突发事件群体决策演化

Group Decision-making Evolution in Emergency

首先建立突发事件群体决策的Q学习模型,给出决策者的Q值表,以及智能体行动规则;然后基于Netlogo软件平台,编程实现突发事件群体决策的元胞自动机仿真系统;再对决策群体应激能力相同和不同两种情况研究其群体决策,模拟其演化过程。最后,分析这两种情况的仿真结果,并得出结论:对于一个应急决策团队而言,在决策者风险偏好不同的情况下,决策者应激能力差异越大,其群体决策集结的效果越差。

This paper first establishes a Q-Learning model of group decision-making in emergency,including the Q value table of decision-makers and the agent acting principle.Then based on the Netlogo software platform,it programs to develop a cellular automata simulation system on the group decision-making in emergency.In the next step,it studies the group decision-making based on the groups with same or different stress abilities and simulates its evolution process.Finally,the paper analyzes the simulation results of these two cases and draws a conclusion that for an emergency decisionmaking group,in the case where the decision-makers in it have different risk preferences,the greater the difference between the decision-maker exits,the worse the group decision-making aggregation will be.