基于信息耦合度的群集系统自组织分群方法

针对“速度平均”协同机制不能表征群集系统应激分群运动的问题，基于信息熵定义融合邻居速度、距离、数量及自身感知半径的信息耦合度指标，提出一种“min-max”形式的速度协同策略，结合“近距排斥-远距吸引”的位置协同，实现群集系统的自组织应激分群运动。数值仿真分析表明，基于该速度协同机制的群集能够完成一种概率意义上的等规模分群，且其组群效率优于传统基于速度平均机制的群集。

Information coupling degree based approach for self-organized fission behavior in flocking system

For the problem that the traditional flocking approaches with “averaged” velocity coordination strategy can not realize the fission behavior from a coherent flock into multiple sub-groups under external conflicting stimulus, the information entropy is deployed to formulate the information coupling degree(ICD) index with the information of neighbor’s velocity, distance, number and its own perception range. Then, an ICD-based “min-max” velocity coordination strategy is established. Together with the “long range attraction/short range repulsion” position regulation method, self-organized fission behavior is achieved under external stimulus. Simulation results show that the flocking system under this motion law is able to achieve the fission behavior with equal size from a probabilistic perspective. In addition, this approach has better fusion performance than traditional flocking methods.