Abstract: Real-time data of emergencies are the basis of emergency decision making. Better capacity of data handling, exacting entropy weights of emergency attributes, and filtering critical attributes which reflect the development trend of emergencies are keys to improving efficiency and accuracy of decision making. First, this paper shows a way that uses common knowledge model to structurally represent emergency and its attributes. Then the monitoring data of emergency attributes are converted to interval numbers with the method of interval multi-attribute decision making referred. Consequently, dimension of the data decreases while this way keeps temporal order information. Besides, the emergency attributes are transformed into cost-type attributes by means of defining the distance of data and interval, and entropy weights of emergency attributes are calculated by utilizing the method of interval multi-attribute decision-making based on entropy weight. The higher the entropy weight, the more the information which contains the trend of emergency, and this kind of attributes is more important. Finally, an example is given to illustrate effectiveness and practicality of the method.

Key words: emergency, common knowledge model, interval multiple attribute decision making, entropy weight

摘要： 突发事件的实时数据是应急决策的依据,提高对数据的处理能力,确定突发事件属性的熵权,筛选出反映事件发展趋势的重要属性,是提高决策效率和准确性的关键问题。利用共性知识模型结构化表示突发事件和属性,参考区间型多属性决策方法,把事件属性监测值转换为区间数型,在保持时序信息的同时降低数据维数,并通过定义精确数与区间的距离,使突发事件属性集均转化为成本型属性;继而利用基于熵权的区间型多属性决策方法计算事件的属性熵权值,权值越大,表示包含事件演化趋势的信息越多,在决策时就应被重点关注。最后,通过实例说明此方法的有效性和实用性。

关键词: 突发事件, 共性知识模型, 区间型多属性决策, 熵权