Predictive Control of a Smart Grid: A Distributed Optimization Algorithm with Centralized Performance Properties*

Size: px
Start display at page:

Download "Predictive Control of a Smart Grid: A Distributed Optimization Algorithm with Centralized Performance Properties*"

Transcription

1 Predctve Contro of a Smart Grd: A Dstrbuted Optmzaton Agorthm wth Centrazed Performance Propertes* Phpp Braun, Lars Grüne, Chrstopher M. Keett 2, Steven R. Weer 2, and Kar Worthmann 3 Abstract The authors recenty proposed severa mode predctve contro MPC) approaches to managng resdenta eve energy generaton and storage, ncudng centrazed, dstrbuted, and decentrazed schemes. As expected, the dstrbuted and decentrazed schemes resut n a oss of performance but are scaabe and more fexbe wth regards to network topoogy. n ths paper we present a dstrbuted optmzaton approach whch asymptotcay recovers the performance of the centrazed optmzaton probem performed n MPC at each tme step. Smuatons usng data from an Austraan eectrcty dstrbuton company, Ausgrd, are provded showng the beneft of a varabe step sze n the agorthm and the mpact of an ncreasng number of partcpatng resdenta energy systems. Furthermore, when used n a recedng horzon scheme, smuatons ndcate that teratng the teratve dstrbuted optmzaton agorthm before convergence does not resut n a sgnfcant oss of performance.. NTRODUCTON Wth the proferaton of resdenta rooftop soar photovotacs and the ncreasng avaabty of cost-effectve resdenta-scae energy storage soutons e.g., batteres or fue ces), there s a need to coordnate the storage charge/dscharge schedues so as to avod arge demand peaks or troughs. n [4], [3], the authors proposed three dfferent mode predctve contro MPC) schemes to smooth the energy demand of a coecton of resdences. These MPC schemes nvoved a centrazed approach, requrng fu communcaton of a reevant system varabes, a dstrbuted approach, requrng mted communcaton of reevant system varabes, and a decentrazed approach, requrng no communcaton of system varabes. Whe a three approaches succeeded n smoothng the aggregate energy demand, unsurprsngy the centrazed approach acheved better performance when compared to the dstrbuted and decentrazed approaches, but suffered from an nabty to scae to a arge number of resdenta systems. n ths paper, we present a dstrbuted optmzaton agorthm wth the goa of recoverng the performance of the centrazed MPC scheme whst remanng scaabe. n other *C.M. Keett s supported by ARC Future Feowshp FT746. L. Grüne s supported by the Deutsche Forschungsgemenschaft, Grand GR 569/3-. P. Braun and L. Grüne are wth the Mathematca nsttute, Unverstät Bayreuth, 9544 Bayreuth, Germany, e-ma: {phpp.braun, ars.gruene}@un-bayreuth.de. 2 C. M. Keett and S. R. Weer are wth the Schoo of Eectrca Engneerng and Computer Scence at the Unversty of Newcaste, Caaghan, New South Waes 238, Austraa, e-ma: {chrs.keett, steven.weer}@newcaste.edu.au. 3 K. Worthmann s wth the nsttute for Mathematcs, Technsche Unverstät menau, menau, Germany, e-ma: kar.worthmann@tumenau.de. words, we focus on the souton of a snge, fnte tme horzon, optmzaton probem mpemented n a dstrbuted fashon. At east n the contro terature, the fed of dstrbuted optmzaton traces ts roots to the thess of Tstsks [] see aso [2]). Much of the recent work n ths fed has nvoved mut-agent systems tryng to optmze a goba objectve functon under dfferent condtons; see for exampe [5], [7], [8], [9], [5] and the references theren. A common feature n many of these references s the assumpton that the goba cost functon can be decomposed as a sum of the cost functons for each ndvdua agent. However, the cost functon naturay used to sove the probem of smoothng the energy demand s not decomposabe n ths way. n [4], a cosey reated probem s soved where an eectrcty retaer ams to mze the cost due to dscrepances between the power the retaer bds to use and what ts customers actuay use. Agan, ths gves rse to a dfferent cost functon to that whch we propose. The paper s organzed as foows. n Secton we ntroduce the mathematca mode of the Resdenta Energy System RES) and defne the desred performance metrcs. The centrazed MPC approach s presented n Secton and our proposed dstrbuted computaton agorthm s descrbed n Secton V-A. A bref comparson wth prma/dua decomposton s provded n Secton V-B. A smuaton study usng data from an Austraan eectrcty dstrbuton company, Ausgrd, s undertaken n Secton V. n partcuar, we demonstrate the beneft of a varyng step-sze n the dstrbuted optmzaton agorthm Secton V-A), we exae the mpact of ncreasng the number of systems Secton V-B), and the effect of eary teraton of the dstrbuted optmzaton agorthm s ustrated Secton V- C). Concudng remarks are provded n Secton V.. THE RESDENTAL ENERGY SYSTEM Let N be the number of RESs connected n the oca area under consderaton. We summarze a smpe mode of RES, {,..., }, presented n [3] x k + ) x k) + T u k), ) z k) w k) + u k) where x s the state of charge of the battery n [kwh], u s the battery charge/dscharge rate n [kw], w s the statc oad us the oca generaton n [kw], and z s the power supped by/to the grd n [kw]. Here, T represents the ength of the sampng nterva n [h] hours); e.g., T.5[h] corresponds to 3 utes. Whe the system dynamcs )

2 s autonomous, the performance output 2) depends on the tme varyng quantty w ). The RES network s then defned by the foowng dscrete-tme system xk + ) fxk), uk)), zk) huk), wk)) where x, u, w, z R, and the defntons of f and h are gven componentwse by ) and 2), respectvey. For each RES {,..., }, the constrants on the battery capacty and charge/dscharge rates are descrbed by the constants C, u R > and u R <,.e., x k) C and u u k) u k N. 3) Our goa s to fatten the performance output z. We ntroduce two reevant performance metrcs. To ths end, et Πk) : z k) denote the average power demand at tme k and et N denote the number of sampes comprsng a smuaton ength. The performance metrc of peak-to-peak PTP) varaton of the average demand of a RESs s gven by ) ) max Πk) k {,...,N } Πk) k {,...,N }. PTP) The second performance metrc of the root-mean-square RMS) devaton from the average s defned as N N k Πk) Υ wth the average demand Υ : N N k w k).. MODEL PREDCTVE CONTROL APPROACHES RMS) We reca a mode predctve contro MPC) agorthm for the contro of a network of RESs ntroduced n [3] and [4], respectvey. Ths approach s a centrazed MPC CMPC) scheme, n whch fu communcaton of a reevant varabes for the entre network as we as a known mode of the network are requred. n Secton V-A we present a dstrbuted optmzaton agorthm whch s based on oca optmzaton probems, keepng the fexbty of the network topoogy, whe mantanng optmaty wth respect to the CMPC approach. A correspondng proof of convergence s gven n the Appendx. MPC teratvey mzes an optmzaton crteron wth respect to predcted trajectores and mpements the frst part of the resutng optma contro sequence unt the next optmzaton s performed see, e.g., [] or [6]). To ths end, we assume that we have predctons of the resdenta oad and generaton some tme nto the future that s concdent wth the horzon of the predctve controer. n other words, gven a predcton horzon N N, we assume knowedge of w j) for j {k,..., k + N }, where k N s the current tme. A. Centrazed Mode Predctve Contro CMPC) To mpement the CMPC agorthm, we compute the network-wde average demand at every tme step k over the predcton horzon by ζk) : N k+ jk w j) 4) and then mze the jont cost functon k+ V xk); k) : ζk) w j) + û j)) û ) } {{ } jk ẑ j) 5) wth respect to the predcted contro nputs û ) û ), û 2 ),..., û )) T wth û ) û j)) k+ jk, {, 2,..., }, subject to the system dynamcs ), the current state xk) x k),..., x k)) T, and the constrants 3) for {,..., }. The vector of the predcted performance output ẑ ) s defned n the same way as the predcted contro û ). To smpfy the notaton, the current tme k s dropped when t does not dever extra nformaton. Addtonay we use the notaton uj) u j),..., u j)) T for a fxed tme j N. The same hods for the other varabes x, w and z. n Fgure the aggregated energy profe and the aggregated battery profe for a smuaton of one week N 336, T.5[h]) for RESs, nta condtons x ).5[kWh], constrants u u.3[kw] and battery capacty C 2[kWh] for a {,..., } are vsuazed. The oad and generaton data for ths smuaton was coected by an Austraan eectrcty dstrbuton company, Ausgrd, as part of ther Smart Grd, Smart Cty project. The fgures compare the uncontroed system dynamcs wth the cosed oop dynamcs of CMPC. z n [KW] Uncontroed CMPC Tme n hours x n [KWh] Tme n hours Fg.. Performance of CMPC for a smuaton ength of one week and RES. The eft fgure shows the average power demand whe the rght fgure shows the average state of charge of the batteres. V. CENTRALZED MPC WTH DSTRBUTED COMPUTATON n ths secton, we propose a herarchca dstrbuted mode predctve contro DMPC) approach where each RES can communcate wth a centra entty to acheve the performance of the CMPC agorthm,.e., a network-wde objectve whe keepng fexbty. The optma vaue returned by the dstrbuted optmzaton agorthm concdes wth the optma vaue of the mzaton probem 5) cf. the Appendx for a proof).

3 A. The Dstrbuted Optmzaton Agorthm The dstrbuted optmzaton agorthm s based on the cost functon 5) ntroduced n the centrazed settng. nstead of sovng one mzaton probem, severa teratons are performed at every tme step k n whch every RES mzes ony over ts own contro varabes. The centra entty communcates the aggregated performance output between the systems and computes an approprate step sze θ n every teraton. At tme step k, the agorthm s ntazed wth ζ : ζk) cf. Equaton 4)), w j) : w k + j), j,..., N, {, 2,..., }, and x) : xk). Agorthm Dstrbuted Optmzaton Agorthm nput: RES, {, 2,..., }: nta state of charge x ), predcton horzon N, energy profe w j)), and ζ. Centra Entty: Number of RESs, N, ζ, maxma teraton number max N { }, desred precson ε. ntazaton: RES, {, 2,..., }: defne and transmt ẑ j)) and ẑ j)). Centra Entty: Set the teraton counter and V, receve ẑ j)), {.2...., }. Phase Centra Entty): ncrement the teraton counter. Then, receve ẑ j)),, 2,...,. Compute the step sze θ as arg θ [,] ζ θẑ j) + θ)ẑ j) 6) Compute ẑ + j) : θ ẑ j) + θ )ẑ j) and the predcted average demand Π j) : ẑ+ j) for j {,,..., N }. Then, evauate the performance ndex V + : ζ Π j). 7) f V + V < ε or max hods, terate the agorthm. Otherwse transmt θ and Π j)) Phase 2 RES, {, 2,..., }): Receve θ and Π j)) For j,,..., N compute ẑ + j) : θ ẑ j) + θ )ẑ j) 8) Sove the oca) mzaton probem û ) ζ Π j) + ẑ+ j) w j) + û j) subject to the system dynamcs ), ˆx ) x ), and the constrants 3) to obtan the unque mzer ẑ + j)) : w j) + û + j)). Transmt ẑ + j)). Note that Π ) ony depends on ẑ + ). The ndex s chosen n such a way that n teraton, the predcted average Π ) has to be transmtted. A feasbe ntazaton of RES, {,..., } s for exampe gven by ẑ j) ẑ j) : w j), whch corresponds to the choce u ) and can be repaced by any other admssbe ntazaton. Agorthm s terated ether after a fxed number of teratons max or f the stoppng crtera V + V < ε s satsfed. The communcaton structure of Agorthm s vsuazed n Fgure 2. teraton `, Phase teraton `, Phase 2 CE Compute ` Update ẑ`+ Compute ` ẑ?` ẑ?` `, ` ẑ?`+ RES Update ẑ`+ Compute ẑ?`+ `, ` Fg. 2. Communcaton structure of Agorthm. The nput uk) s defned by the update rue of Equaton 8),.e., as a convex combnaton of the ast two computed nputs. Snce the constrants defne a convex set t s ensured that ẑ + ) corresponds to an admssbe nput sequence n every teraton. Theorem.4 n the Appendx ensures that the vaue V converges to the unque optma vaue f the teraton ndex tends to nfnty. Moreover, note that sovng the mzaton probem 6) s equvaent to a smpe functon evauaton as proven n the foowng proposton. Proposton 4.: f ẑ j)) ẑ j)), the parameter θ n teraton s gven by the projecton of θ : ζ ẑ j) )) ẑ j) ẑ j)) ẑ j) ẑ j))) to the nterva [, ],.e., θ max{, { θ, }}. Proof: n order to show the asserton, we defne the functon F θ) : ζ ζ θẑ j) + θ)ẑ j) ẑ j) θ ẑ j) ẑ j) ) Snce F s strcty convex, the asserton foows by sovng F θ) and projectng the souton on the nterva [, ]. Hence, showng that θ soves F ˆθ) competes the proof.

4 Ths foows by computng 2 /2 F θ): [ ) ] ζ ẑ j) θ j) j) ζ ẑ j) )) j) θ j) wth j) ẑ j) ẑ j). Remark 4.2: Aternatvey to the varabe step sze θ computed n Equaton 6), the fxed step sze θ / eads to a decrease of the optma vaue V n every teraton and convergence to the optma vaue of the CMPC mzaton probem whch s an mmedate consequence of the proof of Lemma.2. n Secton V the mpact of a fxed and a varabe step sze s ustrated by smuatons. n every teraton, the centra entty communcates N vaues the average consumpton at each tme wthn the predcton wndow) and the parameter θ to a RESs. n the reverse drecton, each RES transmts N vaues n each teraton. Hence, the amount of data transmtted by the centra entty s ndependent of the number of systems and the nformaton can be broadcast. Snce the optmzaton probems are soved by the RESs ndvduay, the compexty of the agorthm does not grow wth the number of systems. The centra entty does not make use of the constrants 3). Changng system dynamcs, constrants or addng/removng snge systems can be acheved easy on a oca eve, makng the agorthm ncey scaabe n contrast to CMPC. B. Comparson to prma and dua decomposton n ths secton we compare Agorthm wth prma and dua decomposton agorthms descrbed n [3]. Decomposton approaches descrbe methods to break a snge optmzaton probem nto severa optmzaton probems whch are easer to sove. Prma decomposton refers to the decomposton of the orgna probem whe dua decomposton manpuates the dua formuaton. Consder the mzaton probem v,y s.t. fv, y) v, y) P gven n [3]. Here f denotes a convex functon and P a poyhedron of sutabe dmenson. Assume that the functon f and the poyhedron P can be spt such that the mzaton probem 9) can be equvaenty wrtten as v,y f v, y) 9) s.t. v P,..., y P y ) wth convex functons f and poyhedra P y and P for {,..., }. Hence the objectve functon s decouped wth respect to the varabes v, and for a fxed vaue y P y, one can sove the mzaton probems v f v, y) s.t. v P ) separatey. Ths technque of rewrtng 9) as severa probems of the form ) s caed prma decomposton. To sove the probem n a dstrbuted way, ) s soved for a {,..., } and a fxed vaue y P y. Afterwards, the optmzaton varabe y s updated and the process s repeated unt an optma souton s found. n our case, the mzaton probem 5) can be wrtten as v,...,v f v,..., v ) s.t. v P,..., where v u and fv,..., v ) ξj) v j) wth constant vaues ξj). Observe that due to the square, the functon f s not separabe wth respect to the varabes v,..., v. Addtonay, an anaog of the varabe y does not exst n our settng. Nevertheess, t s possbe to fnd smartes between prma decomposton and Agorthm. We defne the vaues y j) ξj) ṽ j) j; j for gven vaues ṽ j). Then we can defne the functons fv, y ) y j) v j) and the correspondng mzaton probems v f v, y ) s.t. v P whch are separated for constant vaues y or constant vaues ṽ, respectvey. Hence, the mzaton probems can be soved n a dstrbuted manner by teratvey updatng ṽ. One way of updatng ṽ s gven by Agorthm. n contrast to prma decomposton, however, we pont out that n our case y s not an optmzaton varabe and we need an ndvdua y for every f. n dua decomposton, the mzaton probem ) s wrtten n the form v,y f v, y ) s.t. v P,..., y P y,..., y y j, j,...,. 2) nstead of fxng the parameter y, y s used as an addtona optmzaton varabe. The optmzaton probem 2) can be separated by ookng at the Lagrangan and fxng the Lagrange varabes. n dua decomposton, the mzaton probems are soved for the unknowns x, y ) and fxed Lagrange varabes for the next teraton, the Lagrange varabes are updated unt a souton s found. As emphaszed above, the varabe y does not exst n our objectve functon and hence, dua decomposton s not appcabe n our context.

5 V. A NUMERCAL CASE STUDY A numerca case study s presented n order to show the beneft of DMPC compared to CMPC. Ths case study s based on anonymzed oad and generaton profes of resdenta customers provded by an Austraan eectrcty dstrbuton company, Ausgrd, based n New South Waes. The numerca experments are conducted usng the nteror pont sover POPT [2] and the HSL mathematca software brary [] to sove the underyng mzaton probems and near systems of equatons, respectvey. For a numerca experments we fx the nta vaues x ).5[kWh] and the constrants C 2[kWh], u u.3[kw] for a {,..., }. A. Choce of the Step Sze θ n ths subsecton we nvestgate the roe of the step ength θ. To ths end, 2 RESs are smuated for a duraton of 3 days N 44, T.5[h]). n Fgure 3 we vsuaze the number of teratons unt a certan accuracy V k) V k), {, 2,..., 5}, s reached. a) Varabe θ b) Fxed θ / Number of teratons Tme ndex k Number of teratons Tme ndex k Fg. 3. Number of teratons to obtan a certan accuracy ε for,..., 5 at tme step k,.e., V k) V k) ε where V denotes the souton of the centrazed MPC agorthm. constant ne after approxmatey 2 teratons s due to the optmzaton accuracy of POPT. N N k V k) V k) teraton Fg. 4. Average speed of convergence of the dstrbuted optmzaton agorthm wth fxed θ / back) and varabe θ bue). B. mpact of the Number of Systems Next, we anayze the dependence of the average) number of teratons on the number of RESs. To ths end, the number of RESs,, s vared wthn the set {, 2,..., 3}. Then, the number of teratons s counted unt the accuracy V V 2 s obtaned both for varabe and fxed step sze θ. n Fgure 5, we observe a near growth n the number of teratons for fxed θ whe ths number s sgnfcanty smaer and seems to grow subneary n the case of varabe θ. n concuson, the number of teratons stays moderate for varabe θ whe t may become too arge for θ / to make the agorthm appcabe for a very) arge number of RESs. f a fxed step sze θ / s used nstead of a varabe θ accordng to Proposton 4. the requred number of teratons s, on average, twce as arge, see Tabe. Accuracy average no. maxmum no. mum no. θ / varabe / varabe / varabe ε ε ε ε ε TABLE Average, mum, and maxmum number of teratons to acheve a certan accuracy for varabe and fxed θ. Average number of teratons Number of RES n Fgure 4 the average devaton n teraton from the N benchmark CMPC souton,.e., N k V k) V k), s vsuazed. The average s taken wth respect to each sampng nstant k wth smuaton ength N 44. Hence, the convergence speed of the dstrbuted optmzaton agorthm wth step sze θ n accordance wth Proposton 4. ceary outperforms ts counterpart usng constant θ /. The Fg. 5. Average number of teratons needed to ensure the accuracy V V 2 n dependence of the number of RES wth fxed θ / red) and wth varabe θ bue). The dashed nes show the maxma and ma number of teratons. C. mperfect Optmzaton Agorthm needs about 42 teratons on average to obtan an accuracy of 2 n the settng of RESs and

6 varabe θ, cf. Fgure 5. However, n practce, t may be necessary to terate the agorthm after a fxed number of teratons; e.g., due to a fxed aowabe computaton tme. We exae two ssues. The frst s merey the performance of Agorthm wth a fxed number of teratons. The second s the cosed oop performance of Agorthm wth a fxed number of teratons when used n a recedng horzon fashon. We frst compute the devaton V k) V k) at each tme nstant k wthn the smuaton wndow and, then, we anayze the MPC cosed oop performance. f the step sze θ s chosen such that 6) s soved n each teraton the tota devaton s st arge after teratons, but the cosed oop performance aready ooks convncng, see Fgure 6. V corresponds to a arge sma) devaton from the average ζ. Therefore, we use the absoute error nstead of the reatve error V V ɛ V V ɛ V as a quatatve measure of the resuts. f V s sma the performance wth respect to our metrcs s good even f the reatve error mght st be arge. The choce ε 2 for most of the numerca smuatons seems to be reasonabe for our appcaton, but can be repaced by any other vaue. V k) V k) teratons 5 teratons teratons Tme ndex k z n [KW] CMPC DMPC Tme n hours Fg. 6. Devaton and MPC cosed oop evouton for RESs usng varabe θ and ncompete optmzaton teratons). On the contrary, the cosed oop performance s not satsfactory for fxed θ as seen n Fgure 7. V k) V k) Tme ndex k z n [KW] Tme n hours Fg. 7. Devaton and MPC cosed oop evouton for RESs usng fxed θ and ncompete optmzaton teratons). The same concusons can be drawn for even smaer teraton numbers see Tabe ). Number of teratons varabe θ θ / DMPC - CMPC PTP RMS PTP RMS TABLE Devaton of Dstrbuted MPC wth ncompete optmzaton and CMPC for RES n dependence of the step sze θ. Remark 5.: For the consdered data set n ths secton,.e., the 44 sampes and a varabe number of RESs, the vaues of V are n the nterva [.54,.85]. A arge sma) V. CONCLUSON n ths paper we have presented a dstrbuted optmzaton agorthm for the appcaton to the probem of smoothng energy consumpton n a resdenta eectrcty network where resdences have sma scae generaton e.g., rooftop soar photovotac panes) and storage e.g., a battery). Ths teratve message-passng agorthm asymptotcay recovers the optma vaue of the centrazed optmzaton probem. Va a smuaton study, the dstrbuted optmzaton agorthm has been shown to scae we wth the number of systems and, when used n an MPC scheme, to retan good performance when the agorthm s terated after a fxed number of teratons. Furthermore, we have demonstrated the beneft of mpementng a varabe step sze. APPENDX n ths secton, we prove convergence of Agorthm to the optma vaue of 5),.e., we show that the mt V : m V correspondng to Agorthm concdes wth the optma vaue V of the mzaton probem ẑ ),...,ẑ ) s.t. ζ ẑj) x) ˆx) x j + ) x j) + T u j) z j) w j) + u j) u u j) u x j + ) C, j) {,..., } {,..., N } 3) whch has to be soved n every tme step of CMPC. To ths end, we defne the functons v ẑ ); ) : ζ Π j) + ẑ j) ẑ j) ) 4) and rewrte the oca mzaton probem from Phase 2 of

7 Agorthm ẑ ) vẑ ); ) s.t. x ) ˆx ) x j + ) x j) + T u j) z j) w j) + u j) u u j) u x j + ) C j {,..., N } 5) for {,..., }. The constrants of 5) defne a convex and compact set. The functon v s strcty convex and contnuous n ẑ ) and n the parameters ζ Π j)+ẑ j)/, j {,..., N }. Hence the optma vaue v ẑ ); ), where ẑ ) denotes the unque mzer of the oca mzaton probem, depends contnuousy on the parameters ζ Π j) + ẑ j)/, j {,..., N }. Snce we w use ths resut n the foowng we w state t n a Lemma. Lemma.: The optma vaue v ẑ ); ) of the oca mzaton probem 5) of RES {,..., } s contnuous wth respect to the parameters ζ Π j) + ẑ j)/, j {,..., N }. Before we can prove the convergence of the sequence V ) N we show the weaker resut of monotoncty. Lemma.2: The sequence V ) N generated by Agorthm s monotoncay decreasng,.e., V + V hods for a N. f, addtonay, ẑ ) ẑ ), then V + < V hods. Hence, the sequence V ) N s strcty monotoncay decreasng unt Agorthm stops. Proof: Snce θ [, ] s chosen such that F θ) attans ts mum, see Remark 4., repacng θ by yeds a arger vaue V + V. ζ Π j) ζ ẑ j) + ζ Π j) + θ θ ζ Π j) + ẑ j) ẑ ẑ j) ẑ ẑ j) ẑ ) j) ) j) ) ) j) ζ Π j) + ) ẑ j) ẑ j) } {{ } v ẑ ); ) v ẑ );) ζ Π j) The frst nequaty foows wth θ /. The second nequaty foows from the defnton of convex functons or Jensen s nequaty),.e., M ) M M f α x α fx ), α, α m m m apped to fx) x 2. The thrd nequaty s a drect consequence of the optmaty of ẑ ). Snce v ; ) s strcty convex we obtan v ẑ ); ) < v ẑ ); ) f there exsts an ndex, j) {, 2,..., } {,,..., } such that ẑ j) ẑ j) hods. The proof of Lemma.2 shows that / s a possbe, fxed, choce for θ n Agorthm. Hence, the convergence aso hods f the optma) step sze n Agorthm s repaced by the step sze /. Coroary.3: For the sequence V ) N R of Agorthm converges,.e., m V V R. Proof: Snce V and V ) N s monotoncay decreasng by Lemma.3, V ) N converges to ts nfmum V. n Lemma.2 and Coroary.3 we have shown that the sequence V ) N s convergng. What s eft to show, s the convergence aganst the vaue correspondng to the mzaton probem 3) whch w be done next. Theorem.4: The mt V of the sequence V ) N generated by Agorthm concdes wth the optma vaue V of the mzaton probem 3). Proof: Let z ) denote the souton of Probem 3). Snce the cost functon s contnuous and defned on a compact set, there exsts an admssbe) accumuaton pont z ) of the sequence ẑ )) satsfyng the equaty ζ 2 z j)) V. We frst assume that the mt ẑ ) s obtaned n fntey many teratons,.e., there exsts a j N such that ẑ j ) ẑ ). We defne the functon F : [, ] R as F θ) : ζ ζ θ )z j) + θ z j) ) z j) To show the asserton, we assume θ z j) z j)) F ) V < V F ). 6) Snce F ) s convex, ts drectona dervatve n wth respect to θ s ess than zero,.e., > grad F ), F θ ). 7) nequaty 6) mpes the exstence of an ndex {,..., } such that z ) z F ) and, thus, > θ ) hods. However, then the -th RES updates ẑ ), cf. 4) a contradcton to the assumpton that V s the mt of V ) N accordng to Lemma.2 snce the update ẑ j + ) j + eads to a better vaue V < V. f the accumuaton pont ẑ ) s not reached n fntey many steps then there exsts a subsequence j k ) k N such that m k ẑ j k ) ẑ ). Then, due to the contnuty of the optma vaue functon c.f. Lemma.) there exsts a k N such that

8 V j k + < V whch agan contradcts the propertes of V. REFERENCES [] HSL Mathematca Software Lbrary. A coecton of Fortran codes for arge-scae scentfc computaton, [2] D. P. Bertsekas and J. N. Tstsks. Parae and Dstrbuted Computaton: Numerca Methods. Athena Scentfc, Bemont, MA, USA, 989. [3] S. Boyd, L. Xao, A. Mutapcc, and J. Mattngey. Notes on decomposton methods. Technca report, Stanford Unversty, 27. [4] T.-H. Chang, A. Nedć, and A. Scagone. Dstrbuted constraned optmzaton by consensus-based prma-dua perturbaton method. EEE Transactons on Automatc Contro, 596): , 24. [5] J. C. Duch, A. Agarwa, and M. J. Wanwrght. Dua averagng for dstrbuted optmzaton: Convergence anayss and network scang. EEE Transactons on Automatc Contro, 573):592 66, 22. [6] L. Grüne and J. Pannek. Nonnear Mode Predctve Contro. Theory and Agorthms. Sprnger London, 2. [7] D. Jakovetcć, J. Xaver, and J. M. F. Moura. Fast dstrbuted gradent methods. EEE Transactons on Automatc Contro, 595):3 46, 24. [8] A. Nedć and A. Ozdagar. Dstrbuted subgradent methods for mut-agent optmzaton. EEE Transactons on Automatc Contro, 54):48 6, 29. [9] A. Nedć, A. Ozdagar, and P. A. Paro. Constraned consensus and optmzaton n mut-agent networks. EEE Transactons on Automatc Contro, 554): , 2. [] J. B. Rawngs and D. Q. Mayne. Mode Predctve Contro: Theory and Desgn. Nob H Pubshng, 29. [] J. N. Tstsks. Probems n Decentrazed Decson Makng and Computaton. PhD thess, MT, Cambrdge, MA, USA, 984. [2] A. Wächter and L. T. Beger. On the mpementaton of a prmadua nteror pont fter ne search agorthm for arge-scae nonnear programg. Mathematca Programg, 6):25 57, 26. [3] K. Worthmann, C. M. Keett, P. Braun, L. Grüne, and S. R. Weer. Dstrbuted and decentrazed contro of resdenta energy systems ncorporatng battery storage. EEE Transactons on Smart Grd, 25. Do:.9/TSG [4] K. Worthmann, C. M. Keett, L. Grüne, and S. R. Weer. Dstrbuted contro of resdenta energy systems usng a market maker. n 9th FAC Word Congress, South Afrca, pages , 24. [5] M. Zhu and S. Martínez. On dstrbuted convex optmzaton under nequaty and equaty constrants. EEE Transactons on Automatc Contro, 57):5 64, 22.

TCP/IP Interaction Based on Congestion Price: Stability and Optimality

TCP/IP Interaction Based on Congestion Price: Stability and Optimality TCP/IP Interacton Based on Congeston Prce: Stabty and Optmaty Jayue He Eectrca Engneerng Prnceton Unversty Ema: jhe@prncetonedu Mung Chang Eectrca Engneerng Prnceton Unversty Ema: changm@prncetonedu Jennfer

More information

Approximation Algorithms for Data Distribution with Load Balancing of Web Servers

Approximation Algorithms for Data Distribution with Load Balancing of Web Servers Approxmaton Agorthms for Data Dstrbuton wth Load Baancng of Web Servers L-Chuan Chen Networkng and Communcatons Department The MITRE Corporaton McLean, VA 22102 chen@mtreorg Hyeong-Ah Cho Department of

More information

Multi-agent System for Custom Relationship Management with SVMs Tool

Multi-agent System for Custom Relationship Management with SVMs Tool Mut-agent System for Custom Reatonshp Management wth SVMs oo Yanshan Xao, Bo Lu, 3, Dan Luo, and Longbng Cao Guangzhou Asan Games Organzng Commttee, Guangzhou 5063, P.R. Chna Facuty of Informaton echnoogy,

More information

Dynamic Virtual Network Allocation for OpenFlow Based Cloud Resident Data Center

Dynamic Virtual Network Allocation for OpenFlow Based Cloud Resident Data Center 56 IEICE TRANS. COMMUN., VOL.E96 B, NO. JANUARY 203 PAPER Speca Secton on Networ Vrtuazaton, and Fuson Patform of Computng and Networng Dynamc Vrtua Networ Aocaton for OpenFow Based Coud Resdent Data Center

More information

A Simple Congestion-Aware Algorithm for Load Balancing in Datacenter Networks

A Simple Congestion-Aware Algorithm for Load Balancing in Datacenter Networks A Smpe Congeston-Aware Agorthm for Load Baancng n Datacenter Networs Mehrnoosh Shafee, and Javad Ghader, Coumba Unversty Abstract We study the probem of oad baancng n datacenter networs, namey, assgnng

More information

An Efficient Job Scheduling for MapReduce Clusters

An Efficient Job Scheduling for MapReduce Clusters Internatona Journa of Future Generaton ommuncaton and Networkng, pp. 391-398 http://dx.do.org/10.14257/jfgcn.2015.8.2.32 An Effcent Job Schedung for MapReduce usters Jun Lu 1, Tanshu Wu 1, and Mng We Ln

More information

Predicting Advertiser Bidding Behaviors in Sponsored Search by Rationality Modeling

Predicting Advertiser Bidding Behaviors in Sponsored Search by Rationality Modeling Predctng Advertser Bddng Behavors n Sponsored Search by Ratonaty Modeng Hafeng Xu Centre for Computatona Mathematcs n Industry and Commerce Unversty of Wateroo Wateroo, ON, Canada hafeng.ustc@gma.com Dy

More information

Asymptotically Optimal Inventory Control for Assemble-to-Order Systems with Identical Lead Times

Asymptotically Optimal Inventory Control for Assemble-to-Order Systems with Identical Lead Times Asymptotcay Optma Inventory Contro for Assembe-to-Order Systems wth Identca ead Tmes Martn I. Reman Acate-ucent Be abs, Murray H, NJ 07974, marty@research.be-abs.com Qong Wang Industra and Enterprse Systems

More information

An Ensemble Classification Framework to Evolving Data Streams

An Ensemble Classification Framework to Evolving Data Streams Internatona Journa of Scence and Research (IJSR) ISSN (Onne): 39-7064 An Ensembe Cassfcaton Framework to Evovng Data Streams Naga Chthra Dev. R MCA, (M.Ph), Sr Jayendra Saraswathy Maha Vdyaaya, Coege of

More information

Swing-Free Transporting of Two-Dimensional Overhead Crane Using Sliding Mode Fuzzy Control

Swing-Free Transporting of Two-Dimensional Overhead Crane Using Sliding Mode Fuzzy Control Swng-Free Transportng of Two-Dmensona Overhead Crane Usng Sdng Mode Fuzzy Contro Dantong Lu, Janqang, Dongn Zhao, and We Wang Astract An adaptve sdng mode fuzzy contro approach s proposed for a two-dmensona

More information

Recurrence. 1 Definitions and main statements

Recurrence. 1 Definitions and main statements Recurrence 1 Defntons and man statements Let X n, n = 0, 1, 2,... be a MC wth the state space S = (1, 2,...), transton probabltes p j = P {X n+1 = j X n = }, and the transton matrx P = (p j ),j S def.

More information

8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by

8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by 6 CHAPTER 8 COMPLEX VECTOR SPACES 5. Fnd the kernel of the lnear transformaton gven n Exercse 5. In Exercses 55 and 56, fnd the mage of v, for the ndcated composton, where and are gven by the followng

More information

On the Interaction between Load Balancing and Speed Scaling

On the Interaction between Load Balancing and Speed Scaling On the Interacton between Load Balancng and Speed Scalng Ljun Chen, Na L and Steven H. Low Engneerng & Appled Scence Dvson, Calforna Insttute of Technology, USA Abstract Speed scalng has been wdely adopted

More information

The Dynamics of Wealth and Income Distribution in a Neoclassical Growth Model * Stephen J. Turnovsky. University of Washington, Seattle

The Dynamics of Wealth and Income Distribution in a Neoclassical Growth Model * Stephen J. Turnovsky. University of Washington, Seattle The Dynamcs of Weath and Income Dstrbuton n a Neocassca Growth Mode * Stephen J. Turnovsy Unversty of Washngton, Seatte Ceca García-Peñaosa CNRS and GREQAM March 26 Abstract: We examne the evouton of the

More information

On the Interaction between Load Balancing and Speed Scaling

On the Interaction between Load Balancing and Speed Scaling On the Interacton between Load Balancng and Speed Scalng Ljun Chen and Na L Abstract Speed scalng has been wdely adopted n computer and communcaton systems, n partcular, to reduce energy consumpton. An

More information

DEFINING %COMPLETE IN MICROSOFT PROJECT

DEFINING %COMPLETE IN MICROSOFT PROJECT CelersSystems DEFINING %COMPLETE IN MICROSOFT PROJECT PREPARED BY James E Aksel, PMP, PMI-SP, MVP For Addtonal Informaton about Earned Value Management Systems and reportng, please contact: CelersSystems,

More information

Expressive Negotiation over Donations to Charities

Expressive Negotiation over Donations to Charities Expressve Negotaton over Donatons to Chartes Vncent Contzer Carnege Meon Unversty 5000 Forbes Avenue Pttsburgh, PA 523, USA contzer@cs.cmu.edu Tuomas Sandhom Carnege Meon Unversty 5000 Forbes Avenue Pttsburgh,

More information

GRADIENT METHODS FOR BINARY INTEGER PROGRAMMING

GRADIENT METHODS FOR BINARY INTEGER PROGRAMMING Proceedns of the 4st Internatona Conference on Computers & Industra Enneern GRADIENT METHODS FOR BINARY INTEGER PROGRAMMING Chen-Yuan Huan¹, Ta-Chun Wan² Insttute of Cv Avaton, Natona Chen Kun Unversty,

More information

On-Line Trajectory Generation: Nonconstant Motion Constraints

On-Line Trajectory Generation: Nonconstant Motion Constraints 2012 IEEE Internatona Conference on Robotcs and Automaton RverCentre, Sant Pau, Mnnesota, USA May 14-18, 2012 On-Lne Trajectory Generaton: Nonconstant Moton Constrants Torsten Kröger Abstract A concept

More information

Support Vector Machines

Support Vector Machines Support Vector Machnes Max Wellng Department of Computer Scence Unversty of Toronto 10 Kng s College Road Toronto, M5S 3G5 Canada wellng@cs.toronto.edu Abstract Ths s a note to explan support vector machnes.

More information

Off-line and on-line scheduling on heterogeneous master-slave platforms

Off-line and on-line scheduling on heterogeneous master-slave platforms Laboratore de Informatque du Paraésme Écoe Normae Supéreure de Lyon Unté Mxte de Recherche CNRS-INRIA-ENS LYON-UCBL n o 5668 Off-ne and on-ne schedung on heterogeneous master-save patforms Jean-Franços

More information

USING EMPIRICAL LIKELIHOOD TO COMBINE DATA: APPLICATION TO FOOD RISK ASSESSMENT.

USING EMPIRICAL LIKELIHOOD TO COMBINE DATA: APPLICATION TO FOOD RISK ASSESSMENT. Submtted to the Annas of Apped Statstcs USING EMPIRICA IKEIHOOD TO COMBINE DATA: APPICATION TO FOOD RISK ASSESSMENT. By Amée Crépet, Hugo Harar-Kermadec and Jessca Tressou INRA Mét@rs and INRA COREA Ths

More information

The Development of Web Log Mining Based on Improve-K-Means Clustering Analysis

The Development of Web Log Mining Based on Improve-K-Means Clustering Analysis The Development of Web Log Mnng Based on Improve-K-Means Clusterng Analyss TngZhong Wang * College of Informaton Technology, Luoyang Normal Unversty, Luoyang, 471022, Chna wangtngzhong2@sna.cn Abstract.

More information

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic Lagrange Multplers as Quanttatve Indcators n Economcs Ivan Mezník Insttute of Informatcs, Faculty of Busness and Management, Brno Unversty of TechnologCzech Republc Abstract The quanttatve role of Lagrange

More information

SUPPORT VECTOR MACHINE FOR REGRESSION AND APPLICATIONS TO FINANCIAL FORECASTING

SUPPORT VECTOR MACHINE FOR REGRESSION AND APPLICATIONS TO FINANCIAL FORECASTING SUPPORT VECTOR MACHINE FOR REGRESSION AND APPICATIONS TO FINANCIA FORECASTING Theodore B. Trafas and Husen Ince Schoo of Industra Engneerng Unverst of Okahoma W. Bod Sute 4 Norman Okahoma 739 trafas@ecn.ou.edu;

More information

Feasibility of Using Discriminate Pricing Schemes for Energy Trading in Smart Grid

Feasibility of Using Discriminate Pricing Schemes for Energy Trading in Smart Grid Feasblty of Usng Dscrmnate Prcng Schemes for Energy Tradng n Smart Grd Wayes Tushar, Chau Yuen, Bo Cha, Davd B. Smth, and H. Vncent Poor Sngapore Unversty of Technology and Desgn, Sngapore 138682. Emal:

More information

Module 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module LOSSLESS IMAGE COMPRESSION SYSTEMS Lesson 3 Lossless Compresson: Huffman Codng Instructonal Objectves At the end of ths lesson, the students should be able to:. Defne and measure source entropy..

More information

Automated information technology for ionosphere monitoring of low-orbit navigation satellite signals

Automated information technology for ionosphere monitoring of low-orbit navigation satellite signals Automated nformaton technology for onosphere montorng of low-orbt navgaton satellte sgnals Alexander Romanov, Sergey Trusov and Alexey Romanov Federal State Untary Enterprse Russan Insttute of Space Devce

More information

benefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ).

benefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ). REVIEW OF RISK MANAGEMENT CONCEPTS LOSS DISTRIBUTIONS AND INSURANCE Loss and nsurance: When someone s subject to the rsk of ncurrng a fnancal loss, the loss s generally modeled usng a random varable or

More information

THE DISTRIBUTION OF LOAN PORTFOLIO VALUE * Oldrich Alfons Vasicek

THE DISTRIBUTION OF LOAN PORTFOLIO VALUE * Oldrich Alfons Vasicek HE DISRIBUION OF LOAN PORFOLIO VALUE * Oldrch Alfons Vascek he amount of captal necessary to support a portfolo of debt securtes depends on the probablty dstrbuton of the portfolo loss. Consder a portfolo

More information

Loop Parallelization

Loop Parallelization - - Loop Parallelzaton C-52 Complaton steps: nested loops operatng on arrays, sequentell executon of teraton space DECLARE B[..,..+] FOR I :=.. FOR J :=.. I B[I,J] := B[I-,J]+B[I-,J-] ED FOR ED FOR analyze

More information

Increasing Supported VoIP Flows in WMNs through Link-Based Aggregation

Increasing Supported VoIP Flows in WMNs through Link-Based Aggregation Increasng Supported VoIP Fows n WMNs through n-based Aggregaton J. Oech, Y. Hamam, A. Kuren F SATIE TUT Pretora, South Afrca oechr@gma.com T. Owa Meraa Insttute Counc of Scentfc and Industra Research (CSIR)

More information

Luby s Alg. for Maximal Independent Sets using Pairwise Independence

Luby s Alg. for Maximal Independent Sets using Pairwise Independence Lecture Notes for Randomzed Algorthms Luby s Alg. for Maxmal Independent Sets usng Parwse Independence Last Updated by Erc Vgoda on February, 006 8. Maxmal Independent Sets For a graph G = (V, E), an ndependent

More information

ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING

ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING Matthew J. Lberatore, Department of Management and Operatons, Vllanova Unversty, Vllanova, PA 19085, 610-519-4390,

More information

v a 1 b 1 i, a 2 b 2 i,..., a n b n i.

v a 1 b 1 i, a 2 b 2 i,..., a n b n i. SECTION 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS 455 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS All the vector spaces we have studed thus far n the text are real vector spaces snce the scalars are

More information

A Resources Allocation Model for Multi-Project Management

A Resources Allocation Model for Multi-Project Management A Resources Aocaton Mode for Mut-Proect Management Hamdatou Kane, Aban Tsser To cte ths verson: Hamdatou Kane, Aban Tsser. A Resources Aocaton Mode for Mut-Proect Management. 9th Internatona Conference

More information

Clustering based Two-Stage Text Classification Requiring Minimal Training Data

Clustering based Two-Stage Text Classification Requiring Minimal Training Data OI: 10.2298/CSIS120130044Z Custerng based Two-Stage Text Cassfcaton Requrng Mnma Tranng ata Xue Zhang 1,2 and Wangxn Xao 3,4 1 Key Laboratory of Hgh Confdence Software Technooges, Mnstry of Educaton, Pekng

More information

Research on Single and Mixed Fleet Strategy for Open Vehicle Routing Problem

Research on Single and Mixed Fleet Strategy for Open Vehicle Routing Problem 276 JOURNAL OF SOFTWARE, VOL 6, NO, OCTOBER 2 Research on Snge and Mxed Feet Strategy for Open Vehce Routng Probe Chunyu Ren Heongjang Unversty /Schoo of Inforaton scence and technoogy, Harbn, Chna Ea:

More information

Cardiovascular Event Risk Assessment Fusion of Individual Risk Assessment Tools Applied to the Portuguese Population

Cardiovascular Event Risk Assessment Fusion of Individual Risk Assessment Tools Applied to the Portuguese Population Cardovascuar Event Rsk Assessment Fuson of Indvdua Rsk Assessment Toos Apped to the Portuguese Popuaton S. Paredes, T. Rocha, P. de Carvaho, J. Henrques, J. Moras*, J. Ferrera, M. Mendes Abstract Cardovascuar

More information

Brigid Mullany, Ph.D University of North Carolina, Charlotte

Brigid Mullany, Ph.D University of North Carolina, Charlotte Evaluaton And Comparson Of The Dfferent Standards Used To Defne The Postonal Accuracy And Repeatablty Of Numercally Controlled Machnng Center Axes Brgd Mullany, Ph.D Unversty of North Carolna, Charlotte

More information

J. Parallel Distrib. Comput.

J. Parallel Distrib. Comput. J. Parallel Dstrb. Comput. 71 (2011) 62 76 Contents lsts avalable at ScenceDrect J. Parallel Dstrb. Comput. journal homepage: www.elsever.com/locate/jpdc Optmzng server placement n dstrbuted systems n

More information

Application of Quasi Monte Carlo methods and Global Sensitivity Analysis in finance

Application of Quasi Monte Carlo methods and Global Sensitivity Analysis in finance Applcaton of Quas Monte Carlo methods and Global Senstvty Analyss n fnance Serge Kucherenko, Nlay Shah Imperal College London, UK skucherenko@mperalacuk Daro Czraky Barclays Captal DaroCzraky@barclayscaptalcom

More information

Branch-and-Price and Heuristic Column Generation for the Generalized Truck-and-Trailer Routing Problem

Branch-and-Price and Heuristic Column Generation for the Generalized Truck-and-Trailer Routing Problem REVISTA DE MÉTODOS CUANTITATIVOS PARA LA ECONOMÍA Y LA EMPRESA (12) Págnas 5 38 Dcembre de 2011 ISSN: 1886-516X DL: SE-2927-06 URL: http://wwwupoes/revmetcuant/artphp?d=51 Branch-and-Prce and Heurstc Coumn

More information

Logistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification

Logistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification Lecture 4: More classfers and classes C4B Machne Learnng Hlary 20 A. Zsserman Logstc regresson Loss functons revsted Adaboost Loss functons revsted Optmzaton Multple class classfcaton Logstc Regresson

More information

Stability, observer design and control of networks using Lyapunov methods

Stability, observer design and control of networks using Lyapunov methods Stablty, observer desgn and control of networks usng Lyapunov methods von Lars Naujok Dssertaton zur Erlangung des Grades enes Doktors der Naturwssenschaften - Dr. rer. nat. - Vorgelegt m Fachberech 3

More information

1 Example 1: Axis-aligned rectangles

1 Example 1: Axis-aligned rectangles COS 511: Theoretcal Machne Learnng Lecturer: Rob Schapre Lecture # 6 Scrbe: Aaron Schld February 21, 2013 Last class, we dscussed an analogue for Occam s Razor for nfnte hypothess spaces that, n conjuncton

More information

Solving Factored MDPs with Continuous and Discrete Variables

Solving Factored MDPs with Continuous and Discrete Variables Solvng Factored MPs wth Contnuous and screte Varables Carlos Guestrn Berkeley Research Center Intel Corporaton Mlos Hauskrecht epartment of Computer Scence Unversty of Pttsburgh Branslav Kveton Intellgent

More information

A heuristic task deployment approach for load balancing

A heuristic task deployment approach for load balancing Xu Gaochao, Dong Yunmeng, Fu Xaodog, Dng Yan, Lu Peng, Zhao Ja Abstract A heurstc task deployment approach for load balancng Gaochao Xu, Yunmeng Dong, Xaodong Fu, Yan Dng, Peng Lu, Ja Zhao * College of

More information

Causal, Explanatory Forecasting. Analysis. Regression Analysis. Simple Linear Regression. Which is Independent? Forecasting

Causal, Explanatory Forecasting. Analysis. Regression Analysis. Simple Linear Regression. Which is Independent? Forecasting Causal, Explanatory Forecastng Assumes cause-and-effect relatonshp between system nputs and ts output Forecastng wth Regresson Analyss Rchard S. Barr Inputs System Cause + Effect Relatonshp The job of

More information

BERNSTEIN POLYNOMIALS

BERNSTEIN POLYNOMIALS On-Lne Geometrc Modelng Notes BERNSTEIN POLYNOMIALS Kenneth I. Joy Vsualzaton and Graphcs Research Group Department of Computer Scence Unversty of Calforna, Davs Overvew Polynomals are ncredbly useful

More information

When Network Effect Meets Congestion Effect: Leveraging Social Services for Wireless Services

When Network Effect Meets Congestion Effect: Leveraging Social Services for Wireless Services When Network Effect Meets Congeston Effect: Leveragng Socal Servces for Wreless Servces aowen Gong School of Electrcal, Computer and Energy Engeerng Arzona State Unversty Tempe, AZ 8587, USA xgong9@asuedu

More information

A Lyapunov Optimization Approach to Repeated Stochastic Games

A Lyapunov Optimization Approach to Repeated Stochastic Games PROC. ALLERTON CONFERENCE ON COMMUNICATION, CONTROL, AND COMPUTING, OCT. 2013 1 A Lyapunov Optmzaton Approach to Repeated Stochastc Games Mchael J. Neely Unversty of Southern Calforna http://www-bcf.usc.edu/

More information

"Research Note" APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES *

Research Note APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES * Iranan Journal of Scence & Technology, Transacton B, Engneerng, ol. 30, No. B6, 789-794 rnted n The Islamc Republc of Iran, 006 Shraz Unversty "Research Note" ALICATION OF CHARGE SIMULATION METHOD TO ELECTRIC

More information

How Sets of Coherent Probabilities May Serve as Models for Degrees of Incoherence

How Sets of Coherent Probabilities May Serve as Models for Degrees of Incoherence 1 st Internatonal Symposum on Imprecse Probabltes and Ther Applcatons, Ghent, Belgum, 29 June 2 July 1999 How Sets of Coherent Probabltes May Serve as Models for Degrees of Incoherence Mar J. Schervsh

More information

Prediction of Success or Fail of Students on Different Educational Majors at the End of the High School with Artificial Neural Networks Methods

Prediction of Success or Fail of Students on Different Educational Majors at the End of the High School with Artificial Neural Networks Methods Predcton of Success or Fa of on Dfferent Educatona Maors at the End of the Hgh Schoo th Artfca Neura Netors Methods Sayyed Mad Maznan, Member, IACSIT, and Sayyede Azam Aboghasempur Abstract The man obectve

More information

Answer: A). There is a flatter IS curve in the high MPC economy. Original LM LM after increase in M. IS curve for low MPC economy

Answer: A). There is a flatter IS curve in the high MPC economy. Original LM LM after increase in M. IS curve for low MPC economy 4.02 Quz Solutons Fall 2004 Multple-Choce Questons (30/00 ponts) Please, crcle the correct answer for each of the followng 0 multple-choce questons. For each queston, only one of the answers s correct.

More information

PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 12

PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 12 14 The Ch-squared dstrbuton PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 1 If a normal varable X, havng mean µ and varance σ, s standardsed, the new varable Z has a mean 0 and varance 1. When ths standardsed

More information

Calculating the high frequency transmission line parameters of power cables

Calculating the high frequency transmission line parameters of power cables < ' Calculatng the hgh frequency transmsson lne parameters of power cables Authors: Dr. John Dcknson, Laboratory Servces Manager, N 0 RW E B Communcatons Mr. Peter J. Ncholson, Project Assgnment Manager,

More information

Conversion between the vector and raster data structures using Fuzzy Geographical Entities

Conversion between the vector and raster data structures using Fuzzy Geographical Entities Converson between the vector and raster data structures usng Fuzzy Geographcal Enttes Cdála Fonte Department of Mathematcs Faculty of Scences and Technology Unversty of Combra, Apartado 38, 3 454 Combra,

More information

Neural Network-based Colonoscopic Diagnosis Using On-line Learning and Differential Evolution

Neural Network-based Colonoscopic Diagnosis Using On-line Learning and Differential Evolution Neura Networ-based Coonoscopc Dagnoss Usng On-ne Learnng and Dfferenta Evouton George D. Magouas, Vasss P. Paganaos * and Mchae N. Vrahats * Department of Informaton Systems and Computng, Brune Unversty,

More information

The Application of Fractional Brownian Motion in Option Pricing

The Application of Fractional Brownian Motion in Option Pricing Vol. 0, No. (05), pp. 73-8 http://dx.do.org/0.457/jmue.05.0..6 The Applcaton of Fractonal Brownan Moton n Opton Prcng Qng-xn Zhou School of Basc Scence,arbn Unversty of Commerce,arbn zhouqngxn98@6.com

More information

POLYSA: A Polynomial Algorithm for Non-binary Constraint Satisfaction Problems with and

POLYSA: A Polynomial Algorithm for Non-binary Constraint Satisfaction Problems with and POLYSA: A Polynomal Algorthm for Non-bnary Constrant Satsfacton Problems wth and Mguel A. Saldo, Federco Barber Dpto. Sstemas Informátcos y Computacón Unversdad Poltécnca de Valenca, Camno de Vera s/n

More information

Fault tolerance in cloud technologies presented as a service

Fault tolerance in cloud technologies presented as a service Internatonal Scentfc Conference Computer Scence 2015 Pavel Dzhunev, PhD student Fault tolerance n cloud technologes presented as a servce INTRODUCTION Improvements n technques for vrtualzaton and performance

More information

A hybrid global optimization algorithm based on parallel chaos optimization and outlook algorithm

A hybrid global optimization algorithm based on parallel chaos optimization and outlook algorithm Avalable onlne www.ocpr.com Journal of Chemcal and Pharmaceutcal Research, 2014, 6(7):1884-1889 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCPRC5 A hybrd global optmzaton algorthm based on parallel

More information

PERRON FROBENIUS THEOREM

PERRON FROBENIUS THEOREM PERRON FROBENIUS THEOREM R. CLARK ROBINSON Defnton. A n n matrx M wth real entres m, s called a stochastc matrx provded () all the entres m satsfy 0 m, () each of the columns sum to one, m = for all, ()

More information

Dynamic optimization of the LNG value chain

Dynamic optimization of the LNG value chain Proceedngs of the 1 st Annual Gas Processng Symposum H. Alfadala, G.V. Rex Reklats and M.M. El-Halwag (Edtors) 2009 Elsever B.V. All rghts reserved. 1 Dynamc optmzaton of the LNG value chan Bjarne A. Foss

More information

Improved SVM in Cloud Computing Information Mining

Improved SVM in Cloud Computing Information Mining Internatonal Journal of Grd Dstrbuton Computng Vol.8, No.1 (015), pp.33-40 http://dx.do.org/10.1457/jgdc.015.8.1.04 Improved n Cloud Computng Informaton Mnng Lvshuhong (ZhengDe polytechnc college JangSu

More information

Enabling P2P One-view Multi-party Video Conferencing

Enabling P2P One-view Multi-party Video Conferencing Enablng P2P One-vew Mult-party Vdeo Conferencng Yongxang Zhao, Yong Lu, Changja Chen, and JanYn Zhang Abstract Mult-Party Vdeo Conferencng (MPVC) facltates realtme group nteracton between users. Whle P2P

More information

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign PAS: A Packet Accountng System to Lmt the Effects of DoS & DDoS Debsh Fesehaye & Klara Naherstedt Unversty of Illnos-Urbana Champagn DoS and DDoS DDoS attacks are ncreasng threats to our dgtal world. Exstng

More information

Least Squares Fitting of Data

Least Squares Fitting of Data Least Squares Fttng of Data Davd Eberly Geoetrc Tools, LLC http://www.geoetrctools.co/ Copyrght c 1998-2016. All Rghts Reserved. Created: July 15, 1999 Last Modfed: January 5, 2015 Contents 1 Lnear Fttng

More information

Section 5.3 Annuities, Future Value, and Sinking Funds

Section 5.3 Annuities, Future Value, and Sinking Funds Secton 5.3 Annutes, Future Value, and Snkng Funds Ordnary Annutes A sequence of equal payments made at equal perods of tme s called an annuty. The tme between payments s the payment perod, and the tme

More information

Forecasting the Direction and Strength of Stock Market Movement

Forecasting the Direction and Strength of Stock Market Movement Forecastng the Drecton and Strength of Stock Market Movement Jngwe Chen Mng Chen Nan Ye cjngwe@stanford.edu mchen5@stanford.edu nanye@stanford.edu Abstract - Stock market s one of the most complcated systems

More information

SIMPLIFYING NDA PROGRAMMING WITH PROt SQL

SIMPLIFYING NDA PROGRAMMING WITH PROt SQL SIMPLIFYING NDA PROGRAMMING WITH PROt SQL Aeen L. Yam, Besseaar Assocates, Prnceton, NJ ABSRACf The programmng of New Drug Appcaton (NDA) Integrated Summary of Safety (ISS) usuay nvoves obtanng patent

More information

Research Article Enhanced Two-Step Method via Relaxed Order of α-satisfactory Degrees for Fuzzy Multiobjective Optimization

Research Article Enhanced Two-Step Method via Relaxed Order of α-satisfactory Degrees for Fuzzy Multiobjective Optimization Hndaw Publshng Corporaton Mathematcal Problems n Engneerng Artcle ID 867836 pages http://dxdoorg/055/204/867836 Research Artcle Enhanced Two-Step Method va Relaxed Order of α-satsfactory Degrees for Fuzzy

More information

An Evaluation of the Extended Logistic, Simple Logistic, and Gompertz Models for Forecasting Short Lifecycle Products and Services

An Evaluation of the Extended Logistic, Simple Logistic, and Gompertz Models for Forecasting Short Lifecycle Products and Services An Evaluaton of the Extended Logstc, Smple Logstc, and Gompertz Models for Forecastng Short Lfecycle Products and Servces Charles V. Trappey a,1, Hsn-yng Wu b a Professor (Management Scence), Natonal Chao

More information

Downlink Power Allocation for Multi-class. Wireless Systems

Downlink Power Allocation for Multi-class. Wireless Systems Downlnk Power Allocaton for Mult-class 1 Wreless Systems Jang-Won Lee, Rav R. Mazumdar, and Ness B. Shroff School of Electrcal and Computer Engneerng Purdue Unversty West Lafayette, IN 47907, USA {lee46,

More information

A Secure Password-Authenticated Key Agreement Using Smart Cards

A Secure Password-Authenticated Key Agreement Using Smart Cards A Secure Password-Authentcated Key Agreement Usng Smart Cards Ka Chan 1, Wen-Chung Kuo 2 and Jn-Chou Cheng 3 1 Department of Computer and Informaton Scence, R.O.C. Mltary Academy, Kaohsung 83059, Tawan,

More information

Project Networks With Mixed-Time Constraints

Project Networks With Mixed-Time Constraints Project Networs Wth Mxed-Tme Constrants L Caccetta and B Wattananon Western Australan Centre of Excellence n Industral Optmsaton (WACEIO) Curtn Unversty of Technology GPO Box U1987 Perth Western Australa

More information

NON-CONSTANT SUM RED-AND-BLACK GAMES WITH BET-DEPENDENT WIN PROBABILITY FUNCTION LAURA PONTIGGIA, University of the Sciences in Philadelphia

NON-CONSTANT SUM RED-AND-BLACK GAMES WITH BET-DEPENDENT WIN PROBABILITY FUNCTION LAURA PONTIGGIA, University of the Sciences in Philadelphia To appear n Journal o Appled Probablty June 2007 O-COSTAT SUM RED-AD-BLACK GAMES WITH BET-DEPEDET WI PROBABILITY FUCTIO LAURA POTIGGIA, Unversty o the Scences n Phladelpha Abstract In ths paper we nvestgate

More information

Period and Deadline Selection for Schedulability in Real-Time Systems

Period and Deadline Selection for Schedulability in Real-Time Systems Perod and Deadlne Selecton for Schedulablty n Real-Tme Systems Thdapat Chantem, Xaofeng Wang, M.D. Lemmon, and X. Sharon Hu Department of Computer Scence and Engneerng, Department of Electrcal Engneerng

More information

Optimal resource capacity management for stochastic networks

Optimal resource capacity management for stochastic networks Submtted for publcaton. Optmal resource capacty management for stochastc networks A.B. Deker H. Mlton Stewart School of ISyE, Georga Insttute of Technology, Atlanta, GA 30332, ton.deker@sye.gatech.edu

More information

INVESTIGATION OF VEHICULAR USERS FAIRNESS IN CDMA-HDR NETWORKS

INVESTIGATION OF VEHICULAR USERS FAIRNESS IN CDMA-HDR NETWORKS 21 22 September 2007, BULGARIA 119 Proceedngs of the Internatonal Conference on Informaton Technologes (InfoTech-2007) 21 st 22 nd September 2007, Bulgara vol. 2 INVESTIGATION OF VEHICULAR USERS FAIRNESS

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

Method for Production Planning and Inventory Control in Oil

Method for Production Planning and Inventory Control in Oil Memors of the Faculty of Engneerng, Okayama Unversty, Vol.41, pp.20-30, January, 2007 Method for Producton Plannng and Inventory Control n Ol Refnery TakujImamura,MasamKonshandJunIma Dvson of Electronc

More information

Risk-based Fatigue Estimate of Deep Water Risers -- Course Project for EM388F: Fracture Mechanics, Spring 2008

Risk-based Fatigue Estimate of Deep Water Risers -- Course Project for EM388F: Fracture Mechanics, Spring 2008 Rsk-based Fatgue Estmate of Deep Water Rsers -- Course Project for EM388F: Fracture Mechancs, Sprng 2008 Chen Sh Department of Cvl, Archtectural, and Envronmental Engneerng The Unversty of Texas at Austn

More information

Application of Multi-Agents for Fault Detection and Reconfiguration of Power Distribution Systems

Application of Multi-Agents for Fault Detection and Reconfiguration of Power Distribution Systems 1 Applcaton of Mult-Agents for Fault Detecton and Reconfguraton of Power Dstrbuton Systems K. Nareshkumar, Member, IEEE, M. A. Choudhry, Senor Member, IEEE, J. La, A. Felach, Senor Member, IEEE Abstract--The

More information

An interactive system for structure-based ASCII art creation

An interactive system for structure-based ASCII art creation An nteractve system for structure-based ASCII art creaton Katsunor Myake Henry Johan Tomoyuk Nshta The Unversty of Tokyo Nanyang Technologcal Unversty Abstract Non-Photorealstc Renderng (NPR), whose am

More information

Data Broadcast on a Multi-System Heterogeneous Overlayed Wireless Network *

Data Broadcast on a Multi-System Heterogeneous Overlayed Wireless Network * JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 24, 819-840 (2008) Data Broadcast on a Mult-System Heterogeneous Overlayed Wreless Network * Department of Computer Scence Natonal Chao Tung Unversty Hsnchu,

More information

1. Fundamentals of probability theory 2. Emergence of communication traffic 3. Stochastic & Markovian Processes (SP & MP)

1. Fundamentals of probability theory 2. Emergence of communication traffic 3. Stochastic & Markovian Processes (SP & MP) 6.3 / -- Communcaton Networks II (Görg) SS20 -- www.comnets.un-bremen.de Communcaton Networks II Contents. Fundamentals of probablty theory 2. Emergence of communcaton traffc 3. Stochastc & Markovan Processes

More information

Comparison of workflow software products

Comparison of workflow software products Internatona Conference on Computer Systems and Technooges - CompSysTech 2006 Comparson of worfow software products Krasmra Stoova,Todor Stoov Abstract: Ths research addresses probems, reated to the assessment

More information

7.5. Present Value of an Annuity. Investigate

7.5. Present Value of an Annuity. Investigate 7.5 Present Value of an Annuty Owen and Anna are approachng retrement and are puttng ther fnances n order. They have worked hard and nvested ther earnngs so that they now have a large amount of money on

More information

Price Competition in an Oligopoly Market with Multiple IaaS Cloud Providers

Price Competition in an Oligopoly Market with Multiple IaaS Cloud Providers Prce Competton n an Olgopoly Market wth Multple IaaS Cloud Provders Yuan Feng, Baochun L, Bo L Department of Computng, Hong Kong Polytechnc Unversty Department of Electrcal and Computer Engneerng, Unversty

More information

Optimal Scheduling in the Hybrid-Cloud

Optimal Scheduling in the Hybrid-Cloud Optmal Schedulng n the Hybrd-Cloud Mark Shfrn Faculty of Electrcal Engneerng Technon, Israel Emal: shfrn@tx.technon.ac.l Ram Atar Faculty of Electrcal Engneerng Technon, Israel Emal: atar@ee.technon.ac.l

More information

2) A single-language trained classifier: one. classifier trained on documents written in

2) A single-language trained classifier: one. classifier trained on documents written in Openng the ega terature Porta to mutngua access E. Francescon, G. Perugne ITTIG Insttute of Lega Informaton Theory and Technooges Itaan Natona Research Counc, Forence, Itay Te: +39 055 43999 Fax: +39 055

More information

The Stock Market Game and the Kelly-Nash Equilibrium

The Stock Market Game and the Kelly-Nash Equilibrium The Stock Market Game and the Kelly-Nash Equlbrum Carlos Alós-Ferrer, Ana B. Ana Department of Economcs, Unversty of Venna. Hohenstaufengasse 9, A-1010 Venna, Austra. July 2003 Abstract We formulate the

More information

What is Candidate Sampling

What is Candidate Sampling What s Canddate Samplng Say we have a multclass or mult label problem where each tranng example ( x, T ) conssts of a context x a small (mult)set of target classes T out of a large unverse L of possble

More information

QoS-Aware Spectrum Sharing in Cognitive Wireless Networks

QoS-Aware Spectrum Sharing in Cognitive Wireless Networks QoS-Aware Spectrum Sharng n Cogntve reless Networks Long Le and Ekram Hossan Abstract e consder QoS-aware spectrum sharng n cogntve wreless networks where secondary users are allowed to access the spectrum

More information

Distributed Multi-Target Tracking In A Self-Configuring Camera Network

Distributed Multi-Target Tracking In A Self-Configuring Camera Network Dstrbuted Mult-Target Trackng In A Self-Confgurng Camera Network Crstan Soto, B Song, Amt K. Roy-Chowdhury Department of Electrcal Engneerng Unversty of Calforna, Rversde {cwlder,bsong,amtrc}@ee.ucr.edu

More information

Testing and Debugging Resource Allocation for Fault Detection and Removal Process

Testing and Debugging Resource Allocation for Fault Detection and Removal Process Internatonal Journal of New Computer Archtectures and ther Applcatons (IJNCAA) 4(4): 93-00 The Socety of Dgtal Informaton and Wreless Communcatons, 04 (ISSN: 0-9085) Testng and Debuggng Resource Allocaton

More information