切换拓扑下测量受限多智能体系统一致性迭代学习控制

Iterative learning control for consensus of measurement-constrained multi-agent systems under switching topology

中文关键词  多智能体系统  输出一致性  测量受限  迭代学习控制  切换拓扑

英文关键词  multi-agent systems  output consensus  measurement constraint  iterative learning control  switching topology

中文摘要

      针对通信拓扑至少含有一个沿迭代轴的联合生成树且同时沿有限时间轴和无限迭代轴切换的情况, 文本研究了存在测量受限的连续线性多智能体系统输出一致性迭代学习控制问题. 首先, 文章采用迭代学习控制方法设 计了一种基于跟随者局部信息的分布式输出一致性协议. 然后, 给出了系统可解输出一致性问题的两个充分性条件, 其中之一可使跟随者实时获取迭代学习增益, 避免了全局信息对学习增益设计的影响, 且保证了算法的分布式 实现. 接着, 利用λ范数理论和圆盘定理严格证明了所设计算法的收敛性. 最后, 通过实例仿真验证了所得结论的有效性.

英文摘要

      Aiming at the case that the communication topology contains at least one joint spanning tree along the iteration axis and simultaneously switches along the finite time axis and the infinite iteration axis, the output consensus problem of continuous linear multi-agent systems with measurement constraint based on the iterative learning control is studied. Firstly, a distributed output consensus protocol based on the local information available to the follower is designed by using iterative learning control method. Then two sufficient conditions for the solvable output consensus problem of the system are given, one of which can make the follower obtain the iterative learning gain in real time, avoid the influence of global information on the design of learning gain, and ensure the distributed implementation of the algorithm. Next, the convergence of the designed algorithm is strictly proved by using the λ norm theory and the disk theorem. Finally, the validity of the conclusions is verified by an example simulation.