具有信道噪声的多智能体量化迭代学习控制  

Quantized Iterative Learning Control for Multi-agent System with Channel Noise

作　　者：李辰龙 方勇[1] 

机构地区：[1]上海先进通信与数据科学研究院,特种光纤与光接入网重点实验室,上海大学通信与信息工程学院,上海200444

出　　处：《工业控制计算机》2020年第5期23-25,共3页Industrial Control Computer

基　　金：国家自然科学基金资助(61673253)。

摘　　要：迭代学习控制是实现多智能体编队的有效技术手段之一。针对多智能体系统在同时受到信道噪声以及数据量化产生的量化误差的共同影响下,研究多智能体迭代学习控制编队实现问题。提出信道噪声和量化误差下的分布式迭代学习控制算法,并在范数意义上对算法进行收敛性分析,推导出收敛性充分条件。分析仿真结果可得,信道噪声和量化误差均会降低迭代学习控制算法的收敛速度,且两者同时作用时会加剧这一影响,但随着迭代次数的增加,轨迹误差渐近收敛,所提算法仍能实现多智能体编队。Iterative learning control(ILC)is one of the effective technical methods to achieve multi-agent formation.This paper studies ILC for multi-agent formation when the multi-agent system(MAS)is affected by channel noise and quantization error simultaneously.A distributed ILC algorithm under channel noise and quantization error is proposed.The algorithm is analyzed and sufficient condition is derived in this paper.Simulation results show that both channel noise and quantization error will reduce the convergence speed of the ILC algorithm.Nevertheless,as the number of iterations increases,the trajectory error asymptotically converges,and the proposed algorithm achieves multi-agent formation.

关 键 词：分布式迭代学习控制 分布式多智能体系统 信道噪声 量化误差 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]