基于双重服务的复杂系统决策流程算法与模型研究

项目名称： 基于双重服务的复杂系统决策流程算法与模型研究

项目编号： No.71301075

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 管理科学

项目作者： 刘健

作者单位： 南京理工大学

项目金额： 20.5万元

中文摘要： 针对现有决策支持系统在使用过程中人是系统服务被动的接受者，系统产生的结果与使用者理想中的决策方案大相径庭，以及系统的可靠性不强等问题，拟基于人是决策系统的服务对象及提升决策效能的关键要素的观点，提出基于双重服务（计算机提供的服务和人需求的服务）的复杂系统决策流程算法与模型研究，探讨融入人多种偏好的偏好技术模型的特点、描述方式、参数及参数体系，研究决策任务驱动的动态分解策略与求解流程及不同偏好决策群体的协调策略，解决偏好约束下的决策指标约简和决策对象推荐算法，利用过程挖掘技术提取决策流程与决策者社会关系网络，通过网络推理建立偏好模型的自学习自适应技术，进一步优化偏好模型，基于冲突分析原理研究基于群体满意度最大、整体最优的多级协调复杂信息集结模型，切实提高决策系统的人性化服务功能与决策可靠性，为大型复杂产品研制过程中最优质量管理控制方案的优选决策提供支持。

中文关键词： 双重服务；风险偏好；决策流程；信息集结；策略选择

英文摘要： Several problems exist in the current decision support system, including human always as the passive recipients during the system service, results produced by the system differing much from the user's ideal decision-making scheme, and the relatively weakness of the system's reliability, etc. To address these problems, as human are the service object of the decision-making system and to improve the performance for the key elements, we propose a study for decision-making process algorithm and model in complex system based on dual service (services provided by computers and required by human). We analyze the features, descriptions, and parameter system of the proposed model involving human preferences. Task-driven decision-making decomposition strategies,decsion process, and the coordination strategy in decision group with different preferences. Followed by a description of preference constrained decision-making index reduction and decision-making object recommendation algorithm. We then extract decision-making procedures and decision-makers social networking model by employing process mining techniques. To further optimize the preference model, network inference and reasoning are used to establish a self-learning adaptive approach for it. To effectively improve human service functionality and decision-making relia

英文关键词： Dual Services；Risk Preference；Decision-Making Process；Information Aggregation；Strategy Choosing