strategy and Efficiency Performance of the Life insurance. Industry in Taiwan. Abstract

Distribution Channel Strategy and Efficiency Performance of the Life insurance. Industry in Taiwan. Abstract Dsrbuon Channel Sraegy and Effcency Performance of he Lfe nsurance Indusry n Tawan Absrac Changes n regulaons and laws he pas few decades have afeced Tawan s lfe nsurance ndusry and caused many nsurers

More information

A multi-item production lot size inventory model with cycle dependent parameters

A multi-item production lot size inventory model with cycle dependent parameters INERNAIONA JOURNA OF MAHEMAICA MODE AND MEHOD IN APPIED CIENCE A mul-em producon lo ze nvenory model wh cycle dependen parameer Zad. Balkh, Abdelazz Foul Abrac-In h paper, a mul-em producon nvenory model

More information

Global supply chain planning for pharmaceuticals

Global supply chain planning for pharmaceuticals chemcal engneerng research and desgn 8 9 (2 0 1 1) 2396 2409 Conens lss avalable a ScenceDrec Chemcal Engneerng Research and Desgn journal homepage: www.elsever.com/locae/cherd Global supply chan plannng

More information

Optimization of Nurse Scheduling Problem with a Two-Stage Mathematical Programming Model

Optimization of Nurse Scheduling Problem with a Two-Stage Mathematical Programming Model Asa Pacfc Managemen Revew 15(4) (2010) 503-516 Opmzaon of Nurse Schedulng Problem wh a Two-Sage Mahemacal Programmng Model Chang-Chun Tsa a,*, Cheng-Jung Lee b a Deparmen of Busness Admnsraon, Trans World

More information

A Hybrid Method for Forecasting Stock Market Trend Using Soft-Thresholding De-noise Model and SVM

A Hybrid Method for Forecasting Stock Market Trend Using Soft-Thresholding De-noise Model and SVM A Hybrd Mehod for Forecasng Sock Marke Trend Usng Sof-Thresholdng De-nose Model and SVM Xueshen Su, Qnghua Hu, Daren Yu, Zongxa Xe, and Zhongyng Q Harbn Insue of Technology, Harbn 150001, Chna Suxueshen@Gmal.com

More information

Modèles financiers en temps continu

Modèles financiers en temps continu Modèles fnancers en emps connu Inroducon o dervave prcng by Mone Carlo 204-204 Dervave Prcng by Mone Carlo We consder a conngen clam of maury T e.g. an equy opon on one or several underlyng asses, whn

More information

Social security, education, retirement and growth*

Social security, education, retirement and growth* Hacenda P úblca Espa ñola / Revsa de Econom ía P úblca, 198-(3/2011): 9-36 2011, Insuo de Esudos Fscales Socal secury, educaon, reremen and growh* CRUZ A. ECHEVARR ÍA AMAIA IZA** Unversdad del Pa ís Vasco

More information

Optimal Taxation. 1 Warm-Up: The Neoclassical Growth Model with Endogenous Labour Supply. β t u (c t, L t ) max. t=0

Optimal Taxation. 1 Warm-Up: The Neoclassical Growth Model with Endogenous Labour Supply. β t u (c t, L t ) max. t=0 Opmal Taxaon Reference: L&S 3rd edon chaper 16 1 Warm-Up: The Neoclasscal Growh Model wh Endogenous Labour Supply You looked a lle b a hs for Problem Se 3. Sudy planner s problem: max {c,l,,k +1 } =0 β

More information

The US Dollar Index Futures Contract

The US Dollar Index Futures Contract The S Dollar Inde uures Conrac I. Inroducon The S Dollar Inde uures Conrac Redfeld (986 and Eyan, Harpaz, and Krull (988 presen descrpons and prcng models for he S dollar nde (SDX fuures conrac. Ths arcle

More information

Kalman filtering as a performance monitoring technique for a propensity scorecard

Kalman filtering as a performance monitoring technique for a propensity scorecard Kalman flerng as a performance monorng echnque for a propensy scorecard Kaarzyna Bjak * Unversy of Souhampon, Souhampon, UK, and Buro Informacj Kredyowej S.A., Warsaw, Poland Absrac Propensy scorecards

More information

CONTROLLER PERFORMANCE MONITORING AND DIAGNOSIS. INDUSTRIAL PERSPECTIVE

CONTROLLER PERFORMANCE MONITORING AND DIAGNOSIS. INDUSTRIAL PERSPECTIVE Copyrgh IFAC 5h Trennal World Congress, Barcelona, Span CONTROLLER PERFORMANCE MONITORING AND DIAGNOSIS. INDUSTRIAL PERSPECTIVE Derrck J. Kozub Shell Global Soluons USA Inc. Weshollow Technology Cener,

More information

Auxiliary Module for Unbalanced Three Phase Loads with a Neutral Connection

Auxiliary Module for Unbalanced Three Phase Loads with a Neutral Connection CODEN:LUTEDX/TEIE-514/1-141/6 Indusral Elecrcal Engneerng and Auomaon Auxlary Module for Unbalanced Three Phase Loads wh a Neural Connecon Nls Lundsröm Rkard Sröman Dep. of Indusral Elecrcal Engneerng

More information

MODEL-BASED APPROACH TO CHARACTERIZATION OF DIFFUSION PROCESSES VIA DISTRIBUTED CONTROL OF ACTUATED SENSOR NETWORKS

MODEL-BASED APPROACH TO CHARACTERIZATION OF DIFFUSION PROCESSES VIA DISTRIBUTED CONTROL OF ACTUATED SENSOR NETWORKS MODEL-BASED APPROACH TO CHARACTERIZATION OF DIFFUSION PROCESSES IA DISTRIBUTED CONTROL OF ACTUATED SENSOR NETWORKS Kevn L. Moore and YangQuan Chen Cener for Self-Organzng and Inellgen Sysems Uah Sae Unversy

More information

Optimization Design of the Multi-stage Inventory Management for Supply Chain

Optimization Design of the Multi-stage Inventory Management for Supply Chain Inernaonal Journal of u- an e- Servce, Scence an echnology Vol.8, o. 25, pp.35-366 hp://x.o.org/.4257/uness.25.8..34 Opmzaon Desgn of he ul-sage Invenory anagemen for Supply Chan Hongxng Deng*, ng Yuan

More information

(Im)possibility of Safe Exchange Mechanism Design

(Im)possibility of Safe Exchange Mechanism Design (Im)possbly of Safe Exchange Mechansm Desgn Tuomas Sandholm Compuer Scence Deparmen Carnege Mellon Unversy 5 Forbes Avenue Psburgh, PA 15213 sandholm@cs.cmu.edu XaoFeng Wang Deparmen of Elecrcal and Compuer

More information

Event Based Project Scheduling Using Optimized Ant Colony Algorithm Vidya Sagar Ponnam #1, Dr.N.Geethanjali #2

Event Based Project Scheduling Using Optimized Ant Colony Algorithm Vidya Sagar Ponnam #1, Dr.N.Geethanjali #2 Inernaonal Journal of Compuer Trends and Technology (IJCTT) Volume 18 Number 6 Dec 2014 Even Based Projec Schedulng Usng Opmzed An Colony Algorhm Vdya Sagar Ponnam #1, Dr.N.Geehanjal #2 1 Research Scholar,

More information

Applying the Theta Model to Short-Term Forecasts in Monthly Time Series

Applying the Theta Model to Short-Term Forecasts in Monthly Time Series Applyng he Thea Model o Shor-Term Forecass n Monhly Tme Seres Glson Adamczuk Olvera *, Marcelo Gonçalves Trenn +, Anselmo Chaves Neo ** * Deparmen of Mechancal Engneerng, Federal Technologcal Unversy of

More information

ON THE INTEGRATED PRODUCTION, INVENTORY AND DISTRIBUTION ROUTING PROBLEM. By Shuguang Liu. A Dissertation submitted to the. Graduate School-Newark

ON THE INTEGRATED PRODUCTION, INVENTORY AND DISTRIBUTION ROUTING PROBLEM. By Shuguang Liu. A Dissertation submitted to the. Graduate School-Newark ON THE INTEGRATED PRODUCTION INVENTORY AND DISTRIBUTION ROUTING PROBLEM By Shuguang Lu A Dsseraon submed o he Graduae School-Newark Rugers The Sae Unersy of New Jersey n paral fulfllmen of he requremens

More information

Guidelines and Specification for the Construction and Maintenance of the. NASDAQ OMX Credit SEK Indexes

Guidelines and Specification for the Construction and Maintenance of the. NASDAQ OMX Credit SEK Indexes Gudelnes and Specfcaon for he Consrucon and Manenance of he NASDAQ OMX Cred SEK Indexes Verson as of Aprl 7h 2014 Conens Rules for he Consrucon and Manenance of he NASDAQ OMX Cred SEK Index seres... 3

More information

A Reverse Logistics Model for the Distribution of Waste/By-products. Hamid Pourmohammadi, Maged Dessouky*, and Mansour Rahimi

A Reverse Logistics Model for the Distribution of Waste/By-products. Hamid Pourmohammadi, Maged Dessouky*, and Mansour Rahimi A Reverse Logscs Model for he Dsrbuon of Wase/By-producs Hamd Pourmohammad, Maged Dessouy*, and Mansour Rahm Unversy of Souhern Calforna Epsen Deparmen of Indusral and Sysems Engneerng 375 McClnoc Avenue,

More information

Distributed Load Balancing in a Multiple Server System by Shift-Invariant Protocol Sequences

Distributed Load Balancing in a Multiple Server System by Shift-Invariant Protocol Sequences 03 IEEE Wreess Communcaons and Neorkng Conference (WCNC): NETWORS Dsrbued Load Baancng n a Mupe Server Sysem by Shf-Invaran rooco Sequences Yupeng Zhang and Wng Shng Wong Deparmen of Informaon Engneerng

More information

Linear methods for regression and classification with functional data

Linear methods for regression and classification with functional data Lnear mehods for regresson and classfcaon wh funconal daa Glber Sapora Chare de Sasue Appluée & CEDRIC Conservaore Naonal des Ars e Méers 9 rue San Marn, case 44 754 Pars cedex 3, France sapora@cnam.fr

More information

Revision: June 12, 2010 215 E Main Suite D Pullman, WA 99163 (509) 334 6306 Voice and Fax

Revision: June 12, 2010 215 E Main Suite D Pullman, WA 99163 (509) 334 6306 Voice and Fax .3: Inucors Reson: June, 5 E Man Sue D Pullman, WA 9963 59 334 636 Voce an Fax Oerew We connue our suy of energy sorage elemens wh a scusson of nucors. Inucors, lke ressors an capacors, are passe wo-ermnal

More information

A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting*

A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting* journal of compuer and sysem scences 55, 119139 (1997) arcle no. SS971504 A Decson-heorec Generalzaon of On-Lne Learnng and an Applcaon o Boosng* Yoav Freund and Rober E. Schapre - A6 Labs, 180 Park Avenue,

More information

Pricing Rainbow Options

Pricing Rainbow Options Prcng Ranbow Opons Peer Ouwehand, Deparmen of Mahemacs and Appled Mahemacs, Unversy of Cape Town, Souh Afrca E-mal address: peer@mahs.uc.ac.za Graeme Wes, School of Compuaonal & Appled Mahemacs, Unversy

More information

Inventory Management MILP Modeling for Tank Farm Systems

Inventory Management MILP Modeling for Tank Farm Systems 2 h European Sympoum on Compuer Aded Proce Engneerng ESCAPE2 S. Perucc and G. Buzz Ferrar (Edor) 2 Elever B.V. All rgh reerved. Invenory Managemen MILP Modelng for Tank Farm Syem Suana Relva a Ana Paula

More information

The Performance of Seasoned Equity Issues in a Risk- Adjusted Environment?

The Performance of Seasoned Equity Issues in a Risk- Adjusted Environment? The Performance of Seasoned Equy Issues n a Rsk- Adjused Envronmen? Allen, D.E., and V. Souck 2 Deparmen of Accounng, Fnance and Economcs, Edh Cowan Unversy, W.A. 2 Erdeon Group, Sngapore Emal: d.allen@ecu.edu.au

More information

APPLICATION OF CHAOS THEORY TO ANALYSIS OF COMPUTER NETWORK TRAFFIC Liudvikas Kaklauskas, Leonidas Sakalauskas

APPLICATION OF CHAOS THEORY TO ANALYSIS OF COMPUTER NETWORK TRAFFIC Liudvikas Kaklauskas, Leonidas Sakalauskas The XIII Inernaonal Conference Appled Sochasc Models and Daa Analyss (ASMDA-2009) June 30-July 3 2009 Vlnus LITHUANIA ISBN 978-9955-28-463-5 L. Sakalauskas C. Skadas and E. K. Zavadskas (Eds.): ASMDA-2009

More information

t φρ ls l ), l = o, w, g,

t φρ ls l ), l = o, w, g, Reservor Smulaon Lecure noe 6 Page 1 of 12 OIL-WATER SIMULATION - IMPES SOLUTION We have prevously lsed he mulphase flow equaons for one-dmensonal, horzonal flow n a layer of consan cross seconal area

More information

International Journal of Mathematical Archive-7(5), 2016, 193-198 Available online through www.ijma.info ISSN 2229 5046

International Journal of Mathematical Archive-7(5), 2016, 193-198 Available online through www.ijma.info ISSN 2229 5046 Inernaonal Journal of Mahemacal rchve-75), 06, 9-98 valable onlne hrough wwwjmanfo ISSN 9 506 NOTE ON FUZZY WEKLY OMPLETELY PRIME - IDELS IN TERNRY SEMIGROUPS U NGI REDDY *, Dr G SHOBHLTH Research scholar,

More information

Developing a Risk Adjusted Pool Price in Ireland s New Gross Mandatory Pool Electricity Market

Developing a Risk Adjusted Pool Price in Ireland s New Gross Mandatory Pool Electricity Market 1 Developng a Rsk Adjused Pool Prce n Ireland s New Gross Mandaory Pool Elecrcy Marke Déaglán Ó Dónáll and Paul Conlon Absrac-- The Sngle Elecrcy Marke (SEM) Programme, whch esablshed for he frs me a gross

More information

Attribution Strategies and Return on Keyword Investment in Paid Search Advertising

Attribution Strategies and Return on Keyword Investment in Paid Search Advertising Arbuon Sraeges and Reurn on Keyword Invesmen n Pad Search Adversng by Hongshuang (Alce) L, P. K. Kannan, Sva Vswanahan and Abhshek Pan * December 15, 2015 * Honshuang (Alce) L s Asssan Professor of Markeng,

More information

Anomaly Detection in Network Traffic Using Selected Methods of Time Series Analysis

Anomaly Detection in Network Traffic Using Selected Methods of Time Series Analysis I. J. Compuer Nework and Informaon Secury, 2015, 9, 10-18 Publshed Onlne Augus 2015 n MECS (hp://www.mecs-press.org/) DOI: 10.5815/jcns.2015.09.02 Anomaly Deecon n Nework Traffc Usng Seleced Mehods of

More information

An Ensemble Data Mining and FLANN Combining Short-term Load Forecasting System for Abnormal Days

An Ensemble Data Mining and FLANN Combining Short-term Load Forecasting System for Abnormal Days JOURNAL OF SOFTWARE, VOL. 6, NO. 6, JUNE 0 96 An Ensemble Daa Mnng and FLANN Combnng Shor-erm Load Forecasng Sysem for Abnormal Days Mng L College of Auomaon, Guangdong Unversy of Technology, Guangzhou,

More information

UNIFICATION OF OVERHEAD LINES IN THE CONDITIONS OF THE MARKET OF TWO-PARTY AGREEMENTS AND BALANCING ELECTRIC ENERGY MARKET

UNIFICATION OF OVERHEAD LINES IN THE CONDITIONS OF THE MARKET OF TWO-PARTY AGREEMENTS AND BALANCING ELECTRIC ENERGY MARKET ENERGETIS AND ELETRIAL ENGINEERING P. D. Lezhnuk, Dc. Sc. (Eng.), Prof.; M. M. heremsn, and. Sc. (Eng.), Prof.; V. V. herkashna, and. Sc. (Eng.), Ass. Prof. UNIFIATION OF OVERHEAD LINES IN THE ONDITIONS

More information

Currency Exchange Rate Forecasting from News Headlines

Currency Exchange Rate Forecasting from News Headlines Currency Exchange Rae Forecasng from News Headlnes Desh Peramunelleke Raymond K. Wong School of Compuer Scence & Engneerng Unversy of New Souh Wales Sydney, NSW 2052, Ausrala deshp@cse.unsw.edu.au wong@cse.unsw.edu.au

More information

Case Study on Web Service Composition Based on Multi-Agent System

Case Study on Web Service Composition Based on Multi-Agent System 900 JOURNAL OF SOFTWARE, VOL. 8, NO. 4, APRIL 2013 Case Sudy on Web Servce Composon Based on Mul-Agen Sysem Shanlang Pan Deparmen of Compuer Scence and Technology, Nngbo Unversy, Chna PanShanLang@gmal.com

More information

Long Run Underperformance of Seasoned Equity Offerings: Fact or an Illusion?

Long Run Underperformance of Seasoned Equity Offerings: Fact or an Illusion? Long Run Underperformance of Seasoned Equy Offerngs: Fac or an Illuson? 1 2 Allen D.E. and V. Souck 1 Edh Cowan Unversy, 2 Unversy of Wesern Ausrala, E-Mal: d.allen@ecu.edu.au Keywords: Seasoned Equy Issues,

More information

This research paper analyzes the impact of information technology (IT) in a healthcare

This research paper analyzes the impact of information technology (IT) in a healthcare Producvy of Informaon Sysems n he Healhcare Indusry Nrup M. Menon Byungae Lee Lesle Eldenburg Texas Tech Unversy, College of Busness MS 2101, Lubbock, Texas 79409 menon@ba.u.edu The Unversy of Illnos a

More information

Contract design and insurance fraud: an experimental investigation *

Contract design and insurance fraud: an experimental investigation * Conrac desgn and nsurance fraud: an expermenal nvesgaon * Frauke Lammers and Jörg Schller Absrac Ths paper nvesgaes he mpac of nsurance conrac desgn on he behavor of flng fraudulen clams n an expermenal

More information

Decentralized Model Reference Adaptive Control Without Restriction on Subsystem Relative Degrees

Decentralized Model Reference Adaptive Control Without Restriction on Subsystem Relative Degrees 1464 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 44, NO 7, JULY 1999 [5] S Kosos, Fne npu/oupu represenaon of a class of Volerra polynoal syses, Auoaca, vol 33, no 2, pp 257 262, 1997 [6] S Kosos and D

More information

[ ] Econ4415 International trade. Trade with monopolistic competition and transportation costs

[ ] Econ4415 International trade. Trade with monopolistic competition and transportation costs Econ445 Inernaonal rade Trade wh monopolsc compeon and ransporaon coss We have nroduced some basc feaures of he Dx-Sglz modellng approach. We shall now also nroduce rade coss n he Dx-Sglz framework. I

More information

Fundamental Analysis of Receivables and Bad Debt Reserves

Fundamental Analysis of Receivables and Bad Debt Reserves Fundamenal Analyss of Recevables and Bad Deb Reserves Mchael Calegar Assocae Professor Deparmen of Accounng Sana Clara Unversy e-mal: mcalegar@scu.edu February 21 2005 Fundamenal Analyss of Recevables

More information

A Common Neural Network Model for Unsupervised Exploratory Data Analysis and Independent Component Analysis

A Common Neural Network Model for Unsupervised Exploratory Data Analysis and Independent Component Analysis A Common Neural Nework Model for Unsupervsed Exploraory Daa Analyss and Independen Componen Analyss Keywords: Unsupervsed Learnng, Independen Componen Analyss, Daa Cluserng, Daa Vsualsaon, Blnd Source

More information

A 3D Model Retrieval System Using The Derivative Elevation And 3D-ART

A 3D Model Retrieval System Using The Derivative Elevation And 3D-ART 3 Model Rereal Sysem Usng he erae leaon nd 3-R Jau-Lng Shh* ng-yen Huang Yu-hen Wang eparmen of ompuer Scence and Informaon ngneerng hung Hua Unersy Hsnchu awan RO -mal: sjl@chueduw bsrac In recen years

More information

Using Cellular Automata for Improving KNN Based Spam Filtering

Using Cellular Automata for Improving KNN Based Spam Filtering The Inernaonal Arab Journal of Informaon Technology, Vol. 11, No. 4, July 2014 345 Usng Cellular Auomaa for Improvng NN Based Spam Flerng Faha Bargou, Bouzane Beldjlal, and Baghdad Aman Compuer Scence

More information