多品質特性系統應用智慧型實驗設計於小型水族箱養殖環境最佳化之研究

(1)

кݡኳপّր௚ᑕϡംᇊݭ၁រనࢍٺ̈ݭͪ୉ቐዳ തᒖဩ౵ָ̼̝ࡁտ

ᖎځᇇ ች࿭ᆸ

ϖ྿ԫఙጯੰ࿪፟̍඀ր/機械工程系

ͳˠನ

৸ࡗэϲ̂ጯOneonta ̶७௚ࢍăࢍზ፟ࡊጯăᇴጯր

ၡ! ࢋ

ώࡁտ೩ֻ˘჌ᖎಏăѣड़ăԣిă˫གྷᑻ۞၁រనࢍ̶ژڱĄώࡁտͽ ϣ˾͞ڱࠎૄᖂĂඕЪѷᙯᓑ̶ژᄃሀቘଠטጡనࢍ൴णംᇊݭкݡኳዋ̼̝

পّր௚Ąѩր௚Ӏϡѷᙯᓑ̶ژĂι࣌ѣ఍நሀݭ̙ቁؠᄃ͌ᇴᇴፂ۞প

ّĂᄃሀቘଠט̝ଯநྋՙЧݡኳপّ̝࠹၆ࢦࢋّĄ

ώ ၁ រ ᑕ ϡ ሀ ቘ ଠ ט ጡ ଯ ኢ ଀ ז ྵ މ ៍ ۞ ტ Ъ ّ к ݡ ኳ প ّ ޽ ᇾ (MPCI)Ăפ΃̍඀र۞͹៍གྷរҿᕝĂ҃ͽ࠹၆ݡኳ̝ࢦࢋّүࠎ౵ָ௡ЪĂ ቁܲ၁រ໤ቁޘĄ၁រඕڍགྷϤតளᇴ̶ژĂΞቁᄮ໢ޘଠטጡ̝ણᇴтΐ໢

ഔᇴณĂᕭጡ྅ཉҜཉĂΐ໢ॡมᄃͪ஄፩ޘࠎ̈ݭͪ୉ቐٸዳᒖဩ̝ព඾ଠ טЯ̄Ă΁ࣇ၆ፋវតள̝੒ᚥޘ̂ࡗҫ62%Ă౵ዋͪ໤ቁᄮ၁រ̝ѷᙯᓑк ࢦݡኳপّ޽ᇾགྷሀቘଠטጡଯኢඕڍ࠹༊ତܕٺ1Ă҃࣎Ҿݡኳপّ໢ޘĂ

ͧࢦĂPH ࣃ̶Ҿତܕ࿰ؠϫᇾࣃĂͷᄃЧݡኳϫᇾࣃᄱमѺ̶ͧӮҲٺ 2%Ą ඕڍពϯٙ೩΍ͽϣ˾͞ڱࠎૄᖂĂඕЪѷᙯᓑ̶ژᄃሀቘଠטጡనࢍ͞ڱĂ གྷዋ̼ޢΞѣड़྿ז႕ຍ۞ඕڍĄ

ᙯᔣෟĈѷᙯᓑ̶ژăሀቘទᏭଠטጡăϣ˾͞ڱă౵ָ̼Ą

INTELLIGENT DESIGNS OF EXPERIMENT FOR MULTIPLE CHARACTERISTICS OPTIMIZING A SMALL-SCALE AQUACULTURE

ENVIRONMENT

Ming-Der Jean Jen-Shen Tsai

Department of Electrical Engineering / Department of Mechanical Engineering Yung Ta Institute of Technology and Commerce

PingTung, Taiwan 909, R.O.C.

Jen-Ting Wang

Department of Mathematics, Computer Science, and Statistics State University of New York College at Oneonta

NY 13820, U. S. A.

Key Words: grey relational analysis, fuzzy logic controller, taguchi method, optimization.

(2)

ABSTRACT

The optimization of intelligent multiple performance characteristics can be accomplished by using the Taguchi-based method along with grey relational analysis and a fuzzy logic controller. The grey relational analysis solves the problems for model-uncertainty and data scarceness, while the fuzzy logic controller takes into account the relative importance of each individual quality characteristic. This paper proposes a simple and effective way of developing an efficient and systematic design.

In this study, we apply the proposed procedure to optimize the environment of a small-scale aquarium. Taking the relative importance of the multiple-performance characteristics into consideration, we employ a fuzzy logic controller to generate multiple-performance characteristics indices (MPCI) as an indicator of the overall quality characteristics. From the experimental results, the most significant control factors can be identified as: the water filter set-up, the heating time, the quantity of quartz heaters and the turbidity of the water. These factors account for 62% of the total variance. The confirmation run, at the optimal setting, was conducted and shows that all the individual quality characteristics reach their desired target values with errors well below 2%, and their overall multiple-performance characteristics were achieved satisfactorily.

˘ăჰ ኢ

ͪயዳതߏ˘ਠ૱֍ٺώ࠷Чг۞˘ีயຽĂϤٺዋ

఍ঔफ̚ݭঈ࣏ĂՏ༊΋؞ರ߹ֽ᝚Ă૱ֹЧгͪயຽ۰

̈́छलёͪ୉ቐዳതĂౄјᚑࢦ۞ຫε๋̈́चĄЯѩᅮᖣ Ϥዋ༊۞ΐ໢ܲ޺ዳതѰ۞ͪ໢Ăഴ͌ͪѰ̚໢मĂᔖҺ

ֹዳതͪயЯдͪ໢۞ᆐধត̼˭Ă׎Ϡۏ፟ਕ൑ڱ࢑ఈ

҃Ѫ˸Ă࡭ֹዳത̝х߿தࢫҲĄЯѩĂᘦઉ۞ዳതᒖဩ ၆ዳതຽߏ˘ีࢦࢋኝᗟĄ

၆ٺͪயዳത̝ԫఙĂــ࠰ጴᖣ૞छགྷរĄ൒҃Ă ѣࢨ۞ۢᙊٕགྷរ̏൑ڱ႕֖ኑᗔͪ୉ᒖဩ۞តજĂЯѩ ᅮࢋᖣӄ၁រనࢍ͘ڱΐͽྋՙĄϣ˾͞ڱߏ˘჌జ̍ຽ

ࠧᇃھֹϡ۞၁រనࢍ͞ڱĂ͹ࢋਕԼච็௚̝၁រనࢍ

۞৿ᕇĂд౵͌၁រѨᇴ˭Ăֶ൒ਕԱזତܕዋ̼۞ણᇴĄ

൒҃дٺ఍ந׍кࢦݡኳপّᄃᄬຍតᇴ۞યᗟ˯̪̙Ⴝ ԆචĄтңԼච˯ࢗયᗟ֭ͷ೩΍˘ѣड़ྋՙ͞ڱ˜ߏώ ࡁտ̝ࢦࢋኝᗟ̝˘Ąͽـ˘ֱጯ۰ࡁտ೩΍ྋՙкݡኳ পّયᗟॡĂ૱ᑕϡ̍඀˯࠹ᙯགྷរۢᙊтቢّఢထĂਫ਼ ᕩ̶ژĂݡኳຫεבᇴඈүᝋࢦҿᕝ[1-3]Ą఺ֱࡁտౌѣ

̙Т៍ᕇĂ൒̙҃௲ٺኑᗔ۞ႊზٕ͹៍ّĄૄٺѩࣧЯĂ ώ͛೩΍ͽϣ˾͞ڱࠎૄᖂĂඕЪѷᙯᓑ̶ژᄃሀቘଠט ጡనࢍநኢĂϡͽವՐ౵ָкϫᇾݡኳপّĄ

ϣ˾౾̀ٺ1950 ѐז 1960 ѐ΃ܐഇ൴ण΍ᘦઉనࢍ

۞நه[4]Ă೩΍˞ܫཱིᗔࢰͧ(signal/noise ratioĂS/N ͧ)

۞ ໄ ه Ă ೩ ࣍ ϡ ր ௚ న ࢍ(system design) ă ણ ᇴ న ࢍ (parameter design)̯̈́मనࢍ(tolerance design)ֽԼචயݡ ݡኳĄᑕϡϣ˾͞ڱĂ੫၆ಏ˘ݡኳপّฟ൴౵ָ̼ᄦ඀

ણᇴĂ̏ᙋ၁࠹༊ѣड़ͷјΑĄ̙࿅Ăѩ౵ָ୧І఼૱̙

ዋϡٺ׎ιݡኳপّĄЯѩĂTong ඈˠ[5]Ă݋ߏᑕϡϣ˾

ݡኳຫε۞நهĂ૟࣎Ҿݡኳপّ̝S/N ͧĂӀϡᝋࢦΐ ј۞៍ᕇĂฟ൴кࢦݡኳপّ౵ָ̼୧ІĄWu [6] അͽϣ

˾͞ڱᓁݡኳຫεഴ͌۞͞ڱĂֽฟ൴кࢦݡኳপّ౵ָ

̼ᄦ඀ણᇴĄ

ѷᙯᓑ̶ژ[7]˜ߏವՐր௚̚ЧЯ৵ม۞͹ࢋᙯ

ܼĂ׍ѣ͌ᇴፂ̈́кЯ৵̶ژ۞পّĂ֭ͷਕૉವԱᇆᜩ ݡኳপّ̝Я̄Ăଂ҃ೠ೪ְۏ۞͹ࢋপᇈĂܳซ׶͔ጱ ր௚֝ి҃ѣड़൴णĄ׎͹ࢋᑕϡдր௚ሀݭֽ໚̙ԆБ

۞ଐڶ˭Ăઇր௚۞ᙯᓑ̶ژăሀݭ࿰ീăՙඉᄃଠטĄ ሀቘநኢٺ1965 ѐࢵАϤ Zadeh ౾̀[8]ٙ೩΍Ăߏ ϡֽྋՙ൑ڱځቁؠཌྷ۞ሀቘّໄه۞˘჌ؠณܑ྿̍

׍Ąְ၁˯Ăሀቘநኢߏࠎྋՙৌ၁͵ࠧ̚Ă೼࿆хд۞

ሀቘă̙ቁؠன෪҃൴ण۞˘ܝጯયĄϫ݈дˠ̍ംᇊă ҋજଠטăঈ෪ಡӘăဦညᙊҾăᗁᒚ෧ᕝăՙඉ͚೯ă გநࡊጯăᒖဩෞҤඈЧ჌ᅳા۞ᑕϡ˯ӮѣᖳჇ۞јڍ [9-13]Ą൒҃ĂᑕϡሀቘநኢనࢍವՐྋՙкݡኳপّ௡Ъ યᗟтЧݡኳಏҜ̝̙Тăࢦࢋّ۞मள̈́൑ڱځቁ۞ү

΍ՙඉඈඈЯ৵̙֭к֍Ą

ώ၁រགྷϤۡϹܑ۞నࢍĂሀᑢዳതᒖဩд̙Т۞Я

৵តજ˭Ăٚצкࢦણᇴଠט˭ٸዳᒖဩ۞ͪ໢ă໘উă

ͪኳăݡኳ̈́ٸዳϠۏдᒖဩ۞ត̼˭ĂವՐዋ̼ણᇴ௡

Ъֹͪ୉ቐٸዳϠۏਕૉх߿̝౵ָᘦઉϠܜᒖဩĄͪኳ ᄃ໢ޘ۞ត̼၆̈ͪ୉ቐᘦઉٸዳᒖဩᇆᜩࠤ࿝[14,15]Ă

˵࠹၆ᇆᜩז౦۞х߿தĄ໢ޘ۞̙ᘦؠĂົౄјͪ̚ঊă ൻᅕៃăֲ̈́ൻᅕៃ፧ޘ۞ள૱̿੼Ą఺ֱ੼ّ߲۞̼ጯ ۏ͹ࢋயϠֽ໚ࠎ౦ଵڴۏĄ൑ኢтңĂ౦۞Ϡܜᒖဩх д͉кតᇴĂౄјٸዳᒖဩ̙ٽଠטĄЯѩĂԧࣇӀϡሀ ቘநኢਕ఍நໄهሀቘ̙ቁؠ۞ְۏ۞পّĂ੨Ъ၁រన

ࢍଂ҃ޙϲ˘इംᇊݭٸዳᒖဩዋ̼ր௚Ąѣᙯ͛ᚥዋ̼

(3)

ّࡁտኜт Lin ඈˠ[16]ӀϡሀቘநኢᄃѷҒநኢĂລ੨ ϣ˾͞ڱĂฟ൴кࢦݡኳপّ౵ָ̼ᄦ඀ણᇴĄWang [17-20] ඈ૞छጯ۰೩΍ᖣϤሀቘទᏭ૟౵ָ̼̝кࢦݡ ኳপّᖼೱࠎಏ˘۞ݡኳপّĂ֭ඕЪϣ˾͞ڱᄃሀቘទ Ꮽ൴ण΍׍ѣሀቘّనࢍϫᇾ۞кࢦϫᇾႊზڱĂѩڱ۞

পҒдٺֹϡሀቘទᏭ૟ݡኳপّS/N ͧᖼјᕩᛳבᇴĂ

ֽ૟кࢦݡኳপّᖼ̼ࠎಏ˘ݡኳপّĂ҃଀זЧણᇴͪ

໤۞౵ָྋĄ

ᙯٺͪயዳതᒖဩЯ̝̄ଣ੅̏ѣ̙͌ࡁտĂ൒҃ޝ

͌ಡӘਕૉ೩ֻѣࢍ൪ăѣड़த۞ࡁտĂ͍׎дଣ੅кࢦ ݡኳপّ۞યᗟ˯Ą੫၆ѩયᗟĂώࡁտ೩΍˘इѣր௚ă ड़த੼ăͷགྷᑻ۞͞ڱĂඕЪѷᙯᓑ̶ژᄃሀቘଠטጡన

ࢍᑕϡٺͪயٸዳᒖဩү΍ᘦઉଣ੅Ąዋ̼ણᇴ௡Ъ˜ߏ

࡭˧ٺᏴፄ౵ָଠטЯ̄۞୧ІĂͽܮ૟ͪயዳതᒖဩត ள၆Я̄۞ᇆᜩࢫҌ౵ҲĂЯѩՐ଀ዋ̼۞ણᇴߏԼචͪ

୉Ϡܜᒖဩనࢍ۞ࢦࢋኝᗟĄӀϡ၁រ۞͞ڱĂ૟ৌ၁۞

үຽଐڶĂ౅࿅ሀݭనࢍү΍ሀᑢീྏĂವԱ΍౵ָ࿰ീ

ᄃ၁ᅫ۞ЯڍᙯܼĄЯѩĂώ၁រଳϡϣ˾నࢍĂ֭ፋЪ ѷҒᄃሀቘநኢͽྋՙкࢦݡኳপّ̝યᗟĂ೩ֻྵމ៍

۞ҿᕝֽԱವ΍˘࣎౵ָͷᘦઉ̝ͪ୉ϠܜᒖဩĄ

˟ăᆇጡ̬௜̈́၁រన౯

ώ၁រ̝໢ޘଠטጡࠎጯϠ૞ᗟҋҖనࢍĂѩ໢ޘଠ טሹߏ၁រॡٙࣅᏥ۞Ąܲ໢̈́ΐሤ۞నࢍࠎ८͕ർវ̝

˘Ą၁រ̝̚Чีన౯ֶ౦ཏᇴณ҃ѣ̙Т͎̇ͪ୉ቐĄ ΐሤጡࠎΐ໢ϡĂֹϡ۞჌ᙷߏϮࡻΐሤഔĄᕭጡ݋ࠎ࿅

ᕭͪ୉ቐᆿۏ̝྅ཉĄPH ᇴҜᑭീጡֽീณᅕែࣃ̝ᇴ ፂĄͧࢦࢍߏ౦ᙷٸዳᒖဩͪኳត̼̝ࢦࢋീณᆇጡĂξ ࢬ˯ͧࢦࢍ̶୶ͪ̈́ঔͪ׌჌Ą

ˬă၁រనࢍ

၁រనࢍࠎࡁտᄦ඀ٕր௚̝পؠϫ۞ܑனĂᄦ඀ٕ

ր௚݋ͽЯڍሀݭ(cause and effect)͞ёܑϯĄన౯ă͞ڱ

̈́׎΁ྤ໚Ă౅࿅ᄦ඀௡Ъซˢ఺࣎ր௚ᖼೱјᏮˢᄃᏮ

΍ᙯܼĄ҃ᏮˢੈཱིࠎΞଠטЯ̄(control factors)ᄃ̙Ξଠ טЯ̄(noise factors)̝௡ЪĂ၁រనࢍྻϡ఺ֱЯ̄၆ր

௚ͽਕณԛၗᖼೱјᏮ΍۞ᜩᑕ(response)Ąώ၁រఢထࠎ

ొЊЯ̄၁រĂֶϣ˾ۡϹܑЧҖҋԧπᏊᄃ࠹̢׌Җϒ Ϲ̝পّĂͽ౵͌۞၁រѨᇴݒΞᒔᄃБЯ̄၁រඈТ۞

ྤੈĂЯ҃ᒔ଀ྵ੼۞ड़தĂ֭ᒢྋፋវԫఙ̝ھϡّĄ 1. ϣ˾၁រڱ

˘ਠ۞၁រߏଳϡྏᄱڱĂҭѩ჌үڱٙ଀ඕڍــ

̙ߏ౵ዋ༊۞Ăֹ֭଀ፋ࣎ฟ൴ᄦ඀෱ॡ˫ڀෳĄࠎ੠Ր ᘦઉّ(robustness)̝ᄦ඀నࢍĂϣ˾ϛ˘౾̀(Taguchi)೩

΍˘჌ᖎಏѣड़ͷր௚̼۞၁រనࢍ͞ڱĄ΁۞͹ࢋ੒ᚥ

ܑ˘ ЧଠטЯ̝̄ͪ໤ᇴܑ

ଠטЯ̄

௑ཱི ͪ໤˘ ͪ໤˟ ͪ໤ˬ

A ͪ ኳ ঔͪ ୶ͪ -

B ΐሤጡΑத(w)

ΐሤ࿪߹(A) 100 300 500 C ΐሤ࿪߹(A) 1.181 1.363 2.727 D ͪវ᎕(cm³) 60×30×36 75×45×45 90×45×45

E ΐ໢ഔᇴณ 1 2 3

F ΐሤॡม(min) 10 12 15 G ᕭጡ྅ཉҜཉ غొ ˯ొ ㆘部

H ஄፩ޘ ̈ ̚ ̂

дٺͽ౵ҲᄦౄјώĂჯ޺ᄦౄّਕĂֹ֭ր௚ܑன၆ត ளّЯ̄۞ୂຏّࢫҌ౵ҲĄϣ˾͞ڱ˫׍౯ԣిгăѣ ड़தгăགྷᑻгฟ൴າயݡ̈́Լ։ᄦݡ۞পّĄιܼӀϡ

ۡϹܑనࢍĂͽ͌ᇴ۞၁រֽࡁտிк۞ણᇴតᇴ၆ݡኳ পّ̝ᇆᜩĂᖣϤ၁រᇴፂᖼೱј˘჌າ۞ݡኳ޽ᇾ˘ܫ

ཱིᗔࢰͧ(signal/noise ratio)ֽଯؠݡኳপّĄܫཱིᗔੈͧ

(S/N ratio)ؠཌྷࠎĈѣϡ۞Ꮾ΍(useful output)/ѣच۞Ꮾ΍

(harmful output)Ąயݡ׶ᄦ඀̝ݡኳপّ˜ॲፂ S/N ͧү

ෞҤĈS/N ͧດ̂Ă׎ݡኳດрĄϣ˾౾̀ᄮࠎܫཱིᗔੈ

ͧĂߏᘦઉّ઱˘ෞҤ޽ᇾĄ൒҃Ăᄦ඀൴ण࠰གྷ።ˬ࣎

ล߱(ր௚నࢍăણᇴనࢍă̯मనࢍ)Ăώ၁រᑕϡϣ˾

͞ڱ͹ࢋ඾ࢦણᇴనࢍ(parameter design)Ąણᇴనࢍૄώү ڱࠎࢫҲயݡצଠטЯ̄̈́ᗔੈЯ̝̄ᇆᜩĂֹតளࢫҌ

౵ҲĂ֭Ա΍౵ָણᇴ௡ЪĂֹր௚Ꮾ΍ᔌܕϫᇾࣃĂᖣ ϤҲјώ۞నࢍ྿זயݡݡኳ̝౵ָ̼Ą

2. Я̄੨ཉᄃۡϹܑ

дϣ˾͞ڱ่̚ᅮޝ͌၁រүീྏĂֹϡۡϹܑĂߏ ԓ୕ਕͽ౵͌၁រѨᇴଐԛ˭Ăᒔ଀౵ָ۞Гனّ၁រ୧ І௡ЪĂΞ̂ณഴ͌၁រѨᇴĄ၁រ੨ཉͽ L(2¹×3⁷)۞ۡ

ϹܑଵЕ͞ё੨ཉĂЯ̄ͪ໤ᇴдL18 ஄ЪۡϹܑ̚ੵͪ

ኳ่׌ͪ໤γĂ׎ዶ࠰ࠎˬ჌ͪ໤ĄЯᔖҺ၆ώ၁រயϠ ᄱम߇פҋঔᙝ۞ᇾ໤ঔͪĄώ၁រ۞੓ؕ໢ޘనؠࠎ22 ƨĂЯ၁រีϫᅮࢦኑֹϡͪྤ໚߇ֹϡД๴ֽࢫ໢Ăͽ

྿ј၁រ۞ᇾ໤నؠĂ֭ͷΐి၁រፆү۞ॡड़Ąͪ୉ٸ ዳᒖဩ၁រ୧Іѣ18 ჌௡ЪĂࠎՐ၁រ໤ቁّĂՏ˘Ѩ၁ រࢦᖬซҖ׌ѨീณՐ׎πӮࣃͽഴ͌ᄱमĄ၁រඕڍ̝

ݡኳপّѣͪ໢ޘăPH ࣃᄃͧࢦĄPH ࣃត̼˜Я࿶फ़ᔼ ዳ౦ཏണዶĂ౦ཏ۞ଵڴۏԼតͪ̚ᅕែޘٙ࡭ćϠܜᒖ ဩ̝ͪͧ̚ࢦ˜ࠎᇆᜩ౦ཏјܜ۞Ω˘ࢦࢋតЯĄٙͽన ؠѩˬ჌ࢦࢋݡኳপّࠎᏮ΍ࣃĄࢵАֶፂన౯ăᒖဩă ᄦ඀ඈĂԱٙᅮ̝8 ჌дͪயዳത̙Ξ৿͌۞ࢦࢋଠטЯ

̄Ăтܑ˘ٙϯĄዳതͪயѣ୶ͪᄃঔͪዳത׌჌Ă׎Я г఍ݑΔొ̈́ঔ˯Ăঈ࣏Я৵ౄј໢ޘ̙ᘦؠĂ҃ѣҋજ ଠטΐሤጡన౯ć஄፩ޘֶᔼዳ࿶फ़ࢦณ҃ؠĂ֭ͽٸዳ

(4)

౦ਕٚצ̝टԡࢨޘࠎૄ໤Ąٸዳͪវ᎕۞నࢍֶ౦ཏᇴ ณ۞кဿĂΐሤॡมણ҂͛ᚥ੃ᐂనؠ [14]Ăͪ໢݋ֶ౦

჌۞̙Т҃ؠĄ၁រณീ̝ݡኳপّт໢ޘăPH ࣃᄃͧࢦ

̶ҾซҖѷᙯᓑ඀ޘ̶ژ(grey relational grade analysis)ĄГ ۰Ăࠎ҂ᇋЧݡኳপّ̝࣎ҾࢦࢋّĂͽـࡁտ۰̂кͽ ҋ̎͹៍ҿᕝග̟ᝋࢦࣃĂ֭൑ڱމ៍г޷ݡኳপّ̝ࢦ

ࢋ ّ ү ଣ ੅ Ă ώ ၁ រ ೩ ΍ ሀ ቘ ଠ ט ጡ ሀ ݭ(fuzzy logic controller)ΞྋՙЧݡኳপّ̝࠹၆ࢦࢋّ඀ޘĂГͽ௡Ъ

̝ಏ˘পّѷᙯᓑ޽ᇾ(γ)ĂүࠎᏊณкࢦݡኳপّ۞ّਕ (multiple performance characteristics’ index, MPCI)Ă֭ͽੈ

ཱིᗔੈͧүࠎᏊณፋវݡኳّਕ̝Ω˘޽ᇾĄ 3. ݡኳপّϒఢ̼

кࢦݡኳপّхдኜкӧᕘĂኜтĈಏҜă჌ᙷăࢦ

ࢋّ̝̙ТĂᄃЧݡኳপّϫᇾ̝̙Тт୕̂(larger-the- better) ă ୕ ̈ (smaller-the-better) ă ٕ ୕ ϫ (desired-the- better)ĄЯѩᑕֶ׎̙Тϫᇾ̝পّĂืઇዋ༊൑ЯѨ۞

ϒఢ̼఍நĄώ၁រ่ಶ୕ϫপّϒఢ̼఍நᄲځĂܑϯ т˭Ĉ

)]}

( min[

, )]

( max{max[

)

1 ₍₀₎ ( ₍₀₎

) 0 (

k x x x k x

x k x x

i i

i

i − −

− −

= ) )

)

(1)

) (k

x_i ࠎ୕ϫϒఢ̼̝ᇴࣃćx) ࠎx_i⁽⁰⁾(k)̝ᏴؠࣃĄֶ׎

Чݡኳপّགྷ࿅ϒఢ̼Ă૟ݡኳপّࢨטٺ0 ᄃ 1 ̝มĈ

༊ݡኳপّኳࣃດତܕ1 ܑϯݡኳດрćࡶݡኳপّኳତ ܕٺ0 ܑϯݡኳດमĄ

4. ѷᙯᓑкࢦݡኳႊზ

ѷҒநኢٺ 1982 ѐϤዒჸᐷି଱೩΍Ă͹ࢋࡁտᑕ ϡдր௚ሀݭхд̝̙ቁؠّĂྤफ़̙Ԇፋᄃႊზ̙୻຾

˭Ăү΍ր௚ᙯᓑ̶ژăሀݭ̝ޙϲᄃ࿰ീĄѩநኢ၆ᄦ

඀̝кតณপّଣ੅Ăਕઇѣड़̝఍நĄώ͛Ӏϡѩநኢ ଣ੅к࣎ݡኳপّ̝ม۞ᙯᓑ඀ޘĄ

5. ѷᙯᓑޘࢍზ

ѷᙯᓑޘ̝ҤࢍĂ׎ࢍზӣѣm ࣎ݡኳপّࣃ̝ϫᇾ ԔЕx0 ᄃ p ࣎ͧྵԔЕ(x1, x2,…, xP)ม̝࠹ᙯᓑ඀ޘܼᇴ т˭ܑϯĈ

max )

(

max )) min

( ), (

( ∆ + ∆

∆ +

= ∆

ς ξ ς

k k x k x

oi j

i

i (2)

׎̚i =1, 2,…, m; k = 1, 2,…, n; j∈I ćx0ࠎϫᇾԔЕĂxi

ࠎͧྵԔЕĄ∆_oi= x₀(k)−x_i(k) ࠎԔЕ x0ᄃ xiдௐ k

࣎म۞඗၆ࣃĄ

k oi

i j∈ ∀ ∆

∀

=

∆min ^min ^min

k oi

i j∈ ∀ ∆

∀

=

∆max ^max ^max

ςࠎᏰᙊܼᇴćς∈

[ ]

0,1ĄЯѩĂѷᙯᓑޘΞܑϯт˭Ĉ

= ∑

= n

k i i i i

i w x k x k

1 ξ( ( ), ( ))

γ i =1, 2,…, m (3)

׎̚w ࠎܑௐ i ࣎ݡኳপّ̝ᝋࢦĄЧݡኳপّ̝ᝋࢦֶ_i

ፂώ၁រనࢍ̝ሀቘଠטጡү΍ംᇊّҿᕝĄ 6. ሀቘଠטጡ

ሀቘநኢߏࠎ˞ྋՙৌ၁͵ࠧ̚೼࿆хд۞ሀቘன ෪҃൴ण۞˘ܝጯયĂϡֽܑனߙֱ൑ڱځቁؠཌྷ۞ሀቘ ໄهĂ͍׎ߏдܑனˠᙷᄬ֏পѣ۞ሀቘّன෪[8]Ąሀቘ ଠטߏӀϡሀቘநኢ۞ܕҬଯኢүࠎଠטጡ۞ଯኢ፟ၹٕ

͔ᑜ(inference mechanism)Ă၆ٺኑᗔٕᙱͽϡځቁᇴጯր

௚ೡࢗ۞યᗟĂਕͽۡᛇ׶གྷរࠎૄᖂ۞ଠטĂಶΞᒔ଀

։р۞ଠטड़ڍĄι̙֭ᅮࢋኑᗔ۞ᇴጯሀݭĂӀϡᖎಏ

۞ᄬຍّଠטڱ݋Ăಶਕ྿ј็௚ଠט۞ΑਕજүĄϫ݈

̏ѣ̙͌۞ય͵யݡĂ͍ͽछ࿪ϡݡĂు႙ଳϡሀቘଠט

ֽฟ൴ր௚Ąሀቘଠט̝ሀቘ̼ߏ˘჌ԯ៍ീ۞Ꮾˢ۩ม ၆ᑕזͽቁؠኢા۞ሀቘะЪ͘ᜈĂ఼૱ߏдሀቘଠטᑕ ϡ۞ቑಛ̰Ą៍ീ۞ᇴፂᔵ൒̂ొ̶ౌߏځቁࣃĂҭߏሀ ቘଠט۞ᇴፂგநߏૄٺሀቘะЪநኢซҖ۞ĂЯѩሀቘ

̼఍நߏυࢋ۞ՎូĄሀቘ̼࿅඀Βӣͽ˭ೀ࣎ՎូĈ(1) פ଀ᏮˢតᇴࣃĄ(2)૟Ꮾˢࣃү͎ޘߍड(scale mapping) Ҍኢાቑಛ̰Ă఼૱ֹϡϒఢ̼ሀቘะЪĄ(3)૟ߍड࿅ޢ

۞Ꮾˢࣃ၆ᑕዋ༊۞ሀቘኢાĂӀϡሀቘבᇴ૟Ꮾˢ۞ྤ

फ़ᖼјዋ༊۞ᄬຍࣃͽֻሀቘଯநྻზֹϡĄ 7. ሀቘଯኢ

ሀቘଯኢ(fuzzy inference)˜ॲፂۢᙊऱ̝ሀቘఢ݋

(rules)˭ซҖሀቘநኢ۞ЪјྻზĂߏፋ࣎ሀቘଠטጡ۞

८͕ĄιߏֶፂܕҬଯந(approximate reasoning)۞ໄه൴ ण΍ֽ۞Ăྵ็௚ଯኢ۞ᚑϒଯந(exact reasoning)ՀЪந

˵Հ׍ᇅّĄٙᏜܕҬଯኢڱĂಶߏАԯఢ݋ऱ྆ٙѣ۞

ఢ݋ĂͽሀቘᙯܼR ܑϯĂГԯ R ᄃְ၁ A'ઇሀቘᙯܼ۞

Ъјྻზ଀זඕኢ B'Ąሀቘఢ݋Ъјଯኢ(compositional rule of inference)˜ॲፂ Zadeh д 1975 ѐ೩΍۞݈Шଯኢ

͞ڱĄЪјଯኢ۞જүΞϡ̳ё(4)ೡࢗĂ̳ё̚۞Œ΃ܑ

ሀቘᙯܼR ᄃᏮˢณ A'۞ЪјႊზĄϺӈĂЪјႊზ۞ඕ ڍߏགྷϤ̳ё(4)ଯኢĂࢍზ΍ඕኢ B ۞ᕩᛳޘu_B_'(v)Ą

( ) ( ,v)

' '

' A R A u u

B = o = o _A→_B

{

⁽ ⁾ ( )^,

}

{ ⁽ ⁾

[

⁽ ⁾ ⁽ ⁾

]

}

)

( ' _'

' v u u u uv u u u u u v

u _A _A _B

B n A A

B =∨n ∧ _→ =∨ ∧ →

(4)

∨ ࠎ maxć ∧ ࠎ minĄ˯ࢗሀቘᇴጯଯኢ͔ᑜࠎሀቘଠט ጡ۞ࢍზ८͕Ă׎ଯኢ͹ࢋჟৠдٺሀቘଠטጡᖣϤצଠ ၆෪ᕜפ۞៍ീࣃүሀቘ̼ᖼೱ̝ޢĂ็ᅍගሀቘଯኢ͔

ᑜྻϡܕҬଯநĂຩವሀቘۢᙊऱ̰ٙѣ௑Ъ୧І۞ఢ

݋Ă֭ࢍზజᛈજఢ݋۞ૻޘĂགྷᖼೱზ΍ఢ݋үЪјྻ

(5)

Min

(c) 1.0

0.0 Y

X1 is A12

X2 is A22 Y is B2

(b) 1

0

1 0

1

X2 0 Y

X1

Min

X₁ is A₁₁ X₂ is A₂₁ Y is B₁

(a)

IF THEN

1.0 0.0

1.0

X2 0.0 Y

X1

ဦ1 ሀቘଠטጡૄώߛၹ̝ଠטఢ݋

ზĄሀቘଯኢᇃھ۞ᑕϡٺЧ࣎၁ᅫր௚˯Ăυืᑕϡ૞

छགྷរٕ࠹ᙯۢᙊٙ଀۞ఢ݋Ăᖼ̼ࠎ“if…then”۞ఢ݋ԛ ёĄ҃˘ਠ૱ଳϡMamdani ۞౵̂౵̈۞ଯኢڱĂ˵ಶߏ ώ၁រֹϡଯኢڱĂтဦ1 ٙϯĂͽ׌࣎Ꮾˢă˟࣎ሀቘ ଠטఢ݋۞ሀቘଠטጡᄲځ׎ଯኢ࿅඀Ăͽ҂ᇋ׌࣎୧І

ୃࢗĄፋ࣎ଯኢ࿅඀тဦ1 ٙϯĈࢵАĂ૟ݡኳপّࣃϒ ఢ̼ઇࠎᏮˢĂགྷϤᕩᛳבᇴ̟ᄃሀቘ̼јዋ༊۞ᄬຍ ࣃĂ൒ޢ౅࿅ሀቘఢ݋ऱᄃሀቘଯኢ͔ᑜซҖЪјྻზĂ

౵ޢԯགྷ࿅ሀቘଯኢඕڍᖼೱј˘࣎ځቁ۞Ꮾ΍ᇴࣃ (MPCI)Ąဦ 1 ᑕϡଯኢ̳ё(4)ͽ if-then ԛё۞ఢ݋ё(5)

ܑனĂϡͽܑ྿ր௚۞ᏮˢᄃᏮ΍̝ม۞ଯኢᙯܼĂ׎ܑ

ϯт˭Ĉ

R₁Ĉif x is ₁ A and ₁₁ x is ₂ A then y is ₂₁ B ₁ R2Ĉif x is ₁ A₁₂ and x is ₂ A then y is ₂₂ B (5) ₂

׎̚A₁_iăA₂_i׶B ۞࠹၆ᑕᕩᛳבᇴࠎ_i µ_A₁_iăµ_A₂_i ׶

Bi

µ Ąώ͛ଳϡ Mamdani ۞ max-min ڱซҖሀቘଯኢྻ

ზĂ׎ሀቘଯኢᏮ΍ܑϯࠎĈ

)]}

( ), ( [ min { max )

(y _i₁ x₁ _i₂ x₂

i A A

i i

B µ µ

µ = , i =1, 2, …, M (6)

҃ྋሀቘ̼݋ߏͽࢦ͕ڱ(center of gravity)ֽՐ΍ሀቘଯ ኢඕڍ۞ౚᇆࢬ᎕ࢦ͕ొ̶Ăё(7)ࠎࢦ͕ڱ̝ᇴጯёĂӈ ߏ૟ሀቘଯኢᏮ΍µ_B_oᖼೱјࠎځቁᏮ΍ࣃ

∑

∑ ⋅

=

= k

i B i

k

i B i i

o y

y y

o o

1 1

) (

) ( µ µ

µ (7)

µoĄ׎̚_µ_B₍_y_i₎ܑϯy ᛳٺሀቘะЪ B ۞ᕩᛳࣃĂ_i µ0ࠎ кࢦݡኳّਕ޽ᇾ(MPCI)ࣃĄͽώ၁រٙүሀቘଠטጡన

ࢍ۞ሀቘଠטఢ݋ֽ࠻ĂᏮˢតᇴѣͪ໢ăͧࢦͽ̈́ PH

VS S M

wariahle PH

L VL

ဦ2 ͧࢦă໢ޘᄃ PH ᅕែޘ̝Ꮾˢሀቘבᇴ

T VS S SM M

variahle MPCJ

LM L VL H

ဦ3 ѷᙯᓑ޽ᇾᏮ΍̝ሀቘבᇴ

ࣃĂ҃Ꮾ΍តᇴࠎტЪّݡኳপّ޽ᇾ(MPCI)Ăఢ݋ؠཌྷ

ֶፂ૞छ̈́ԫఙಡӘޙᛉĄώ၁រՎូֶೈ݈ࢬ೩̈́ࣧ

நĂᖣϤሀቘଯநྻүĂА૟׌࣎ݡኳপّүࠎሀቘଠט ጡ̝ᏮˢᄬຍតᇴĂᏮ΍តᇴࠎ׌࣎ݡኳপّ̝Ꮚณ޽ᇾ (T-S)Ă૟ѩᏊณ޽ᇾГᄃௐˬ࣎ݡኳপّүࠎ˟Ѩሀቘଠ טጡ̝ᏮˢᄬຍតᇴĂᏮ΍តᇴࠎፋ࣎ݡኳপّ̝Ꮚณ޽

ᇾ(MPCI)Ąώ၁រଳϡୗԛᕩᛳבᇴซҖᄬຍតᇴ̝ሀቘ

̶౷тဦ2ăဦ 3 ٙϯĂ׎ˬᏮˢੈ̶ཱིҾπӮ̶੨ј̣

࣎ሀቘะЪ(fuzzy sets)ĂтĈޝ̈(VS)ă̈(S)ă̚(M)ă̂

(L)ăޝ̂(VL)ĂᏮ΍ੈཱི࠹၆ྵࠎ௟̝˝࣎ሀቘะЪĂтĈ ໂ̈(T)ăޝ̈(VS)ă̈(S)ă̈̚(SM)ă̚(M)ă̂̚(ML)ă

̂(L)ăޝ̂(VL)ăໂ̂(H)Ąሀቘଠטጡᖎဦтဦ 4 ٙϯĂ

׎̚FI ܑϯሀቘ̼ࠧࢬĂ̚ม͞๴ࠎଯኢ͔ᑜᄃۢᙊऱĂ ࠎώ၁រ८͕ĄDFI ܑϯྋሀቘ̼ࠧࢬĄώ၁រࠎ҂ณˬ

࣎Ꮾˢݡኳপّ۞˘࡭Ăֶݡኳপّઇϒఢ̼఍நĂѩన

ࢍଠטጡ่ࠎ25 ୧ሀቘఢ݋Ăֶဦ 1 ଯኢఢ݋ޙϲĄࢵА

૟໢ޘᄃͧࢦ׌࣎ݡኳপّགྷሀቘଠטጡଯኢ଀זT-S ଯ ኢඕڍĂГ̟ௐˬ࣎ݡኳপّPH ࣃᏮˢሀቘଠטጡГ˘

ѨүЪјଯኢĄώ၁រሀቘଯኢ͞ڱଳϡ౵̂౵̈ڱĂྋ ሀቘ۞͞ڱ݋ֹϡࢦ͕ڱፆүĄ

8. кࢦݡኳܫཱིᗔੈͧࢍზ

ॲፂ˯༼̝ኢࢗĂкࢦݡኳপّགྷ࿅ѷᙯᓑ̶ژᄃሀ ቘទᏭྻზඕڍĂֶݡኳ୕̂(larger-the-better)পّ۞ܫཱི

ᗔੈͧࢍზт˭Ĉ η =-10 [ 1]

∑1

= n

i i

Log γ ⁽⁸⁾

(6)

ܑ˟ ણᇴ௡Ъᇴፂᄃቁᄮ၁រ̝ϒఢ̼

ݡኳপّ ϒఢ̼

EXP A B C D E F G H ໢ޘ ͧࢦ PH ໢ޘ ͧࢦ PH 1 1 1 1 1 1 1 1 1 22.5 1.11 7.65 0.0541 0.3333 0.7222 2 1 1 2 2 2 2 2 2 25.5 1.15 7.85 0.8649 0.0000 0.9074 3 1 1 3 3 3 3 3 3 26 1.12 7.63 1.0000 0.2500 0.6852 4 1 2 1 1 2 2 3 3 24.5 1.08 7.72 0.5946 0.5833 0.8519 5 1 2 2 2 3 3 1 1 24.1 1.075 7.68 0.4865 0.6250 0.7778 6 1 2 3 3 1 1 2 2 24.9 1.05 7.58 0.7027 0.8333 0.5926 7 1 3 1 2 1 3 2 3 25.25 1.13 7.56 0.7973 0.1667 0.5556 8 1 3 2 3 2 1 3 1 25.5 1.095 7.6 0.8649 0.4583 0.6296 9 1 3 3 1 3 2 1 2 24.2 1.12 7.52 0.5135 0.2500 0.4815 10 2 1 1 3 3 2 2 1 28.14 1.022 8.06 0.4216 0.9333 0.5185 11 2 1 2 1 1 3 3 2 25.55 1.024 8.26 0.8784 0.9500 0.1481 12 2 1 3 2 2 1 1 3 27.34 1.023 8.24 0.6378 0.9458 0.1852 13 2 2 1 2 3 1 3 2 22.3 1.025 8.18 0.0000 0.9458 0.2963 14 2 2 2 3 1 2 1 3 23.58 1.023 8.2 0.3459 0.9417 0.2593 15 2 2 3 1 2 3 2 1 24.67 1.025 7.98 0.6405 0.9458 0.6667 16 2 3 1 3 2 3 1 2 25.3 1.025 8.1 0.8108 0.9625 0.4444 17 2 3 2 1 3 1 2 3 24.83 1.025 8.34 0.6838 0.9542 0.0000 18 2 3 3 2 1 2 3 1 26.7 1.023 7.64 0.8108 0.9417 0.7037 19 2 2 3 2 2 3 3 1 25.86 1.046 7.71 0.9622 0.8667 0.8333

γiࠎѷᙯᓑ޽ᇾĂη ࠎкࢦݡኳܫཱིᗔੈͧĄᓁ҃֏̝Ă

̶ژкࢦݡኳّਕĂΪࢋܫཱིᗔੈͧດ̂Ăಶܑϯፋវݡ ኳّਕດрĄ

αă၁រඕڍᄃ੅ኢ 1. кࢦݡኳ̶ژ

ֶ໰ٙనͪயዳത໢ޘ̝ଠטЯ̄ᄃЧ჌ͪ໤˭ซ Җ18 ௡ྏរ̝ඕڍтܑ˟ٙϯĄॲፂણ҂͛ᚥᄃ૞छಡӘ [14]Ă଀ۢͪயዳത୧Іഇ୕ࣃ̝໢ޘࠎ 26ƨĂͧࢦࠎ 1.03ĂPH ࣃࠎ 7.8 ॡĂߏࠎ̈ͪቐዳത۞౵ָ୧ІĄᑕϡ

୕ϫপّ̝ഇ୕ࣃүϒఢ̼఍நĂពϯٺܑ˟ĂΞ൴னྏ

រ3 ̝ͪቐ̝໢ޘତܕநຐࣃĂ׎ᇴࣃତܕ 1Ă൒҃Т˘

ྏរീྏͧࢦᅈᗓϫᇾࣃĂ࠹၆г׎ᇴࣃಶྵତܕٺ 0Ă Ξଯീѩ୧І̙֭ዋЪٺٸዳᒖဩĄྏរ 18 ̝ͪቐ̝໢

ޘĂͧࢦăPH ࣃ࠰ޝତܕநຐࣃĂ׎ᇴࣃତܕٺ 1Ăពϯ གྷ࿅ሀቘଠטጡଯኢ̝ѷᙯᓑّਕ޽ᇾ(MPCI)ᇴࣃϺତ ܕٺ 1Ă׎ពϯܫཱིᗔੈͧѣྵ੼۞౥ШĄፋវݡኳّਕ

޽ᇾ΍னд18 ௡̚ࠎ౵ָ۰ពϯٺܑˬĄ൒҃Ăώ၁រགྷ ϣ˾͞ڱనࢍĂፋЪѷᙯᓑᄃሀቘט̶ژᒔ଀˘௡ણᇴ௡

ЪĂѩ௡Ъүቁᄮ၁រٺௐ19 ௡тܑˬĄдѩ௡Ъ̚ពϯ кࢦݡኳّਕᏊณ޽ᇾ౵ତܕ1Ăͷˬ࣎ݡኳপّ(໢ޘă

ͧࢦăPH ࣃ)ТॡତܕٺϫᇾࣃĂЯѩΞЪநଯҤ׎ፋវ ݡኳّਕͧ׎΁18 ௡ྏរࠎָĂдѩ௡Ъ୧І˭Ξޙၹᘦ ઉٸዳᒖဩĄ

ܑˬ! ણᇴ௡Ъ̝ѷᙯᓑ඀ޘᄃкࢦݡኳপّܫཱིᗔͧ

ϫᇾࣃᗓमࣃ ѷᙯᓑܼᇴ EXP ∆T ∆S ∆PH ξ1 ξ2 ξ3

S/N Ratio 1 0.9459 0.6667 0.2778 0.3458 0.4607 0.7059 -2.5026 2 0.1351 1.0000 0.0926 0.7872 0.3583 1.0000 -1.0846 3 0.0000 0.7500 0.3148 1.0000 0.4300 0.6667 -1.8310 4 0.4054 0.4167 0.1481 0.5522 0.5864 0.8889 -1.3787 5 0.5135 0.3750 0.2222 0.4933 0.6143 0.7742 -1.8776 6 0.2973 0.1667 0.4074 0.6271 0.8062 0.5854 -1.8177 7 0.2027 0.8333 0.4444 0.7115 0.4031 0.5581 -2.4565 8 0.1351 0.5417 0.3704 0.7872 0.5160 0.6154 -2.0412 9 0.4865 0.7500 0.5185 0.5068 0.4300 0.5106 -3.2975 10 0.5784 0.0667 0.4815 0.4637 0.9485 0.5333 -2.0482 11 0.1216 0.0500 0.8519 0.8043 0.9773 0.3692 -2.0482 12 0.3622 0.0542 0.8148 0.5799 0.9699 0.3810 -2.4109 13 1.0000 0.0542 0.7037 0.3333 0.9699 0.4211 -2.7084 14 0.6541 0.0583 0.7407 0.4333 0.9627 0.4068 -2.5104 15 0.3595 0.0542 0.3333 0.5818 0.9699 0.6486 -1.4935 16 0.1892 0.0375 0.5556 0.7255 1.0000 0.4898 -1.6304 17 0.3162 0.0458 1.0000 0.6126 0.9847 0.3288 -2.4565 18 0.1892 0.0583 0.2963 0.7255 0.9627 0.6857 -1.0846 19 0.0378 0.1333 0.1667 0.9296 0.8487 0.8571 -0.8672

Control Rule Sets Inference Mechanismand T

S PH

MPCI DFI

FI

ဦ4! ሀቘଠטጡ߹඀ဦ

2. Я̄ड़ᑕᄃ࿰ീ

ଂۡϹܑણᇴ௡Ъ၁រ̚Ҥࢍ̝ͅᑕဦᄃͅᑕܑĂ׎

࿅඀ϯຍπӮࣃड़ᑕĂΞͅߍЧଠטͪ໤ม࠹၆̝ࢦࢋ

ّĄҌٺЧણᇴͪ໤Ҥზтё(9)ࢍზĄᓝώ၁រ L18 ۡϹ

ܑ̚ણᇴB ࠎּĂB1ĂB2 ᄃ B3 ̝ѷᙯᓑ඀ޘࣃ̶ҾࠎĈ

) 6(

1

) 6(

1

) 6(

1

18 , 17 , 16 9

, 8 , 7 3

15 , 14 , 13 6

, 5 , 4 2

12 , 11 , 10 3

, 2 , 1 1

MPCI MPCI

MPCI

MPCI MPCI

MPCI

MPCI MPCI

MPCI

B B B

+

=

+

=

+

=

(9)

1

MPCI ࠎ B ણᇴௐ˘ͪ໤˭̝πӮࣃĂB MPCI₁_,₂_,₃ࠎB ણ ᇴˬ࣎ീྏ௡Т˘ͪ໤˭̝πӮࣃĄ҃B ણᇴ̝౵ָͪ໤

݋ͽௐˬͪ໤၆ᑕ̝ѷᙯᓑܫཱིᗔੈͧ޽ᇾࣃࠎ౵̂۰Ą Ϥ౵̂౵̈ଯኢඕڍΞۢĂΐ໢ഔᇴณણᇴड़ᑕ౵ព඾Ă

҃ͪኳҾણᇴड़ᑕࠎ౵̈Ă׎ซ˘Վពϯણᇴ၆ᄦ඀̝࠹

(7)

ܑα! Чણᇴͪ໤௡ЪܫཱིᗔࢰͧπӮࣃͅᑕࣃ A B C D E F G H

ͪ໤1 -2.03 -1.98 -2.12 -2.19 -2.07 -2.32 -2.37 -1.84

ͪ໤2 -1.98 -1.96 -2.00 -1.93 -1.67 -1.90 -1.89 -2.09

ͪ໤3 -2.16 -1.98 -1.97 -2.36 -1.88 -1.84 -2.17 ड़! ᑕ 0.05 0.19 0.13 0.25 0.69 0.43 0.52 0.33 ଵ! Щ 8 6 7 5 1 3 2 4

ܑ̣ តளᇴ̶ژܑ

ଠט Я̄

តજ S

ҋϤޘ f

តள V

តளͧ

F

੒ᚥޘ P(%) A 0.0115 1 0.01147 0.013 0.20 B 0.1387 2 0.06935 0.078 2.39 C 0.0627 2 0.03136 0.035 1.08 D 0.2315 2 0.11574 0.130 3.99 E 1.4654 2 0.73267 0.823 25.28 F 0.7324 2 0.36618 0.411 12.63 G 1.0091 2 0.50452 0.567 17.41 H 0.3646 2 0.18231 0.206 6.29

ᄱमe 1.78086 2 0.89042 30.72

ᓁ׶T 5.796694 17 100.00

၆ࢦࢋّĂड़ᑕດ̂΃ܑ၆யݡݡኳດѣᇆᜩ˧Ą၁រે

ҖඕڍтܑαٙϯĂព඾ᇆᜩώ၁រݡኳّਕ̝ણᇴࠎĈ ΐ໢ഔᇴณĂΐሤॡมĂᕭጡ྅ཉҜཉĂᄃͪ஄፩ޘĄ࠹

ྵ̝˭ĂͪኳҾĂΐሤጡᏮ΍ΑதĂΐሤ࿪߹Ăᄃͪቐវ

᎕ඈ̝ᇆᜩ඀ޘྵ̙ព඾ĄЧણᇴᇆᜩѷᙯᓑܫཱིᗔੈͧ

ࣃࢦࢋّ̪ืགྷតளᇴ̶ژᑭរĂͽ˭ᖎࢋᄲځĄ 3. តளᇴ̶ژ(ANOVA)

តளᇴ̶ژ͹ࢋࠎ߄Ᏼᇆᜩᄦ඀̝ࢦࢋણᇴĂᔵ൒ଂ

ͅᑕဦΞ࠻΍ѷᙯᓑܫཱིᗔੈͧ޽ᇾ౵ָͪ໤௡ЪĂҭѩ ड़ᑕ֭൑ڱՙؠЧણᇴ၆ፋវݡኳّਕ̝ᇆᜩ඀ޘĂ҃υ

ืᖣϤតளᇴ̶ژܑүᅃӄҿᕝĄតளᇴ̶ژܑ௡јΒӣ ᓁតள(sum square of total)ăҋϤޘ(degree of freedom)ăត ளᇴ(variance)ăតளͧ(F-ratio)ă৷តள(pure variation)ᄃ

੒ᚥத(contribution rate)ඈтܑ̣ٙϯĄ׎̚ EăFăG ᄃ H ۞តளͧࣃྵ׎΁ଠטЯ̄ࠎ੼Ă׎࠹၆ፋ࣎ݡኳّਕ

۞੒ᚥޘ̶Ҿࠎ25.28%ă12.63%ă17.41%̈́ 6.29%ĄϤѩ Ξଯᕝѩα۰ࠎᇆᜩѷᙯᓑܫཱིᗔੈͧ޽ᇾ̝ព඾Я̄Ă ၆ፋវតளྋᛖ̂ࡗҫ62%ĄЯѩдү౵ָ̼˭̝ѷᙯᓑ ܫཱིᗔੈͧ޽ᇾ۞ҤࢍॡĂ่҂ᇋព඾Я̄۞ड़ᑕĂ଺ୢ

׎ዶЯ̄ĄགྷϤតளᇴ̶ژ̶ܑژٙ଀ՀΐОᙋͅᑕဦड़ ᑕᄃ̝˘࡭Ą

4. ቁᄮ၁រ

ଂЯ̄ड़ᑕͅᑕဦ̈́តளᇴܑΞۢкࢦݡኳপّ̝

ੈཱིᗔੈͧ޽ᇾࠎ౵ዋͪ໤௡ЪࠎA2ăB2ăC3ăD2ăE2ă

-1.50 -1.70 -1.90 -2.10 -2.30 -2.50

A1 A2 B1 B2 B3 C1 C2 C3 D1 D2 D3 E1 E2 E3 F1 F2 F3 G1 G2 G3 H1 H2 H3

ဦ5 Чણᇴͪ໤௡ЪੈཱིᗔࢰͧπӮࣃͅᑕဦ

12 34 56 78 109 1112 1314 1516 1718 1920 2122 2324 25

T=0.93 S=0.849 T-S=0.853

0 1 0 1 0 1

ဦ6 ዋ̼ણᇴ௡Ъ˭̝׌ݡኳপّ໢ޘᄃͧࢦᏮˢሀቘ ଠטጡ̝ଯኢඕڍ

F3ăG3ăH1Ăӈͪኳࠎ୶ͪĂᏮ΍Αதࠎ 300WĂΐሤ࿪

߹ࠎ2.727AĂͪ୉ቐ̝វ᎕ࠎ 75×45×45cm³Ă׌࣎ΐ໢ഔĂ 15 ̶ᛗΐ໢ॡมĂᕭጡ྅ཉҜཉࠎ˭ొ̈́ͪኳืྵ̈஄፩ ޘĄ׎̚ͽΐ໢ഔᇴณĂΐሤॡมĂᕭጡ྅ཉҜཉᄃͪኳ

஄፩ޘࠎᇆᜩώ၁រ౵ࢦࢋЯ̄ĄѩγĂ၆ዋ̼ણᇴ௡Ъ үቁᄮ၁រࠎௐ19 ௡Ă૟׌࣎ݡኳপّࣃ˘໢ޘᄃͧࢦĂ གྷϒఢ̼఍நޢүሀቘଠטጡЪјྻზĂዋ̼̝໢ޘࣃᛈ

൴ଠטఢ16,17,18, 19,20,21,22,23,24,25 ඈ˩୧ఢ݋Ă҃ͧ

ࢦࣃᛈ൴ௐ4,5,9,10,14, 15,19,20,24,25 ̝˩୧ఢ݋Ăགྷ࿅ሀ ቘଠטጡଯኢᄃЪјྻზޢĂ่யϠௐ19,20, 24,25 ୧ඈα ୧ఢ݋ĂӔனٺဦ6 Π͞ຳҒొЊĂГགྷྋሀቘ̼ᒔ଀ T-S ࣃдဦ6 Π˭֎ࡓҒؠҜᕇҜཉĂ׎ࣃࠎ 0.853ĄТநĂГ

૟T-S ᄃ PH ࣃࢦᖬௐ˘ѨሀቘଠטጡଯኢĂЪјྻზޢĂ ྋሀቘ̼ᒔ଀ѷᙯᓑܫཱིᗔੈͧ޽ᇾࠎ0.819Ăтဦ 7Ąགྷ Ϥܑˬ̝౵ዋͪ໤̝ௐ19 ௡࠹ྵٺ 18 ௡ྏរ۞ͧྵĂΞ

̂࡭ଯᕝ౵ዋͪ໤ቁᄮ၁រ̝кࢦݡኳপّѷᙯᓑ޽ᇾͧ

׎΁18 ௡ྏរՀତܕٺ 1ĂтܑˬٙϯĄଂͽ˯ኢࢗΞᙋ ၁ĂඕЪѷᙯᓑޘ̶ژᄃሀቘଠט̝ଯኢĂቁ၁Ξ྿ז࿰

ഇ۞ड़ڍĄ׎ඕڍ˵ܑனٺ౵ዋͪ໤ቁᄮ၁រ̝࣎Ҿݡኳ পّĂ׎໢ޘăͧࢦăPH ᅕែࣃ̶Ҿࠎ 25.86ƨă1.046ă 7.71Ąᖣѩ៍၅׎ᄃЧݡኳϫᇾࣃᄱमѺ̶̶ͧҾࠎĈ 0.5384%ă1.5534%׶ 1.1538%ĄЧݡኳᄱमѺ̶ͧӮҲٺ 2%ĂϺពϯ׎ତܕٺϫᇾࣃĂՀЪந۞ᙋځѷᙯᓑሀቘሀ ݭᑕϡٺ̈ݭͪ୉ᒖဩкࢦݡኳপّր௚̝јड़ᅲࠎព

඾Ą

(8)

12 34 56 78 109 1112 1314 1516 1718 1920 2122 2324 25

T-S=0.853 PH=0.857 MPCI=0.819

0 1 0 1 0 1

ဦ7 ዋ̼ણᇴ௡Ъ˭̝ T-S ᄃ PH ࣃགྷሀቘଠטጡ̝ଯ ଯኢඕڍ

̣ăඕ ኢ

ώ၁រᑕϡѷᙯᓑ̶ژᄃሀቘଠטጡሀᑢ੨Ъዳത ᒖဩᅮՐĂϡ̈ͪ୉ቐ̝ዳതᒖဩ̝၁រವՐٸዳᒖဩ̝

౵ָଠטણᇴĄΩγĂͽϣ˾͞ڱࠎૄᖂĂ̙ҭΞ଀ז౵

ָ੨ཉĂ֭ਕ༼࠷నࢍ۞ॡมĂᆧΐड़தĂֹዳതᒖဩΞ ͽ଀זᐹ̼Ąώ၁រ଀זඕڍт˭Ĉ

1. ώ၁រ̝ᆇጡࠎጯϠ૞ᗟᄦүనࢍ̝໢ޘଠטጡĂ੨Ъ

ͪ୉ٸዳᒖဩ၁រĂͽϣ˾͞ڱࠎૄᖂΞү΍ߊགྷᑻͷ ड़த੼ր௚̶̼ژĄ

2. ͪ୉ᒖဩ̝၁រ̚Ăѣᙯкݡኳপّ޽ᇾ̝Чݡኳপّ

۞௡ЪᝋࢦĂώ၁រᑕϡሀቘଠטጡפ΃ͽـጴᖣ̍඀

र۞͹៍གྷរҿᕝĂ҃ͽ࠹၆ݡኳ̝ࢦࢋّĂүࠎ౵ָ

௡ЪĂቁܲ၁រ໤ቁޘĄ

3. གྷ࿅តளᇴ̶ܑژĂᇆᜩͪ୉ᒖဩ̝ព඾ଠטЯ̄ࠎΐ

໢ഔᇴณĂΐሤॡมĂᕭጡ྅ཉҜཉᄃͪ஄፩ޘĂ׎ᓁ

੒ᚥޘତܕ62%Ą

4. གྷϤѷᙯᓑޘ̶ژᄃሀቘଠט̝ଯኢᙋ၁Ξᒔ଀ໂָ

۞࿰ീјڍĄ׎ඕڍܑனٺ౵ዋͪ໤ቁᄮ၁រ̝кࢦݡ ኳপّ޽ᇾତܕٺ1Ă҃࣎Ҿݡኳপّ໢ޘĂͧࢦĂPH ࣃ̶ҾତܕϫᇾࣃĄ׎ᄃЧݡኳϫᇾࣃᄱमѺ̶̙ͧ෹

࿅2%Ą

ણ҂͛ᚥ

1. Elsayed, E. A., and Chen, A., “Optimal Levels of Process Parameters for Products with Multiple Characteristics,”

International Journal of Production Research, Vol. 31, No.

5, pp. 1117-1132 (1993).

2. Tong, L. I., Su, C. T., and Wang, C. H., “The Optimization of Multi-Response Problems in Taguchi Method,”

International Journal Quality and Reliability Manage- ment, Vol. 14, No. 4, pp. 367-380 (1997).

3. Kumar, P., Barua, P. B., and Gaindhar, J. L., “Quality

Optimization Through Taguchi’s Technique Utility and Concept,” Quality and Reliability Engineering Interna- tional, Vol. 16, No. 6, pp. 475-485 (2000).

4. Yuin, W., and Alan, W., Taguchi Methods for Robust Ddesign, ASME, USA (2000).

5. Tong, L. I., Su, C. T., and Wang, C. H., “The Optimization of Multi-Response Problems in the Taguchi Method,”

International Journal of Quality and Reliability Manage- ment, Vol. 14, No. 4, pp. 367-380 (1997).

6. Wu, F. C., “Optimization of Multiple Quality Charac- teristics Based on Reduction Percent of Taguchi’s Quality Loss,” International Journal of Advanced Manufacturing Technology, Vol. 20, No. 11, pp. 749-753 (2002).

7. ዒჸᐷĂĶѷҒր௚ૄώ͞ڱķĂර̚ந̍̂ጯ΍ۍۤĂ ڠپ(1987)Ą

8. Zadeh, L. A., “Fuzzy Set,’’ Information and Control, Vol.

8, No. 3, pp. 338-353 (1965).

9. Lee, C. C., “Fuzzy Logic Control Systems: Fuzzy Logical Controller-Part I,” IEEE Transactions on Systems, Man and Cybernetics, Vol. 20, No. 2, pp. 404-418 (1990).

10. Zadeh, L. A., “Outline of a New Approach to the Analysis of Complex Systems and Decision Processes,” IEEE Transactions on Systems, Man and Cybernetics, Vol. 3, No. 1, pp. 28-44 (1973).

11. Zimmermann, H. J., Fuzzy Set Theory and Its Applications, 2nd Edition. Kluwer Academic Publishers, USA (1991).

12. Lin, C. T., and Lee, C. S., Neural Fuzzy Systems: A Neuro-Fuzzy Synergism to Intelligent Systems, Prentice- Hall, Simon and Schuster, NJ, USA (1996).

13. Ross, T. J., Fuzzy Logic with Engineering Applications, McGraw-Hill, Inc, Singapore (1995).

14. ج̡ܲăڒϖ͈ĂĶૄᖂͪயજۏጯķĂ㝱ߐۤࣧϠᅦĂ

͟ώĂௐ15-29 ࢱ(1998)Ą

15. ҥ౎ЪϠă˯౎߶˘ࢡĂĶೈᒖͪ̍඀۞ᙯᔣԫఙķĂ

ͪய΍ۍۤĂέΔĂௐ103-159 ࢱ(1992)Ą

16. Lin, C. L., Lin, J. L., and Ko, T. C., “Optimisation of the EDM Process Based on the Orthogonal Array with Fuzzy Logic and Grey Relational Analysis Method,” Interna- tional Journal of Advanced Manufacturing Technology, Vol. 19, No. 4, pp. 271-277 (2002).

17. Tarng, Y. S., Yang, W. H., and Juang, S. C., “The Use of Fuzzy Logics in the Taguchi Method for the Optimization of the Submerged arc Welding Process,” International Journal of Advanced Manufacturing Technology, Vol. 16, No. 9, pp. 688-694 (2000).

18. Tong, L. I., and Su, C. T., “Optimizing Multi-Response Problems in the Taguchi method by Fuzzy Multiple

(9)

Attribute Decision Making,” Quality and Reliability Engineering International, Vol. 13, No. 1, pp. 25-34 (1997).

19. Lin, J. L., Wang, K. S., Yan, B. H., and Tarng, Y. S.,

“Optimization of the Electrical Discharge Machining Process Based on the Taguchi Method with Fuzzy Logics,” Journal of Materials Processing Technology, Vol.

102, No. 1, pp. 48-55 (2000).

20. Wang, J. T., and Jean, M. D., “Optimization of

Cobalt-Based Hardfacing in Carbon Steel Using the Fuzzy Analysis for the Robust Design,” International Journal of Advanced Manufacturing Technology, Accepted, March (2004).

2003 ѐ 12 ͡ 31 ͟! ќቇ 2004 ѐ 03 ͡ 31 ͟! ܐᆶ 2004 ѐ 08 ͡ 26 ͟! ኑᆶ 2004 ѐ 10 ͡ 18 ͟! ତצ