多智能体不确定性系统的自适应一致性问题研究

项目名称： 多智能体不确定性系统的自适应一致性问题研究

项目编号： No.61273183

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 俞辉

作者单位： 三峡大学

项目金额： 81万元

中文摘要： 多智能体系统的一致性问题是分布式多智能体协调控制的一个关键问题，在多机器人编队控制、传感器网络信息融合、通信网络拥塞控制等方面都有广泛应用。本项目研究具有未知非线性动态的多智能体不确定性系统的自适应一致性问题，通过对未知非线性动态的线性参数化，拟引入局部一致性误差，利用状态（输出）反馈与局部一致性误差反馈，设计分散式控制器，基于不同的网络拓扑研究：1）系统的自适应一致性与分散控制；2）系统的自适应部分状态一致性与分散控制；3）系统的自适应有限时间一致性、自适应有限时间部分状态一致性与分散控制；4）系统存在外部扰动时上述算法的鲁棒性，给出合理的误差界估计。其成果旨在探寻系统信息不完备情况下，系统渐近稳定性、有限时间稳定性、参数收敛及有限时间参数收敛的充分条件，揭示系统的复杂动力学行为，建立多智能体不确定性系统自适应一致性问题的基础理论。因此，本项目的研究具有重要的理论及实际意义。

中文关键词： 多智能体不确定性系统；一致性问题；自适应控制；有限时间稳定性；参数收敛性

英文摘要： Consensus problem of multi-agent systems is one of key problems of distributed cooperative control of multi-agent. Its broad applications include formation control of multi-robot, information fusion in sensor networks and congestion control in communication networks, etc. In this project, adaptive consensus problem of uncertain multi-agent systems with unknown nonlinear dynamics is studied. The decentralized controllers are designed using feedback of states (outputs) and local consensus errors by linear parameterizations of unknown nonlinear dynamics and introducing local consensus errors. Based on different network topologies, 1)adaptive consensus and decentralized control of systems; 2)adaptive partial states consensus and decentralized control of systems; 3)adaptive finite-time consensus, adaptive finite-time partial states consensus and decentralized control of systems; 4)robustness of the above algorithms in the case of existing external disturbances and then obtaining reasonable estimations of error bounds, are studied. The objectives of this research are exploring the sufficient conditions of asymptotical stability, finite-time stability, parameter convergence and finite-time parameter convergence of the considered systems, revealing complex dynamical behaviors of systems and esteblishing foundmental theo

英文关键词： Multi-agent uncertain system；Consensus problem；Adaptive control；Finite-time Stability；Parameter convergence