基于Agent与元胞自动机的无人机集群混合式控制

高效有序的集群控制方式是集群顺利完成作战任务的前提。针对无人机集群控制问题，结合集中式与分布式2种控制方式，提出基于Agent与元胞自动机的集群混合式控制。从无人机集群作战流程出发构建了无人机集群控制体系框架、通信拓扑结构及集群控制规则，将集群个体由上至下分为中心长机、小组长机、个体无人机3个层次，高层级对低层级采用自上而下的集中式控制，同层级采用自下而上的分布式控制。在此基础上，利用Agent模型的层次性与元胞自动机模型的同质性，设计了基于Agent与元胞自动机的集群混合式控制模型，实现2种控制方式有效结合，元胞自动机模型实现集群基本的聚合、分离、速度一致规则，Agent模型实现不同层级个体间的协同交互规则。在编队集结与保持任务背景下，对分布式、集中式与混合式3种控制进行对比仿真，结果表明:基于混合式控制的集群在编队可控性、跟随性、一致性以及降低通信负载等方面具有明显优势，验证了混合式集群控制方法的有效性。 

Hybrid control for UAV swarms based on Agent and cellular automata

The efficient and orderly swarms control mode is the prerequisite for the swarms to successfully complete the combat mission. Aimed at the problem of UAV swarms control, a hybrid swarms control based on Agent and cellular automata is proposed by combining the centralized and distributed control modes. Based on the analysis of UAV swarms operation flow, the framework of UAV swarms control system, communication topology and swarms control rules are constructed. The swarms' individuals are divided from top to bottom into three levels: center leader, group leader, and individual UAV. The upper level adopts top-down centralized control to the lower level, and the same level adopts bottom-up distributed control. On this basis, using the hierarchy of Agent model and the homogeneity of cellular automata model, a swarms hybrid control model based on Agent and cellular automata is designed to realize the effective combination of the two control modes. The cellular automata model realizes the basic swarms rules of aggregation, separation and uniform speed, and the Agent model realizes the cooperative interaction rules among individuals at different levels. Under the background of formation assembly and maintenance task, three kinds of control: distributed, centralized and hybrid, are compared and simulated. The simulation results show that the swarms based on hybrid control have obvious advantages in formation controllability, following, consistency and reducing communication load, which verifies the effectiveness of the hybrid swarms control method.