基于CBR的应急情报智能决策支持系统研究

Research on Emergency Intelligence Intelligent Decision Support System based on Case-based Reasoning

[目的/意义]针对应急情报智能决策支持系统(IDSS)自学习能力较差等问题,提出了一种基于案例推理的应急情报智能决策支持系统。[方法/过程]首先,分析了应急情报IDSS的特点和局限性;其次,分析了基于案例推理(CBR)系统和贝叶斯信念网(BBN)在应急情报智能决策中的应用;再次,使用人工神经网络(ANN)算法代替CBR系统中原有检索算法,使用BBN训练分类器,建立了基于ANN-BBN-CBR的应急情报IDSS模型;最后,给出了该系统的工作流程,并通过收集到的数据验证了该系统有效性和可行性。[结果/结论]该系统模型的使用不仅使大型案例库的检索和匹配速度及准确率得到了大幅度的提高,并且为案例相似性的度量提供了适当的统计信息,给案例库的组织管理也带来了方便,更为应急情报收集、整合、分析、传递、反馈等工作提供了研究思路。

[Purpose/Significance] Aiming at solving the problem of poor self-learning ability of emergency intelligence intelligent deci- sion support system （ IDSS）, an emergency intelligent IDSS based on case-based reasoning （CBR） is proposed. [ Method/Process] Firstly, the characteristics and limitations of the emergency intelligence IDSS are analyzed. Secondly, the application of CBR and Bayesian Be- lief Network（BBN） in emergency intelligence intelligent decision is analyzed. Thirdly, the Artificial Neural Network （ANN） algorithm is used to replace the retrieval algorithm in CBR system and the BBN trains model, and the model of emergency intelligence IDSS based on ANN-BBN-CBR is established. Finally, the working flow of the system is analyzed, and its effectiveness and feasibility are verified by the data collected. [ Result/Conclusion] The use of the system model not only makes the retrieval and matching speed and accuracy of the large case database greatly improved, provides the appropriate statistical information for the similarity measure, and brings the convenience to the organization and management of the case database, but also provides research ideas for emergency information work.