网络化多智能体主从式预测编队控制

为解决通讯延时对编队控制造成的不利影响,研究网络化多智能体在通讯延时情形下的主从式预测编队控制问题,提出一种主从式预测编队控制架构.在该架构中,所有智能体都基于延时状态预测自身当前时刻状态,用于主动补偿反馈通道延时.主智能体将自身未来预测状态发送给各从智能体,从而主动补偿主从智能体间的通讯延时.仿真结果验证了所提出主从式预测编队控制架构的可行性和灵活性.

Master-slave predictive formation control of networked multi-agents

To solve the negative effects caused by time-delay, the master-slave formation control problem of networked multi-agents with communication delay is considered,and a master-slave predictive formation control architecture is proposed. Each individual agent estimates its current state based on delayed states with the proposed architecture to compensate time-delay in the feedback channel actively. Master agent estimates its future state and transmits it to each slave agent to compensate the time-delay between master and slave agents actively. Simulations results show the feasibility and flexibility of the proposed architecture.