基于复杂偏好大数据分析的风险性动态大群体应急决策模型研究

In the background of outsize emergent event, its particularity makes the response decision-mkaing become dynamic interaction process under high time pressure. The involved decision-making group has the large-scale, heterogeneity, high complexity and dynamics, which determines the emergency decision-making is a high risk decision under complex data environment. In view of the above problems, from the of large data structure analysis of complex preference the project systematically study the system of risk dynamic large group emergency decision-making models. The specific research contents are as follows: (1) By use of the case of Tianjin port "8.12" outsize explosion fire etc., the characteristics of emergency decision-making complex preference large data and its distribution structure are analyzed. The analysis framework of risk large group emergency decision-making is set up; (2) The risk measure models of large group emergency decision-making are structured; (3) The risk state transfer model and the dynamic risk assessment method of coping process based on the analyzing for complex Preferences big data are designed. The dynamic risk eliminating models of large group emergency decision-making are constructed; (4) Oriented the emergency decision-making theme solving, the risk measure model and risk eliminating model are organically jointed to construct the risk eliminating coordination mechanism of large group emergency decision-making; (5) The risk large group emergency decision-making experimental platform based on large data analysis is developed. The simulating application is performed in the cases of Tianjin port super fire etc., which provides a scientific basis for the emergency management of super unexpected events.

以特大突发事件为背景，其特殊性使得应对决策成为高时间压力下动态交互过程，涉及的决策群体具有大规模性、异质性、高复杂性和动态性，决定了这种应急决策是一种复杂大数据环境下的高风险性决策，本项目针对上述问题从复杂偏好大数据结构分析入手，系统地研究风险性动态大群体应急决策模型体系。具体研究内容:(1)利用天津港“8•12”特大爆炸火灾等案例深入分析应急决策复杂偏好大数据特征和分布结构，建立风险性大群体应急决策分析框架；(2)构建基于复杂偏好大数据分析的大群体应急决策风险测度模型；(3)设计应对过程风险状态转移模型和动态风险评价方法，构建大群体应急决策动态风险消解模型；(4)围绕应急决策主题求解，将风险测度模型和风险消解模型有机结合构建大群体应急决策风险消解协调机制；(5)研发基于大数据分析的风险性大群体应急决策实验平台，以天津港特大火灾等案例大数据进行模拟应用，为特大突发事件应急管理提供科学依据。

本项目以天津港“8•12”特大爆炸火灾等重大突发事件为研究背景，该事件的特殊性使得应对决策成为高时间压力下动态交互过程，参与应急决策的群体应具有大规模性、异质性、高复杂性和动态性，决定了这种应急决策是一种复杂大数据环境下的高风险性决策，本项目针对上述问题从复杂偏好大数据结构分析入手，在大群体决策辅助支持技术、基于偏好冲突的大群体应急决策方法、大群体应急决策风险因素识别与表示、大群体应急决策风险测度建模及应用、大群体应急决策风险感知与演化、大数据环境下大群体风险性应急决策方法、多种不确定偏好混合形式大群体决策方法等主要方面获得较好的研究进展。项目发表学术论文41篇、其中SCI/SSCI期刊论文21篇、EI期刊论文7篇，在科学出版社出版学术专著1部，举办国际学术会议1次、参加国际学术会议2次，参与撰写政策建议2份。培养毕业博士研究生2人、毕业硕士研究生15人，为特大突发事件应急管理决策提供科学依据。