空地多智能体围捕系统体系结构设计与实现

Aerial and Ground Multi-agent Pursuit System Architecture Design and Implement

针对空地多智能体围捕问题,提出一种层次化的体系结构与任务实现方法。通过系统模态设计实现任务调度与分配,利用改进型人工势场法的路径规划和碰撞预测算法实现追捕者的自主导航,使用基于人工视觉的辅助定位方式完成惯性器件的修正。系统试验表明,该系统可稳定完成围捕任务。

With regard to the problem of aerial and ground pursuit-evasion game, an architecture system and task implement method is proposed. The task scheduling and assignment can be realized by using system mode design. The improved artifieial potential field and collision prediction arithmetic can be used in the pursuers＇ automatic navigation. The inertial device is modified by use of assistance location based on artificial vision. The system experiment indicates that the system can settle the pursuit problem.