计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2015年
11期
22-26,30
,共6页
孙永河%段万春%许成磊%谢晖
孫永河%段萬春%許成磊%謝暉
손영하%단만춘%허성뢰%사휘
网络分析法%收益,机会,成本,风险%群组决策%投票排序
網絡分析法%收益,機會,成本,風險%群組決策%投票排序
망락분석법%수익,궤회,성본,풍험%군조결책%투표배서
analytic network process%Benefits,Opportunities,Costs and Risks(BOCR)%group decision%voting ranking
传统ANP-BOCR方法(即从收益(B)、机会(O)、成本(C)和风险(R)视角分别构造ANP(网络分析法)子网络,再将BOCR子网络下的方案评价值进行综合集成)被认为是一种处理复杂系统问题的有效方法。然而,一方面,该方法不仅尚未考虑隶属于不同BOCR子网络中元素之间的关联关系,而且在对BOCR方案评价值进行集成时也会陷入简单还原论的思维误区。另一方面,该方法集成群组专家意见时通常会损失部分专家的偏好判断信息,并硬性要求所有专家提供各方案的全偏好判断信息。为克服上述缺陷,通过构建复杂问题ANP-BOCR的新分析结构,提出基于DEA(数据包络分析)投票排序的ANP-BOCR群组决策新方法。新方法不仅实现了方法论和整体论的有机融合,而且还可保证群组专家信息集成过程中的信息无损。案例应用结果表明:新方法是行之有效的,有较强的实践应用推广价值。
傳統ANP-BOCR方法(即從收益(B)、機會(O)、成本(C)和風險(R)視角分彆構造ANP(網絡分析法)子網絡,再將BOCR子網絡下的方案評價值進行綜閤集成)被認為是一種處理複雜繫統問題的有效方法。然而,一方麵,該方法不僅尚未攷慮隸屬于不同BOCR子網絡中元素之間的關聯關繫,而且在對BOCR方案評價值進行集成時也會陷入簡單還原論的思維誤區。另一方麵,該方法集成群組專傢意見時通常會損失部分專傢的偏好判斷信息,併硬性要求所有專傢提供各方案的全偏好判斷信息。為剋服上述缺陷,通過構建複雜問題ANP-BOCR的新分析結構,提齣基于DEA(數據包絡分析)投票排序的ANP-BOCR群組決策新方法。新方法不僅實現瞭方法論和整體論的有機融閤,而且還可保證群組專傢信息集成過程中的信息無損。案例應用結果錶明:新方法是行之有效的,有較彊的實踐應用推廣價值。
전통ANP-BOCR방법(즉종수익(B)、궤회(O)、성본(C)화풍험(R)시각분별구조ANP(망락분석법)자망락,재장BOCR자망락하적방안평개치진행종합집성)피인위시일충처리복잡계통문제적유효방법。연이,일방면,해방법불부상미고필대속우불동BOCR자망락중원소지간적관련관계,이차재대BOCR방안평개치진행집성시야회함입간단환원론적사유오구。령일방면,해방법집성군조전가의견시통상회손실부분전가적편호판단신식,병경성요구소유전가제공각방안적전편호판단신식。위극복상술결함,통과구건복잡문제ANP-BOCR적신분석결구,제출기우DEA(수거포락분석)투표배서적ANP-BOCR군조결책신방법。신방법불부실현료방법론화정체론적유궤융합,이차환가보증군조전가신식집성과정중적신식무손。안례응용결과표명:신방법시행지유효적,유교강적실천응용추엄개치。
Traditional ANP-BOCR method(i.e., sub-networks of Analytic Network Process(ANP)are constructed from the viewpoints of Benefits, Opportunities, Costs and Risks(BOCR)merits, respectively. Then, the composite priorities of alternatives under BOCR are finally synthesized), is regarded as an effective approach to deal with the issues of complex systems. However, first, the complex relation between these elements lying different BOCR sub-networks is ignored in tra-ditional ANP-BOCR method. Furthermore, when the priorities of alternatives are synthesized in ANP-BOCR, simple reductionism thinking cannot be avoided. Second, when group decisions are used, the method would lose some expert preference information, and all information of alternatives is asked to give. To overcome the aforementioned drawbacks, a new group decision method of ANP-BOCR is presented by using Data Envelopment Analysis(DEA)to vote ranking, based on a new ANP-BOCR analysis structure. In the new method, reductionism and holism thinking is well integrated. Also, information lossless is realized. Finally, the new method is validated to be feasible and can be widely applied in the real-world.