Bahan Ajar Minggu 1 Simsis | Simulation

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Documents

Published:

Views: 9 | Pages: 9

Extension: DOCX | Download: 0

Share
Related documents
Description
Teknik Industri
Transcript
Bahan Ajar Minggu Ke-1 Simulasi Sistem Bahan Ajar Minggu Ke-1 Tujuan Instruksional UmumSetelah menyelesaikan mata kuliah ini mahasiswa semester 6 mampu menganalisis output simulasi dengan Promodel dari sistem nyata.Tujuan Instruksional KhususMenjelaskan pengertian sistem, model, dan simulasi. Introduction (1/2) 1.Systems And Models Modeling is the enterprise of deising a simplified representation of a!omple system with the goal of proiding predi!tions of the system#s performan!e measures \$metri!s% of interest. Su!h a simplified representation is!alled a model. & model is designed to !apture !ertain 'ehaioral aspe!ts of the modeled system(those that are of interest to the analyst)modeler(inorder to gain knowledge and insight into the system#s 'ehaior \$Morris *+6%.Modeling !alls for a'stra!tion and simplifi!ation. In fa!t, if eery fa!et of thesystem under study were to 'e reprodu!ed in minute detail, then the model!ost may approa!h that of the modeled system, there'y militating against!reating a model in the first pla!e.The modeler would simply use the -real system or 'uild an e perimental oneif it does not yet e ist(an e pensie and tedious proposition. Models aretypi!ally 'uilt pre!isely to aoid this unpalata'le option. More spe!ifi!ally,while modeling is ultimately motiated 'y e!onomi! !onsiderations, seeralmotiational strands may 'e dis!erned/ ã 0aluating system performan!e under ordinary and unusual s!enarios. &model may 'e a ne!essity if the routine operation of the real1life systemunder study !annot 'e disrupted without seere !onse2uen!es \$e.g.,attempting an upgrade of a produ!tion line in the midst of filling !ustomer orders with tight deadlines%. In other !ases, the e treme s!enario modeled 1  Bahan Ajar Minggu Ke-1 Simulasi Sistem is to 'e aoided at all !osts \$e.g., think of modeling a !rash1aoidingmaneuer of manned air!raft, or !ore meltdown in a nu!lear rea!tor%. ã Predi!ting the performan!e of e perimental system designs. 3hen theunderlying system does not yet e ist, model !onstru!tion \$andmanipulation% is far !heaper \$and safer% than 'uilding the real1life systemor een its prototype. 4orror stories appear periodi!ally in the media on proje!ts that were rushed to the implementation phase, without proper erifi!ation that their design is ade2uate, only to dis!oer that the systemwas flawed to one degree or another \$re!all the !ase of the 'rand newairport with faulty luggage transport%. ã 5anking multiple designs and analying their tradeoffs. This !ase is relatedto the preious one, e !ept that the e!onomi! motiation is een greater. Itoften arises when the re2uisition of an e pensie system \$with detailedspe!ifi!ations% is awarded to the 'idder with the 'est !ost7'enefit metri!s.Models !an assume a ariety of forms/ ã & physi!al model is a simplified or s!aled1down physi!al o'je!t \$e.g.,s!ale model of an airplane%. ã & mathemati!al or analyti!al model is a set of e2uations or relationsamong mathemati!al aria'les \$e.g., a set of e2uations des!ri'ing theworkflow on a fa!tory floor%. ã & !omputer model is just a program des!ription of the system. &!omputer model with random elements and an underlying timeline is!alled a Monte 8arlo simulation model \$e.g., the operation of amanufa!turing pro!ess oer a period of time%. 2.Analytical ersus Simulation Modeling & simulation model is implemented in a !omputer program. It is generally arelatiely ine pensie modeling approa!h, !ommonly used as an alternatie toanalyti!al modeling. The tradeoff 'etween analyti!al and simulation modelinglies in the nature of their -solutions, that is, the !omputation of their  performan!e measures as follows/*.&n analyti!al model !alls for the solution of a mathemati!al pro'lem, andthe deriation of mathemati!al formulas, or more generally, algorithmi! 2  Bahan Ajar Minggu Ke-1 Simulasi Sistem  pro!edures. The solution is then used to o'tain performan!e measures of interest.9.& simulation model !alls for running \$e e!uting% a simulation program to produ!e sample histories. & set of statisti!s !omputed from these historiesis then used to form performan!e measures of interest.To !ompare and !ontrast 'oth approa!hes, suppose that a produ!tion line is!on!eptually modeled as a 2ueuing system. The analyti!al approa!h would!reate an analyti!al 2ueuing system \$represented 'y a set of e2uations% and pro!eed to sole them. The simulation approa!h would !reate a !omputer representation of the 2ueuing system and run it to produ!e a suffi!ient num'er of sample histories. Performan!e measures, su!h as aerage work in thesystem, distri'ution of waiting times, and so on, would 'e !onstru!ted fromthe !orresponding -solutions as mathemati!al or simulation statisti!s,respe!tiely.The !hoi!e of an analyti!al approa!h ersus simulation is goerned 'y generaltradeoffs. :or instan!e, an analyti!al model is prefera'le to a simulation modelwhen it has a solution, sin!e its !omputation is normally mu!h faster than thatof its simulation1model !ounterpart. Unfortunately, !omple systems rarelylend themseles to modeling ia suffi!iently detailed analyti!al models.;!!asionally, though rarely, the numeri!al !omputation of an analyti!alsolution is a!tually slower than a !orresponding simulation. In the majority of !ases, an analyti!al model with a tra!ta'le solution is unknown, and themodeler resorts to simulation.3hen the underlying system is !omple , a simulation model is normally prefera'le, for seeral reasons. :irst, in the unlikely eent that an analyti!almodel !an 'e found, the modeler#s time spent in deriing a solution may 'ee !essie. Se!ond, the modeler may judge that an attempt at an analyti!alsolution is a poor 'et, due to the apparent mathemati!al diffi!ulties. :inally,the modeler may not een 'e a'le to formulate an analyti!al model withsuffi!ient power to !apture the system#s 'ehaioral aspe!ts of interest. In!ontrast, simulation modeling !an !apture irtually any system, su'je!t to any 3  Bahan Ajar Minggu Ke-1 Simulasi Sistem set of assumptions. It also enjoys the adantage of dispensing with the la'or attendant to finding analyti!al solutions, sin!e the modeler merely needs to!onstru!t and run a simulation program. ;!!asionally, howeer, the effortinoled in !onstru!ting an ela'orate simulation model is prohi'itie in termsof human effort, or running the resultant program is prohi'itie in terms of !omputer resour!es \$8PU time and memory%. In su!h !ases, the modeler mustsettle for a simpler simulation model, or een an inferior analyti!al model.&nother way to !ontrast analyti!al and simulation models is ia the!lassifi!ation of models into des!riptie or pres!riptie models. <es!riptiemodels produ!e estimates for a set of performan!e measures !orresponding toa spe!ifi! set of input data. Simulation models are !learly des!riptie and inthis sense sere as performan!e analysis models. Pres!riptie models arenaturally geared toward design or optimiation \$seeking the optimal argumentalues of a pres!ri'ed o'je!tie fun!tion, su'je!t to a set of !onstraints%.&nalyti!al models are pres!riptie, whereas simulation is not. Morespe!ifi!ally, analyti!al methods !an sere as effe!tie optimiation tools,whereas simulation1'ased optimiation usually !alls for an e haustie sear!hfor the optimum.;erall, the ersatility of simulation models and the feasi'ility of their solutions far outstrip those of analyti!al models. This a'ility to sere as an initro la', in whi!h !ompeting system designs may 'e !ompared and !ontrastedand e treme1s!enario performan!e may 'e safely ealuated, renderssimulation modeling a highly pra!ti!al tool that is widely employed 'yengineers in a 'road range of appli!ation areas. In parti!ular, the !omple ity of industrial and seri!e systems often for!es the issue of sele!ting simulation asthe modeling methodology of !hoi!e. !.Simulation Modeling And Analysis 4
Recommended

11 pages

9 pages

26 pages

46 pages

9 pages

7 pages

13 pages

5 pages

55 pages

24 pages

page

10 pages

12 pages

9 pages

Bahan Ajar Minggu 1 Simsis.docx

9 pages

View more...

ged_scanning.pdf | Image Scanner | Media Technology

We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks