区间直觉联系指数势决策模型的研究

An interval multiple attribute decision-making model based on intuitionistic fuzzy sets and connection exponential potential

文章针对区间多属性决策问题指标和权重具有区间未确知特点,建立了区间多属性决策的直觉联系指数势模型。该模型基于直觉模糊集与联系数理论耦合方法,首先应用待选方案评价指标构成决策矩阵,通过规范化处理转换成直觉模糊评价矩阵,进而构建待选方案对最优方案的隶属度和非隶属度的线性规划求解模型,并应用单纯形法求解后转换为联系数,以计算联系指数势,确定最优方案。最后将模型应用于空调系统方案优选,以验证模型的可靠性和实用性。实例计算结果方案1~5的联系指数势分别为0.558 6、0.518 9、0.550 2、0.595 0、0.562 7,评价结果与MAEA法、可能度法和TOPSIS法的结果基本吻合,表明该模型应用于空调系统方案优选是有效可行的,能真实反映实际评价指标和权重的区间与未确知特点,并克服了传统区间多属性决策排序难的缺点,也为类似方案优选问题提供了新的参考。

Aiming at the interval and uncertainty characteristics of indicators and weights in multiple attribute decision-making problems,a decision-making model based on intuitionistic-connection exponential potential was presented.In this model,based on intuitionistic fuzzy sets and connection number theory,the decision matrix was set up by evaluation indicators.Then it was transformed into the evaluation matrix expressed in form of intuitionistic fuzzy sets by normalized processing.Next,linear programming models used to depict the relative closeness of the evaluation scheme and the optimal scheme were constructed.And the linear programming method was used to obtain the membership and non-membership degree.The membership and non-membership degree were transferred into the identity and contrary degrees of the connection number in order to determine the optimal scheme by connection exponential potential.Finally,the model was applied to the air conditioning system scheme optimization to verify the reliability and practicability of the model.Connection exponential potentials of evaluation schemes 1 5 of the case study were 0.558 6,0.518 9,0.550 2,0.595 0 and 0.562 7,respectively.The selected optimal scheme by the connection exponental potential method was consisted with that by the MAEA method,possibility degree method and TOPSIS method.Thus,the proposed model is effective and feasible in selecting optimal scheme of air conditioning system,which overcomes the sorting disadvantages of conventional interval multiple attribute decision-making,and can reflect real interval and uncertainty characteristics of evaluation indicators and weights.It provides a new reference for similar scheme optimization problems.