A Mobile Robots Experimental Environment with Event-Based Wireless Communication

Size: px
Start display at page:

Download "A Mobile Robots Experimental Environment with Event-Based Wireless Communication"

Transcription

1 Sensos 2013, 13, ; do: /s Atcle OPEN ACCESS sensos ISSN A Moble Robots Expemental Envonment wth Event-Based Weless Communcaton Maía Gunaldo 1, *, Enesto Fábegas 1, Gonzalo Faas 2, Sebastán Domdo-Canto 1, Dctno Chaos 1, JoséSánchez 1 and Sebastán Domdo Depatment of Compute Scence and Automatc Contol, UNED, Juan del Rosal 16, Madd 28040, Span; E-Mals: efabegas@bec.uned.es (E.F.); sebas@da.uned.es (S.D-C.); dchaos@da.uned.es (D.C.); jsanchez@da.uned.es (J.S.); sdomdo@da.uned.es (S.D.) School of Electcal Engneeng, Pontfca Unvesdad Católca de Valpaaso, Av. Basl 2147, Valpaaso, Chle; E-Mal: gonzalo.faas@ucv.cl * Autho to whom coespondence should be addessed; E-Mal: mgunaldo@da.uned.es; Tel.: ; Fax: Receved: 15 Ma 2013; n evsed fom: 12 Jul 2013 / Accepted: 18 Jul 2013 / Publshed: 22 Jul 2013 Abstact: An expemental platfom to communcate between a set of moble obots though a weless netwok has been developed. The moble obots get the poston though a camea whch pefoms as senso. The vdeo mages ae pocessed n a PC and a Waspmote cad sends the coespondng poston to each obot usng the ZgBee standad. A dstbuted contol algothm based on event-tggeed communcatons has been desgned and mplemented to bng the obots nto the desed fomaton. Each obot communcates to ts neghbos onl at event tmes. Futhemoe, a smulaton tool has been developed to desgn and pefom expements wth the sstem. An example of usage s pesented. Kewods: dstbuted contol sstems; moble obots; mult-agent sstems; weless communcatons; embedded sstems 1. Intoducton Sgnfcant developments n the felds of communcaton technolog, weless technolog, embedded devces, and man othes, have enabled the development of autonomous a, gound, o

2 Sensos 2013, undewate vehcles. Goups of such vehcles, efeed to as agents, can be utlzed to solve a vaet of poblems ve effcentl, such as fo example exploaton and montong tasks. One of the man poblems n these sstems s when the contol task s centalzed. Ths centalzaton demands a ve poweful communcaton netwok to send state nfomaton n a tmel manne and detaled models of the moble obot nteactons. Both of these equements can geatl lmt the scalablt of centalzed appoaches n netwoked contol sstems. An altenatve wa s to use a dstbuted appoach [1,2], whee each moble obot decdes ts actons based onl on hs own state and the dect neghbos nfomaton, whle guaanteeng that the whole goup fulflls a common goup objectve. One example of a goup objectve fo mult-agent sstems s state ageement o consensus,.e., all agents ae supposed to convege to a common pont o state. Such consensus poblems have a vaet of applcatons n flockng, atttude snchonzaton n satellte swams, dstbuted senso netwoks, congeston contol n communcaton netwoks, o fomaton contol [3]. We ae patculal nteested n the last feld of applcaton snce achevng a stable fomaton s analogous to eachng consensus. In ths consensus poblem the agents communcate though a netwok exchangng nfomaton to each the desed fomaton. Specfcall, we popose to ncopoate weless lnks between the moble obots to suppot the communcatons, the embedded sstems, and the contol. Moeove, all eal netwoks have bandwdth lmtatons that can cause delas n message delve poducng a majo mpact on oveall sstem stablt [4]. In ths sense, t s desable to know how fequentl moble obots should communcate such that the sstem mantans a desed level of pefomance. In ths lght, event-based communcatons ae an altenatve to tme-scheduled contol [5], and the have been n the focus of man eseaches n ecent eas such as [6 11], to menton just a few. In the lteatue, seveal smplfcatons ae usuall done n both the obot dnamcs and the netwok models of a mult-agent obot sstem. Howeve, a eal obot has ts own phscal pats, ncludng sensos, pocessos, actuatos, and communcaton devces. Theefoe, ts lmtatons must be consdeed, such as those elated to computatonal powe, senso eos, and actuatng and communcatng devces. Ths pape pesents a mult-agent appoach though the use of a decentalzed contol wth event-tggeed communcaton n a moble obot fomaton contol wth a consensus algothm. The decentalzed contol s less vulneable to a node falue and ease to scale than the centalzed appoaches pesented n [12 14]. In [12] the fomaton contol of a mult-khepea obot sstem s solved wthout communcaton between obots. Howeve, each obot executes an algothm peodcall that eques the knowledge of the complete gaph of the fomaton, so t s computatonall ntensve fo lage numbe of obots. Centalzed appoaches ae poposed n [13,14]. In [13] all complext of sensng, contol, and computaton s elegated to the leade of a goup of obots, and [13] teats the obots fomaton as a sngle vtual gd stuctue. A decentalzed appoach fo coopeatve contol s pesented n [15] and a detaled hadwae descpton s gven and seveal expemental examples of mult-vehcle contol fo TXT-1 monste tucks fom Tama, Inc, ae gven fo both communcaton and vson fomaton contol. In [16] a hadwae platfom fo a sngle vehcle s developed and a PID event-based contol s mplemented. Howeve, none of these woks consde dstbuted event-based contol n a mult-obot envonment. Hence, the man contbuton of ths pape s the mplementaton of dstbuted event-tggeed

3 Sensos 2013, algothms n a eal moble obots platfom snce, to the best knowledge of the authos, all the esults n ths feld ae manl theoetcal and onl pesent smulaton esults. The expemental esults show the goodness of the event-tgge appoach compaed to tme-dven contol, educng the communcaton between agents. Futhemoe, n contast to othe woks, as fo example [12,15], n whch a costl platfom fo ndoo o outdoo expementaton s used, we gve a detaled descpton of ou low-cost hadwae platfom, and how t was mplemented n LEGO obots, n ode to help eades who ae consdeng buldng a smla sstem fo eseach, educatonal puposes [17], o to mmc eal wold exploato applcatons [18]. An nteactve smulaton, whch s avalable at s also pesented as a complement to the expemental setup. Ths smulaton helped n the desgn and tal of dffeent expements wth vaable sstem and netwok condtons. The pape s oganzed as follows. Secton 2 descbes n detal the expemental envonment and ts components. Secton 3 gves a backgound on event-based contol and mult-agent sstems, and the extenson of these to the fomaton contol of a mult-obot sstem. Secton 4 descbes the smulaton tool and an example of usage. Secton 5 dscusses expemental esults and compaes them to smulaton esults and expeences based on tme-dven contol. Fnall, Secton 6 ends the pape wth conclusons and suggestons of futue woks. 2. Test Platfom Descpton We next descbe the expemental envonment desgned wth low-cost components. Ths ma help eades to buld smla sstems. It can also be scaled up o down to equp bgge o smalle obots, espectvel [19,20] Platfom Achtectue and Components The developed tool s a test platfom fo mult-moble obot sstems. It conssts of two man components: the vson and the communcaton sstems. Dffeent softwae tools exchange nfomaton to tansfom the mages captued b the camea nto postonng data, whch ae sent to the obots va weless communcatons. Fgue 1 shows the block dagam of the achtectue and the components of the platfom, whch ae the followng: 1. A pesonal compute (PC) to calculate and send the poston and oentaton to the obots. 2. A CCD camea (chage-coupled devce) connected to the PC and equpped wth wde angle optcs. The camea s placed n the celng of the oom, as descbed n [21]. 3. A Waspmote Gatewa cad connected to the PC [22]. 4. A goup of LEGO Mndstoms NXT obots wth the HTechnc IR RC Ca confguaton [23,24]. 5. A Waspmote cad connected to each NXT LEGO Mndstoms obot b I 2 C pot [25].

4 Sensos 2013, A contol and communcaton applcaton, developed n LabVIEW, fo each NXT LEGO obot. 7. An applcaton, wtten n C++, fo each Waspmote cad. 8. An applcaton (unnng on the PC), developed n LabVIEW, fo the nteacton wth the camea [21]. 9. A Java applcaton, bult wth Eas Java Smulatons (EJS) [26], fo communcaton between the obots and the PC. 10. A Java XBee API to send and eceve data though the Waspmote Gatewa cad [27]. 11. The JIL seve [28] to lnk LabVIEW wth EJS. Fgue 1. Block dagam of the platfom achtectue Vson Sstem Though LEGO platfoms povde an nteestng goup of sensos such as ultasonc, touch, o lght sensos, thee s no devce fo measung absolute postons. Ths makes expementaton n obot fomaton contol dffcult. One opton to solve the pevous dawback s the use of a GPS as the man localzaton senso as n [29]. Howeve, when the expements ae pefomed ndoos, GPS s mpactcal. Anothe altenatve s to ncopoate a vson sstem to get the obots poses. The ecent appeaance of RGB-D cameas has made possble 3D mappng fom a sngle mage and state estmaton. Howeve, unnng the obot eal-tme state estmaton algothms on boad would eque a poweful compute, whch would ncease the cost of the platfom consdeabl. Thus, a sngle camea on the celng obseves the aena fom an ovehead poston. Each obot can be dstngushed thanks to thee hgh

5 Sensos 2013, bghtness LEDs. These lght spots fom an sosceles tangle wth dffeent aspect ato fo each obot as shown n Fgue 2a. In Fgue 2b, the ponts Pa, Pb, and Pc epesent the LEDs. The oentaton s denoted b θ. The poston of each obot s calculated usng the cente of mass (Xcm; Ycm). Fgue 2. (a) The NXT LEGO obot, (b) Isosceles tangle fomed b the LEDs. (a) (b) The Basle A631fc CCD camea (block 2 of Fgue 1) s confgued n monochome mode fo mage acquston n ode to maxmze the esoluton and shutte speed. Unde these condtons, we obtan mages wth a esoluton of 1,388 1,038 pxels at 18.7 fps. The mages captued b the camea ae pocessed b the LabVIEW Camea App (block 8 of Fgue 1), whch computes the obots poses. The executon of the applcaton mples the followng steps: Bael dstoton coecton: The mage magnfcaton deceases wth dstance fom the optcal axs. The appaent effect means that an mage has been mapped aound a sphee (o bael). Pespectve tansfomaton: Tansfoms the coodnates of the object to the camea coodnates, makng otaton and tanslaton tansfomatons. Calculaton of postons and oentatons: Fom the efeence ponts on the gound, we calculate the poston and oentaton takng nto account the mopholog of the obtaned ponts. Tansfomaton of coodnates: Fom pxels to metes. Calbaton: Gets the pespectve calbaton paametes. The font panel of the LabVIEW Camea App s shown n Fgue 3. The values of the confguaton of the camea (gan, shutte, and fame ate) and the peod (T) ae equed to compute the beangs, and shown on the font panel. The applcaton sends the absolute poston and oentaton to the obots eve 200 mllseconds. Futhe detals about ths applcaton can be found n [21].

6 Sensos 2013, Fgue 3. Use nteface of the Camea LabVIEW App Communcaton Sstem Afte the pocessng of the obots poses, the ae sent to the obots. Each one eceves ts poston and oentaton, and decdes when to boadcast ths nfomaton to ts neghbos based on the local data, updatng the contol law aftewads. Thus, the contol s dstbuted n the sense that each obot takes b tself the decson of when to tansmt ts state and the contol law s computed locall, n contast to [16], whee the contol law s computed on the cental compute and sent to the obots. Waspmote boads (block 5 of Fgue 1) ae used fo the weless communcatons. Ths cad has a modula achtectue and s based on the Atmel ATMEGA 1281 mcocontolle. The Waspmote also ntegates the Dg XBee RF (ZgBee standad) module fo communcaton n the ISMB bands (Industal Scentfc Medcal Band, RF-868 MHz) [30]. Ths module communcates wth the mcocontolle va nteuptons usng the UART at the speed of 38,400 bps. It allows the defnton of dffeent netwok confguatons (e.g. p2p, tee, o mesh) and uses the potocol wth a fequenc of 2.4 GHz [22]. Fo weless netwok settngs, each cad uses the MAC addess of ts coespondng XBee. Each obot talks to ts neghbos gven b the communcaton gaph (see Sectons 3 and 4.2 fo futhe detals). The powe suppl fo the Waspmote s a lthum batte. The eneg sstem has fou opeatonal modes (on, sleep, deep sleep, and hbenate). In nomal opeatng mode (on), the consumpton s 9 ma. Eve 125 ns the mcocontolle uns a outne of low level (machne language) nstuctons [31,32], whch s fast enough to the tme equements of ou sstem. The I 2 C bus ntefaces the NXT LEGO wth the Waspmote. The NXT LEGO uns as maste and the Waspmote as slave. The Waspmote cad uns an applcaton (block 7 Fgue 1) that allows the LEGO NXT to communcate wth othe obots and the PC. Eve 100 ms the followng tasks ae executed:

7 Sensos 2013, It checks f a packet has been eceved b the XBee ado. Packets can be sent fom the PC o fom anothe obot. If the answe s postve, the eceved data s stoed n a buffe untl the NXT LEGO eques t. 2. It checks f a flag s tue. If the answe s postve, t ndcates that the NXT LEGO eques the sendng of a packet to othe obots (ts neghbos) b Xbee ado. Once the packet s sent the flag s set to false. The NXT LEGO tgges an nteupton outne (whch sets the flag to tue) when a packet must be sent. In event-based contol, the flag s set to tue at the event occuence tme. Each NXT LEGO uns a LabVIEW App (block 6 of Fgue 1) to nteface wth the Waspmote cad b I 2 C (see Secton 2.4 fo futhe detals). Ths applcaton equests data to the Waspmote eve 200 ms. If the eceved packets ae fom the camea applcaton (own poston and oentaton), the poston s compaed to the last boadcasted poston,.e., to the state whch geneated the last event. If the dffeence s geate than a theshold, then the cuent poston s sent to ts neghbos b usng the I 2 C nteface wth the Waspmote. B contast, f the souce s an obot n the neghbohood, the eceved data s then used to update the contol law as explaned late n Secton 3. The communcaton between the PC and the Waspmote s caed out b the Waspmote s Gatewa (block 3 of Fgue 1) connected to the PC though the USB pot. The Waspmote s Gatewa has dves fo Java but not fo LabVIEW. Fo ths eason, t s necessa to use an applcaton developed n Java to send the data to each coespondng obot usng the gatewa. In ths case, the JIL Seve applcaton [28] (block 11 of Fgue 1) s used as nteface between the camea applcaton and the EJS Robots App (block 9 of Fgue 1). Ths latte applcaton sends the poston and oentaton to the obots b usng XBee-Java-API (block 10 of Fgue 1). The data packet sent fom the PC and between the obots has a sze of 16 btes. Ths s the maxmum sze allowed b NXT LEGO Mndstoms fo the communcatons b I 2 C pot [33]. Fgue 4 shows the packet stuctue, whch s as follows: Bte 0 (S): The souce whee the packet comes fom. The value s 0 fo the obot 0, 1 fo obot 1, and so on. Btes 1 4 (X 1 X 4 ): 4 btes coespondng to the X coodnate of the obot poston. Btes 5 8 (Y 1 Y 4 ): 4 btes coespondng to the Y coodnate of the obot poston. Btes 9 12 (T 1 T 4 ): 4 btes coespondng to the obot oentaton (θ). Btes 13 15: Not used n ths platfom. Fgue 4. Data packet stuctue Usng I 2 C n NXT LEGO The LabVIEW toolkt fo LEGO has buffes to wte and ead n each pot, and thee sstem call methods to access to these buffes. The wte opeaton begns wth the coespondng sstem call method, whch consttutes the stat of an asnchonous tansacton between the NXT bck and the

8 Sensos 2013, Waspmote. If the etun value s zeo, the method stats a communcaton tansacton. Afte a wte tansacton s stated t s necessa to use a sstem call functon to check the status of the pot. To ensue the success of the tansmsson, thee s a sstem call functon to check the status pot. The ead opeaton of a devce though I 2 C pot s caed out n two stages. Fst, a wte opeaton must be executed wth the numbe of btes that the slave has to send. If the count of btes avalable n the ead buffe s not zeo, the ead opeaton s executed and the data s obtaned n the extenal buffe. 3. Event-Based Contol and Communcaton n Mult-Agent Sstems 3.1. Mult-Agent Sstems and the Consensus Poblem The smplest model to epesent the communcaton topolog of a mult-agent sstem s a gaph G = {V,E}, whee the nodes V coespond to agents and the edges E epesent communcaton lnks between nodes. Accodng to [3], a smple consensus algothm to each an ageement egadng the state of N sngle ntegatos of the fom x u can be expessed as an nth ode lnea sstem on a gaph: x j x ( t x ) jn The dnamcs of the goup of agents can be wtten as: (1) x Lx( t) (2) whee L s the laplacan of the gaph, and N the set of neghbos of the node [34]. Based on analtcal tools fom algebac gaph theo, t can be shown that f the gaph s connected, then thee s an unque equlbum state fo Equaton (2) of the fom x 1, whee 1 N N 1 x ) (0 and 1 = (1 1) T. Consensus algothms can be extended to fomaton contol f the fomaton s epesented b vectos of elatve postons of neghbong agents. In patcula, let us denote b j the desed nte-vehcle elatve poston vecto. The followng contol law: x x j x j j N elds the goup to acheve the objectve of the fomaton. The consensus poblem has been extendng egadng the agents dnamcs to hghe ode ntegatos [35,36], o lnea sstems [37] Model of Non-Holonomc Moble Robots Sngle o double ntegatos do not descbe popel the dnamcs of most of commecal moble obots, snce these cannot move n an decton nstantaneousl. In obotcs, holonomct efes to the elatonshp between the contollable and total degees of feedom of a gven obot. If the contollable degees of feedom ae less than the total degees of feedom the vehcle s non-holonomc. To avod the non-holonomc constant, Refeence [38] defnes the dnamcs n tems of the font wheels coodnates x, as: eq (3)

9 Sensos 2013, whee x cos d sn v sn d cos 0 1 x x d cos, d sn. The tple (x,, θ ) denotes the poston and the oentaton of the vehcle (see Fgue 2b, x = Xcm, = Ycm fo each vehcle), and v and ω ae the longtudnal and angula veloctes, espectvel. Let us consde as contol nputs fo each obot u = (v, ω ) Tme Schedule Contol In a dstbuted contol appoach, each agent collects nfomaton fom ts neghbong nodes and updates the contol law accodng to some ules. In a netwoked sstem, collectng nfomaton means to tansmt t though the communcaton channel. The nfomaton can be tansmtted though the netwok n a peodc o an event-based fashon. The fst case nvolves the tadtonal appoach of samplng at pe-specfed tme nstances, usuall sepaated b a specfc samplng peod. Howeve, the selected peod must assue adequate sstem pefomance ove a wde ange of opeatng condtons. Ths can eld to a consevatve choce and esult n sgnfcant ove-povsonng of the communcaton-netwok. Event-tggeng epesents one wa of geneatng spoadc tansmssons acoss netwok channels [39]. The communcaton s nvoked onl when somethng sgnfcant has occued n the sstem. The poblem of mult-agent sstems wth event-based communcatons has been ecentl addessed [40,41]. Wheeas the contolle desgn of [41] s centalzed, [40] pesents a dstbuted appoach, n whch the agents ae modeled b sngle and double ntegatos. The contol law s based on the states that the agents boadcast to the neghbos though the netwok at event tmes. The aveage consensus s eached pesevng convegence popetes and the expemental esults show a sgnfcant educton n the numbe of tansmssons espect to a peodc appoach. Dstbuted event-based contol has been also addessed n [42], whee phscal couplng ma exst between the subsstems. The contol sgnals n each obot ae computed based on the boadcasted nfomaton, whch means that ae not updated contnuousl, not even at each measuement of ts own state eceved b the camea (see Secton 2.2). The ule to detemne when to boadcast s gven b the tgge functon. A tgge functon detemnes the nstants of tme { t k} at whch each agent boadcasts the state to ts neghbos. In geneal, these boadcastng tmes ae not equdstant as n peodc communcatons. Moeove, the poposed tgge functons depend on local nfomaton onl. If we denote the state of an agent as x (t) and xˆ as the boadcasted state of agent to ts neghbos, whch s a pecewse constant functon, an event s tggeed fo an agent when: f ( ), ˆ ( ), ˆ ( ) x t x t x j t 0 (5) jn s fulflled, whee N s the neghbohood of agent. The tgge functons ae sad to depend on local nfomaton because the decson of when to e-compute the contol law depends on the last update values and the neghbos states, and each agent does not have to monto the states of ts neghbos contnuousl n ode to evaluate the tggeng condton, n contast to the tgge ules poposed n [10]. (4)

10 Sensos 2013, Specfc tgge functons based on the eo e ˆ x x ae poposed. Fo example, statc tgge functons have the fom f ( e ) e c, whee c 0 0 s a constant value. The contol law u = (v, ω ) to each the desed fomaton s computed as: whee: v cos sn d sn d cos 1 L 0 0 xˆ x L ˆ T xˆ and ŷ ae the stack vectos of the boadcasted postons,.e. xˆ xˆ,..., xˆ 1 N T ˆ ˆ,..., ˆ. 1 N (6) and x and ae the stack vectos of the desed postons of the agents wth efeence to the cente of the goup. Combnng Equatons (4) and (6) t elds that: x L xˆ x L ˆ (7) 3.4. Stablt Anaalss Note that defnng the contol law gven n Equaton (6), two decoupled equatons ae obtaned and the coespond to the equaton of the event-tggeed consensus poblem of sngle ntegatos studed n [40], but extended to each a fomaton. Thee t was poved that the consensus was eached, so n ths case the agents acheve the desed fomaton tanslated nto the XY plane b a vecto whch s the aveage of the ntal postons (aveage consensus). Hence, the stablt esults sum up n the followng coolla to the Theoem 4 n [40], n whch tme-dependent tgge functons of the fom: f ( 1 t e ) e ( c0 c e ) (8) ae consdeed. Remak 1. In tgge functons gven n Equaton (8) t s assumed that the ntal condtons ae t 0 = 0. Othewse, Equaton (8) can be ewtten as: f ( e ) e ( c 0 c e ( tt 0 ) 1 whee t 0 efes to the ntal tme of the expement. Thus, the theshold deceases exponentall and takes the value c 0 n the lmt when tme goes to nfnt. Coolla 2. If tgge functons gven n Equaton (8) ae defned fo the eos n the x- and - coodnates, such that: e e x,, xˆ ˆ x, the fomaton gven b x, s eached wth an eo that depends on the value of paamete c 0 of the tgge functon, and the cente of the fomaton s the aveage of the ntal postons: ) (9) (10)

11 Sensos 2013, Poof. Let us denote: a a x x 1 N 1 N N 1 N 1 x x (0) (0). x a 1 x a 1 (11) (12) the dsageement vectos n the x- and -coodnates, espectvel, whee 1 = (1 1) T egenvecto of the laplacan L [3]. Thus, t holds that x,, and t elds: x L x ex L e, x s an (13) whee e x = (e x,1 e x,n ) T and e = (e,1 e,n ) T. Hence, applng the esults of Theoem 4 n [40], t elds the dsageement vectos, convege to the followng egons, espectvel: x L x ( G) 2 L ( G) 2 Nc 0 Nc, whee ( ) s the algebac connectvt of the gaph G (second smallest egenvalue of L). 2 G Remak 2. Note that the convegence of the dsageement vectos x, poves that the goup of obots eaches the fomaton at ( ax1 x, a1 ) wth an eo gven b Equaton (14). Despte the fact that the exstng theo that poves the obustness of the poposed contol stateg and the smulaton esults that show a sgnfcant mpovement as fa as the numbe of tansmssons between agents s concened, thee ae nethe expemental esults on a eal platfom no a smulaton tool to pefom a wde set of expements of fomaton contol of moble obots. The followng secton descbes the developed softwae and all ts capabltes. 4. Smulaton Tool The man contol and sstem confguaton optons needed to smulate a mult-agent sstem demand a gaphcal use nteface (GUI) wth a hgh degee of flexblt. The GUI s ntended to make apd pototpng and smulaton of weless autonomous agents whch execute dstbuted contol algothms and pefom event-based communcatons. Nevetheless, the GUI has been made keepng the nteacton wth the use elatvel smple and ntutve n ode to be used as a pedagogcal tool fo advanced engneeng contol couses. The smulaton tool s avalable on lne at See [43] fo futhe detals Descpton of the Smulato The nteface of the applcaton, shown n Fgue 5, has fve man panels, a menu ba, and a small task ba. The two uppe panels of the nteface povde a quck vew of the mult-agent sstem and a tme plot of the outputs and contol sgnals. The top left panel shows an anmaton of the complete 0 (14)

12 Sensos 2013, mult-agent sstem. Each agent s numbeed and shows a tace of ts fome postons. The lowe left panel, named Agent Paametes, allows uses to set the numbe of agents n the sstem, as well as to add and emove a patcula agent. Fgue 5. Illustatve example of the smulaton tool. A mult-agent sstem wth fou agents s smulated. Note that the contol actons ae constant between two communcaton events. The tme plots on the top ght panel, whch ae gouped n the Sstem Sgnals tab, dspla the elatve dstance to the desed fomaton as well as the contol actons of each agent. Thee ae also plots gouped n the Netwok Sgnals tab, whch povde nfomaton about the measued dela of the sent packets. The lowe panel, named Netwok Paametes, s devoted to settng up the netwok dela and also the netwok packet loss. Uses can choose ths panel to set fxed o andom dela and loss pobablt functons. The lowe ght panel Contol Paametes s used to specf the communcaton stateg whch tgges the sendng of packets fom the agents n ode to update the contol actons. The nteface s completed wth a top task ba that povdes buttons to stat, pause, and eset the smulaton. The components of the nteface descbed eale povde the basc functonalt equed to opeate the applcaton. A menu ba povdes some addtonal featues such as the possblt to specf the dnamc model of the agents and to select a pedefned mult-agent sstem confguaton and the expement to be pefomed. The use can also save a confguaton, mak t as default o load a pevous confguaton A Pactcal Example Ths secton pesents a usage example of the smulato descbed above. Dffeent expements ae pefxed. We select one of them n the top menu, fo example, Expement 2, n whch the communcaton

13 Sensos 2013, topolog fo fou vehcles s gven b the gaph n Fgue 6, and the desed fomaton espect to the cente of the goup s: , x (11) Fgue 6. Communcaton gaph descbng one-dmensonal netwok topolog. The use can fnd n the help menu all the nfomaton about these expements. Random ntal condtons ae geneated, but we ma be nteested n compang the pefomance of dffeent contol stateges (peodc and event-based, fo example) fo a pefxed ntal coodnates. We can do that b ght-clckng n the dawng panel and selectng the opton Settng cuent state as default. Fgue 5 shows the smulaton output when the tgge functon s of the fom 0.2t f ( e ) e e. The use can change the paametes of f whle the smulaton s unnng o swtch to a peodc communcaton and check onlne the effect of these changes n the sstem. 5. Expemental Results 5.1. Compason of Smulaton and Expemental Results Fgue 7a,b shows the second component of the contol sgnals (ω ) and the dstances to the 2 2 fomaton d x x ) (, espectvel, fo the fou agents when the expement descbed (,, ) n Secton 4.2 s pefomed ove the model (sold lnes) and the eal sstem (dashed lnes). Fgue 7. Compaatve of smulaton and expemental esults: (a) Contol sgnal, (b) Dstance to fomaton. (a) (b)

14 Sensos 2013, We obseve that the majo dscepances ae at the begnnng of the expement. Note that the contol sgnals ae constant pecewse functons and change the value onl at event tmes, whle Equaton (6) eads that the change contnuousl wth the oentaton θ even though x ˆ x and ˆ eman constant n the nte-event tmes. The eason fo ths s the constant mposed b the hadwae n the eal sstem. Because one agent cannot monto ts state contnuousl (the camea sends measuements eve 200 ms, whch s the mnmum peod allowed b the Waspmote cads), and the slow vaaton of the angle, the computaton equed to mplement the contnuous contol law gven n Equaton (6) s not woth. Thus, the angle s consdeed to eman constant between events. Moeove, the selected tgge functons decease wth tme. Ths means that the theshold to detect an event s geate fo small values of tme, and fewe events ae geneated consequentl. When the sstem appoaches to the desed fomaton, the theshold deceases and makes the eo abta small. Note that d takes the same value fo all = 1,,4 and the fnal state onl depends on the ntal condtons, as pedcted b the consensus poblem theo (see Secton 3). Howeve, the fnal state fo the eal expement dffes slghtl fom the esult gven b the model. The eason fo that mght be sensng eos, dead-zone n the actuato and/o tansmsson delas. Fnall, Fgue 8 shows the vew of the expement at t = 0, 6, 12, and 24 s. Fgue 8. Vew of the eal expement at t = 0, 6, 12, and 24 s Event-Tgge Vesus Tme-Tgge Fgue 9 compaes the esults obtaned fom event-tgge and tme-tgge expements. In the last case, each agent eceves data fom the camea eve 200 ms, t pocesses the nfomaton, boadcasts ts state to ts neghbos, and computes the contol law eve two samples (400 ms). In the lowe gaph of Fgue 9 the boadcasts n both cases ae depcted fo each agent (labeled fom 1 to 4). Fo the peodc case, the numbe of packets sent though the netwok dung the expement should be 50s 4agents / Howeve, fo the event-based contol ths numbe s b fa lowe: 83 events n total.

15 Sensos 2013, Fgue 9. Compason of the esults obtaned fo event-tgge and tme-tgge expements. In the uppe gaph of Fgue 9 the dstance to the fomaton s shown. Note that even though the desed fomaton and the ntal condtons ae the same, the fnal state (whch depends on the aveage of the ntal postons) s dffeent. As emaked befoe, these dffeences come out because t s dffcult n expementaton to exactl epoduce the same condtons. Fo nstance, netwok delas cannot be contolled, o sensng eos ma occu moe fequentl n one case than n the othe. Howeve, obseve that the tme at whch the fomaton s acheved (the statona) s ve smla fo both tme-tgge and event-tgge. 6. Conclusons We have descbed an expemental envonment fo moble obots. The hadwae platfom was developed usng dffeent components and softwae to ovecome the lmtatons of the LEGO Mndstoms NXT obots. A dstbuted contol appoach has been used n fomaton contol. Each obot communcates to ts neghbohood n an event-tggeng fashon fo netwok savng puposes. A smulaton tool wth hgh degee of flexblt has been pesented. An example has demonstated ts usablt, and smulated and expemental esults have been pesented showng how the mult-agent sstem eaches the desed fomaton wth few communcatons meanwhle the sstem pefomance s peseved. Futue woks wll nclude the mplementaton of collson and obstacle avodance contol and fomaton contol wth specfed paths. Explong self-tggeng technques [44] to pedct the next event occuence n ode to optmze the communcaton between the camea and the obots s also a futue lne of eseach. Acknowledgments Ths wok was suppoted n pat b the Spansh Mnst of Econom and Compettveness unde Pojects DPI and DPI C Conflct of Inteest The authos declae no conflct of nteest.

16 Sensos 2013, Refeences 1. Fanch, L.; Feda, G.O.; Vendttell, M. The senso-based andom gaph method fo coopeatve obot exploaton. IEEE ASME Tans. Mechaton. 2009, 14, Magnenat, S.; Rétonaz, P.; Bonan, M.; Longchamp, V.; Mondada. F. ASEBA: A modula achtectue fo event-based contol of complex obots. IEEE ASME Tans. Mechaton. 2011, 6, Olfat-Sabe, R.; Fax, J.; Mua, R. Consensus and coopeaton n netwoked mult-agent sstems. IEEE Poc. 2007, 95, Lan, F.-L.; Mone, J.; Tlbu, D. Netwok desgn consdeaton fo dstbuted contol sstems. IEEE Tans. Contol Sst. Technol. 2002, 10, Astöm, K.; Benhadsson, B. Compason of Remann and Lebesgue Samplng fo Fst Ode Stochastc Sstems. In Poceedngs of IEEE Confeence on Decson and Contol, Las Vegas, NA, USA, Decembe 2002; pp Tabuada, P. Event-tggeed eal-tme schedulng of stablzng contol tasks. IEEE Tans. Autom. Contol 2007, 52, Rab, M.; Johansson, K.H.; Johansson, M. Optmal Stoppng fo Event-Tggeed Sensng and Actuaton. In Poceedngs of 47th IEEE Confeence on Decson and Contol, Cancun, Mexco, 9 11 Decembe 2008; pp Dmaogonas, D.V.; Fazzol, E.; Johansson, K.H. Dstbuted event-tggeed contol fo mult-agent sstems. IEEE Tans. Autom. Contol 2012, 57, Mazo, M.; Tabuada, P. Decentalzed event-tggeed contol ove weless senso/actuato netwoks. IEEE Tans. Autom. Contol 2011, 56, Wang, X.; Lemmon, M.D. Event-tggeng n dstbuted netwoked contol sstems. IEEE Tans. Autom. Contol 2011, 56, Machand, N.; Duand, S.; Gueeo-Castellanos, J.F. A geneal fomula fo event-based stablzaton of nonlnea sstems. IEEE Tans. Autom. Contol 2013, 58, Macdonald, E.A. Mult-Robot Assgnment and Fomaton Contol. M.Sc. Thess, Geoga Insttute of Technolog, Atlanta, GA, USA, 7 June Huang, J.; Fato, S.M.; Qad, A.; Goddad, S. Localzaton and follow-the-leade contol of a heteogeneous goup of moble obots. IEEE ASME Tans. Mechaton. 2006, 11, Mehjed, H.; Saad, M.; Ghommam, M.J. Heachcal fuzz coopeatve contol and path followng fo a team of moble obots. IEEE ASME Tans. Mechaton. 2011, 16, Cuz, D.; Mcclntock, J.; Peteet, B.; Oqueda, O.; Cao, Y.; Feo, R. A multvehcle platfom fo eseach n netwoked embedded sstems. IEEE Contol Sst. 2007, 27, Duand, S.; Mnet, J.; Gueeo, J.F.; Machand, N. ASYNCAR, a Rado-Contolled Vehcle fo Asnchonous Expements Implementaton of an Event-Based Cuse Contol. In Poceedngs of 8th Intenatonal Confeence on Electcal Engneeng, Computng Scence and Automatc Contol, Meda Ct, Mexco, Octobe 2011; pp Casan, M.; Gaull, A.; Ganntapan, A.; Vcno, A. A LEGO Mndstoms Mult-Robot Setup n the Automatc Contol Telelab. In Poceedngs of 18th Intenatonal Fedeaton of Automatc Contol (IFAC) Wod Congess, Mlano, Ital, 28 August 2 Septembe 2011; pp

17 Sensos 2013, Maze, M.; Wan, Y.; Namudu, K.; Vaanas, M. A LEGO Mndstoms NXT-Based Test Bench fo Cohesve Dstbuted Mult-Agent Exploato Sstems; Moblt and Coodnaton Poc.: Gaden Gove, CA, USA, June 2012; pp Fu, G.; Coad, P.; Mencass, A.; Dao, P. An ntegated tangulaton lase scanne fo obstacle detecton of mnatue moble obots n ndoo envonment. IEEE ASME Tans. Mechaton. 2011, 16, Espnosa, F.; Santos, C.; Maón-Romea, M.; Pzao, D.; Valdés, F.; Dongl, J. Odomet and lase scanne fuson based on a dscete extended kalman flte fo obotc platoonng gudance. Sensos 2011, 11, Chaos, D. Contol no lneal de vehículos subactuados manos no-holonómcos. Ph.D. Thess, Unvesdad Naconal de Educacón a Dstanca (UNED), Madd, Span, 10 Decembe Lbelum Comuncacones Dstbudas S.L., Waspmote Datasheet. Avalable onlne: lbelum.com/downloads/documentaton/waspmote_datasheet.pdf (accessed on 18 June 2013). 23. LEGO Goup. LEGO Mndstoms NXT Use Gude. Avalable onlne: en-us/suppot/buldngnstuctons/8547/8547%20use%20gude%20englsh.aspx (accessed on 18 June 2013). 24. HTechnc IR RC Kat Buldng Instuctons. Avalable onlne: fle.php?f=503-htrckat.pdf (accessed on 18 June 2013). 25. Robot Electoncs, Usng the I2C Bus. Avalable onlne: acatalog/i2c_tutoal.html (accessed on 4 Apl 2013). 26. Esquembe, F. Eas Java Smulatons: A softwae tool to ceate scentfc smulatons n Java. Comput. Phs. Commun. 2004, 156, Raap, XBee-API A Java API fo Dg XBee/XBee-Po OEM RF Modules. Avalable onlne: (accessed on 6 Apl 2013). 28. Vagas, H.; Sánchez, J.; Salzmann, C.; Esquembe, F.; Gllet, D.; Domdo, S. Web-enabled emote scentfc envonments. Comput. Sc. Eng. 2009, 11, Vgua, A.; Howad, M. An Integated appoach fo achevng multobot task fomatons. IEEE ASME Tans. Mechaton. 2009, 14, Dg Intenatonal, Inc. Poduct Manual v1.xex Potocol. Avalable onlne: (accessed on 7 Apl 2013). 31. Lbelum Comuncacones Dstbudas S.L. Waspmote Techncal Gude. Avalable onlne: (accessed on 7 Apl 2013). 32. Guascon, D. Long Range Multpotocol Weless Senso Netwoks. In Poceedngs of Best Demo Awad at the Sensng Technolog 2010, Bmngham, UK, Natonal Instuments, LabVIEW Toolkts fo LEGO Mndstoms NXT Pogammng Gude. Avalable onlne: Pogammng_Gude.pdf (accessed on 18 June 2013). 34. Godsl, G.R. Algebac Gaph Theo; Spnge: New Yok, NY, USA, Ren, W.; Atkns, E. Dstbuted mult-vehcle coodnated contol va local nfomaton exchange. Int. J. Robust Nonlnea Contol 2007, 17,

18 Sensos 2013, Ren, W.; Mooe, K.; Chen, Y. Hgh-Ode Consensus Algothms n Coopeatve Vehcle Sstems. In Poceedngs of Intenatonal Confeence on Netwokng, Sensng and Contol, Ft. Laudedale, FL, USA, Apl 2006; pp Seo, J.H.; Shm, H.; Back, J. Consensus of hgh-ode lnea sstems usng dnamc output feedback compensato: Low gan appoach. Automatca 2009, 45, Lawton, J.R.T.; Bead, R.W.; Young, B.J. A decentalzed appoach to fomaton maneuves. IEEE Tans. Robot. Autom. 2003, 19, Lemmon, M. Event-Tggeed Feedback n Contol, Estmaton, and Optmzaton. In Netwoked Contol Sstems; Spnge Velag: Hedelbeg, Geman, Seboth, G.; Dmaogonas, D.; Johansson, K. Contol of Mult-Agent Sstems va Event-Based Communcaton. In Poceedngs of 18th Intenatonal Fedeaton of Automatc Contol (IFAC) Wold Congess, Mlano, Ital, 28 August 2 Septembe 2011; pp Dem, O.; Lunze, J. Coopeatve Contol of Mult-Agent Sstems wth Event-Based Communcaton. In Poceedngs of Amecan Contol Confeence, Monteal, Canada, June 2012; pp Gunaldo, M.; Dmaogonas, D.; Johansson, K.; Sanchez, J.; Domdo, S. Dstbuted Event-Based Contol fo Inteconnected Lnea Sstems. In Poceedngs of 50th IEEE Confeence on Decson and Contol, Olando, FL, USA, Decembe 2011; pp Gunaldo, M.; Faas, G.; Fabegas, E.; Sánchez, J.; Domdo-Canto, S.; Domdo, S. An nteactve smulato fo netwoked moble obots. IEEE Netw. Mag. 2012, 26, Mazo, M.; Anta, A.; Tabuada, P. An ISS self-tggeed mplementaton fo lnea contolles. Automatca 2010, 46, b the authos; lcensee MDPI, Basel, Swtzeland. Ths atcle s an open access atcle dstbuted unde the tems and condtons of the Ceatve Commons Attbuton lcense (

Experiment 26 Refraction

Experiment 26 Refraction Expement 26 Refacton Advanced Readng: (Knght 4 th Edton) Chapte 34 secton 34-3 Equpment: 1 Plex-Ray 2 cokboads 2 potactos 2 30 cm ules Objectve: The object of ths expement s the phenomena of efacton. Theoy:

More information

An Accurate and Fast WLAN User Location Estimation Method Based on Received Signal Strength

An Accurate and Fast WLAN User Location Estimation Method Based on Received Signal Strength An Accuate and Fast WLAN Use Locaton Estmaton Method Based on Receved Sgnal Stength Mnghua Zhang and Shensheng Zhang Depatment of Compute Scence and Engneeng, Shangha Jao Tong Unvesty, Chna {zhangmh, sszhang}@sjtu.edu.cn

More information

Power Electronics (25) Please prepare your student ID card (with photo) on your desk for the attendance check.

Power Electronics (25) Please prepare your student ID card (with photo) on your desk for the attendance check. Depatment of Powe Electoncs and Electcal Dves Pof D ng Joachm Böcke Powe Electoncs 31032015 Suname Student numbe Fst name Couse of study Task (Ponts) 1 2 3 4 Total (100) Mak Duaton 120 mnutes Pemtted esouces

More information

Robust frequency offset estimator for OFDM over fast varying. multipath channel

Robust frequency offset estimator for OFDM over fast varying. multipath channel Robust equency oset estmato o OFDM ove ast vayng multpath channel L We, Youyun Xu, Yuemng Ca and Xn Xu Ths pape pesents a obust cae equency oset(cfo) estmaton algothm sutable o ast vayng multpath channels.

More information

A Control Method for SVG Based on Differential Geometry Nonlinear Control

A Control Method for SVG Based on Differential Geometry Nonlinear Control Enegy and Powe Engneeng,, 3, 7-75 do:.436/epe..3334 Publshed Onlne July (http://www.scp.og/jounal/epe) A Contol Method fo SVG Based on Dffeental Geomety Nonlnea Contol Abstact Q-Yong Pan, Jn Wu, Y-Hua

More information

Frequency Reuse Distance of A 2-hop Virtual Cellular Network

Frequency Reuse Distance of A 2-hop Virtual Cellular Network oceedngs of ACC8 copyght 8 IEICE 8 SB 8 Fequency Reuse Dstance of A -hop Vtual Cellula etwok Esuke KUDOH and Fumyuk ADACHI ohoku Unvesty Senda, 98-8579, Japan Abstact hee have been stong demands fo hghe

More information

Research on the Electromagnetic Interference of Antennas on the Satellite

Research on the Electromagnetic Interference of Antennas on the Satellite PIERS ONLINE, VOL. 4, NO. 3, 28 336 Reseach on the Electomagnetc Intefeence of Antennas on the Satellte Bn Zhou 1, Qzhong Lu 2, and Xnyang He 1 1 Reseach Offce of Mcowave Technology X an Insttute of Space

More information

HIGH FREQUENCY BOOST CONVERTER EMPLOYING SOFT SWITCHING AUXILIARY RESONANT CIRCUIT

HIGH FREQUENCY BOOST CONVERTER EMPLOYING SOFT SWITCHING AUXILIARY RESONANT CIRCUIT HIGH FREQUENCY BOOST CONERTER EMPOYING SOFT SWITCHING AUXIIARY RESONANT CIRCUIT G. NARESH GOUD, Y.AKSHMI DEEPA, 3 G.DII BABU, 4 P.RAJASEKHAR & 5 N.GANGADHER Abstact - A new soft-swtchng boost convete s

More information

Simulation of ZVS H-Bridge Inverter Using Soft Switching Boost Converter

Simulation of ZVS H-Bridge Inverter Using Soft Switching Boost Converter K. Seekanth Reddy Smulaton of ZS H-Bdge Inete Usng Soft Swtchng Boost Conete K. Seekanth Reddy ABSTRACT: Ths pape pesents a two stage zeo-oltageswtchng H-Bdge nete usng soft swtchng boost conete. The conentonal

More information

3 THE INTERFERENCE PATTERN GNSS-R

3 THE INTERFERENCE PATTERN GNSS-R Chapte 3 The Intefeence Patten GNSS-R Technque 3 THE INTERFERENCE PATTERN GNSS-R TECHNIQUE 3. Intoducton GNSS-R technques have been used n ocean scenaos wth the man goal of educng eos n sea suface salnty

More information

Journal of Applied Research and Technology ISSN: Centro de Ciencias Aplicadas y Desarrollo Tecnológico.

Journal of Applied Research and Technology ISSN: Centro de Ciencias Aplicadas y Desarrollo Tecnológico. Jounal of Appled Reseach and Technology ISSN: 1665-6423 jat@aleph.cnstum.unam.mx Cento de Cencas Aplcadas y Desaollo Tecnológco Méxco Dn, I.; Km, J.; Km, H. Jont Powe and Beam Angle Contol fo Enegy Effcency

More information

Recommendation ITU-R M (03/2006)

Recommendation ITU-R M (03/2006) ecommendaton ITU- M.64- (3/26) A methodology fo co-channel ntefeence evaluaton to detemne sepaaton dstance fom a system usng hgh-alttude platfom statons to a cellula system to povde IMT-2 sevce M ees Moble,

More information

Resonant DC Link in PWM AC Chopper

Resonant DC Link in PWM AC Chopper Intenatonal Jounal of Electcal and Compute Engneeng Resonant DC Lnk n PWM AC Choppe Apnan Auasopon Abstact Ths pape poposes a esonant dc lnk n PWM ac choppe. Ths can sole the spke poblems and also educe

More information

Measurement of thickness of dielectric materials from microwave reflection data

Measurement of thickness of dielectric materials from microwave reflection data st Intenatonal Confeence on Sensng Technology Novembe -, 5 Palmeston Noth, New Zealan Measuement of thckness of electc mateals fom mcowave eflecton ata Abstact Kalash Thaku an Wayne Holmes Electomagnetc

More information

Real Time Voltage Stability Prediction based on Modal Analysis for Supporting Wide Area Monitoring System

Real Time Voltage Stability Prediction based on Modal Analysis for Supporting Wide Area Monitoring System Real Tme Voltage Stablty Pedcton based on Modal Analyss fo Suppotng Wde Aea Montong System Lesnanto.Multa Putanto, Naebboon. Hoonchaeon Abstact In powe system tansmsson netwok, voltage stablty poblem can

More information

A Dynamic Networking Substrate for Distributed MMOGs

A Dynamic Networking Substrate for Distributed MMOGs Receved 31 Octobe 2013; evsed 9 Apl 2014; accepted 9 June 2014. Date of publcaton 00 xxxx 0000; date of cuent veson 00 xxxx 0000. Dgtal Object Identfe 10.1109/TETC.2014.2330520 A Dynamc Netwokng Substate

More information

EWFQ: Enhanced Weighted Fair Queuing Scheme ensuring accurate inter-packets spacing

EWFQ: Enhanced Weighted Fair Queuing Scheme ensuring accurate inter-packets spacing EWFQ: Enhanced Weghted Fa Queung Scheme ensung accuate nte-packets spacng Mudass TUFAIL and Benad COUSIN IRISA, Campus de Beauleu 35042 Rennes Cedex, Fance. fmtufal, bcousng@sa.f tel: (33) 2 99 84 71 00

More information

Optimal Resource Allocation in Wireless Multiaccess Video Transmissions

Optimal Resource Allocation in Wireless Multiaccess Video Transmissions Optmal Resouce Allocaton n Weless Multaccess Vdeo Tansmssons Cong Shen and Mhaela van de Schaa Electcal Engneeng Depatment Unvesty of Calfona, Los Angeles Los Angeles, CA 90095-1594 Emal: {congshen,mhaela@ee.ucla.edu

More information

Implementing a Real-Time Locating System Based on Wireless Sensor Networks and Artificial Neural Networks to Mitigate the Multipath Effect

Implementing a Real-Time Locating System Based on Wireless Sensor Networks and Artificial Neural Networks to Mitigate the Multipath Effect 14th Intenatonal Confeence on Infomaton Fuson Chcago, Illnos, USA, July 5-8, 2011 Implementng a Real-Tme Locatng System Based on Weless Senso Netwoks and Atfcal Neual Netwoks to Mtgate the Multpath Effect

More information

Latency-Rate Servers: A General. Model for Analysis of Trac. Dimitrios Stiliadis. Anujan Varma UCSC-CRL July 18, 1995.

Latency-Rate Servers: A General. Model for Analysis of Trac. Dimitrios Stiliadis. Anujan Varma UCSC-CRL July 18, 1995. Latency-Rate Seves: A Geneal Model fo Analyss of Tac Schedulng Algothms Dmtos Stlads Anujan Vama UCSC-CRL-95-38 July 18, 1995 Baskn Cente fo Compute Engneeng & Infomaton Scences Unvesty of Calfona, Santa

More information

POWER FLOW ANALYSIS USING NR METHOD

POWER FLOW ANALYSIS USING NR METHOD POWER FLOW ANALYSIS USING NR METHOD Adtya Shaa 1, Manyu San 2 and MdIfan Ahed 3 1 UG Student, 2,3 Assstant Pofesso, Electcal Engneeng, Caee Pont Unvesty, Kota, Raasthan, (Inda) ABSTRACT Powe s geneated

More information

Dynamic Optimization. Assignment 1. Sasanka Nagavalli January 29, 2013 Robotics Institute Carnegie Mellon University

Dynamic Optimization. Assignment 1. Sasanka Nagavalli January 29, 2013 Robotics Institute Carnegie Mellon University Dynamc Optmzaton Assgnment 1 Sasanka Nagavall snagaval@andrew.cmu.edu 16-745 January 29, 213 Robotcs Insttute Carnege Mellon Unversty Table of Contents 1. Problem and Approach... 1 2. Optmzaton wthout

More information

Cross Entropy Optimization for Optimal Design of Water Distribution Networks

Cross Entropy Optimization for Optimal Design of Water Distribution Networks Intenatonal Jounal of Compute Infomaton Systems and Industal Management Applcatons. ISSN 250-7988 Volume 5 (203) (202) pp. 308-36 MIR Labs, www.mlabs.net/csm/ndex.html Coss Entopy Optmzaton fo Optmal Desgn

More information

Efficient VLSI Layouts for Homogeneous Product Networks

Efficient VLSI Layouts for Homogeneous Product Networks 17 IEEE TRANSACTIONS ON COMPUTERS, VOL. 46, NO. 1, OCTOBER 1997 Effcent VLSI Layouts fo Homogeneous Poduct Netwoks Antono enández and Kemal Efe, Membe, IEEE Abstact In ths pape, we develop genealzed methods

More information

where and are polynomials with real coefficients and of degrees m and n, respectively. Assume that and have no zero on axis.

where and are polynomials with real coefficients and of degrees m and n, respectively. Assume that and have no zero on axis. function whee is an unknown constant epesents fo the un-modeled dynamics The pape investigates the position contol of electical moto dives that can be configued as stuctue of Fig 1 This poblem is fomulated

More information

Robust 3D Camera Motion Parameter Estimation for Video Coding

Robust 3D Camera Motion Parameter Estimation for Video Coding 96 IJCSNS Intenatonal Jounal of Compute Scence and Netwo Secuty, VOL.7 No., Febuay 7 Robust 3D Camea Moton Paamete Estmaton fo Vdeo Codng Eung-ae Km, Koea Polytechnc Unvesty, Sheung-Cty, Kyongg-Do, Koea

More information

Performance Analysis of Channel Estimation and Adaptive Equalization in Slow Fading Channel

Performance Analysis of Channel Estimation and Adaptive Equalization in Slow Fading Channel Pefomance Analyss of Channel Estmaton and Adaptve Equalzaton n Slow Fadng Channel Zhfeng Chen, Electcal and Compute Engneeng, Unvesty of Floda ABSTRACT In ou poject, we fst uld up a weless communcaton

More information

Low-Complexity Time-Domain SNR Estimation for OFDM Systems

Low-Complexity Time-Domain SNR Estimation for OFDM Systems Low-Complexity Time-Domain SR Estimation fo OFDM Systems A. jaz, A.B. Awoseyila and B.G. Evans A low-complexity SR estimation algoithm fo OFDM systems in fequency-selective fading channels is poposed.

More information

Geometric Factors in Target Positioning and Tracking

Geometric Factors in Target Positioning and Tracking th Intenatonal Confeence on Infomaton Fuson Seattle, WA, USA, July 6-9, 9 Geometc Factos n aget Postonng and ackng Chun Yang Sgtem echnology, Inc. San Mateo, CA 944 chunynag@sgtem.com Ek Blasch A Foce

More information

Desgin of On-line Monitoring Device for MOA (Metal Oxide Arrestor) Based on FPGA and C8051F

Desgin of On-line Monitoring Device for MOA (Metal Oxide Arrestor) Based on FPGA and C8051F Sensos & Tansduces, Vol. 8, Issue, Octobe 24, pp. 56-6 Sensos & Tansduces 24 by IFSA Publshng, S. L. http://www.sensospotal.com Desgn of On-lne Montong Devce fo MOA (Metal Ode Aesto) Based on FPGA and

More information

SENSORED and SENSORLESS CONTROL OF PERMANENT MAGNET SYNCHRONOUS MOTORS USING MATRIX CONVERTERS

SENSORED and SENSORLESS CONTROL OF PERMANENT MAGNET SYNCHRONOUS MOTORS USING MATRIX CONVERTERS SENSORED and SENSORLESS CONTROL OF PERMANENT MAGNET SYNCHRONOUS MOTORS USING MATRIX CONVERTERS A. Aas and D. Saltve *Dept. d'engnyea Electònca. Unvestat Poltècnca de Catalunya. C. Colom 1. 8222 Teassa.

More information

N2-1. The Voltage Source. V = ε ri. The Current Source

N2-1. The Voltage Source. V = ε ri. The Current Source DC Cicuit nalysis The simplest cicuits to undestand and analyze ae those that cay diect cuent (DC). n this note we continue ou study of DC cicuits with the topics of DC voltage and cuent souces, the idea

More information

Online Signature Verification Using Angular Transformation for e-commerce Services

Online Signature Verification Using Angular Transformation for e-commerce Services Wold Academy of Scence, Engneeng and Technology Intenatonal Jounal of Electcal and Compute Engneeng Vol:4, No:1, 21 Onlne Sgnatue Vefcaton Usng Angula Tansfomaton fo e-commece Sevces Peeapong Uthansakul

More information

Unified Power Quality Conditioner (UPQC) with Voltage Dips and Over-voltages Compensation Capability

Unified Power Quality Conditioner (UPQC) with Voltage Dips and Over-voltages Compensation Capability Unfed Powe Qualty Condtone UPQC wth oltage Dps and Oe-oltages Compensaton Capablty ícto M. Moeno, Albeto Pgazo, Maco Lsee and Antono Dell Aqula Depatamento de Electonca y Computadoes Unesdad de Cantaba

More information

Real-time scheduling for systems with precedence, periodicity and latency constraints

Real-time scheduling for systems with precedence, periodicity and latency constraints Real-tme schedulng fo systems wth pecedence, peodcty and latency constants lana ucu, Rémy Kock, Yves Soel () INRI Rocquencout, BP 05-853 e hesnay edex, Fance llana.cucu@na.f, yves.soel@na.f (2) ESIEE,

More information

Applying Computational Methods to Test Site and Antenna Design

Applying Computational Methods to Test Site and Antenna Design Applyng Computatonal Methods to Test Ste and Antenna Desgn By Glen Dash, Ampyx LLC, GlenDash at alum.mt.edu Copyght 000, 005 Ampyx LLC In the past, mathematcs and expementaton wee the only tools engnees

More information

Node-Disjoint On-demand Multipath Routing with Route Utilization in Ad-Hoc Networks

Node-Disjoint On-demand Multipath Routing with Route Utilization in Ad-Hoc Networks Node-Dsjont On-demand Multpath Routng wth Route Utlzaton n Ad-Hoc Netwos Mona Gupta Student- M.E Neeaj Kuma Assstant Pofesso CSED, Thapa UnvestyCSED, Thapa Unvesty Patala, Inda Patala, Inda ABSTRACT Moble

More information

New Distributed QoS Control Scheme for IEEE Wireless Access Networks

New Distributed QoS Control Scheme for IEEE Wireless Access Networks New Dstbuted QoS Contol Scheme fo IEEE 802.16 Weless Access Netwoks Xaofeng Ba 1, Abdallah Sham 1, Khalm Amjad Meeja 1 and Chad Ass 2 1 The Unvesty of Westen Ontao, Canada, xba6@uwo.ca, asham@eng.uwo.ca,

More information

Theoretical And Experiment Researches On A Multi-Temperature Quasi-Cascade Refrigerating System

Theoretical And Experiment Researches On A Multi-Temperature Quasi-Cascade Refrigerating System Pudue Unvesty Pudue e-pubs Intenatonal Refgeaton and A Condtonng Confeence School of Mechancal Engneeng 2010 Theoetcal And Expement Reseaches On A Mult-Tempeatue Quas-Cascade Refgeatng System Yuhu Lang

More information

Varney Land Surveys (Dick Varney) th Avenue Sacramento, CA 95822

Varney Land Surveys (Dick Varney) th Avenue Sacramento, CA 95822 CTY OF SACRAMENTO DEPARTMENT OF PLANNNG & DEVELOPMENT ZONNG ADMNSTRATOR 23 Steet Sacamento CA 9584 ACTON OF THE ZONNG ADMNSTRATOR On Thusday Decembe 9 2002 the Zonng Admnstato appoved wth condtons a pacel

More information

Customized Beam Forming at the Allen Telescope Array

Customized Beam Forming at the Allen Telescope Array ATA Memo #51 Customzed Beam Fomng at the Allen Telescope Aay G. R. Hap August 11, 22 Abstact One of the exctng pospects fo lage N aays s the potental fo custom beam fomng when opeatng n phased aay mode.

More information

Electrical Modeling of Wirebonds in Stacked ICs Using Cylindrical Conduction Mode Basis Functions

Electrical Modeling of Wirebonds in Stacked ICs Using Cylindrical Conduction Mode Basis Functions Electcal Modelng of Webonds n Stacked ICs Usng Cylndcal Conducton Mode Bass Functons K Jn Han, Madhavan Swamnathan, and Ege Engn School of Electcal and Compute Engneeng, Geoga Insttute of Technology Fest

More information

A VECTOR CONTROLLED INDUCTION MOTOR DRIVE WITH NEURAL NETWORK BASED SPACE VECTOR PULSE WIDTH MODULATOR

A VECTOR CONTROLLED INDUCTION MOTOR DRIVE WITH NEURAL NETWORK BASED SPACE VECTOR PULSE WIDTH MODULATOR A ECOR CONROLLED INDUCION MOOR DRIE WIH NEURAL NEWORK BAED PACE ECOR PULE WIDH MODULAOR Rajesh Kuma, R. A. Gupta, Rajesh. ujuse Reade., Depatment of Electcal Engneeng, MNI, Japu, Inda- Pof., Depatment

More information

Tensor Product Model Transformation-based Sliding Surface Design

Tensor Product Model Transformation-based Sliding Surface Design Acta Polytechnca Hungaca Vol. 3, No. 4, 6 Tenso Poduct Model Tansfomaton-based Sldng Suface Desgn Péte Koond Depatment of Automaton and Appled Infomatcs Budapest Unvesty of Technology and Economcs PO.

More information

Design and Optimization of VoD schemes with Client Caching in Wireless Multicast Networks

Design and Optimization of VoD schemes with Client Caching in Wireless Multicast Networks Desgn and Optmzaton of VoD schemes wth Clent Cachng n Weless Multcast Netwoks Hao Feng, Student Membe, IEEE, Zhyong Chen, Membe, IEEE, and Hu Lu, Fellow, IEEE axv:78.366v [cs.it] Aug 27 Abstact Due to

More information

Efficient Power Control for Broadcast in Wireless Communication Systems

Efficient Power Control for Broadcast in Wireless Communication Systems Efficient Powe Contol fo Boadcast in Wieless Communication Systems A. T. Chonopoulos Compute Science Depatment Univesity of Texas at San Antonio San Antonio, TX Email:atc@cs.utsa.edu P. Cotae Depatment

More information

Geometric Factors in Target Positioning and Tracking

Geometric Factors in Target Positioning and Tracking th Intenatonal Confeence on Infomaton Fuson Seattle, WA, USA, July 6-9, 9 Geometc Factos n aget Postonng and ackng Chun Yang Sgtem echnology, Inc. San Mateo, CA 944 chunynag@sgtem.com Ek Blasch A Foce

More information

Demosaicking using Adaptive Bilateral Filters

Demosaicking using Adaptive Bilateral Filters Demosaicking using Adaptive Bilateal Filtes Ekine Akuiyibo Timothy Atobatele EE 362, Winte 2006 0. Abstact Digital cameas sample the continuous spectum using an aay of colo filtes such that each pixel

More information

A Novel Phase Shifted DC-DC Converter with Adaptive Soft Switching to Improve Efficiency under wide Load Range

A Novel Phase Shifted DC-DC Converter with Adaptive Soft Switching to Improve Efficiency under wide Load Range ntenatonal Electcal Engneeng Jounal (EEJ) Vol. 6 (05) No.3, pp. 8-87 A Novel Phase Shfted DC-DC Convete wth Adaptve Soft Swtchng to mpove Effcency unde wde Load Range Sudha Bansal, Lalt Mohan San bansal.sudha@gmal.com

More information

DISTRICT USES; SAID ZONING MAP AND ORDINANCE BEING NUMBERED REPEALING ALL ORDINANCES OR PARTS OF ORDINANCES IN CONFLICT

DISTRICT USES; SAID ZONING MAP AND ORDINANCE BEING NUMBERED REPEALING ALL ORDINANCES OR PARTS OF ORDINANCES IN CONFLICT ORDNANCE 55-28 AN ORDNANCE OF THE CTY OF GRAND PRARE, TEXAS, AMENDNG THE ZONNG MAP AND ORDNANCE TO REZONE THE SOUTHERN HALF OF LOT 5, BLOCK 38, DALWORTH PARK ADDTON, CTY OF GRAND PRARE, DALLAS COUNTY,

More information

Decoupled Vector Control Scheme for Dual Three-Phase Permanent Magnet Synchronous Machines

Decoupled Vector Control Scheme for Dual Three-Phase Permanent Magnet Synchronous Machines 1 Decoupled Vecto Contol Scheme fo Dual Thee-Phase Pemanent Magnet Synchonous Machnes Juss Kattunen, Samul Kallo, Pas Peltonem, Pett Slventonen, and Oll Pyhönen. Abstact Dual thee-phase electc machnes

More information

Massive Data Delivery in Unstructured Peer-to-Peer Networks with Network Coding

Massive Data Delivery in Unstructured Peer-to-Peer Networks with Network Coding Yu, Shu and L, hongwen 7, Massve data delvey n unstuctued pee-to-pee netwoks wth netwok codng, n IEEE/ACIS Intenatonal Confeence on Compute and Infomaton Scence (ICIS 7), IEEE Compute Socety, Los Alamtos,

More information

Distributed Instruments for Control and Diagnosis Applied to a Water Distribution System

Distributed Instruments for Control and Diagnosis Applied to a Water Distribution System IEEE Instumentaton and Measuement echnology Confeence Anchoage, AK, USA, 21-23 May 2002 Dstbuted Instuments fo Contol and Dagnoss Appled to a Wate Dstbuton System Dde GEORGES 1, Ec BENOI 2, Andé CHOVIN

More information

School of Electrical and Computer Engineering, Cornell University. ECE 303: Electromagnetic Fields and Waves. Fall 2007

School of Electrical and Computer Engineering, Cornell University. ECE 303: Electromagnetic Fields and Waves. Fall 2007 School of Electical and Compute Engineeing, Conell Univesity ECE 303: Electomagnetic Fields and Waves Fall 007 Homewok 1 Due on Nov. 8, 007 by 5:00 PM Reading Assignments: i) Review the lectue notes. ii)

More information

TAKAGI-SUGENO MODEL FOR QUADROTOR MODELLING AND CONTROL USING NONLINEAR STATE FEEDBACK CONTROLLER

TAKAGI-SUGENO MODEL FOR QUADROTOR MODELLING AND CONTROL USING NONLINEAR STATE FEEDBACK CONTROLLER ntenatonal Jounal of Contol Theo and Compute Modellng (JCTCM) Vol., No., Ma TAKAG-SUGENO MODEL FOR QUADROTOR MODELLNG AND CONTROL USNG NONLNEAR STATE FEEDBACK CONTROLLER Fouad Yacef, Oma Bouhal, Hcham

More information

10! !. 3. Find the probability that a five-card poker hand (i.e. 5 cards out of a 52-card deck) will be:

10! !. 3. Find the probability that a five-card poker hand (i.e. 5 cards out of a 52-card deck) will be: MATH 0(001 Fall 2018 Homewok 2 Solutions Please infom you instucto if you find any eos in the solutions 1 Suppose that thee ae duck huntes, each with a pefect shot A flock of ducks fly ove, and each hunte

More information

An Efficient Control Approach for DC-DC Buck-Boost Converter

An Efficient Control Approach for DC-DC Buck-Boost Converter 2016 Published in 4th Intenational Symposium on Innovative Technologies in Engineeing and Science 3-5 Novembe 2016 (ISITES2016 Alanya/Antalya - Tukey) An Efficient Contol Appoach fo DC-DC Buck-Boost Convete

More information

Fuzzy Control of a Supercritical Once-Through Boiler

Fuzzy Control of a Supercritical Once-Through Boiler Fuzzy Contol of a Supectcal Once-hough Bole S. Al A. Moosavan Al Ghaffa 2 Al Chabakhsh 3 Depatment of Mechancal Engneeng. N. oos Unv. of echnology, ehan, Ian, P.O. Box 635-996, Fax: (+98 2-733-4338 Emal:

More information

Design of A Circularly Polarized E-shaped Patch Antenna with Enhanced Bandwidth for 2.4 GHz WLAN Applications

Design of A Circularly Polarized E-shaped Patch Antenna with Enhanced Bandwidth for 2.4 GHz WLAN Applications VNU Jounal of Science: Comp. Science & Com. Eng., Vol. 31, No. 2 (2015) 1-7 Design of A Ciculaly Polaized E-shaped Patch Antenna with Enhanced Bandwidth fo 2.4 GHz WLAN Applications Hong Van Tam 1, Luong

More information

ABSTRACTT FFT FFT-' Proc. of SPIE Vol U-1

ABSTRACTT FFT FFT-' Proc. of SPIE Vol U-1 Phase econstuction stategies in phase-locking system based on multi-apetue wave font sensos P.A. Semenov, S.D. Pol skikh Shvabe-Reseach, Moscow, Russian Fedeation; e-mail: pite@bk.u ABSTRACTT System of

More information

SOFT-SWITCHING ACTIVE POWER FILTER FOR REDUCING HARMONIC CURRENT IN SYSTEM WITH NON-LINIER LOAD

SOFT-SWITCHING ACTIVE POWER FILTER FOR REDUCING HARMONIC CURRENT IN SYSTEM WITH NON-LINIER LOAD VO. 2, NO. 8, SEPTEMBER 207 ISSN 89-6608 ARPN Jounal of Engneeng and Appled Scences 2006-207 Asan Reseach Publshng Netwok ARPN). All ghts eseved. www.apnjounals.com SOFT-SWITHING ATIVE POWER FITER FOR

More information

Near-field emission prediction of CESAME

Near-field emission prediction of CESAME Nea-field emission pediction of CESAME E. Sicad, S. Ben Dhia, E. Lamoueux INSA-GEI, 135 Av de Rangueil 31077 Toulouse Fance B. Vignon, L. Couau ST Micoelectonics, Cental R&D Colles, Fance Contact : etienne.sicad@insa-toulouse.f

More information

Design of FIR Filter using Filter Response Masking Technique

Design of FIR Filter using Filter Response Masking Technique Design of FIR Filte using Filte Response Masking Technique Sandeep Shivastava, Alok Jain, Ram Kuma Soni Abstact- In this pape autho is tying to implement Fequency esponse masking (FRM) technique. In this

More information

Kalman-Filter-Based Integer Ambiguity Resolution Strategy for Long-Baseline RTK with Ionosphere and Troposphere Estimation

Kalman-Filter-Based Integer Ambiguity Resolution Strategy for Long-Baseline RTK with Ionosphere and Troposphere Estimation Kalman-Flte-Based Intege Ambguty Resoluton Stategy fo Long-Baselne RK wth Ionosphee and oposphee Estmaton omoj aasu, Ao Yasuda oyo Unvesty of Mane Scence and echnology, Japan evsed 2010-10-03 BIOGRAPHY

More information

Optimum Air-Gap Flux Distribution with Third Harmonic Rotor Flux Orientation Adjustment for Five-Phase Induction Motor

Optimum Air-Gap Flux Distribution with Third Harmonic Rotor Flux Orientation Adjustment for Five-Phase Induction Motor J Elect Eng Technol.018; 13(1): 315-35 http://do.og/10.5370/jeet.018.13.1.315 ISSN(Pnt) 1975-010 ISSN(Onlne) 093-743 Optmum A-Gap Flux Dstbuton wth Thd Hamonc Roto Flux Oentaton Adjustment fo Fve-Phase

More information

COMPARISON OF GRAPH-BASED METHODS OF KINEMATICAL ANALYSIS OF PLANETARY GEARS

COMPARISON OF GRAPH-BASED METHODS OF KINEMATICAL ANALYSIS OF PLANETARY GEARS Jóef Dewnak, Stansław Zawślak Compason of gap-based metods of knematcal analyss of planetay geas COMPARISON OF GRAPH-BASED METHODS OF KINEMATICAL ANALYSIS OF PLANETARY GEARS Jóef DREWNIAK *, Stansław ZAWIŚLAK

More information

Section 5. Signal Conditioning and Data Analysis

Section 5. Signal Conditioning and Data Analysis Secton 5 Sgnal Condtonng and Data Analyss 6/27/2017 Engneerng Measurements 5 1 Common Input Sgnals 6/27/2017 Engneerng Measurements 5 2 1 Analog vs. Dgtal Sgnals 6/27/2017 Engneerng Measurements 5 3 Current

More information

Investigation. Name: a About how long would the threaded rod need to be if the jack is to be stored with

Investigation. Name: a About how long would the threaded rod need to be if the jack is to be stored with Think Unit bout 6 This Lesson Situation 1 Investigation 1 Name: Think about the design and function of this automobile jack. Use the uto Jack custom tool to test ou ideas. a bout how long would the theaded

More information

Spectrum Sharing between Public Safety and Commercial Users in 4G-LTE

Spectrum Sharing between Public Safety and Commercial Users in 4G-LTE 1 Spectum Shaing between Public Safety and Commecial Uses in 4G-LTE Haya Shajaiah, Ahmed Abdel-Hadi and Chales Clancy Badley Depatment of Electical and Compute Engineeing Viginia Tech, Alington, VA, 22203,

More information

Optimal Design of Smart Mobile Terminal Antennas for Wireless Communication and Computing Systems

Optimal Design of Smart Mobile Terminal Antennas for Wireless Communication and Computing Systems Optimal Design of Smat Mobile Teminal Antennas fo Wieless Communication and Computing Systems Autho Lu, Junwei, Yang, Shiyou Published 2007 Confeence Title 2007 4th Intenational Symposium on Electomagnetic

More information

ISSN: [Reddy & Rao* et al., 5(12): December, 2016] Impact Factor: 4.116

ISSN: [Reddy & Rao* et al., 5(12): December, 2016] Impact Factor: 4.116 IJESRT INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY SIMULATION COMPARISONS OF INDUCTION MOTOR DRIVE WITH ESTIMATOR AND PLL V. Nasi Reddy *, S. Kishnajuna Rao*, S.Nagenda Kuma * Assistant

More information

RESOURCE allocation is an important research topic in

RESOURCE allocation is an important research topic in 36 IEEE TRANSACTIONS ON WIRELESS COMMNICATIONS, VOL. 6, NO., OCTOBER 27 tlty-based Resoce Allocaton n Weless Netwoks Wen-Hsng Ko and Wanjn Lao Abstact In ths pape, we stdy tlty-based maxmzaton fo esoce

More information

Multiagent Reinforcement Learning Dynamic Spectrum Access in Cognitive Radios

Multiagent Reinforcement Learning Dynamic Spectrum Access in Cognitive Radios Sensos & Tansduces 204 by IFSA Publishing, S L http://wwwsensospotalcom Multiagent Reinfocement Leaning Dynamic Spectum Access in Cognitive Radios Wu Chun, 2 Yin Mingyong, 2 Ma Shaoliang, Jiang Hong School

More information

Receive Antenna Selection for MIMO Systems in Correlated Channels

Receive Antenna Selection for MIMO Systems in Correlated Channels Receve Antenna Selecton fo IO Systems n oelated hannels n Da, Sana Sfa, and Khaled B. etaef ente fo Weless Infomaton Technology The Hong Kong Unvesty of Scence & Technology lea Wate Bay, HONG KONG eedaln@ust.hk,

More information

Overlapping community detection in signed networks

Overlapping community detection in signed networks Ovelapping community detection in signed netwoks Yi Chen 1 Xiao-long Wang 1, 2 Bo Yuan 1 1 (Depatment of Compute Science and Technology, Shenzhen Gaduate School, Habin Institute of Technology, Shenzhen

More information

IN recent years there has been extensive research on the

IN recent years there has been extensive research on the IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. X, NO. X, OCTOBER 2006 1 On the Teletaffc Capacty of Optcal CDMA Shaon Goldbeg, and Paul R. Pucnal, Fellow, IEEE Abstact The capacty of an optcal CDMA (OCDMA netwok

More information

Fwd: CPC CU-DBSPR ( W. Olympic Boulevard, Los Angeles)

Fwd: CPC CU-DBSPR ( W. Olympic Boulevard, Los Angeles) 3/8/209 Cty of Los Angeles Mal Fwd: CPC2063662CUDBSPR (370388 W Olympc Boulevad, Los Angeles) Fwd: CPC2063662CUDBSPR (370388 W Olympc Boulevad, Los Angeles) Anna Matnez To: Clek Councl

More information

Abstract. 1. Introduction. Brunel University 1 Anglia Ruskin University 2

Abstract. 1. Introduction. Brunel University 1 Anglia Ruskin University 2 A ctcal assessment fo RINE Data fom OS NET fo GPS Accuacy Imovement M. AL Nabhan 1, S. AL Mas 2,W. Balachandan 1,. Hunat 2 m.alnabhan@bunel.ac.u, s.almas@angla.ac.u, w.balachandan@bunel.ac.u, zad.hunat@angla.ac.u

More information

Analysis of the optimized low-nonlinearity lateral effect sensing detector

Analysis of the optimized low-nonlinearity lateral effect sensing detector Jounal of hysics: Confeence Seies Analysis of the optimized low-nonlineaity lateal effect sensing detecto To cite this aticle: Saeed Olyaee et al J. hys.: Conf. Se. 76 4 Related content - Neual netwok

More information

Maximum Power Control for Photovoltaic Power based on Fuzzy Takagi-Sugeno Model

Maximum Power Control for Photovoltaic Power based on Fuzzy Takagi-Sugeno Model Jounal of Electcal Engneeng www.jee.o Maxmum Powe Contol fo Photovoltac Powe based on Fuzzy akag-sugeno Model N. HARRABI * M. SOUISSI ** * Gaduate School of Engneeng of Sfax, unsa, nozha.haab@gmal.com

More information

Figure Geometry for Computing the Antenna Parameters.

Figure Geometry for Computing the Antenna Parameters. Spheical Coodinate Systems Definitions Figue 1.2.1 Geomety fo Computing the Antenna Paametes. Antenna Radiation Patten: The distibution of adiated enegy fom an antenna ove a suface of constant adius centeed

More information

GRADE 6 FLORIDA. Division WORKSHEETS

GRADE 6 FLORIDA. Division WORKSHEETS GRADE 6 FLORIDA Division WORKSHEETS Mental division stategies invese opeations As we know, multiplication and division ae invese opeations. 8 9 = 7 This means they do the evese of each othe: 7 9 = 8 We

More information

TECHNICAL REPORT: CVEL Maximum Radiated Emission Calculator: Power Bus EMI Algorithm. Chentian Zhu and Dr. Todd Hubing. Clemson University

TECHNICAL REPORT: CVEL Maximum Radiated Emission Calculator: Power Bus EMI Algorithm. Chentian Zhu and Dr. Todd Hubing. Clemson University TECHNICAL REPORT: CVEL-13-053 Maximum Radiated Emission Calculato: Powe Bus EMI Algoithm Chentian Zhu and D. Todd Hubing Clemson Univesity Octobe 12, 2013 Abstact The Maximum Radiated Electomagnetic Emissions

More information

Using On-Line Parameter Estimation to Improve Efficiency of IPM Machine Drives

Using On-Line Parameter Estimation to Improve Efficiency of IPM Machine Drives Usng On-Lne Paamete Estmaton to Impove Effcency of IPM Machne Dves Hyunbae Km and Robet D. Loenz Dept. of Mech. Eng. and Dept. of Elect. & Comp. Eng. Unvesty of Wsconsn Maon Unvesty Avenue Maon, WI 7 UA

More information

GPS Attitude Determination for a JPALS Testbed: Integer Initialization and Testing

GPS Attitude Determination for a JPALS Testbed: Integer Initialization and Testing GPS Atttude Detemnaton fo a JPALS estbed: Intege Intalzaton and estng Davd S. De Loenzo, Santago Alban, Jennfe Gaute, Pe Enge Depatment of Aeonautcs and Astonautcs, Stanfod Unvesty Duand Buldng, 496 Lomta

More information

Fault Detection and Classification using Kalman Filter and Hybrid Neuro-Fuzzy Systems

Fault Detection and Classification using Kalman Filter and Hybrid Neuro-Fuzzy Systems Intenatonal Jounal of Compute Applcatons (975 8887) Volume 45 No., May Fault Detecton and Classfcaton usng Kalman Flte and Hybd Neuo-Fuzzy Systems A. Khoukh Dept. Systems Engneeng KFUPM Dhahan, KSA, 36

More information

Localization Algorithm for a mobile robot using igs

Localization Algorithm for a mobile robot using igs Poceedings of the 17th Wold Congess The Intenational Fedeation of Automatic Contol Seoul, Koea, July 6-11, 008 Localization Algoithm fo a mobile obot using igs DaeGeun Seo *, JangMyung Lee** * Electical

More information

Priority based Dynamic Multiple Robot Path Planning

Priority based Dynamic Multiple Robot Path Planning 2nd Internatonal Conference on Autonomous obots and Agents Prorty based Dynamc Multple obot Path Plannng Abstract Taxong Zheng Department of Automaton Chongqng Unversty of Post and Telecommuncaton, Chna

More information

Antenna fundamentals: With answers to questions and problems (See also Chapter 9 in the textbook.)

Antenna fundamentals: With answers to questions and problems (See also Chapter 9 in the textbook.) adio Technology Metopolia/A. Koivumäki Antenna fundamentals: With answes to questions and poblems (See also Chapte 9 in the textbook.) 1. a) Make up a definition fo the tem "antenna". Answe: One definition:

More information

ONE-WAY RADAR EQUATION / RF PROPAGATION

ONE-WAY RADAR EQUATION / RF PROPAGATION ONE-WAY RADAR EQUATION / RF PROPAGATION The one-way (tansmitte to eceive) ada equation is deived in this section. This equation is most commonly used in RWR o ESM type of applications. The following is

More information

High Speed ADC Sampling Transients

High Speed ADC Sampling Transients Hgh Speed ADC Samplng Transents Doug Stuetzle Hgh speed analog to dgtal converters (ADCs) are, at the analog sgnal nterface, track and hold devces. As such, they nclude samplng capactors and samplng swtches.

More information

Minimizing Ringing and Crosstalk

Minimizing Ringing and Crosstalk Minimizing Ringing and Cosstalk By Glen Dash, Ampyx LLC, GlenDash at alum.mit.edu Copyight 1998, 26 Ampyx LLC When viewed on a schematic, a wie is just a wie. Howeve, when isetimes shink to a few nanoseconds

More information

UNCERTAINTY ESTIMATION OF SIZE-OF-SOURCE EFFECT MEASUREMENT FOR 650 NM RADIATION THERMOMETERS

UNCERTAINTY ESTIMATION OF SIZE-OF-SOURCE EFFECT MEASUREMENT FOR 650 NM RADIATION THERMOMETERS XIX IMEKO Wold Congess Fundamental and Applied Metology Septembe 6 11, 29, Lisbon, Potugal UNCERTAINTY ESTIMATION OF SIZE-OF-SOURCE EFFECT MEASUREMENT FOR 65 NM RADIATION THERMOMETERS Fumihio Sakuma, Laina

More information

WIRELESS SENSORS EMBEDDED IN CONCRETE

WIRELESS SENSORS EMBEDDED IN CONCRETE 7th Euopean Wokshop on Stuctual Health Monitoing July 8-11, 2014. La Cité, Nantes, Fance Moe Info at Open Access Database www.ndt.net/?id=17296 WIRELESS SENSORS EMBEDDED IN CONCRETE Amal Abbadi 1 Phd Student,

More information

Postprint.

Postprint. http://www.dva-potal.og Postpnt Ths s the accepted veson of a pape publshed n IEEE Tansactons on Vehcula Technology. Ths pape has been pee-evewed but does not nclude the fnal publshe poof-coectons o jounal

More information

Measurements of Rainfall Rates From Videos

Measurements of Rainfall Rates From Videos Measuements of Ranfall Rates Fom Vdeos Dong, R., Lao, J., L, B., Zhou, H., & Cookes, D. (07). Measuements of Ranfall Rates Fom Vdeos. In 0th Intenatonal Congess on Image and Sgnal Pocessng, BoMedcal Engneeng

More information

A Simple Technique to Handle Fields Singularities in the Finite Element Time Domain Method

A Simple Technique to Handle Fields Singularities in the Finite Element Time Domain Method 1 A Smple Technque to Handle Felds Sngulates n the Fnte Element Tme oman Method alkelly Schmdke, lson A. Atuz, J., Membe, IEEE Abstact The use of dvegence-confomng edge bass functons of fst ode s poposed

More information

Diagnosis method of radiated emission from battery management system for electric vehicle

Diagnosis method of radiated emission from battery management system for electric vehicle Available online at www.sciencediect.com ScienceDiect Enegy Pocedia 88 (2016 ) 662 667 CUE2015-Applied Enegy Symposium and Summit 2015: Low cabon cities and uban enegy systems Diagnosis method of adiated

More information

Capacity and Reaction Functions of Joint-Stock Firms

Capacity and Reaction Functions of Joint-Stock Firms Intenatonal Jounal of Management, Accountng and Economcs ISSN 2383-2126 (Onlne) Authos, All Rghts Reseed www.mae.com Capacty and Reacton Functons of Jont-Stock Fms Kazuho Ohnsh 1 Insttute fo Basc Economc

More information

Implementation of Expert System for Power System Voltage Stability Improvement

Implementation of Expert System for Power System Voltage Stability Improvement Intenatonal Jounal of Electcal Enegy, ol., No. 1, June 017 Ipleentaton of Expet Syste fo Powe Syste oltage Stablty Ipoveent Abaha Lo and. Yud Lpaptono Depatent of Electcal Engneeng, Insttut Tenolog Nasonal,

More information