关键词: 突发事件管理, 应急团队组建, 双阶段匹配, 正态云模型

Abstract: Recently, the emergencies deeply damaged the social development, which has attracted the attentions from the whole world. Scholars from diverse areas have studied how to deal with the emergency’s disasters, like the emergency supplies, emergency rescue, and post-disaster reconstruction. However, one of key issues, which means the building teams for emergency, has been neglected, which influences the performance of rescuing. Lots of examples have verified the importance of emergency response team (ERT), like the Fukushima nuclear meltdown in Japan and Australian bushfire. We would like to discuss the reasons for ERT’s importance. On one hand, because of the burstiness of emergency, the ERT temporarily forms without mature mechanism. On the other hand, generally, experts from different areas are required to deal with the emergence together. Different experts from diverse areas always have different opinions, and consensus reaching process in a short time is always difficult. These mentioned problems could be solved through selecting the team member. For the reasons, we could design some rules to select echelon with mature architecture and easily form the consensus. Obviously, how to determine the selecting rules are vital for ERT.
After full investigations, we propose two-layer selecting mechanism for ERT. For the proposed two-layer selecting mechanism, considering the importance of the leader, we select the leader in the first layer. For the process of selecting the leader, because each task has his own features, we select the leader according to the main feature of the emergency task. Also, considering the confusion of multiple leaders, each emergency task has only one leader. Certainly, in the first layer, we select one leader for one emergency task based on the key feature of task. Then, for the second layer, the objective of the second layer is serving the leader, and we should design the rules for cooperating. The first rule is that the major of ERT’s team member should be different, and the all majors of the whole team could meet the requirements of the emergency tasks. As the emergency task requires several experts from different areas to cooperate together,the leader only has the knowledge of key requirement. The remained majors should be finished by the ERT team member. For the team member of ERT, it is necessary to make up the disadvantages of the leader of ERT team. Certainly, the first rule is making up the deletion of majors. Next, the second rule is the complementarity of ability. To better finish the emergency task, the whole team should cooperate together, and the complementarity is important. The complementarity of ability means the leader’s weakness should be made up for by the team member. We could perform the rule through the differences between the value of different indexes. Then, the third rule is the consistency. We discuss the cooperation from the perspective of indexes, and the cooperation intention has not been discussed. The consistency indicates the cooperation intension. The scale, which is used to measure the cooperation intensions, is used to represent the cooperation ability of the whole team members. Also, the objective is maximizing the whole cooperation intensions. All in all, the second layer is aiming at selecting the team members, and the three rules, including the covering of majors, the complementarity of ability, and the cooperation intention. In these rules, the first two rules could be realized through the constraints, and the last rule could be realized by the objective or the constraint. Finally, after introducing the two-layer mechanism, how to deal with the mechanism and select the final teams is the important. Because the two layers are connected and the second layer could only be selected after determining the first layer, the two layers have precedence relationships. To better perform the relationship, the two-level programming is proposed to solve the problem. Also, considering the difficulties of solving the problem, the genetic algorithm is proposed to solve the two-level programming model.
To better verify the effectiveness of proposed method, the case study, indicating the aircraft fire, is used. The results of case study prove the advantages of our proposed methods.

Key words: emergency management, emergency team building, two-stage matching, cloud model