计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2013年
18期
120-124
,共5页
前景理论%量子进化算法%区间直觉模糊数%多属性决策
前景理論%量子進化算法%區間直覺模糊數%多屬性決策
전경이론%양자진화산법%구간직각모호수%다속성결책
prospect theory%quantum evolution algorithm%interval intuitionistic fuzzy number%multi-attribute decision-making
针对属性权重不完全确定且属性偏好值为区间直觉模糊数的多属性决策问题,提出一种基于前景理论和量子进化算法的模糊多属性决策方法。该方法根据前景理论及模糊数距离公式,定义区间直觉模糊数的前景价值函数,同时将决策者对方案的风险偏好纳入决策行为中,以此来构建方案综合前景值最大化的非线性规划模型。通过引入量子进化算法,求解模型得出最优权重向量。最终根据方案前景值确定出方案的排序。该方法适用于模糊决策环境,能满足决策者不提供确定属性权重的要求,并充分考虑决策者风险心理因素对决策行为的影响,具有广泛的应用价值。数值算例说明了该方法的有效性和可行性。
針對屬性權重不完全確定且屬性偏好值為區間直覺模糊數的多屬性決策問題,提齣一種基于前景理論和量子進化算法的模糊多屬性決策方法。該方法根據前景理論及模糊數距離公式,定義區間直覺模糊數的前景價值函數,同時將決策者對方案的風險偏好納入決策行為中,以此來構建方案綜閤前景值最大化的非線性規劃模型。通過引入量子進化算法,求解模型得齣最優權重嚮量。最終根據方案前景值確定齣方案的排序。該方法適用于模糊決策環境,能滿足決策者不提供確定屬性權重的要求,併充分攷慮決策者風險心理因素對決策行為的影響,具有廣汎的應用價值。數值算例說明瞭該方法的有效性和可行性。
침대속성권중불완전학정차속성편호치위구간직각모호수적다속성결책문제,제출일충기우전경이론화양자진화산법적모호다속성결책방법。해방법근거전경이론급모호수거리공식,정의구간직각모호수적전경개치함수,동시장결책자대방안적풍험편호납입결책행위중,이차래구건방안종합전경치최대화적비선성규화모형。통과인입양자진화산법,구해모형득출최우권중향량。최종근거방안전경치학정출방안적배서。해방법괄용우모호결책배경,능만족결책자불제공학정속성권중적요구,병충분고필결책자풍험심리인소대결책행위적영향,구유엄범적응용개치。수치산례설명료해방법적유효성화가행성。
This paper presents a multi-attribute sorting method applied to problem whose information on weights is partly known and attribute values are interval intuitionistic fuzzy numbers. The method is based on prospect theory and quantum evolu-tion algorithm. According to prospect theory and fuzzy numbers’distance formula, it defines interval intuitionistic fuzzy numbers’ prospect value function. Meanwhile, it brings decision makers’risk preference of project into decision-making behavior. The nonlinear programming model based on projects’comprehensive prospect value maximization can be constructed. By putting quantum evolution algorthm, the optimal weight vector of model can be solved. The projects’comprehensive prospect values deter-mine the order. This method is applied to fuzzy decision making environment, can meet decision makers who do not provide deter-mining attribute weights, and fully consider risk psychological factor’s influence on decision behavior. Thus it has wide range of application value. At the end, a numerical example shows the effectiveness and feasibility of the method.