论文部分内容阅读
针对应急案例调整中存在不确定信息的问题,开发一种基于证据推理(ER)的应急方案调整方法。首先,计算目标案例与历史案例间的相似度,进而运用相似度阈值构建相似历史案例集。其次,定义方案属性的基本置信度评价等级,并通过集结规则得到目标案例的置信度分布形式。然后,根据期望效用值计算目标案例的方案属性值。最后,将该方法运用于重大交通事故应急响应实践中,得到重大交通事故的应急方案为出动人力135人,出动车辆[26,35]辆;当历史案例C21的出动人力改为缺失值时,得到出动人力为127人。这一结果表明,用该方法能够对存在不确定甚至无知信息的应急方案进行调整,而且案例调整精度也高于其他传统的案例调整方法。
Aiming at the problem of uncertain information in the adjustment of emergency cases, a contingency plan adjustment method based on evidence reasoning (ER) is developed. First, the similarity between the target case and the historical case is calculated, and then the similar historical case set is constructed by using the similarity threshold. Secondly, we define the basic level of confidence rating of program attributes and get the distribution of the confidence level of the target case through the aggregation rules. Then, calculate the scenario attribute value of the target case based on the expected utility value. Finally, this method is applied to the practice of emergency response to major accidents. The emergency plan to get major traffic accidents is 135 people dispatched, and vehicles [26,35] dispatched. When the dispatch manpower in historical case C21 is changed to missing value, Get mobilized to 127 people. This result shows that this method can be used to adjust contingency plans with uncertain or even ignorant information, and the accuracy of case adjustment is higher than other traditional case adjustment methods.