Maintenance Scheduling and Production Control of Multiple-Machine Manufacturing Systems
|
|
- Charlene Alexander
- 5 years ago
- Views:
Transcription
1 Coputers & Industral Engneerng 48 (2005) do:0.06/.ce Mantenance Schedulng and Producton Control of Multple-Machne Manufacturng Systes A. Gharb a and J.-P. Kenné b a Autoated Producton Engneerng Departent, Unversty of Quebec, École de technologe supéreure, Producton Systes Desgn and Control Laboratory 00, Notre Dae Street West, Montreal (Quebec), Canada, H3C K3. b Mechancal Engneerng Departent, Unversty of Quebec, École de technologe supéreure, 00, Notre Dae Street West, Montreal (Quebec), Canada, H3C K3. Abstract Ths paper deals wth the producton and preventve antenance control proble for a ultple-achne anufacturng syste. The obectve of such a proble s to fnd the producton and preventve antenance rates for the achnes so as to nze the total cost of nventory/backlog, repar and preventve antenance. A two-level herarchcal control odel s presented, and the structure of the control polcy for both dentcal and non-dentcal anufacturng systes s descrbed usng paraeters, referred to here as nput factors. By cobnng analytcal forals wth sulaton-based statstcal tools such as experental desgn and response surface ethodology, an approxaton of the optal control polces and values of nput factors are deterned. The results obtaned extend those avalable n exstng lterature to cover non-dentcal achne anufacturng systes. A nuercal exaple and a senstvty analyss are presented n order to llustrate the robustness of the proposed approach. The extenson of the proposed producton and preventve antenance polces to cover large systes (ultple achnes, ultple products) s dscussed. Key words: Preventve Mantenance, Herarchcal Control, Flexble Manufacturng Systes, Sulaton, Experental Desgn, Response Surface Methodology (RSM).. Introducton The proble of controllng anufacturng systes wth unrelable achnes was forulated as a stochastc control proble by Older and Sur (980). Falure and repar processes were supposed to be descrbed usng hoogeneous Markov processes. The related optal control odel falls under the category of probles studed prevously by Rshel (975). Slar nvestgatons have resulted n the analytcal soluton of the one-achne one-product anufacturng syste control proble obtaned by Akella and Kuar (986). In the case of nonhoogeneous Markov processes nvolvng states and control-dependent transton rates, the control proble becoes ore coplex. In ths sphere, Boukas and Haure (990) consdered the Ths anuscrpt verson s ade avalable under the CC-BY-NC-ND 4.0 lcense
2 fact that the falure probabltes of a achne depend on ts age, and they added the possblty of perforng preventve antenance to the exstng odels. The related age-dependent set of dynac prograng equatons were solved nuercally for a gven anufacturng syste. However, wth the nuercal schee presented by Boukas and Haure (990), t reans dffcult to obtan a general structure for the optal control of a large class of anufacturng systes. A potental way of copng wth such a dffculty s to develop heurstcal ethods based on the reducton of the sze of the consdered control proble. Hence, dfferent approaches have been proposed n the exstng lterature wth a vew to dervng sple near-optal control polces for anufacturng systes. The concept of hedgng pont polcy, ntroduced by Kea and Gershwn (983), s one of the sple ways avalable for fndng suboptal control polces n the producton plannng and antenance schedulng of anufacturng systes. For further detals on ths concept, we refer the reader to the age-dependent hedgng pont concept presented by Boukas et al. (995) and by Kenne and Gharb (999). Because of the coputaton of threshold levels, the dervaton of suboptal polces based on ths concept sees to be dffcult for a large class of anufacturng systes. Another approach s to develop herarchcal control ethods based on the partcular structure of the syste. Ths can be done by usng the sngular perturbaton approach. Such an approach anly nvolves reducng the sze of the control proble accordng to the dscrepancy between the te scales of events nvolved. By replacng fast processes wth ther respectve ean values, one can construct a deternstc ltng proble, whch s coputatonally ore tractable. Detals on ths approach can be found n Kokotovc et al. (986), Lehoczky et al. (99), Seth and Zhang (994) and Soner (993). In ths paper, we wll frst defne the structure of the optal control polces, both for dentcal and for non-dentcal achne anufacturng systes. Based on such structures, we wll then extend the producton and antenance rates control odel presented n Kenne and Boukas (2003) n order to deterne the control polcy n a ore general case ncludng non-dentcal achne anufacturng systes. The resultng structure s descrbed through a set of paraeters we call nput factors. We resort to a cobnaton of analytcal and sulaton-based experental approaches to fnd an approxaton of the optal control polces for producton and preventve antenance by deternng the values of nput factors. In the proposed approach, the paraeterzed near-optal control polcy s used as an nput for the sulaton odel. For each entry consstng of a cobnaton of paraeters, the cost ncurred s obtaned. It s fro ths relatonshp that the best control factor values are deterned and a relatonshp between nput factors and such a cost s gven. The applcaton of such an approach s otvated by the works of Kenne and Gharb (999) and Gharb and Kenne (2000). We refer the reader to these works for a lterature revew on the applcatons of sulaton and statstcal ethods such as experental desgn and RSM n the sphere of anufacturng systes control. The reander of the paper s organzed as follows: In secton 2, the optal control proble s descrbed both for dentcal and non-dentcal achne systes. The proposed control approach s descrbed n secton 3. The logc of the sulaton odel s descrbed n secton 4. In secton 5, the experental desgn approach and response surface ethodology are outlned. A nuercal exaple and a senstvty analyss are also presented n secton 5. Concludng rearks are presented n secton 6. 2
3 2. Proble Stateent In ths secton, we present an explct forulaton of the stochastc optal control proble related to the producton control and preventve antenance schedulng of anufacturng systes wth non-dentcal achnes. Based on the large sze of the optalty condtons obtaned, we next present a sngular perturbaton for of the control odel for ultple dentcal achne anufacturng systes. The structure of the control polcy n such a stuaton s extended to defne that polcy n the case of non-dentcal achnes, for whch optalty condtons are dffcult to solve. The syste under study conssts of achnes producng n dfferent part types. The operatonal ode of achne can be descrbed by a stochastc process t. Such a achne s avalable when t s operatonal t and unavalable when t s under repar t 2 or under preventve antenance t 3. We then have t, 2,3. We can descrbe the anufacturng syste ode by the rando vector t t,, t wth values n. Let a a,, a and, antenance rates respectvely. The process t, be the vectors of achne ages and preventve s odelled by a contnuous te Markov chan defned by achne ages and control dependent transton rates atrx wth The transton rates processes t,,, M. Q a, a, a, 0,,,M, wth M card Q defned as follows: are derved fro the cobnaton of those of the dependent Our approach s used when the rate of change n the achne states s uch hgher than the rate at whch the cost s dscounted. In ths paper, we assue a constant deand rate wthout any loss of generalty. Two te scales are then consdered: the dscountng cost event and the achne state process te scales. When the dfference between the two te scales s very large, the te can be splt. Thus, the transton rates for the syste can be expressed as q, where q and the dscount rates are of the sae agntude. The sngular perturbaton paraeter s used here to express the herarchcal structure of the proposed approach. Wth, an equvalent deternstc proble can be derved fro the forulaton of the ntal stochastc proble. Both control probles (stochastc and deternstc) are descrbed by dynac prograng equatons (DPE) presented later n ths secton. In order to ncrease the syste capacty or the avalablty of the achnes, we assue that the transton rate fro the operatonal ode to the preventve antenance ode for each achne s a control varable called t,,,. The syste behavour s descrbed by a hybrd state coprsng both a dscrete and a contnuous coponent. The dscrete coponent conssts of the dscrete event stochastc process 3
4 t, whle the contnuous coponent conssts of contnuous varables x,, x n a,, x and a a correspondng to the nventory/backlog of products and the cuulatve ages of achnes. These state varables are descrbed by the followng dfferental equatons: x t u t d x 0 () x t ut a0 a a f (2), where x, a and d are gven ntal surplus or backlog, ntal achne ages and deand rates vectors respectvely. Let x~ x,a and ~ u u,. Let u,, u n producton rates. The set of feasble control polces, s gven by: u denote the vector of K and 0 n u t, t, u t 0, u t t ax n p p z, u p u p 0 (3), where p s the processng te of the part type p, and ax s the axu preventve antenance rate of each achne. Let G, x~, u~ be the nstantaneous cost defned as follows: x~,u ~ G, c x c x c, (4), B where c + and c - are costs ncurred per unt produced parts for postve nventory and backlog respectvely, x + = (ax (0, x ),, ax (0, x n )), x - = (ax (-x, 0),, ax (-x n, 0)) and c are gven constants used here for preventve antenance and repar actvty costs. Our obectve s to control the producton rate u and the preventve antenance rate or the control polcy u~ so as to nze the expected dscounted cost gven by: x~,u pt J, e G, x~,u dt x0 x, a0 a, 0 0 (5), subect to constrants gven by equatons () to (4). The value functon of such a proble s: x~, nf u~ J, B x~,u ~ (6) The value functon gven by equaton (6) s locally Lpschtz, convex and s the unque vscosty soluton of the followng HJB equatons (see Kenne (997) and Seth and Zhang (994)), x n. x, x G, x, u q, x u (7), 4
5 where, and v x, x~ u d f u s the gradent of v n ~ x. d p T T n up lo n 3 2 (8), h where T up, Tlo and h are the upper value, the lower value and the dscrete step of the state varable T. For each product, =,, n, the producton rate u, =,,, has three possble ~ values 0, d, U ; ths corresponds to 3 +n ponts for the set of u. For each achne, the ax preventve antenance rate s chosen between two values, 0 and ; ths gves 2 ponts for the achnes. The denson d, as n equaton (8), s very large for a ultple-part, ultpleachne anufacturng syste. Gven that there s no way to solve HJB equatons (7) analytcally, nuercal ethods based on d are usually used to characterse the optal control polcy. A sngular perturbaton approach s used to defne a herarchcal control schee based on ltng probabltes and deternstc optalty condtons, as presented here n the case of dentcal achnes. Such an approach s unusable n the case of non-dentcal achne anufacturng systes. 0 For dentcal achnes, let t, t the nuber of operatonal achnes at te t wth values n 0,,, control the producton rate u(t) and the preventve antenance rate or the control polcy u~ u so as to nze the expected dscounted cost gven by: J x~,u pt, e G, x~,u dt x0 x, a0 a, 0 0 The value functon of the control proble consdered s gven by: be a fnte-state stochastc process correspondng to. Our obectve s to (9), x~ nf J u~, x~,u ~ (0) Such a functon s also locally Lpschtz, convex and s the unque vscosty soluton of the followng HJB equatons:, x~ n u d x~,u ~ q, ~ x, x~ f u a, x~ G, x () u~ 0 The optal control polcy, u s the soluton of HJB equatons descrbed by (). Gven the large sze of () for ultple-achne, ultple-part type anufacturng systes, t s necessary here to proceed to the reducton of the syste sze through the ltng control proble. Such a proble s based on the statonary dstrbuton of the stochastc process, whch s coputed here n ters of the ean values of the achne ages and preventve antenance rates denoted by a M and respectvely. If a and ω, =,, are replaced n the odel by a M and, a constant transton rates atrx a, Q M s obtaned. The fnte state Markov chan assocated 5
6 to such a atrx s then hoogeneous, wth statonary or ltng probabltes gven by:, 2,, vq. 0 and v (2) Wth these probabltes, the convergence property of the ntal stochastc control proble to the ltng proble when 0 s establshed n Soner (993). Based on the herarchcal odel presented n Kenne (997) and Kenne and Boukas (2003) statng that ltng probabltes are obtaned for gven, the overall optzaton proble can be descrbed by the followng deternstc HJB equatons: x~ n 0 kax n uk n x~ k d x~ G x~, a u d x k. M (3) The structure of the stochastc control presented n Boukas and Haure (990) and Kenne (997) provde a achne age-dependent control polcy through the soluton of (3). The constructon of the correspondng stochastc control polcy structure provdes the followng producton and preventve antenance polces: Producton Control Polcy: The structure of the optal producton polcy could be gven by: t t t u ax s x u x~ d s x (4), 0 s x where Z s the threshold value of product and u ax u ax wth u ax descrbng the axu producton rate of product on achne for a ultple-achne, ultple-product anufacturng syste. Preventve Mantenance Polcy: The structure of the optal achne age-dependent preventve antenance polcy depends on the ean age value a M of the achne and a paraeter, and s defned as follows: f a t a ax M x~ (5) 0 otherwse Note that f 0, preventve antenance actons are perfored before the achne age reaches a M ; otherwse they are perfored after a M. The values and sgn of depend anly on the nstantaneous cost paraeters defned n equaton (4) for a gven anufacturng syste. For such a syste, a achne ean age value a M s deterned usng sulaton through off-lne experents. The followng achne age-dependent breakdown probablty dstrbuton s used to descrbe the dynacs of any achne. 6
7 a exp k a t (6), where k,,,, a s defned as the nuber of parts produced snce the last nterventon on the achne (repar or preventve antenance). A achne age ncreasng falure rate (IFR) s gven by (6), wth k. A well-known IFR, such as Webull, can be recovered by choosng sutable values for k. Note that the optal control polcy s well defned by paraeters and, whch we call here nput factors, for a gven a M. are gven constants and the achne age t In the next secton, we present a heurstcal control approach used here to estate the optal values of such factors. The proposed approach s based on a cobnaton of analytcal and sulaton odels, experental desgn and response surface ethodology. 3. Control Approach Results obtaned fro tradtonal ethods of producton and preventve antenance schedulng of ultple-achne anufacturng systes are not generally enough to provde a cofortable level of desred perforance. To prove these ethods, the descrptve capactes of conventonal sulaton odels are cobned wth analytcal odels, experental desgn and response surface ethodology. Ths approach has been successfully used n the cases of sngleachne and ultple-dentcal-achne anufacturng systes (see Kenne and Gharb (999) and Gharb and Kenne (2000)). A block dagra of the resultng control approach s depcted n Fg.. FMS Control Proble Dscrete event Sulaton Model Perforances evaluaton (costs of: nventory, backlogs and antenance) 5 2 Analytcal odel (Herarchcal approach & optalty condtons) Experental desgn (ANOVA & factors effects) 6 3 Nuercal ethods structure of the optal control polcy paraeterzed near optal control polces Response Surface Methodology Regresson analyss Optzng the estated cost functon 7 4 Control Factors Z and δ =, ; =, n Near-optal control polcy u(z * ) and ω(δ * ) 8 Fg.: Proposed control approach 7
8 The structure of the proposed control approach presented n Fgure, conssts of the followng sequental steps:. The Control proble stateent of the anufacturng syste, as n secton 2, conssts of the representaton of the producton plannng and antenance schedulng proble through a stochastc optal control odel based on control theory. Hence, the proble of the optal flow control for the anufacturng syste consdered s descrbed n ths frst step, whch contans a specfcaton of the obectve of the study. That obectve s to fnd the control varables (u, ω) called the producton and preventve antenance rates n order to prove the related output (.e., the ncurred cost). 2. The optalty condtons, descrbed by the HJB equatons, are obtaned fro the proble stateent of the prevous step. It s shown n ths step that the value functon, representng the ncurred cost, s the soluton of the HJB equatons, and the correspondng control polcy (producton and preventve antenance rates) s optal. When the rate of change n the achne states s uch hgher than the rate at whch the cost s dscounted, the te can be splt, and the sngular perturbaton approach used to develop optalty condtons for both stochastc and deternstc control probles. The control polcy of the stochastc proble s constructed fro that one of the correspondng deternstc proble. 3. The nuercal ethods are used n ths step to solve the HJB equatons of the ltng proble, gven that there s no way of solvng the equatons analytcally when t coes to real anufacturng systes (ultple-achne, ultple parts). 4. The control factors Z, =, n for producton plannng and δ, =, for preventve antenance schedulng, descrbe the nuercal control polcy obtaned, extended to the stochastc proble. 5. The sulaton odel uses the near optal control polcy defned n the prevous step as nput for conductng experents n order to evaluate the perforances of the anufacturng syste. Hence, for gven values of the control factors, the cost ncurred s obtaned fro the sulaton odel presented n secton The experental desgn approach defnes how the control factors can be vared n order to deterne the effects of the an factors and ther nteractons (.e., analyss of varance or ANOVA) on the cost through a nal set of sulaton experents. 7. The response surface ethodology s then used to obtan the relatonshp between the ncurred cost and sgnfcant an factors and nteractons gven n the prevous step. The obtaned regresson odel s then optzed n order to deterne best values of factors called here Z * for producton, and δ * for preventve antenance schedulng. 8. The near-optal control polcy (u(z * ), ω(δ * )) s then an proved age-dependent hedgng pont polcy to be appled to the anufacturng syste. The applcaton of the proposed control approach gves the producton and preventve antenance rates descrbed by equatons (4) and (5) respectvely for best values of factors Z * and δ *. 4. Sulaton Model A dscrete event sulaton odel that descrbes the dynacs of the syste ()-(2), s developed usng the Vsual SLAM language (Prtsker and O Relly, 999). Ths odel conssts of several networks, each of whch descrbes a specfc task n the syste (.e., deand generaton, control polcy, states of the achnes, nventory control..., etc.). The dagra of the proposed sulaton odel s shown n Fg. 2 wth the followng notaton block descrptons: 8
9 Fg. 2: Dagra of sulaton odel. The INITIALIZATION block ntalzes the varables (current surplus, producton rates, ncurred cost, etc) 2. The Deand Arrval block perfors the arrval of a deand for product at each d - unt of te. A verfcaton s then perfored on the nventory value of product, and the nventory or the backorder s updated. 3. The CONTROL POLICY segent block s defned n the prevous secton (see equaton (4) for the achne producton rates). The control polcy s defned by the output of the FLAG block. Ths block s used to peranently verfy the varaton n the stock level x (t). If x (t) > Z, then the producton rate s set to a zero value; otherwse the producton rate s set to the deand rate (x (t) = Z ) or to the axu producton rate (x (t) < Z ). 4. The PARTS PRODUCTION block perfors the producton of fnshed goods. 5. The update the nventory block perfors the varaton of the nventory level when a fnshed goods producton or a deand arrval occurs (.e. producton of fnshed goods ncreases nventory f there s no backorder or t satsfes the cuulatve deands, and hence decreases backorders). Off-lne runs of the sulaton odel, for a two-dfferent-achne, one-part type anufacturng syste, usng control polcy descrbed by (4) for Z =20 s llustrated n Fg. 3 for a product stock traectory. 9
10 STOCK TIME Fg. 3: A product stock traectory (Z =20). It s nterestng to note that: () the nventory level ncreases to Z and reans at ths value; () the nventory level decreases durng repar or preventve antenance tes; () the decreasng level of the nventory depends on the repar and the preventve antenance tes, whch are dfferent for each achne. 6. The falure-repar block perfors two functons: t defnes the te-to-falure of each achne, and repars broken ones. Usng the breakdown probablty dstrbuton defned by (6), n Fg. 4, we present the achne age traectores for two achnes (k = 0-4, k 2 = ), obtaned fro off-lne runs of the sulaton odel. For exaple, achne, fro ntal te to the frst up te (where the breakdown occurs), the achne age ncreases fro zero to 65. The achne age s then set to zero durng the repar te. When the achne s repared, t produces parts and ts age ncreases agan a M 20 Produced parts Fgure 4: Machne age traectory. a M2 Mach. Mach Te Fg. 4: Machne age traectores 0
11 7. The preventve antenance block defnes the te at whch we should send each achne out for preventve antenance. The ean value of the ages at whch achne breakdowns occur s the ean age of achne denoted here as (a M ). The preventve antenance on that achne should be done around that age (see equaton (5) for the achne preventve antenance rates). We deterne a M through off-lne sulaton runs. For the exaple llustrated n Fg. 4, these values are: a M = 28 and a M2 = 89. The falure-repar, the preventve antenance and the update the nventory blocks update the ncurred cost block. 8. The updates the ncurred cost block calculates n a real te the cost of nventory, backlogs and correctve and preventve antenance. 5. Experental Desgn and Response Surface Methodology To llustrate the approach presented n ths paper, we consder a two-achne, one-product anufacturng syste. The optal flow control for the anufacturng syste consdered s forulated as n secton 2, wth optalty condtons gven by HJB equatons () and (3) for the two levels of the proposed herarchcal approach. The optal control polcy s approxated by a heurstc control polcy defned n ters of desgn factors, as n equatons (4)-(5) for the achne producton and preventve antenance rates. The obectve of the proposed approach s to fnd the best paraeters of the control varables u(.) and (.), (.e., Z, δ and δ 2 ), n order to prove the related output (.e., the ncurred cost). The sulaton odel descrbes the dynacs of the syste usng the control polcy paraeterzed by the factors Z, δ and δ 2. These factors are consdered as the nput of such a odel, and the correspondng ncurred cost s defned as ts output. Fro the values of the nput factors and the correspondng cost values, the experental desgn approach deternes nput factors and/or ther nteractons that have sgnfcant effects on the output. Sgnfcant factors or nteractons are then consdered as nput of a response surface ethodology, n order to ft the relatonshp between the cost and the nput factors. The optal values of the nput factors, called Z *, δ * and δ 2 * are deterned fro ths estated relaton. The related achne age-dependent odfed hedgng pont polcy s then an proved hedgng pont polcy to be appled to the anufacturng syste. Due to the convexty of the value functon (see secton 2), we selected a 3 3 response surface desgn. The experental desgn s used to study and understand the effects that soe paraeters, naely Z, δ and δ 2, for the anufacturng syste, have on the perforance easure (.e., the cost). 5. Nuercal Exaple The followng are the nuercal values of the constants used prevously: d=2; U ax =.5; 2 U ax =.6; c + =; c - =0; c 2 = 60; c 2 2= 50; c 3 = 00; c 3 2=80; q 2 =0.05; q 2 2 = 0.022; q 2 =0.045; q 2 2 = 0.042; q 3 =0.8; q 2 3 =0.67; k = 0-4 ; k 2 = Note that, n the case of non-dentcal achnes, c 2, and c 3 are used as repar and antenance costs of achne, respectvely. In
12 addton, q 2 s the breakdown rate of achne, and q 2 and q 3 are respectvely the correctve and preventve antenance rates for achne. Based on off-lne sulaton runs, where the nu and the axu values of the factors were observed, the ndependent varable levels were chosen as n Table. Table : Level of ndependent varables Factor Low Level Center Hgh Level Descrpton Z Stock level δ Mantenance for M δ Mantenance for M 2 Three replcatons were conducted for each cobnaton of the factors, and therefore, 8 (3 3 x 3) sulaton runs were ade. To reduce the nuber of replcatons, we used a varance reducton technque called coon rando nubers (Law and Kelton (2000)). We conducted soe prelnary sulaton experents usng 3 replcatons, and notced that the varablty allows the effects to be dstngushed. 5.2 Result Analyss The statstcal analyss of the sulaton data conssts of the ultfactor analyss of varance (ANOVA). Ths s done usng a statstcal software applcaton, such as STATGRAPHICS, to provde the effects of the three ndependent varables (Z, δ and δ 2 ) on the dependant varable (Cost). The ANOVA table correspondng to the generated data s llustrated n Table 2. Fro Table 2, as all the p-values are less than 5%, we conclude that the an factors Z, δ and δ 2, ther quadratc effects, as well as ther nteractons are sgnfcant at the 0.05 level. The R-squared value of fro the ANOVA table, states that 94% of the total varablty s explaned by the odel (Montgoery (200)). Table 2: ANOVA Table Source Su of squares Df Mean Square F-Rato P-Value A:Z ,0000 B: d C:d AA AB BB BC CC Blocks Total error Total (corr.) R-squared = percent 2
13 The resdual analyss was used to verfy the adequacy of the odel. A resdual versus predcted value plot and noral probablty plot were used to test the hoogenety of the varances and the resdual noralty, respectvely. We conclude that the odel s satsfactory, and there s no need for the transforaton of response varables or for addtonal replcatons for the sulaton odel. The second order odel s then gven by: Cost = Z δ δ Z Z * δ Z * δ δ δ * δ δ 2 2 (7) d Z d Fg. 5: Estated response surface Z The near-optal control polcy to be appled to the anufacturng syste consdered s defned by the nu of the cost functon (7) located at Z * =22.99, δ * =2.36 and δ 2 * = as shown n Fg. 5. A cost value of 5.76 s obtaned wth such a control polcy. To crosscheck the valdty of the soluton, Z * =22.99, δ * =2.36 and δ 2 * = were used as nput to the sulaton odel. The cost value obtaned was 5.6, whch falls n the 95% confdence nterval ( X n n 2 S t = [50.39; 5.93] ), obtaned usng n=0 replcatons of n, n 2 the sulaton odel. Z *, δ * and δ * 2 defne the best-odfed hedgng pont polcy to be appled to the anufacturng syste consdered. Wth the aforeentoned optal values of the ndependent factors or nput paraeters, the cost s nzed and the correspondng control polcy s the best approxaton of the optal control one. The followng control polcy s to be appled to the anufacturng syste presented n ths exaple: If nventory level of product s greater than 23, then the producton rate s set to zero value; If nventory level of product s equal to 23, then produce at the deand rate; If nventory level of product s less than 23, then produce at the axu producton rate. If the age of achne s greater than 5.64 (a M - δ * = ), then send achne to preventve antenance f the nventory level of product s equal to 23. If the age of achne 2 s greater than (a M2 - δ * 2 = 89-(-5.59)), then send achne 2 to preventve antenance f the nventory level of product s equal to 23. 3
14 5.3 Senstvty Analyss. A set of nuercal exaples are consdered on the senstvty of the obtaned control polcy wth respect to nventory, backlog, correctve and preventve antenance costs (.e., c +, c -, c 2, c 3 ). The followng varatons, llustrated n Table 3, are explored and copared to a basc case. Decreasng c + : ths ust result n a tendency to ncrease the stock level n order to avod further backlog costs. Increasng c + : ths ust result n a tendency to decrease the stock level n order to avod further nventory costs. Decreasng c - : ths ust result n a tendency to decrease the stock level n order to avod further nventory costs. Increasng c - : ths ust result n a tendency to ncrease the stock level n order to avod further backlog costs. Decreasng c 2 (correctve antenance cost): ths ust result n a tendency to delay the preventve antenance perod (fewer preventve antenance actvtes). Increasng c 2 (correctve antenance cost): ths ust result n a tendency to advance the preventve antenance perod (ore preventve antenance actvtes). Decreasng c 3 (preventve antenance cost): ths ust result n a tendency to advance the preventve antenance perod (ore preventve antenance actvtes). Increasng c 3 (preventve antenance cost): ths ust result n a tendency to delay the preventve antenance perod (fewer preventve antenance actvtes). Table 3: Senstvty analyss table c + c - c 2 c 2 2 c 3 c 3 2 * Z * δ * δ 2 Cost * Reark Basc case Z * ncreases Z * decreases Z * decreases Z * ncreases δ * decreases δ * ncreases δ * 2 decreases δ * 2 ncreases δ * ncreases δ * decreases δ * 2 ncreases δ * 2 decreases Through the above analyss, t clearly appears that the results obtaned ake sense, and that the proposed approach s robust. In the second part of ths secton, we wll dscuss how to control ore coplex anufacturng systes. For an achne, n products anufacturng syste as n ths paper, we obtan an +n factors experental desgn (.e., one factor for each achne and product). For large values of and n, a ore approprate experental desgn ust be explored, snce the coplete 3 +n s very dffcult to pleent. In such a stuaton, a two-step desgn approach s recoended: 4
15 . Use of two-level fractonal factoral desgns (.e. 2 f-p ) as flter n order to elnate nonsgnfcant factors and/or nteractons (Montgoery, 200). 2. Use of experental desgn related to sgnfcant factors or nteractons. The Box-Benhken or Box-Wlson central coposte desgns are coonly used at ths level (Montgoery, 200). However, the Box-Wlson desgn s preferred because we reuse all the results of the experents perfored durng the screenng step. The proposed approach sgnfcantly reduces the nuber of sulaton runs, and should gves rse to near-optal control polces for ore coplex anufacturng systes. 6. Concluson In ths paper, we have extended the concept of hedgng pont polcy to the producton and preventve antenance control proble of a ultple, non-dentcal achne anufacturng syste. The proposed approach was based on the cobnaton of the herarchcal control odel, sulaton experents, experental desgn and RSM. Frst, we nvestgated a near-optal control polcy of a achne age dependent Markov process through the constructon of the stochastc control polcy fro one of the dependant deternstc odels. We then assocated to such a polcy paraeters called ndependent varables. A sulaton odel was developed to descrbe the dynac of the producton syste under the proposed odfed hedgng pont polcy. An experental desgn approach was then used to nvestgate the effects of specfc factors on the cost ncurred durng the producton horzon. The proposed approach cobnes the sulaton ethod wth the statstcal ethod to provde the estaton of the cost functon related to the control proble consdered. A response surface ethodology was used to perfor ths functon n ters of sgnfcant an factors and nteractons gven by the experental desgn approach. Fro the estaton of the cost functon, the best values of control paraeters were easly coputed. 7. References Akella R. and Kuar P. R. (986), Optal Control of Producton Rate n a Falure Prone Manufacturng Syste. IEEE Trans. On Autoatc Control, Vol. AC-3, No. 2, pp Boukas, E. K. and Haure A. (990) Manufacturng Flow Control and Preventve Mantenance: A Stochastc Control Approach. IEEE Transactons on Autoatc Control, Vol 33 No 9, Boukas E.K., Kenne J.P. and Zhu Q. (995) Age-dependent Hedgng Pont Polces n Manufacturng Systes. Aercan Autoatc Control Councl, Seattle, Washngton, June Gharb A. and Kenne J. P. (2000) Producton and Preventve Mantenance Rates Control for a Manufacturng Syste: An Experental Desgn Approach. Internatonal Journal of Producton Econocs, Vol 65 No 3, pp Kenne J.P. (997), Planfcaton de la Producton et de la Mantenance des Systèes de Producton: Approche Hérarchsée. Ph.D thess, Ecole Polytechnque de Montréal, Unversté de Montréal. Kenne, J. and P. Boukas, E. K., (2003) Herarchcal Control of Producton and Mantenance Rates n Manufacturng Systes. Journal of Qualty n Mantenance Engneerng, Vol. 9, No.. 5
16 Kenne J.P. and Gharb A. (999) Experental Desgn n Producton and Mantenance Control of a Sngle Machne, Sngle Product Manufacturng Syste. Internatonal Journal of Producton Research, 37(3), Kokotovc, P., Khall, H. P. and O'Relly J. (986) Sngular perturbaton ethods n control: analyss and desgn, Acadec Press. Lehoczky J., Seth S., Soner H.M. and Taksar M. (99) An asyptotc Analyss of Herarchcal Control of Manufacturng Systes Under Uncertanty. Matheatcs of operatons research, 6(3), LAW, A.M. and W.D. Kelton (2000) Sulaton Modelng and Analyss. 3 ed edton, Mc Graw- Hll. Montgoery, D. C. (200) Desgn and analyss of experents. 5 th edton, John Wley & Sons. Older G.J. and Sur R. (980) Te Optal Part-Routng n a Manufacturng Syste wth Falure Prone Machnes. Proc. 9 th Conf. Decs. Contr. Albuquerque, NM. Prtsker A. A. B and O'Relly J. J. and LaVal D. K., Sulaton wth Vsual SLAM and AweS. John Wley & Sons, 999. Rshel Dynac Prograng and Mnu Prncples for Systes wth Jup Markov Dsturbances. SIAM ournal on Control, 3, Seth S.P. and Zhang Q. (994), Herarchcal Control Decson Makng n Stochastc Manufacturng Systes. Brkhauser, Boston. Soner H.M. (993) Sngular Perturbaton n Manufacturng. SIAM J. Control and Optzaton, 3(),
System in Weibull Distribution
Internatonal Matheatcal Foru 4 9 no. 9 94-95 Relablty Equvalence Factors of a Seres-Parallel Syste n Webull Dstrbuton M. A. El-Dacese Matheatcs Departent Faculty of Scence Tanta Unversty Tanta Egypt eldacese@yahoo.co
More informationBAYESIAN CURVE FITTING USING PIECEWISE POLYNOMIALS. Dariusz Biskup
BAYESIAN CURVE FITTING USING PIECEWISE POLYNOMIALS Darusz Bskup 1. Introducton The paper presents a nonparaetrc procedure for estaton of an unknown functon f n the regresson odel y = f x + ε = N. (1) (
More informationExcess Error, Approximation Error, and Estimation Error
E0 370 Statstcal Learnng Theory Lecture 10 Sep 15, 011 Excess Error, Approxaton Error, and Estaton Error Lecturer: Shvan Agarwal Scrbe: Shvan Agarwal 1 Introducton So far, we have consdered the fnte saple
More informationWhat is LP? LP is an optimization technique that allocates limited resources among competing activities in the best possible manner.
(C) 998 Gerald B Sheblé, all rghts reserved Lnear Prograng Introducton Contents I. What s LP? II. LP Theor III. The Splex Method IV. Refneents to the Splex Method What s LP? LP s an optzaton technque that
More informationPROBABILITY AND STATISTICS Vol. III - Analysis of Variance and Analysis of Covariance - V. Nollau ANALYSIS OF VARIANCE AND ANALYSIS OF COVARIANCE
ANALYSIS OF VARIANCE AND ANALYSIS OF COVARIANCE V. Nollau Insttute of Matheatcal Stochastcs, Techncal Unversty of Dresden, Gerany Keywords: Analyss of varance, least squares ethod, odels wth fxed effects,
More informationOur focus will be on linear systems. A system is linear if it obeys the principle of superposition and homogenity, i.e.
SSTEM MODELLIN In order to solve a control syste proble, the descrptons of the syste and ts coponents ust be put nto a for sutable for analyss and evaluaton. The followng ethods can be used to odel physcal
More informationApplied Mathematics Letters
Appled Matheatcs Letters 2 (2) 46 5 Contents lsts avalable at ScenceDrect Appled Matheatcs Letters journal hoepage: wwwelseverco/locate/al Calculaton of coeffcents of a cardnal B-splne Gradr V Mlovanovć
More informationXII.3 The EM (Expectation-Maximization) Algorithm
XII.3 The EM (Expectaton-Maxzaton) Algorth Toshnor Munaata 3/7/06 The EM algorth s a technque to deal wth varous types of ncoplete data or hdden varables. It can be appled to a wde range of learnng probles
More informationLeast Squares Fitting of Data
Least Squares Fttng of Data Davd Eberly Geoetrc Tools, LLC http://www.geoetrctools.co/ Copyrght c 1998-2015. All Rghts Reserved. Created: July 15, 1999 Last Modfed: January 5, 2015 Contents 1 Lnear Fttng
More informationCOMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS
Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS
More informationThree Algorithms for Flexible Flow-shop Scheduling
Aercan Journal of Appled Scences 4 (): 887-895 2007 ISSN 546-9239 2007 Scence Publcatons Three Algorths for Flexble Flow-shop Schedulng Tzung-Pe Hong, 2 Pe-Yng Huang, 3 Gwoboa Horng and 3 Chan-Lon Wang
More informationLeast Squares Fitting of Data
Least Squares Fttng of Data Davd Eberly Geoetrc Tools, LLC http://www.geoetrctools.co/ Copyrght c 1998-2014. All Rghts Reserved. Created: July 15, 1999 Last Modfed: February 9, 2008 Contents 1 Lnear Fttng
More informationSeveral generation methods of multinomial distributed random number Tian Lei 1, a,linxihe 1,b,Zhigang Zhang 1,c
Internatonal Conference on Appled Scence and Engneerng Innovaton (ASEI 205) Several generaton ethods of ultnoal dstrbuted rando nuber Tan Le, a,lnhe,b,zhgang Zhang,c School of Matheatcs and Physcs, USTB,
More informationDesigning Fuzzy Time Series Model Using Generalized Wang s Method and Its application to Forecasting Interest Rate of Bank Indonesia Certificate
The Frst Internatonal Senar on Scence and Technology, Islac Unversty of Indonesa, 4-5 January 009. Desgnng Fuzzy Te Seres odel Usng Generalzed Wang s ethod and Its applcaton to Forecastng Interest Rate
More informationWinter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan
Wnter 2008 CS567 Stochastc Lnear/Integer Programmng Guest Lecturer: Xu, Huan Class 2: More Modelng Examples 1 Capacty Expanson Capacty expanson models optmal choces of the tmng and levels of nvestments
More informationJournal of Global Research in Computer Science A MARKOV CHAIN MODEL FOR ROUND ROBIN SCHEDULING IN OPERATING SYSTEM
Volue 2, No 6, June 20 Journal of Global Research n Coputer Scence RESEARCH AER Avalable Onlne at wwwjgrcsnfo A MARKOV CHAIN MODEL FOR ROUND ROBIN SCHEDULING IN OERATING SYSTEM Deepak Ssoda *, Dr Sohan
More informationReliability estimation in Pareto-I distribution based on progressively type II censored sample with binomial removals
Journal of Scentfc esearch Developent (): 08-3 05 Avalable onlne at wwwjsradorg ISSN 5-7569 05 JSAD elablty estaton n Pareto-I dstrbuton based on progressvely type II censored saple wth bnoal reovals Ilhan
More informationDenote the function derivatives f(x) in given points. x a b. Using relationships (1.2), polynomials (1.1) are written in the form
SET OF METHODS FO SOUTION THE AUHY POBEM FO STIFF SYSTEMS OF ODINAY DIFFEENTIA EUATIONS AF atypov and YuV Nulchev Insttute of Theoretcal and Appled Mechancs SB AS 639 Novosbrs ussa Introducton A constructon
More informationXiangwen Li. March 8th and March 13th, 2001
CS49I Approxaton Algorths The Vertex-Cover Proble Lecture Notes Xangwen L March 8th and March 3th, 00 Absolute Approxaton Gven an optzaton proble P, an algorth A s an approxaton algorth for P f, for an
More informationASYMMETRIC TRAFFIC ASSIGNMENT WITH FLOW RESPONSIVE SIGNAL CONTROL IN AN URBAN NETWORK
AYMMETRIC TRAFFIC AIGNMENT WITH FLOW REPONIVE IGNAL CONTROL IN AN URBAN NETWORK Ken'etsu UCHIDA *, e'ch KAGAYA **, Tohru HAGIWARA *** Dept. of Engneerng - Hoado Unversty * E-al: uchda@eng.houda.ac.p **
More informationDetermination of the Confidence Level of PSD Estimation with Given D.O.F. Based on WELCH Algorithm
Internatonal Conference on Inforaton Technology and Manageent Innovaton (ICITMI 05) Deternaton of the Confdence Level of PSD Estaton wth Gven D.O.F. Based on WELCH Algorth Xue-wang Zhu, *, S-jan Zhang
More informationQuantum Particle Motion in Physical Space
Adv. Studes Theor. Phys., Vol. 8, 014, no. 1, 7-34 HIKARI Ltd, www.-hkar.co http://dx.do.org/10.1988/astp.014.311136 Quantu Partcle Moton n Physcal Space A. Yu. Saarn Dept. of Physcs, Saara State Techncal
More informationAN ANALYSIS OF A FRACTAL KINETICS CURVE OF SAVAGEAU
AN ANALYI OF A FRACTAL KINETIC CURE OF AAGEAU by John Maloney and Jack Hedel Departent of Matheatcs Unversty of Nebraska at Oaha Oaha, Nebraska 688 Eal addresses: aloney@unoaha.edu, jhedel@unoaha.edu Runnng
More informationComputational and Statistical Learning theory Assignment 4
Coputatonal and Statstcal Learnng theory Assgnent 4 Due: March 2nd Eal solutons to : karthk at ttc dot edu Notatons/Defntons Recall the defnton of saple based Radeacher coplexty : [ ] R S F) := E ɛ {±}
More informationMultipoint Analysis for Sibling Pairs. Biostatistics 666 Lecture 18
Multpont Analyss for Sblng ars Bostatstcs 666 Lecture 8 revously Lnkage analyss wth pars of ndvduals Non-paraetrc BS Methods Maxu Lkelhood BD Based Method ossble Trangle Constrant AS Methods Covered So
More informationSolutions for Homework #9
Solutons for Hoewor #9 PROBEM. (P. 3 on page 379 n the note) Consder a sprng ounted rgd bar of total ass and length, to whch an addtonal ass s luped at the rghtost end. he syste has no dapng. Fnd the natural
More informationSolving Fuzzy Linear Programming Problem With Fuzzy Relational Equation Constraint
Intern. J. Fuzz Maeatcal Archve Vol., 0, -0 ISSN: 0 (P, 0 0 (onlne Publshed on 0 Septeber 0 www.researchasc.org Internatonal Journal of Solvng Fuzz Lnear Prograng Proble W Fuzz Relatonal Equaton Constrant
More informationOn Pfaff s solution of the Pfaff problem
Zur Pfaff scen Lösung des Pfaff scen Probles Mat. Ann. 7 (880) 53-530. On Pfaff s soluton of te Pfaff proble By A. MAYER n Lepzg Translated by D. H. Delpenc Te way tat Pfaff adopted for te ntegraton of
More informationThe Parity of the Number of Irreducible Factors for Some Pentanomials
The Party of the Nuber of Irreducble Factors for Soe Pentanoals Wolfra Koepf 1, Ryul K 1 Departent of Matheatcs Unversty of Kassel, Kassel, F. R. Gerany Faculty of Matheatcs and Mechancs K Il Sung Unversty,
More information4 Column generation (CG) 4.1 Basics of column generation. 4.2 Applying CG to the Cutting-Stock Problem. Basic Idea of column generation
4 Colun generaton (CG) here are a lot of probles n nteger prograng where even the proble defnton cannot be effcently bounded Specfcally, the nuber of coluns becoes very large herefore, these probles are
More informationPARAMETER ESTIMATION IN WEIBULL DISTRIBUTION ON PROGRESSIVELY TYPE- II CENSORED SAMPLE WITH BETA-BINOMIAL REMOVALS
Econoy & Busness ISSN 1314-7242, Volue 10, 2016 PARAMETER ESTIMATION IN WEIBULL DISTRIBUTION ON PROGRESSIVELY TYPE- II CENSORED SAMPLE WITH BETA-BINOMIAL REMOVALS Ilhan Usta, Hanef Gezer Departent of Statstcs,
More informationIntegral Transforms and Dual Integral Equations to Solve Heat Equation with Mixed Conditions
Int J Open Probles Copt Math, Vol 7, No 4, Deceber 214 ISSN 1998-6262; Copyrght ICSS Publcaton, 214 www-csrsorg Integral Transfors and Dual Integral Equatons to Solve Heat Equaton wth Mxed Condtons Naser
More informationComparison of Regression Lines
STATGRAPHICS Rev. 9/13/2013 Comparson of Regresson Lnes Summary... 1 Data Input... 3 Analyss Summary... 4 Plot of Ftted Model... 6 Condtonal Sums of Squares... 6 Analyss Optons... 7 Forecasts... 8 Confdence
More informationModule 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur
Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:
More informationOptimal Marketing Strategies for a Customer Data Intermediary. Technical Appendix
Optal Marketng Strateges for a Custoer Data Interedary Techncal Appendx oseph Pancras Unversty of Connectcut School of Busness Marketng Departent 00 Hllsde Road, Unt 04 Storrs, CT 0669-04 oseph.pancras@busness.uconn.edu
More informationChapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems
Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons
More information3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X
Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number
More informationLinear Regression Analysis: Terminology and Notation
ECON 35* -- Secton : Basc Concepts of Regresson Analyss (Page ) Lnear Regresson Analyss: Termnology and Notaton Consder the generc verson of the smple (two-varable) lnear regresson model. It s represented
More informationStudy of the possibility of eliminating the Gibbs paradox within the framework of classical thermodynamics *
tudy of the possblty of elnatng the Gbbs paradox wthn the fraework of classcal therodynacs * V. Ihnatovych Departent of Phlosophy, Natonal echncal Unversty of Ukrane Kyv Polytechnc Insttute, Kyv, Ukrane
More information,..., k N. , k 2. ,..., k i. The derivative with respect to temperature T is calculated by using the chain rule: & ( (5) dj j dt = "J j. k i.
Suppleentary Materal Dervaton of Eq. 1a. Assue j s a functon of the rate constants for the N coponent reactons: j j (k 1,,..., k,..., k N ( The dervatve wth respect to teperature T s calculated by usng
More informationChapter One Mixture of Ideal Gases
herodynacs II AA Chapter One Mxture of Ideal Gases. Coposton of a Gas Mxture: Mass and Mole Fractons o deterne the propertes of a xture, we need to now the coposton of the xture as well as the propertes
More informationChapter 13: Multiple Regression
Chapter 13: Multple Regresson 13.1 Developng the multple-regresson Model The general model can be descrbed as: It smplfes for two ndependent varables: The sample ft parameter b 0, b 1, and b are used to
More informationElastic Collisions. Definition: two point masses on which no external forces act collide without losing any energy.
Elastc Collsons Defnton: to pont asses on hch no external forces act collde thout losng any energy v Prerequstes: θ θ collsons n one denson conservaton of oentu and energy occurs frequently n everyday
More informationOn the number of regions in an m-dimensional space cut by n hyperplanes
6 On the nuber of regons n an -densonal space cut by n hyperplanes Chungwu Ho and Seth Zeran Abstract In ths note we provde a unfor approach for the nuber of bounded regons cut by n hyperplanes n general
More informationIntroducing Entropy Distributions
Graubner, Schdt & Proske: Proceedngs of the 6 th Internatonal Probablstc Workshop, Darstadt 8 Introducng Entropy Dstrbutons Noel van Erp & Peter van Gelder Structural Hydraulc Engneerng and Probablstc
More informationCOMP th April, 2007 Clement Pang
COMP 540 12 th Aprl, 2007 Cleent Pang Boostng Cobnng weak classers Fts an Addtve Model Is essentally Forward Stagewse Addtve Modelng wth Exponental Loss Loss Functons Classcaton: Msclasscaton, Exponental,
More informationVERIFICATION OF FE MODELS FOR MODEL UPDATING
VERIFICATION OF FE MODELS FOR MODEL UPDATING G. Chen and D. J. Ewns Dynacs Secton, Mechancal Engneerng Departent Iperal College of Scence, Technology and Medcne London SW7 AZ, Unted Kngdo Eal: g.chen@c.ac.uk
More informationChapter 11: Simple Linear Regression and Correlation
Chapter 11: Smple Lnear Regresson and Correlaton 11-1 Emprcal Models 11-2 Smple Lnear Regresson 11-3 Propertes of the Least Squares Estmators 11-4 Hypothess Test n Smple Lnear Regresson 11-4.1 Use of t-tests
More informationNUMERICAL DIFFERENTIATION
NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the
More informationChapter 12 Lyes KADEM [Thermodynamics II] 2007
Chapter 2 Lyes KDEM [Therodynacs II] 2007 Gas Mxtures In ths chapter we wll develop ethods for deternng therodynac propertes of a xture n order to apply the frst law to systes nvolvng xtures. Ths wll be
More informationChapter 25: Machining Centers, Machine Tool Structures and Machining Economics
Manufacturng Engneerng echnology n SI Unts, 6 th Edton Chapter 25: Machnng Centers, Machne ool Structures and Machnng Econocs Copyrght 200 Pearson Educaton South Asa Pte Ltd Chapter Outlne 2 Introducton
More informationHandling Overload (G. Buttazzo, Hard Real-Time Systems, Ch. 9) Causes for Overload
PS-663: Real-Te Systes Handlng Overloads Handlng Overload (G Buttazzo, Hard Real-Te Systes, h 9) auses for Overload Bad syste desgn eg poor estaton of worst-case executon tes Sultaneous arrval of unexpected
More informationSlobodan Lakić. Communicated by R. Van Keer
Serdca Math. J. 21 (1995), 335-344 AN ITERATIVE METHOD FOR THE MATRIX PRINCIPAL n-th ROOT Slobodan Lakć Councated by R. Van Keer In ths paper we gve an teratve ethod to copute the prncpal n-th root and
More informationhalftoning Journal of Electronic Imaging, vol. 11, no. 4, Oct Je-Ho Lee and Jan P. Allebach
olorant-based drect bnary search» halftonng Journal of Electronc Iagng, vol., no. 4, Oct. 22 Je-Ho Lee and Jan P. Allebach School of Electrcal Engneerng & oputer Scence Kyungpook Natonal Unversty Abstract
More informationKeywords Unreliable manufacturing systems, economic production quantity, feedback production planning, stochastic dynamic programming, simulation.
Accepted n Internatonal Journal of Producton Research, Vol. 51, No 1, 2013 Optmal producton control polcy n unrelable batch processng manufacturng systems wth transportaton delay B. Bouslah a, A. Gharb
More information2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification
E395 - Pattern Recognton Solutons to Introducton to Pattern Recognton, Chapter : Bayesan pattern classfcaton Preface Ths document s a soluton manual for selected exercses from Introducton to Pattern Recognton
More informationA Hybrid Variational Iteration Method for Blasius Equation
Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method
More informationGadjah Mada University, Indonesia. Yogyakarta State University, Indonesia Karangmalang Yogyakarta 55281
Reducng Fuzzy Relatons of Fuzzy Te Seres odel Usng QR Factorzaton ethod and Its Applcaton to Forecastng Interest Rate of Bank Indonesa Certfcate Agus aan Abad Subanar Wdodo 3 Sasubar Saleh 4 Ph.D Student
More informationStatistics for Economics & Business
Statstcs for Economcs & Busness Smple Lnear Regresson Learnng Objectves In ths chapter, you learn: How to use regresson analyss to predct the value of a dependent varable based on an ndependent varable
More informationStatistical analysis of Accelerated life testing under Weibull distribution based on fuzzy theory
Statstcal analyss of Accelerated lfe testng under Webull dstrbuton based on fuzzy theory Han Xu, Scence & Technology on Relablty & Envronental Engneerng Laboratory, School of Relablty and Syste Engneerng,
More information1. Statement of the problem
Volue 14, 010 15 ON THE ITERATIVE SOUTION OF A SYSTEM OF DISCRETE TIMOSHENKO EQUATIONS Peradze J. and Tsklaur Z. I. Javakhshvl Tbls State Uversty,, Uversty St., Tbls 0186, Georga Georgan Techcal Uversty,
More informationNumerical Heat and Mass Transfer
Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and
More informationGlobal Sensitivity. Tuesday 20 th February, 2018
Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values
More informationAn Optimal Bound for Sum of Square Roots of Special Type of Integers
The Sxth Internatonal Syposu on Operatons Research and Its Applcatons ISORA 06 Xnang, Chna, August 8 12, 2006 Copyrght 2006 ORSC & APORC pp. 206 211 An Optal Bound for Su of Square Roots of Specal Type
More informationRecap: the SVM problem
Machne Learnng 0-70/5-78 78 Fall 0 Advanced topcs n Ma-Margn Margn Learnng Erc Xng Lecture 0 Noveber 0 Erc Xng @ CMU 006-00 Recap: the SVM proble We solve the follong constraned opt proble: a s.t. J 0
More informationx i1 =1 for all i (the constant ).
Chapter 5 The Multple Regresson Model Consder an economc model where the dependent varable s a functon of K explanatory varables. The economc model has the form: y = f ( x,x,..., ) xk Approxmate ths by
More informationITERATIVE ESTIMATION PROCEDURE FOR GEOSTATISTICAL REGRESSION AND GEOSTATISTICAL KRIGING
ESE 5 ITERATIVE ESTIMATION PROCEDURE FOR GEOSTATISTICAL REGRESSION AND GEOSTATISTICAL KRIGING Gven a geostatstcal regresson odel: k Y () s x () s () s x () s () s, s R wth () unknown () E[ ( s)], s R ()
More informationPreference and Demand Examples
Dvson of the Huantes and Socal Scences Preference and Deand Exaples KC Border October, 2002 Revsed Noveber 206 These notes show how to use the Lagrange Karush Kuhn Tucker ultpler theores to solve the proble
More informationLinear Approximation with Regularization and Moving Least Squares
Lnear Approxmaton wth Regularzaton and Movng Least Squares Igor Grešovn May 007 Revson 4.6 (Revson : March 004). 5 4 3 0.5 3 3.5 4 Contents: Lnear Fttng...4. Weghted Least Squares n Functon Approxmaton...
More informationSTAT 3008 Applied Regression Analysis
STAT 3008 Appled Regresson Analyss Tutoral : Smple Lnear Regresson LAI Chun He Department of Statstcs, The Chnese Unversty of Hong Kong 1 Model Assumpton To quantfy the relatonshp between two factors,
More information1 Review From Last Time
COS 5: Foundatons of Machne Learnng Rob Schapre Lecture #8 Scrbe: Monrul I Sharf Aprl 0, 2003 Revew Fro Last Te Last te, we were talkng about how to odel dstrbutons, and we had ths setup: Gven - exaples
More informationDescription of the Force Method Procedure. Indeterminate Analysis Force Method 1. Force Method con t. Force Method con t
Indeternate Analyss Force Method The force (flexblty) ethod expresses the relatonshps between dsplaceents and forces that exst n a structure. Prary objectve of the force ethod s to deterne the chosen set
More informationANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U)
Econ 413 Exam 13 H ANSWERS Settet er nndelt 9 deloppgaver, A,B,C, som alle anbefales å telle lkt for å gøre det ltt lettere å stå. Svar er gtt . Unfortunately, there s a prntng error n the hnt of
More informationNegative Binomial Regression
STATGRAPHICS Rev. 9/16/2013 Negatve Bnomal Regresson Summary... 1 Data Input... 3 Statstcal Model... 3 Analyss Summary... 4 Analyss Optons... 7 Plot of Ftted Model... 8 Observed Versus Predcted... 10 Predctons...
More informationScattering by a perfectly conducting infinite cylinder
Scatterng by a perfectly conductng nfnte cylnder Reeber that ths s the full soluton everywhere. We are actually nterested n the scatterng n the far feld lt. We agan use the asyptotc relatonshp exp exp
More informationFinite Vector Space Representations Ross Bannister Data Assimilation Research Centre, Reading, UK Last updated: 2nd August 2003
Fnte Vector Space epresentatons oss Bannster Data Asslaton esearch Centre, eadng, UK ast updated: 2nd August 2003 Contents What s a lnear vector space?......... 1 About ths docuent............ 2 1. Orthogonal
More informationStructure and Drive Paul A. Jensen Copyright July 20, 2003
Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.
More information1.3 Hence, calculate a formula for the force required to break the bond (i.e. the maximum value of F)
EN40: Dynacs and Vbratons Hoework 4: Work, Energy and Lnear Moentu Due Frday March 6 th School of Engneerng Brown Unversty 1. The Rydberg potental s a sple odel of atoc nteractons. It specfes the potental
More informationThe optimal delay of the second test is therefore approximately 210 hours earlier than =2.
THE IEC 61508 FORMULAS 223 The optmal delay of the second test s therefore approxmately 210 hours earler than =2. 8.4 The IEC 61508 Formulas IEC 61508-6 provdes approxmaton formulas for the PF for smple
More informationCHAPTER 14 GENERAL PERTURBATION THEORY
CHAPTER 4 GENERAL PERTURBATION THEORY 4 Introducton A partcle n orbt around a pont mass or a sphercally symmetrc mass dstrbuton s movng n a gravtatonal potental of the form GM / r In ths potental t moves
More informationCHAPTER 6 CONSTRAINED OPTIMIZATION 1: K-T CONDITIONS
Chapter 6: Constraned Optzaton CHAPER 6 CONSRAINED OPIMIZAION : K- CONDIIONS Introducton We now begn our dscusson of gradent-based constraned optzaton. Recall that n Chapter 3 we looked at gradent-based
More informationThe Impact of the Earth s Movement through the Space on Measuring the Velocity of Light
Journal of Appled Matheatcs and Physcs, 6, 4, 68-78 Publshed Onlne June 6 n ScRes http://wwwscrporg/journal/jap http://dxdoorg/436/jap646 The Ipact of the Earth s Moeent through the Space on Measurng the
More informationAn Accurate Measure for Multilayer Perceptron Tolerance to Weight Deviations
Neural Processng Letters 10: 121 130, 1999. 1999 Kluwer Acadec Publshers. Prnted n the Netherlands. 121 An Accurate Measure for Multlayer Perceptron Tolerance to Weght Devatons JOSE L. BERNIER, J. ORTEGA,
More informationLecture Notes on Linear Regression
Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume
More information1 Definition of Rademacher Complexity
COS 511: Theoretcal Machne Learnng Lecturer: Rob Schapre Lecture #9 Scrbe: Josh Chen March 5, 2013 We ve spent the past few classes provng bounds on the generalzaton error of PAClearnng algorths for the
More informationLECTURE :FACTOR ANALYSIS
LCUR :FACOR ANALYSIS Rta Osadchy Based on Lecture Notes by A. Ng Motvaton Dstrbuton coes fro MoG Have suffcent aount of data: >>n denson Use M to ft Mture of Gaussans nu. of tranng ponts If
More informationPsychology 282 Lecture #24 Outline Regression Diagnostics: Outliers
Psychology 282 Lecture #24 Outlne Regresson Dagnostcs: Outlers In an earler lecture we studed the statstcal assumptons underlyng the regresson model, ncludng the followng ponts: Formal statement of assumptons.
More informationON THE NUMBER OF PRIMITIVE PYTHAGOREAN QUINTUPLES
Journal of Algebra, Nuber Theory: Advances and Applcatons Volue 3, Nuber, 05, Pages 3-8 ON THE NUMBER OF PRIMITIVE PYTHAGOREAN QUINTUPLES Feldstrasse 45 CH-8004, Zürch Swtzerland e-al: whurlann@bluewn.ch
More informationSimultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals
Smultaneous Optmzaton of Berth Allocaton, Quay Crane Assgnment and Quay Crane Schedulng Problems n Contaner Termnals Necat Aras, Yavuz Türkoğulları, Z. Caner Taşkın, Kuban Altınel Abstract In ths work,
More informationMinimization of l 2 -Norm of the KSOR Operator
ournal of Matheatcs and Statstcs 8 (): 6-70, 0 ISSN 59-36 0 Scence Publcatons do:0.38/jssp.0.6.70 Publshed Onlne 8 () 0 (http://www.thescpub.co/jss.toc) Mnzaton of l -Nor of the KSOR Operator Youssef,
More informationABSTRACT
RELIABILITY AND SENSITIVITY ANALYSIS OF THE K-OUT-OF-N:G WARM STANDBY PARALLEL REPAIRABLE SYSTEM WITH REPLACEMENT AT COMMON-CAUSE FAILURE USING MARKOV MODEL M. A. El-Damcese 1 and N. H. El-Sodany 2 1 Mathematcs
More informationRevision: December 13, E Main Suite D Pullman, WA (509) Voice and Fax
.9.1: AC power analyss Reson: Deceber 13, 010 15 E Man Sute D Pullan, WA 99163 (509 334 6306 Voce and Fax Oerew n chapter.9.0, we ntroduced soe basc quanttes relate to delery of power usng snusodal sgnals.
More informationReal-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling
Real-Tme Systems Multprocessor schedulng Specfcaton Implementaton Verfcaton Multprocessor schedulng -- -- Global schedulng How are tasks assgned to processors? Statc assgnment The processor(s) used for
More informationModified parallel multisplitting iterative methods for non-hermitian positive definite systems
Adv Coput ath DOI 0.007/s0444-0-9262-8 odfed parallel ultsplttng teratve ethods for non-hertan postve defnte systes Chuan-Long Wang Guo-Yan eng Xue-Rong Yong Receved: Septeber 20 / Accepted: 4 Noveber
More informationCHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE
CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng
More informationA NOTE ON A PERIODIC REVIEW INVENTORY MODEL WITH UNCERTAIN DEMAND IN A RANDOM ENVIRONMENT. Hirotaka Matsumoto and Yoshio Tabata
Scentae Mathematcae Japoncae Onlne, Vol. 9, (23), 419 429 419 A NOTE ON A PERIODIC REVIEW INVENTORY MODEL WITH UNCERTAIN DEMAND IN A RANDOM ENVIRONMENT Hrotaka Matsumoto and Yosho Tabata Receved September
More informationLecture 12: Discrete Laplacian
Lecture 12: Dscrete Laplacan Scrbe: Tanye Lu Our goal s to come up wth a dscrete verson of Laplacan operator for trangulated surfaces, so that we can use t n practce to solve related problems We are mostly
More informationINPUT-OUTPUT PAIRING OF MULTIVARIABLE PREDICTIVE CONTROL
INPUT-OUTPUT PAIRING OF MULTIVARIABLE PREDICTIVE CONTROL Lng-Cong Chen #, Pu Yuan*, Gu-L Zhang* *Unversty of Petroleu, P.O. Box 902 Beng 00083, Chna # GAIN Tech Co., P.O. Box 902ext.79, Beng 00083, Chna
More informationFermi-Dirac statistics
UCC/Physcs/MK/EM/October 8, 205 Fer-Drac statstcs Fer-Drac dstrbuton Matter partcles that are eleentary ostly have a type of angular oentu called spn. hese partcles are known to have a agnetc oent whch
More informationTemperature. Chapter Heat Engine
Chapter 3 Temperature In prevous chapters of these notes we ntroduced the Prncple of Maxmum ntropy as a technque for estmatng probablty dstrbutons consstent wth constrants. In Chapter 9 we dscussed the
More information