Case-based reasoning is a kind of method that can guide people to probe and solve problems by experience of historical cases.So it more suitable for emergency decision making that needs rapid response to emergency in the case of high information scarcity.In view of a series of differences between unconventional emergency and conventional emergency,so it is very difficult to retrieve a case from emergency case retrieval system which is similar to the current unconventional emergency,which can lead to the failure of emergency case retrieval system.In order to solve this problem,this paper divided the unconventional emergency into several stages according to the time series.There was a particular sub-event at each stage.By using evidential reasoning as the main tool,it designed the mode for retrieving the emergency case matching with the multi-attribute utility to provide technical support for emergency decision making.Finally it gave a numerical example to demonstrate the availability and feasibility of the model.%案例推理是一种依靠历史案例经验来指导人们认识和解决问题的方法,比较适用于信息缺失却需要快速反应的应急决策领域.鉴于非常规突发事件与常规突发事件存在诸多差异,在应急案例库中很难检索到与当前所发生的非常规事件相似的案例,从而可能导致应急案例检索系统失效.为了解决这一难题,首先把非常规突发事件按照时间序列划分成若干阶段,对每一阶段的特定子事件即常规突发事件进行匹配与检索;然后以证据理论为主要工具,构建基于多属性效用匹配的应急案例检索模型并求解算法,为应急决策提供技术支持;最后的实例分析验证了所建模型的有效性和可行性.
展开▼
