面向区域绿色发展水平评价的复杂异构信息多属性决策方法研究

Heterogeneous information has great advantages in expressing the uncertain preferences of decision makers. How to effectively deal with and solve the practical multi-attribute decision-making problems in which the decision data is heterogeneous information is an important and urgent topic to be studied. This project will take the regional green development level evaluation as the research background, and deeply analyzes the key factors that influence the regional green development level and its action mechanism. Accordingly, the hierarchical clustering algorithm based heterogeneous multi-attribute decision-making method and the heterogeneous analytic network process are proposed from the perspectives of index selection and index weighting, respectively. Then, from the perspective of methods of regional green development level evaluation, the heterogeneous dynamic multi-attribute decision-making method with time span is put forward creatively, and the heterogeneous multi-attribute decision-making method based on missing data estimation model is proposed. Also, a heterogeneous behavior multi-attribute decision-making method with consideration of the reference point or the expectation of decision makers is developed, and a heterogeneous priority multi-attribute decision-making method and a heterogeneous interactive multi-attribute decision-making method based on the cooperation mechanism between the relevant parties and decision-makers are proposed, respectively. The multi-attribute decision-making method system which is suitable for the evaluation of regional green development level is then constructed systematically. Finally, the above research results are applied to evaluate the regional green development level in Jiangxi province, to find out the shortcomings of Jiangxi green development, and to put forward a new green development model and policy system with Jiangxi regional characteristics.

复杂异构信息在表达决策者的不确定性偏好时具有极大的优势，如何有效地处理和解决决策数据为复杂异构信息的现实多属性决策问题，是一项重要且亟需研究的课题。本项目将以区域绿色发展水平评价为研究背景，深入分析影响区域绿色发展水平的关键因素及其作用机理；从指标筛选和指标赋权的视角分别提出基于层次聚类算法的复杂异构多属性决策方法和复杂异构网络层次分析法；从区域绿色发展水平评价方法视角创新性地提出带有时间跨度的复杂异构动态多属性决策方法、基于缺失数据推算模型的复杂异构多属性决策方法、具有参照点/预期的复杂异构行为多属性决策方法、基于复杂异构信息的优先多属性决策方法以及基于相关方与决策者协作机制的复杂异构交互式多属性决策方法，系统地构建适合区域绿色发展水平评价的多属性决策方法体系。最后，将上述研究成果应用于评价江西绿色发展水平，找准江西绿色发展中的短板，提出具有江西区域特色的绿色发展新模式与政策体系。

异构信息在描述具有多种数据形式的复杂不确定决策与评价问题方面具有独特优势。本项目提出了基于信息量模型的异构信息融合方法、基于概率语言模型的定性异构信息归一化方法，给出了概率语言信息的新排序方法和距离测度，提出了区间广义正交模糊数的排序方法、融合算子和距离测度等，充实了复杂异构信息理论。同时，本项目提出了基于概率语言偏差模型的复杂异构多专家多属性决策方法、基于优势度模型的复杂异构多属性决策方法、基于公理化设计的复杂异构多属性群决策方法、基于激励机制的区间广义正交模糊动态综合评价方法、区间广义正交模糊PSI-COPRAS多属性群决策方法、区间广义正交模糊QUALIFLEX多属性群决策方法、改进型的广义正交梯形模糊TODIM群决策方法、基于双目标梯形模糊规划模型的决策方法、基于复杂网络和语言信息的交互式大规模群体评价方法和基于共识度融合模型的组合评价方法。最后，本项目对江西绿色发展水平进行了初步的探索，构建了江西绿色发展评价指标体系，从省、市层面分别评估了江西绿色发展状况。