切换拓扑下非线性多智能体系统自适应神经网络一致性

Adaptive NN consensus of nonlinear multi-agent systems under switching topologies

中文关键词  切换拓扑  未知控制系数  神经网络  一致性

英文关键词  switching topologies  unknown control coefficients  neural network  consensus

中文摘要

      本文研究一类具有未知控制系数的非线性多智能体系统自适应神经网络分布式控制策略. 首先, 针对切换拓扑下具有未知控制系数的非线性多智能体系统一致性问题, 提出一类自适应神经网络一致性控制算法. 其中, 采用神经网络函数逼近方法解决系统中的不确定性问题, 并设计一项自适应光滑项处理有界扰动和神经网络函数逼近误差. 随后, 证明了切换拓扑下具有未知控制系数的非线性多智能体系统的一致性, 并保证了闭环系统的有界性. 此外, 本文把相关的一致性算法扩展到了一般有向图含有一个有向生成树的情形. 最后, 通过仿真实例验证了本文所提算法的有效性.

英文摘要

      This paper develops an adaptive neural network (NN) distributed control strategy for the nonlinear multiagent systems (MASs) with nonidentical unknown control coefficients (UCCs). Specifically, the adaptive NN consensus control algorithms are first proposed to deal with the consensus problem of nonlinear MASs with nonidentical UCCs under switching topologies, where the nonlinear system uncertainty is solved by the technique of NN function approximation method, and the bounded disturbance and the approximated error of NN are handled by an adaptive smooth term. Then, it is proved that the consensus of nonlinear MASs with nonidentical UCCs can be achieved with the designed algorithm, and the boundedness of closed-loop systems is ensured. In addition, we extend the correlative consensus algorithm to the case where a general directed graph contains a directed spanning tree. Finally, the simulation examples under different network topologies are presented to verify the effectiveness of the proposed adaptive NN consensus control algorithms.