考虑复杂情形的多方参与决策方法研究

With the increasing complexity of social management practice, the decision making problem of multi-party participation gradually presents some complex characteristics, which demands emerging methods to support decision making. Therefore, it is of great practical significance and frontier to explore the multi-party participation decision under complex situations. By combining the probability statistics, mathematical modelling, uncertainty decision making and intelligent optimization methods, etc., the methods and applications under various complex situations, e.g., the different criteria sets, various aspiration levels and diversified criteria values, are addressed. Following points will be highlighted in this project. (1) A method for screening alternatives with the different criteria is built. (2) The classification and mathematical formulation of various aspiration levels, and the method for constructing satisfactory functions with respect to various aspiration levels are proposed. (3) By analyzing the characteristics of different criteria sets, the integration of different criteria sets, integration of criteria weight and integration of satisfactory degree are investigated. Then the ranking method based on different criteria sets is established. (4) The proposed method is applied to the multiple criteria decision making problem of multi-party participation. A multi-party decision support prototype system is constructed. The accomplishments achieved in this project will provide new solutions for multi-party decision making and provide efficient technical support for solving complex group decision problems in the real world.

随着社会管理实践问题的日趋复杂，多方参与的决策问题逐渐呈现出一些复杂特征，需要新的方法提供决策支持，因此对多种复杂情形下的多方参与决策问题进行研究具有重要的现实意义和前沿性。本项目将基于概率统计、数学建模、不确定决策和智能优化等方法，研究在多方参与决策者具有差异化指标集、多种指标期望水平和指标值形式多样化等多种复杂情形下的决策方法与应用。具体研究包括：（1）考虑差异化指标集的备选方案筛选方法；（2）多种指标期望水平的分类和数学描述，针对不同类型期望水平的满意度函数构建方法；（3）通过分析多方考虑差异化指标集的特点，提出差异化指标的集成、权重的集成和满意度信息的集成方法，并构建基于差异化指标集的方案排序方法；（4）将所提方法应用于现实多方参与的多指标决策问题，并构建多方参与决策支持原型系统。本项目的研究成果将为多方参与的决策问题提供新的解决方案，为解决现实中复杂群决策问题提供有效的技术支持。

随着社会管理实践问题的日趋复杂，多方参与的决策问题逐渐呈现出一些复杂特征，需要新的方法提供决策支持，因此对多种复杂情形下的多方参与决策问题进行研究具有重要的现实意义和前沿性。本项目基于概率统计、数学建模、不确定决策和智能优化等方法，研究了在多方参与决策者具有差异化指标集、多种指标期望水平和指标值形式多样化等多种复杂情形下的决策方法与应用。具体研究包括：（1）多种指标期望水平的分类和数学描述，针对不同类型期望水平的满意度函数构建方法；（2）通过分析多方考虑差异化指标集的特点，提出差异化指标的集成、权重的集成和满意度信息的集成方法，并构建基于差异化指标集的方案排序方法；（3）考虑到决策者的心理行为，研究了基于在线评论的多方参与产品选择方法；（4）将所提方法应用于现实多方参与的多指标决策问题。本项目的研究成果为多方参与的决策问题提供新的解决方案，为解决现实中复杂群决策问题提供了有效的技术支持。