【正文】 System Engineering, National University of Singapore,10 Kent Ridge Crescent, g0202826*ABSTRACTOver the last few decades, a wide variety of multipleattributedecisionmaking (MADM) methods have been developed andreported in the literature. MADM methods provide a useful setof tools to select the best among alternatives. A difficultyfaced by MADM users is the multitude of different techniques,each with distinct advantages and disadvantages. Mostimportantly, different methods can yield different rankings ofalternatives and hence different decisions. Unfortunately, thiscrucial decision making problem which is itself a MADMproblem, had received little attention in the literature and thereis a need to provide a prehensive parison of MADMmethods with regard to the effectiveness and suitability of thetechniques. This paper presents a parative study on anumber of MADM methods based on outrankingmethod/ELECTRE 1. We look into five criteria which affectthe suitability of MADM methods. A sensitivity analysis wasalso carried out to determine the critical parameters.Key Words : Multiattribute decision making (MADM),Multiattribute utility theory (MAUT), Analytic HierarchyProcess (AHP), outranking, ELECTRE 11. IntroductionDecisionmaking plays an important role in a variety of fields,including psychology, business, engineering, and operationsresearch [18]. The main task of decisionmaking is to select oneor more alternatives from a given set. But the process ofselection is usually plicated by the existence of multiple,often conflicting criteria. For example, in the houseselectionproblem, cost and environment are conflicting criteria. Morefortable environment always leads to higher cost. It isdifficult to satisfy both of the criteria at the same time.Over the last few decades, a wide variety of multipleattributedecisionmaking (MADM) methods have been developed andreported in the literature[23] to help decision makers select thebest alternative. A difficulty faced by the MADM users is themultitude of different techniques, each with its distinctstrengths and weaknesses. Most importantly, the application ofdifferent techniques to a problem may lead to differentrankings of alternatives and hence different decisions.Therefore, it is most crucial to the decision makers that themost appropriate MADM method is applied.Unfortunately, this crucial decision making problem which isitself a MADM problem, had received little attention in theliterature and there is a need to provide a prehensiveparison of MADM methods with regard to theeffectiveness and suitability of the techniques.In this paper, we present a parative study on severalMADM methods based on the outranking process ofELECTRE 1[18]. Its purpose is to help decision makers toidentify the strengths and weaknesses of the various MADMmethods. The MADM methods pared in this study areAHP, Outranking/ELECTRE 1 and MAUT. Five criteria usedin this study are ability, simplicity, patibility, softwaresupport and nature of data. Our parison shows that simplermethods such as AH。 系 部：機(jī)械工程系 班 級：013021 姓 名：厲榮 學(xué) 號：01302108 附錄1：啟動窗口標(biāo)題顯示程序IF =0 THEN x1=1 ENDIFIF =1 AND =0 THEN x2=2 ENDIFIF =1 AND =0 THEN x3=3 ENDIFIF =1 AND =0 THEN x4=4 ENDIFIF =1 AND =0 THEN x5=5 ENDIFIF =1 AND =0 THEN x6=6 ENDIFIF =1 AND =0 THEN x7=7 ENDIFIF =1 AND =0 THEN x8=8 ENDIFIF =1 AND =0 THEN x9=9 ENDIFIF =1 AND =0 THEN x10=10 ENDIFIF =1 THEN x1=x1+1x2=x2+1x3=x3+1x4=x4+1x5=x5+1x6=x6+1x7=x7+1x8=x8+1x9=x9+1x10=x10+1 ENDIFIF x10 12 THEN x10=1x11=1 ENDIFIF x9 12 THEN x9=1 ENDIFIF x8 12 THEN x8=1 ENDIFIF x7 12 THEN x7=1 ENDIFIF x6 12 THEN x6=1 ENDIFIF x5 12 THEN x5=1 ENDIFIF x4 12 THEN x4=1 ENDIFIF x3 12 THEN x3=1 ENDIFIF x2 12 THEN x2=1 ENDIFIF x1 12 THEN x1=1 ENDIF附錄2：MCGS循環(huán)策略IF (I14=1 AND I10=0) OR M05=1 OR M06=1 OR M07=1 THEN Q02=1Q06=1 ELSEQ02=0Q06=0 ENDIFIF ( M00=0 AND I13=1 AND I10=0) OR ((M01=1 OR M02=1 OR M03=1) AND (I17=1 OR Q00=1) AND M04=1 AND M05=0 AND M06=0 AND M07=0) THEN Q00=1Q04=1 ELSEQ00=0Q04=0ENDIFIF I17=1 THEN M00=1ELSE M00=0ENDIFIF ((I15=0 AND I16=1) OR M01=1) AND M05=0 AND M02=0 AND M03=0 AND I12=0 THEN M01=1ELSE M01=0ENDIFIF ((I15=1 AND I16=0) OR M02=1) AND M06=0 AND M01=0 AND M03=0 AND I12=0 THEN M02=1ELSE M02=0ENDIFIF ((I15=1 AND I16=1) OR M03=1) AND M07=0 AND M01=0 AND M02=0 AND I12=0 THEN M03=1ELSE M03=0ENDIFIF ( I11=1 OR M04=1 ) AND I12=0 AND I10=1 THEN M04=1ELSE M04=0ENDIFIF ( M01=1 OR M05=1 ) AND I21=1 AND M10=0 AND M04=1 THEN M05=1ELSE M05=0ENDIFIF ( M02=1 OR M06=1 ) AND I22=1 AND M10=0 AND M04=1 THEN M06=1ELSE M06=0ENDIF IF ( M03=1 OR M07=1 ) AND I23=1 AND M10=0 AND M04=1 THEN M07=1ELSE M07=0ENDIF IF M05=1 THEN Q11=1ELSE Q11=0ENDIFIF M06=1 THEN Q12=1ELSE Q12=0ENDIFIF M07=1 THEN Q13=1ELSE Q13=0ENDIF IF ( M05=1 OR M13=1 ) AND I17=0 AND M04=1 THEN M13=1ELSE M13=0ENDIFIF ( M06=1 OR M14=1 ) AND I17=0 AND M04=1 THEN M14=1ELSE M14=0ENDIFIF ( M07=1 OR M15=1 ) AND I17=0 AND M04=1 THEN M15=1ELSE M15=0 ENDIFIF M10=1 AND （M01=1 OR M02=1 OR M03=1） THEN Q01=1 Q05=1ELSE Q01=0 Q05=0ENDIF附錄3：英文資料原文Proceedin