目标导向网络分析方法研究

为克服ANP及相关改进方法"相对于A和甲来比较甲和乙"的特殊判断模式、未能完整描述系统结构的系统分析观、武断随意的元素集自依赖矩阵构造方法和元素集加权方法、违背期望效用理论的不明晰决策分析框架几方面的缺陷与错误，首先在规范性决策框架下对ANP予以理论重构，为理性并坚持偏好绝对独立结构的决策者解决具有网络化系统结构的评价与决策问题，提出面向属性偏好绝对独立的目标导向网络分析方法(TOANP-API)。 在此基础上，以克服ANP在因素偏好依赖上存在的模糊认识为焦点，为理性并坚持偏好序依赖关联的决策者解决具有网络化系统结构的评价与决策问题，提出规范性的面向偏好序依赖的目标导向网络分析方法（TOANP-RPD）。最后，在TOANP-API和TOANP-RPD的基础上，通过纳入决策者选择行为个体差异，在规约性决策框架下，提出更能为决策者所接受的规约性目标导向网络分析方法（P-TOANP）。

The analytic network process (ANP) as well as its relevant improved methods has such defects/mistakes as special judgment mode of comparing A and b with respect to A and X, ill system-analysis-view of incompletely reflecting system relations, arbitrarily constructing and weighting self-dependence matrix of element component, and decision analysis framework against the expected utility theory. To solve these defects, the project first propose a new method of target-oriented analytic network process (TOANP) for rational decision makers with absolute preference independence (API), hereafter TOANP-API, under the normative decision framework. On the basis of TOANP,to solve the evaluation & decision problems with networked system-structure, this project further proposes another TOANP method for rational decision makers with ranked preference-dependence(RPD), hereafter TOANP-RPD, with the focus of overcoming the unclear understanding of ANP on component preference dependence,which is not solved in the TOANP-API. Finally, on the basis of TOANP-API and TOANP-RPD, this project proposes a prescriptive method of TOANP,hereafter P-TOANP, under the prescriptive decision-making framework. The distinguished advantage of the P-TOANP over the TOANP-API or the TOANP-RPD,lies in that the P-TOANP takes the individual-differences of decision-makers' choice-behaviour into consideration, and thus the P-TOANP can be more acceptable to decision makers than the TOANP-API and the TOANP-RPD.