面向复杂决策任务的群体决策研讨系统研究

【作者】 陈志阳；

【导师】 陈晓红；

【作者基本信息】 中南大学 ， 管理科学与工程， 2008， 硕士

【摘要】 综合集成研讨厅体系是中国科学家钱学森院士提出的处理开放复杂巨系统及其相关问题的方法论,是一个人机结合的巨型智能系统和问题求解系统,其核心是将专家群体、统计数据和信息资料、计算机技术这三者结合起来。综合集成研讨厅系统的体系结构、群体专家的意见综合、决策任务分解、满意群体的构建和决策方案的选优等都是综合集成研讨厅体系的关键问题。在分析综合集成研讨厅相关理论的基础上,根据课题需求,提出并设计了一种适合软件实施需求,实现人机结合的决策任务分解、决策群体匹配、并行求解、求解结果合成和层级驱动的面向复杂决策任务的群体决策研讨系统。详细阐述了该系统结构下的群体结构化与管理、决策任务结构化与求解、会议管理、研讨系统支撑、研讨流程控制、知识搜索与可视化等功能模块。引入工作分解结构对复杂决策任务进行线性树形层次分解,通过决策任务的分解有效地降低了决策任务的复杂度,提高群体决策研讨的效率。针对分解后的多任务提出了基于任务特征和群体人员特征的多模型匹配算法,提高了决策的速度和精度,并通过实例验证了算法的有效性。分析了群体决策研讨系统中的研讨过程及专家研讨信息的特点,总结并提出了一种结合人的定性分析能力和计算机定量数据处理能力的群体决策研讨过程,继而总结并提出决策方案的综合选优法,为群体决策研讨系统提供必要的方法和理论。最后,基于上述研究,部分实现了基于网络环境的群体决策研讨系统的功能。详细介绍了系统的开发技术和策略、运行环境、决策任务求解与研讨流程等功能的具体开发步骤。更多还原

【Abstract】 Professor Qian Xuesen proposed the Hall for Workshop of Metasynthetic Engineering (HWME), which is a new methodology to solve complex Problems related to the Open Complex Giant System(OCGS).The HWME is a giant intelligent decision-making and problem-solving supporting system, which is People oriented and man machine integrated. The main idea of HWME is combining expert groups, all sorts of data and information, and the intelligence of computer together. Several key techniques should be studied, such as the architecture of HWME, the method of synthesizing the divergent thinking of experts and organizing discussion information effectively and etc.After analyzing correlative theories of HWME, proposed and designed a hierarchy dirved and complex decision-making task oriented workshop system for group decision-making, which is fit for requirement of software implement, and can do many works, such as disassembling man and machine integrated decision-making task, matching decision-making group, solving concurrent problems and synthesizing result. The paper expatiated functional models of the system structure, includes group structuring and management, decision-making task structuring and solving, conference management, workshop system supporting, workshop procedure controlling, knowledge searching and visioning and so on.Introduced task disassembling structure to disassemble complex decision-making, doing so can reduce the complexity of decision-making task, and improve the efficiency of group decision-making workshop. The paper proposed the multimodal matching arithmetic that based on task character and group member character to solve disassembled tasks, improved the speed and precision of decision-making, and validated validity of the arithmetic by an example. Analyzed the characters of workshop process and experts workshop information of group decision-making workshop system, summarized and designed the group decision-making process that integrated qualitative analysis capability of people and quantificational data disposal capability of compute. Then summarized and proposed the synthesis selective preference method of alternative, provided indispensable method and theory for group decision-making workshop system.Based on above research, implemented function of group decision-making workshop system based on Weber circumstance partly. The paper expatiated technique, strategy and operating circumstance of system development, as well as the developing steps of decision-making task solving and workshop procedure model.更多还原