OptQuest_Optimization

7/30/2019 OptQuest_Optimization

1/28

Opt imizat ion with OptQuestOpt imizat ion with OptQuest


2/28


3/28

Met aheuri st ics is a f ami ly of opt im izat ion appr oachest hat includes scat t er search, genet ic algori t hms,sim ulat ed annealing, Tabu search, and t heir hybrids.

Tabu search uses search hi st ory and m em ory

m anagem ent t o guide t he problem -solving process. In i t ssim plest f orm , m emory prohibi t s t he search f romreinvest igat ing solut ions t hat have already beenevaluat ed. In Opt Quest , m em ory f unct ions encourage

search diversif icat ion and int ensif icat ion. These mem orycom ponent s divert t he search f rom local l y opt im alsolut ions t o f i nd a globall y opt im al solut ions.

Scat t er Search is populat ion-based m et aheuri st ic t hatoperat es on a col l ect ion of ref erence point s w it h t hegoal of f inding high-qual i t y solut ions t o an opt im izat ionproblem.

Metaheurist ics is a f amil y of opt im izat ion approaches

t hat includes scat t er search, genet ic algori t hms,sim ulat ed annealing, Tabu search, and t heir hybri ds.

Tabu search uses search history and memory

m anagement t o guide t he problem-solving process. In i t ssim plest f orm , m emory prohibi t s t he search f romreinvest igat ing solut ions t hat have alr eady beenevaluat ed. In Opt Quest , m emory f unct ions encourage

search diversif icat ion and int ensif icat ion. These mem orycom ponent s divert t he search f rom local l y opt im alsolut ions t o f ind a globall y opt im al solut ions.

Scatter Search is populat ion-based metaheurist ic thatoperat es on a col l ect ion of ref erence point s w it h t hegoal of f inding high-qual i t y solut ions t o an opt im izat ionproblem.

MetaheuristicsMetaheuristics


4/28

The f i r st part of designing an opt im izat ion is t o def i ne t he decisionvar i ables f or t he model.

The main decision variables f or an opt im izat ion can usuall y be chosen by

rest at ing t he problem you w ant t o solve. For exam ple, one problem m aybe: how m any m achines do we need in t his area t o get t he bestthroughput?

This st at ement def ines t he decision var i ables f or t he model, t he maxim umcont ent value of t he processor, and t he t hroughput value of t he model.

Not e t hat t hese t w o var iables have di f f erent ro les. The maxim um cont entis a value we w ant t o change and exper iment w i t h, w hi le t he t hroughput isour f eedback, ref lect ing t he result s of our changes.

Exam ples of decision var iable could be; number of f ork l i f t s requir ed, t hespeed of t he conveyors, bat ch size or r acks capacit y.

The f i r st part of designing an opt im izat ion is t o def ine t he decisionvariables f or t he model .

The main deci sion variables f or an opt im izat ion can usuall y be chosen by

rest at ing t he problem you want t o solve. For exam ple, one problem m aybe: how m any machines do w e need i n t his area t o get t he bestthroughput?

This st at ement def ines t he decision var i ables f or t he model, t he maxim umcont ent value of t he processor, and t he t hroughput value of t he model.

Not e t hat t hese t w o var i ables have di f f erent ro les. The maximum cont entis a value we w ant t o change and exper im ent w i t h, w hi le t he t hroughput is

our f eedback, ref lect ing t he result s of our changes.

Exam ples of decision var i able could be; number of f ork l i f t s requir ed, t hespeed of t he conveyors, bat ch size or r acks capacit y.

Decision VariablesDecision Variables


5/28

The Opt Quest Engine m anipul at es decision var iabl es in search of t heiropt im al values. There are seven t ypes of decision variables:

User-cont rol led: t hey are used as out put vari ables t o get f eedback f orhow w el l d i f f erent scenarios do.

Int eger: a discret e vari able w it h int eger bounds and a st ep size of 1.Discret e: begins at a low er bound and i ncrem ent s by a st ep size up t oan upper bound.

Cont inuous: m ay t ake on any value bet w een a user-specif ied l ow erbound and upper bound.

Binary: a discret e variable wi t h a value of 0 or 1.

The Opt Quest Engine m anipul at es decision vari ables in search of t heiropt im al values. There are seven t ypes of decision vari ables:

User-control led: t hey are used as out put variables t o get f eedback f orhow w el l d i f f erent scenari os do.

Integer: a discret e vari able w it h int eger bounds and a st ep size of 1.

Discrete: begins at a low er bound and i ncrem ent s by a st ep size up t oan upper bound.

Continuous: m ay t ake on any value bet w een a user-specif ied l ow erbound and upper bound.

Binary: a discret e var i able w it h a value of 0 or 1.

Types of Decision VariablesTypes of Decision Variables


6/28

Perm ut at ion: used t o solve sequencing problem s. For examp le, youcould use t his var i able t o det erm ine t he order in w hich paint shouldbe mi xed t o minim ize cleanup t im e bet w een color changes. The value

of a perm ut at ion var i able r epresent s t he order w it h in t he sequence.

Design: Used f or decisions w here value of t he variable r epresent s analt ernat ive, and not a quant it y. Design vari ables are usef ul i n

opt im izat ion pr oblem s w here t he decision vari ables consist ofchoosing t he best al t ernat ive f rom a cat alog, and a larger num ber m aynot im ply t he comm it m ent of m ore resources. Therefore, choice #10m ay not be a more cost ly or bet t er choice t han choice #1. These

vari ables are def ined by a low er bound, an upper bound, and a st epsize t hat cont rols t he num ber of choices avai lable w it h in t he specif iedrange.

Permutat ion: used t o solve sequencing problem s. For exam ple, youcould use t his var i able t o det erm ine t he order in w hich paint shouldbe mi xed t o minim ize cleanup t im e bet w een color changes. The value

of a perm ut at ion var iable represent s t he order w it h in t he sequence.

Design: Used f or deci sions w here value of t he vari able r epr esent s analt ernat ive, and not a quant it y. Design vari ables are usef ul i n

opt im izat ion problem s w here t he decision vari ables consist ofchoosing t he best a l t ernat ive f rom a cat alog, and a larger num ber m aynot im ply t he comm it ment of more r esources. Therefore, choice #10m ay not be a more cost ly or bet t er choice t han choice #1. These

vari ables are def ined by a low er bound, an upper bound, and a st epsize t hat cont rols t he num ber of choices avai l able w it h in t he specif iedrange.

Types of Decision VariablesTypes of Decision Variables


7/28

To add a decision var iable, c l ick on t he Add but t on in t he Vari ablespanel. This adds a new vari able t o t he vari ables t able. Select t his vari ableby select ing any cel l on t he row of t he new var i able, t hen cl ick t he Modi f y but t on. This br ings up a w indow t o edi t t h is new var iable.

To add a decision variable, c l ick on t he Add but t on in t he Var iablespanel. This adds a new vari able t o t he vari ables t able. Select t his vari ableby select ing any cel l on t he row of t he new var iable, t hen cl ick t he Modi fy but t on. This br ings up a w indow t o edi t t h is new var iable.


Each deci sion vari able has an associat ed name,w hich w il l be used by Opt Quest . Each variablealso has an associat ed t ype, l ike cont inuous,

int eger, or user-cont rol l ed. User-cont rol l edvariables are t he "f eedback" variables. Theyare not changed by t he Opt Questexperim ent at ion, but are used as out put

var i ables t o get f eedback for how w el ld i f f erent scenarios do. Al l ot her var i able t ypesw i l l be changed and exper iment ed w i t h dur ingt he opt imi zat ion.

Each decision var iable has an associat ed name,w hich w il l be used by Opt Quest . Each vari ablealso has an associat ed t ype, l ike cont inuous,

int eger, or user-cont rol l ed. User-cont rol l edvariables are t he "f eedback" variables. Theyare not changed by t he Opt Questexperi m ent at ion, but are used as out put

var i ables t o get f eedback for how w el ld i f f erent scenari os do. Al l ot her var i able t ypesw i l l be changed and exper iment ed w i t h dur ingt he opt imi zat ion.


8/28

Once you have speci f ied t hevar i able name and t ype, c l ick ont he Brow se but t on t o associat e

t his var i able w it h a node in t hemodel .

This w il l open a Tree Brow sew indow t o select t he node t hat

holds t h is m axim um cont entvalue.

You must select a node t hat has

number dat a on i t , o r t heopt im izat ion w i l l not w orkproperly.

Once you have speci f ied t hevar i able nam e and t ype, c l ick ont he Brow se but t on to associatet his var i able w it h a node in t hemodel .

This w il l open a Tree Brow sewindow t o select t he node t hat

holds t h is m axim um cont entvalue.

You must select a node t hat has

num ber dat a on i t , o r t heopt im izat ion w i l l not workproper ly .



9/28

Once you have def ined your decision vari ables, you w il l w ant t o defi neconst ra int s f or t he opt im izat ion. Dur ing t he opt im izat ion, t he opt im izerw il l t r y several scenarios on t he decision variables.

Const raint s are used t o nul l i f y cert ain scenari os i f t hese const raint s are notproperl y met , so t hat t he opt im izer doesn't choose an inval id scenari o ast he op t imum.

Each const raint has an exp ression, such as :

Product ion > 1000

IdlePercentage


10/28

The obj ect ive f unct ion is an expression t hat you w ant t o maxim ize orm inim ize. This could be a sim ple expression l ike "Throughput " i f you have adecision variable cal l ed Throughput .

It could also be a revenue vs. cost est im at ion. For exampl e, i f each

product produced yields $5. 00, and t he cost f or each m achine (w eight edby t he lengt h of t he sim ulat ion run) is $50, t hen t he obj ect ive funct ioncould be "(Throughput *5. 00) - (MaxNrof Processors*50. 00)".

Some ot her examp les are:

Maxim ize Prof i t s

Minim ize Picking Tim es

Maximize Production and Minimize WIP

The obj ect ive f unct ion is an expression t hat you w ant t o maxim ize orm inim ize. This could be a sim ple expr ession l ike "Throughput " i f you have adecision variable cal led Throughput .

It could also be a revenue vs. cost est im at ion. For exampl e, i f each

product produced yields $5.00, and t he cost f or each m achine (w eight edby t he lengt h of t he sim ulat ion run) is $50, t hen t he obj ect ive funct ioncould be "(Throughput *5. 00) - (MaxNrof Processors*50. 00)".

Some ot her examp les are:

Maxim ize Prof i t s

Minim ize Picki ng Tim es

Maximize Production and Minimize WIP

Object ive Funct ionObject ive Funct ion


11/28

Maxim um Tim e f or Opt im izat ion This is t hemaximum real t ime t hat t he op t im izer w i l lspend in i t s opt im izat ion.

Aut oSt op If t his box is checked, t heopt im izat ion st ops w hen t he value of t heobj ect ive funct ion st ops im proving. The

Flexsim current set t ing is t o st op w hen t heobj ect ive funct ion value of t he best so lut ionf ound does not vary by at least 0.0001 af t er100 i t erat ions.

Maximum Time f or Opt im izat ion This is t hemaximum real t ime t hat t he op t im izer w i l lspend in i t s opt im izat ion.

AutoStop If t his box is checked, t heopt im izat ion st ops w hen t he value of t heobj ect ive funct ion st ops im proving. TheFlexsim current set t ing is t o st op w hen t heobj ect ive funct ion value of t he best so lut ionf ound does not vary by at l east 0.0001 af t er100 i t erat ions.

Stop ConditionsStop Conditions


12/28

Maxim um Scenari os t his is t he m axim um number of di f f erent scenari ost hat t he opt im izer w i l l t ry . A scenar io is one conf i gurat ion in t heopt im izer 's search.

Curr ent Scenari o t his is t he curr ent scenari o number being t est ed.

Curr ent Solut ion t h is is t he value of t he obj ect ive f unct ion f or t hecurr ent scenari o.

Best Solut ion t h is is t he value of t he obj ect f unct ion f or t he bestscenar io so f ar.

Sim ulat ion Tim e per Scenari o/ Real Tim e per Scenari o t his is t he

m aximum sim ulat ion t im e t hat t he opt im izer w i l l spend for eachscenario. The optimizer stops a scenario as soon as this is met.

Maximum Scenarios t his is t he m axim um number of d i f f erent scenari ost hat t he opt im izer w i l l t ry . A scenar io is one conf i gurat ion in t heopt im izer 's search.

Current Scenario t his is t he curr ent scenari o num ber being t est ed.

Curr ent Solut ion t h is is t he value of t he obj ect ive f unct ion f or t hecurr ent scenari o.

Best Solut ion t h is is t he value of t he obj ect f unct ion for t he bestscenar io so f ar.

Sim ulat ion Tim e per Scenar io/ Real Tim e per Scenari o t his is t he

maximum sim ulat ion t im e t hat t he opt im izer wi l l spend for eachscenario. The optimizer stops a scenario as soon as this is met.

ScenariosScenarios


13/28

If you w ant t o run t he sim ulat ion several t im esf or a given scenario t o increase conf idence oft he mean of t he obj ect ive f unct ion, use t he Repl icat ions panel t o speci f y how m any

repl icat ions t o run.

If you w ant t o run t he sim ulat ion several t im esf or a given scenario t o increase conf idence oft he mean of t he obj ect ive f unct ion, use t he Repl icat ions panel t o specif y how m any

repl i cat ions t o run.

ReplicationsReplications

Perf orm m ult ip l e repl icat ions per scenari o i f t h is box is checked, t heopt im izer w i l l per f orm m ore t hat one repl icat ion per scenar io.

Minim um num ber of repl i cat ions t h is is t he m inim um num ber ofrepl icat ions t o run f or each scenari o. If t here is no Earl y Exit Cri t er i ont hen t he opt im izer w i l l alw ays run t he min im um number of r epl icat ions.

Perf orm m ult ip l e repl i cat ions per scenario i f t h is box is checked, t heopt im izer w i l l per f orm m ore t hat one repl icat ion per scenar io.

Min im um number of repl icat ions t h is is t he min im um number of

repl icat ions t o run f or each scenari o. I f t here is no Earl y Exit Cri t er i ont hen t he opt im izer w i l l alw ays run t he min im um number of r epl icat ions.


14/28

ReplicationsReplicationsMaxim um number of repl icat ions t h is is t he m axim um number ofrepl icat ions t o run f or each scenari o. I f t here is an Earl y Exit Cri t er i on t hent he opt imi zer w i l l run t he scenar io unt i l t he cr i t e r ion is met , up t o t h ism aximum num ber, af t er w hich t he opt im izer w i l l st op t he scenar io.

Ear ly Exit Cr i t er ion t h is al low s t he opt im izer t o st op running f urt herrepl icat ions f or t he same scenari o, based on t he cri t er i a you select . If you'veselect ed "Conf i dence Int erval Sat isf ied" t he opt im izer w i l l st op repl i cat ionsonce i t can det erm ine t he obj ect ive funct ion's t rue mean f or t he scenari o

w it h t he given conf idence and err or per cent ages. For exam ple i f you specif yan 80% confi dence and a 5% err or, t hen t he opt im izer w il l st op runningrepl icat ions as soon as i t is 80% confi dent t hat t he t rue m ean of t heobj ect ive f unct ion is w it hin 5% of t he mean sampl ed. I f "Best Solut ion

Out side Conf idence" is chosen, t hen t he opt im izer w il l st op repl icat ions f or agiven scenario i f i t can det erm ine, w it h in t he given conf i dence andal low able error percent ages, t hat t he best solut ion can never be met w it ht his scenario.

Maximum number of repl icat ions t his is t he maximum number of

repl icat ions t o run for each scenari o. I f t here is an Earl y Exit Cri t er i on t hent he opt imizer w i l l run t he scenar io unt i l t he cr i t e r ion is met , up t o t h ismaximum number, af t er w hich t he opt im izer w i l l st op t he scenar io.

Earl y Exit Cri t er i on t h is al l ows t he opt im izer t o st op running f ur t herrepl icat ions f or t he same scenari o, based on t he cri t er i a you select . I f you'veselect ed "Conf i dence Int erval Sat isf ied" t he opt im izer w i l l st op repl icat ionsonce i t can det erm ine t he obj ect ive funct ion's t rue mean f or t he scenario

w it h t he given conf idence and err or per cent ages. For example i f you specif yan 80% confi dence and a 5% err or, t hen t he opt im izer w il l st op runningrepl icat ions as soon as i t is 80% confi dent t hat t he t rue m ean of t heobj ect ive f unct ion is w it hin 5% of t he mean sampl ed. If "Best Solut ion

Out side Conf i dence" is chosen, t hen t he opt im izer w i l l st op repl i cat ions f or agiven scenari o i f i t can det erm ine, w it h in t he given conf i dence andal l owable error percent ages, t hat t he best solut ion can never be met w it ht his scenario.


15/28

OptimizingOptimizing

Once you have confi gured t he paramet ers, presst he Apply but t on t o app ly your conf igurat ion,t hen press t he Opt imize but t on, and wai tunt i l a message appears t el l ing you t hat t he

opt im izat ion has f ini shed and don't do anyt hingunt i l t he opt im izat ion has f in ished.

Once you have conf igured t he paramet ers, presst he Apply but t on t o app ly your conf igurat ion,t hen press t he Opt imi ze but t on, and wai tunt i l a message appears t el l ing you t hat t he

opt im izat ion has f ini shed and don't do anyt hingunt i l t he opt im izat ion has f in ished.


16/28

How many machines of each

type do I need to accomplishthe maximum prof it?

How many machines of each

type do I need to accomplishthe maximum prof it?

The m odels has 2 processes. The f ir st part of t he pr ocess can use t w odi f f e rent t ype of machines, machines t ype A (up t o 7 o f t hem) orm achines t ype B (up t o 5) . Each kind has a di f f erent process t im eand operat ional cost . The f ast er m achines are al so t he most

expensive.At t he last part of t he process only one kind of m achines could beused, m achines t ype C (up t o 8) .

The invent ory bet w een t he f ir st and t he second process rem ainsunder 31 f low it em s due t o space const raint s. As an assum pt ion al lt he product s are sold w hen f ini shed.

The models has 2 processes. The f ir st part of t he pr ocess can use t w odi f f erent t ype of m achines, m achines t ype A (up t o 7 of t hem ) orm achines t ype B (up t o 5) . Each kind has a di f f erent process t im eand operat ional cost . The f ast er m achines are also t he m ostexpensive.

At t he last part of t he process only one kind of m achines could beused, machines t ype C (up t o 8) .

The invent ory bet w een t he f ir st and t he second process rem ainsunder 31 f low it em s due t o space const raint s. As an assum pt ion al lt he product s are sold w hen f ini shed.


17/28

Type A machines for the fir st process.Process Time= 10 seconds.Operat ion cost per hour = $4,000 pesos.Quantity of machines = 1 to 7.

Type A machines for the fir st process.Process Time= 10 seconds.Operat ion cost per hour = $4,000 pesos.Quantity of machines = 1 to 7.

Type C machines for the final process.Process Time= 13 seg.Operat ion cost per hour = $3,200 pesos.Quanti ty of machines = 1 a 8.

Type C machines for the final process.Process Time= 13 seg.

Operat ion cost per hour = $3,200 pesos.Quanti ty of machines = 1 a 8.

Type B machines for the fir st process.Process Time= 108 seconds.Operat ion cost per hour = $2,600 pesos.Quantity of machines = 1 to 5.

Type B machines for the fir st process.Process Time= 108 seconds.Operat ion cost per hour = $2,600 pesos.Quantity of machines = 1 to 5.

Sale price per unit= $45.Sale price per unit= $45.

The model to opt imizeThe model to opt imize


18/28

Objective Functi on: Maximize Prof itProf it = Incomes Operational CostSale Price per unit = $45.00 pesosOperational Cost : Machine type A = $4,000

Machine type B = $2,600Machine type C = $3,200

Objective Function: Maximize Prof itProfit = Incomes Operational CostSale Price per unit = $45.00 pesosOperational Cost : Machine type A = $4,000

Machine type B = $2,600Machine type C = $3,200

The best solut ion automatically f ound:Use 5 type A machines, 2 t ype B and 8 type C to get a productionof 2,186 units with a prof it of $47,570 pesos.

The best solut ion automatically found:Use 5 type A machines, 2 t ype B and 8 type C to get a productionof 2,186 units with a prof it of $47,570 pesos.

Aft er a few minut es 280 dif ferent scenarios wereanalyzed in order t o find t he biggest profi t :$47,570 (Best Solution)

Aft er a few minut es 280 dif ferent scenari os wereanalyzed in order t o find t he biggest profi t :$47,570 (Best Solution)

The last analyzed scenario showeda lost of $5,540 pesos per hour

The last analyzed scenario showeda lost of $5,540 pesos per hour

Data f rom the last scenari o analyzed.Data f rom the last scenari o analyzed.

Minimum and maximum range of thequant it y of machines to be evaluated.

Minimum and maximum range of thequant it y of machines to be evaluated.


19/28

More opt imizat ion

examples

More opt imizat ion

examples


20/28

Routing by objects namesRouting by objects names


21/28

Model LayoutModel Layout

Global TableGlobal Table


22/28

As you know Flexsim def ault l ogic alw ays sends t he f low it em st o t he next dest inat ion based on t he out put por t num ber, nott he ob j ect name.

Rout ing by obj ect nam e comes handy i f you wi sh t o change t herout ing dest inat ions f or one or m ore obj ect s real ly easy,because you only have t o w ri t e t he nam es of t he dest inat ion on

a global t able using t he sequence you want w it hout having t oconnect or disconnect t he out put port s or change t hedest inat ion por t s num ber on t he obj ect i t se l f .

This m et hod gives you great f lexibi l i t y and could be underst oodand m odif ied by anyone, w i t hout Flexsim know ledge.

As you know Flexsim def ault l ogic alw ays sends t he f low it em st o t he next dest inat ion based on t he out put por t num ber, nott he ob j ect name.

Rout ing by obj ect nam e comes handy i f you w ish t o change t herout ing dest inat ions f or one or m ore obj ect s real l y easy,because you only have t o w ri t e t he nam es of t he dest inat ion on

a global t able using t he sequence you want w it hout having t oconnect or disconnect t he out put port s or change t hedest inat ion por t s num ber on t he obj ect i t se l f .

This m et hod gives you great f lexibi l i t y and could be underst oodand m odif ied by anyone, w i t hout Flexsim know ledge.

Routing by objects namesRouting by objects names


23/28

There are di f f erent pr ocesses in t his m odel , each process has it s ow ninput queue for sending t he f l owi t ems t o t he processors of t heir own group.

The Source and t he out put queue of each process m ust be connect ed t o

al l t he possibl e dest inat ions ( input queues, t he sink f or exit ing t he syst em oreven t o t heir ow n process input queue t o be processed again i f needed).

You could add or delet e pr ocessors f or each process. For adding moreprocessors j ust connect t hem t o t he input queue and out put queue and

def ine t he proper process t im e. No addit ional code is need. To delet e t hemj ust select t hem and press t he Delet e key.

Each process has an out put queue w hich sends t he f low it ems t o t he next

dest inat ion accord ing t o t he rout ing of t he t able. Alw ays w ri t e exact ly t he nam e of t he input queue of each process w hendef in ing t he rout ing sequence on t he global t able.

There are d if f erent pr ocesses in t his m odel , each process has it s ow ninput queue for sending t he f l owi t ems t o t he processors of t heir own group.

The Source and t he out put queue of each process m ust be connect ed t o

al l t he possibl e dest inat ions ( input queues, t he sink f or exit ing t he syst em oreven t o t heir ow n process input queue t o be processed again i f needed).

You could add or delet e pr ocessors f or each process. For adding moreprocessors j ust connect t hem t o t he input queue and out put queue and

def ine t he proper process t im e. No addit ional code is need. To delet e t hemj ust select t hem and press t he Delet e key.

Each process has an out put queue w hich sends t he f low it ems t o t he next

dest inat ion accord ing t o t he rout ing of t he t able. Alw ays w ri t e exact ly t he nam e of t he input queue of each process w hendef in ing t he rout ing sequence on t he global t able.

Explaining the modelExplaining the model


24/28

A w arning m essage w il l be di splayed w hen running t he model i f t hecorr ect dest inat ion can t be found and i t w i l l suggest w here t he m ist akecould be locat ed and t he sim ulat ion wi l l st op.

On t he Global Table, t he colum ns show t he nam e of t he process and t hecolum n next t o t hem shows t he processing t im e.

You can alw ays change t he size of t he global t able f or i ncreasing or

decreasing t he t ypes of product s you require and t he model w i l l w orkwi t hou t f u r t her m od i f i cat ion .

The process t im es are read f rom t he global t able.

If a cert ain product needs t o go t hrough t he same process m ore t hanonce t hen you may def ine a dif f erent process t im e for each st age.

A w arning m essage w il l be di splayed w hen running t he model i f t hecorr ect dest inat ion can t be found and i t w i l l suggest w here t he m ist akecould be locat ed and t he sim ulat ion wi l l st op.

On t he Global Table, t he columns show t he nam e of t he process and t hecolum n next t o t hem shows t he processing t im e.

You can alw ays change t he size of t he global t able f or i ncreasing or

decreasing t he t ypes of product s you require and t he model w i l l w orkwi t hou t fu r t her m od i f i cat ion .

The process t im es are read f rom t he global t able.

If a cert ain product needs t o go t hrough t he same process m ore t hanonce t hen you m ay def ine a dif f erent process t im e f or each st age.



25/28

You can def ine at t he Source t he quant it y and t ypes of product s youw ould l ike t he model t o run w i t h . Each product t ype wi l l show an uniquecolor.

There is a labe l on each f lowi t em cal led st ep creat ed on t he SourceOnCreat ion Trigger in order t o keep t rack of t he curr ent process st ep.

Cust om code is only f ound in t he Send To Port opt ion at t he Flow Tab of

t he Source and t he out put queues. This opt ion is cal l ed Rout ing accordingt o t he global t ab le . You on ly need t o wr i t e t here t he name of t he globalt able w hich should be used f or r out ing.

If you w ould l i ke t o explore t h is cust om code press t he code but t on t hatshow s t he A le t t er , and you w i l l f ind somet h ing as f o l low s

You can def ine at t he Source t he quant it y and t ypes of product s youw ould l ike t he model t o run w i t h . Each product t ype wi l l show an uniquecolor.

There is a labe l on each f lowi t em cal led st ep creat ed on t he SourceOnCreat ion Trigger in order t o keep t rack of t he curr ent process st ep.

Cust om code is only f ound in t he Send To Port opt ion at t he Flow Tab of

t he Source and t he out put queues. This opt ion is cal led Rout ing accordi ngt o t he global t ab le . You on ly need t o wr i t e t here t he name of t he globalt able w hich should be used f or r out ing.

If you w ould l i ke t o explore t h is cust om code press t he code but t on t hatshow s t he A le t t er , and you w i l l f ind somet h ing as f o l low s



26/28


27/28

// text variable for storign the names of the outpot port objects connected to this object:

string compare_name;

// were creating a foor loop which be repetead as many times as the total output ports

// numbers of this object or until the destination output number is found

for(int index=1; index


28/28

Thanks for your at tent ion!Thanks for your at tent ion!

Any addit ional quest ions you mayhave are always welcome:

j orget [email protected]

Any addit ional quest ions you mayhave are always welcome:

j [email protected]

OptQuest_Optimization

Documents