飞行冲突网络演化博弈中牵制机制的研究

Research on the Pinning Scheme of Evolutionary Game on the Aircraft Conflict Networks

为有效解决飞行器冲突网络演化博弈模型背叛诱惑值较大合作行为不理想的问题,构建了飞行器冲突场景的有向网络,在此基础上提出了基于总度的牵制机制和基于聚类系数的牵制机制。在构建的有向网络中分别挑选总度和聚类系数大的节点赋予合作策略,在网络演化博弈的规则下自组织学习,为飞行器产生合作(避让)或背叛(不避让)的策略。仿真分析结果表明,提出的两种方法与随机牵制机制和特定牵制机制比较,能够将合作频率维持在较高水平,以较少的控制代价获得更好的飞行冲突解脱结果。

The cooperative behavior is not ideal in evolutionary game on aircraft conflict networks while the temptation of defection is large.For the purpose of solving the problem,a directed network of an aircraft conflict scene has been built,and new pinning control schemes which are based on totaldegree and clustering coefficient have been proposed in this paper.The procedures are as follows:Firstly,a heterogeneous network in which the radar radiuses of the nodes follow the power-law distribution simulates the actual aircraft conflict network.Secondly,the nodes which are of the largesttotal-degree and clustering coefficient are selected respectivelyupon initialization in the two schemes.Finally,the network realizes self-organizing learning in Evolutionary prisoner’s dilemma game,and then each node obtains a corresponding tactics such as cooperation(collision prevention)or defection(not collision prevention).The pinning control scheme makes it possible that the network cooperative frequency maintains a higher level while the temptation of defection is large.Simulation experiment was conducted by applying this analyzing model,the result shows thatcompared with the randomly pinning scheme and the specially pinning scheme before,the two proposed methods control the cooperative frequency in a higher level,and resolve the flight conflict for less cost.