多智能体系统能控性研究进展

能控性问题是多智能体协调控制领域中一个基本又十分重要的研究课题.本文对多智能体系统能控性问题的研究现状进行综述.介绍了多智能体能控性领域的基本问题和特点,并结合智能体自身动力学与邻居交互协议,从拓扑结构角度对该领域当前的研究热点和前沿进行分析阐述.进一步,对结构能控性的研究成果进行归纳总结,并对能观测性、可镇定性和复杂网络能控性等相关问题进行阐述.最后给出了仍需解决的问题和可能的研究方向.     

单向通信拓扑领航者动态未知线性多智能体系统的协同迭代学习控制

研究单向通信拓扑领航者动态未知线性多智能体系统的协同跟踪问题.基于邻居的相对状态信息,设计了分布式迭代学习控制律实现对领航者的协同跟踪控制,采用Lyapunov-Krasovskii函数分析闭环系统的稳定性与收敛性.进而,将状态反馈结论拓展到输出反馈,通过构造局部观测器估计不可量测的状态信息,采用估计的相对状态信息设计了分布式迭代学习控制器.对于以上两种情形,多智能体系统在通讯拓扑含有生成树的条件下能够实现与领航者的状态同步,同时,所设计的分布式迭代学习律能够对领航者未知输入进行精确估计.仿真实例验证了所提方法的有效性.     

异质多智能体系统滞后一致性跟踪控制

针对一阶、二阶混合异质多智能体系统的滞后一致性问题,提出了一种基于牵制控制的分布式滞后一致性控制协议。首先,将滞后一致性分析转化为稳定性证明;然后,利用图论和Lyapunov稳定性理论分析闭环系统稳定性;最后,分别在固定拓扑和切换拓扑结构下给出了基于线性矩阵不等式（LMI）形式的滞后一致性可解的充分条件,从而实现异质多智能体的领导-跟随者滞后一致性。数值仿真结果表明,所提出的滞后一致性控制方法可以使异质多智能体系统实现领导-跟随者滞后一致性。     

具有领航者的时延多智能体系统的群集运动

计算机技术、网络技术和通信技术的飞速发展,推动着无人驾驶飞行器的编队控制、传感器网络的分布控制、卫星的姿态控制等多智能体系统的建模与应用的逐步深入,也吸引了越来越多的研究者致力于多智能体系统的动态编队控制的研究.研究了具有不同的通信时延和不同的输入时延的移动多智能体算法的群集运动.假设多智能体系统由n个智能体和1个Leader组成,网络连接拓扑是静态有向连通图,智能体Leader为拓扑图的全局可达节点.应用频率域的广义Nyquist判据分析了具有不同的通信时延和不同的输入时延的移动多智能体算法,应用Greshgorin圆盘定理和曲线的曲率理论研究了具有领航者的多智能体算法的群集运动,得到保证系统一致性的收敛条件.该一致性条件是一个应用节点局部信息的分散式条件,只与输入时延有关,而与通信时延无关.最后,通过计算机仿真验证了本文结论的有效性.     

有向网络下非线性多智能体系统的协调跟踪

基于一致性理论,在有向拓扑结构下研究非线性多智能体系统的协调跟踪控制问题．考虑智能体动力学模型为一般且仅满足Lipschitz条件的非线性系统,在仅有部分跟随智能体能获取领航智能体信息的情形下,当领航智能体与跟随智能体之间的拓扑结构具有有向生成树,即存在领航智能体到所有跟随智能体的有向路径时,所设计的分布式控制律可实现所有跟随智能体对领航智能体的跟踪,并指出该拓扑结构是系统实现跟踪的一个必要条件．最后,仿真实验验证了所设计控制算法的有效性．     

线性多智能体系统一致性问题的部分稳定性方法

提出了处理高阶线性多智能体系统一致性问题的线性变换.该线性变换将一致性问题转化为一个部分稳定问题.研究了一般线性协议下线性多智能体系统的三个问题: 1) 寻找一致性收敛判据; 2) 计算一致性函数; 3) 设计线性一致性协议的增益矩阵.具体来说,提出了基于矩阵 Hurwitz 稳定的一致性收敛的充分必要条件,给出了一致性函数的解析表达式,同时建立了一致性协议的增益矩阵与多智能体系统收敛时间和一致性精度的关系,并针对预先给定的收敛时间和精度要求设计了增益矩阵.     

多智能体系统MAS及其应用

分布式人工智能的研究和网络化分布环境的普及,推动了Agent的理论、技术特别是多Agent的理论及其技术的进展.随着计算机科学的发展迅速趋于成熟,多智能体方法和技术在很多领域得到了广泛的应用.针对目前多智能体系统(MAS)的研究现状及存在的问题,运用系统工程的思想,给出了多智能体系统的研究思路与方法.从工程应用的角度出发,详细论述了Agent及MAS的特性、结构模型以及多智能体系统所使用的一种最常用的通讯语言--KQML,重点分析了多智能体技术在几个有代表性领域的应用.最后,对多智能体系统技术的应用前景做出了系统的分析与展望.     

多智能体系统中的协商模型

目前,对多智能体系统的应用多是以开放的、复杂的、变化的网络为平台,为了更好地与环境相匹配,同对提高智能体之间协商的效率,从全局与单个智能体两方面着手,提出了一种新的MAS全局模型,并相应地提出了全局协调Agent的三维模型以及智能体之间的协商模型。通过这一系列模型的模拟实现,证明了此协商模型是正确的与接近自然的。从理论上基本上解决了与当前流行的基于网络的各种先进的智能系统,分布式问题求解等领域的衔接,因而有一定的指导意义并具有广阔的发展前景。     

钢铁工业过程多智能体系统研究

结合钢铁工业过程的特点，通过分析冶金过程工业智能控制的发展状况，提出了采用多Ａｇｅｎｔ技术，建立钢铁工业多智能体系统，从而实现钢铁工业的分布式智能控制，分析了系统建立的必要性和可行性，并给出了该系统的结构和实现方法。     

无速度输入多智能体系统的一致性

本文研究了有无引导者的多智能体系统在非线性协议下的一致性问题.当智能体速度信息无法获知时,分别针对有无引导者的多智能体系统设计了包含辅助系统和智能体相对位移信息的非线性分布式协议.借助图论、Lyapunov稳定性理论、Barbalat引理等方法,推导出有无引导者的多智能体系统在连通无向通讯网络中实现一致的充分条件,其次,设计了一种新的能使引导–追随者多智能体系统在有向通讯网络中实现期望一致的协议.最后,数值仿真验证了结果的正确性.     

Decentralised regulation of nonlinear multi-agent systems with directed network topologies

This paper aims to address the leader–follower regulation problem of multi-agent systems with directed network topologies, where the agents are described by feedforward nonlinearities with the growth rate being unknown a priori. Both the state feedback regulation protocol and the output feedback regulation protocol are delicately constructed such that all the states of followers can converge to the leader state globally. In this paper, a model transformation is firstly performed and the leader–follower regulation problem can be transformed into a general regulation problem. Then, by introducing an appropriate state transformation, the regulation problem can be changed into a parameter determined problem. It is proved that the parameter can be determined by both the properties of M-matrices and the estimates of nonlinear terms. Finally, a numerical example is presented to show the feasibility of designed protocols.     

Leader–follower fixed-time consensus of multi-agent systems with high-order integrator dynamics

The leader–follower fixed-time consensus of high-order multi-agent systems with external disturbances is investigated in this paper. A novel sliding manifold is designed to ensure that the tracking errors converge to zero in a fixed-time during the sliding motion. Then, a distributed control law is designed based on Lyapunov technique to drive the system states to the sliding manifold in finite-time independent of initial conditions. Finally, the efficiency of the proposed method is illustrated by numerical simulations.     

博弈论框架下的二阶多智能体系统领导者选择算法

随着人工智能的发展,多智能体系统中智能体的异质性、工作环境的复杂性、系统目标的多样性,给多智能体系统动态性能的分析带来了挑战.同时,也催生了新的控制策略和优化方法.博弈论作为一种研究社会系统中智能体决策过程的经典工具,如今已被应用到了多智能体系统研究领域.本文主要针对二阶多智能体系统编队过程中领导者选取的两类问题:1)选取k个领导者使系统误差达到最小;2)系统误差在一定范围内,选取最小数量的领导者;提出了一类超模博弈建模方法.在建模过程中设计了各个智能体的效用函数与系统整体的目标函数,使各个智能体在寻求各自效用函数最大化的过程中实现整体目标.而后,运用贪婪算法优化了智能体决策过程.本文分析了所建模型的平衡点存在性和系统稳定性.最后,利用仿真实例对比说明了本文提出的基于超模博弈的二阶多智能体系统领导者选择算法的有效性.     

Controllability of heterogeneous multi-agent systems under directed and weighted topology

This paper considers the controllability problem for both continuous- and discrete-time linear heterogeneous multi-agent systems with directed and weighted communication topology. First, two kinds of neighbour-based control protocols based on the distributed protocol of first-order and second-order multi-agent systems are proposed, under which it is shown that a heterogeneous multi-agent system is controllable if the underlying communication topology is controllable. Then, under special leader selection, the result shows that the controllability of a heterogeneous multi-agent system is solely decided by its communication topology graph. Furthermore, some necessary and/or sufficient conditions are derived for controllability of communication topology from algebraic and graphical perspectives. Finally, simulation examples are presented to demonstrate the effectiveness of the theoretical results.     

Distributed fault detection and isolation in time-varying formation tracking UAV multi-agent systems

The problem of distributed fault detection and isolation (DFDI) in conjunction with time-varying formation control of unmanned aerial vehicle (UAV) multi-agent systems with multiple-leader leader–follower structure is studied, in this paper. It is assumed that the communication data of the agents are noisy, and faults may occur in either leaders or followers. The followers are not aware of the main trajectory and just follow the leaders. The first step in this paper is to reduce the communication noise. Besides, if a fault occurs in one of the leaders causing to exit the leader from the main trajectory, the followers should be able to detect the fault, isolate the leader, use the other leaders' data, and keep the main trajectory. Tracking the trajectory by followers in such a case is the second step of this paper. Preserving the time-varying formation despite a fault occurring in one of the followers is the next step. To reach these goals, a distributed array of Kalman filters is used for noise cancellation and data extraction. Next, the combination of state vector data fusion and some type of $\chi$² test are employed for DFDI, in agents. Besides, a controller is implemented in each agent for formation tracking. In this controller, estimated relative position and velocity vectors of the agents are employed to keep the time varying formation while tracking the leaders. The closed-loop stability is studied via Lyapunov stability theorem, and various simulations are carried out to demonstrate the capability of the proposed strategy.     

二阶非线性多智能体系统领导跟随一致性研究

为减少通信时延对系统一致性的影响,针对有领导者的二阶非线性多智能体系统的领导跟随一致性进行了研究,新颖的提出近似随机脉冲时延的概念并应用于新协议。相比于传统协议,新协议在脉冲时刻通信时延较小时,各智能体基于时延态对自身当前时刻状态进行预测,并以自身未来预测状态取代时延态发送给各邻接智能体同时补偿自身反馈通道时延,从而使系统更快实现一致性。基于Lyapunov稳定性理论,利用一类再推广的Halanay不等式的性质给出两个保证系统实现一致性的充分条件。最后,实例仿真证明了新协议的优越性。     

Adaptive containment control of second-order multi-agent systems with nonlinear dynamics and multiple input-bounded leaders

This paper considers the adaptive containment control problem of second-order multi-agent systems with inherent nonlinear dynamics. In particular, the leaders’ control inputs are nonzero, bounded, and not available to any follower. Based on the relative states among neighbouring agents, a discontinuous adaptive protocol is first proposed to ensure that the containment errors of each follower converge to zero asymptotically, i.e. the states of the followers asymptotically converge to the convex hull spanned by those of the leaders. To eliminate the chattering effect caused by the discontinuous protocol, a continuous adaptive protocol is further designed based on the boundary layer technique and the σ-modification technique. Numerical examples are provided to demonstrate the effectiveness of our theoretical results.     

多Agent系统安全性分析与设计

全面分析了多Agent系统的安全特性、安全风险和安全需求.在此基础上,研究了多Agent系统多层安全模型及安全多Agent系统需要提供的基础安全服务.基于FIPA-OS框架,设计并实现了一个面向Agent的二级认证PKI系统,同时提出一组安全服务Agent,为FIPA-OS框架增加了的身份鉴别、证书管理、安全审计、资源访问控制及Agent社区管理等多种安全机制.     

Finite-time consensus for multi-agent systems with globally bounded convergence time under directed communication graphs

In this paper, we study the finite-time consensus problems with globally bounded convergence time also known as fixed-time consensus problems for multi-agent systems subject to directed communication graphs. Two new distributed control strategies are proposed such that leaderless and leader-follower consensus are achieved with convergence time independent on the initial conditions of the agents. Fixed-time formation generation and formation tracking problems are also solved as the generalizations. Simulation examples are provided to demonstrate the performance of the new controllers.     

Stability and convergence analysis of multi-agent consensus with information reuse

In this article, we study multi-agent consensus algorithms with information reuse by intentionally introducing the outdated state information into the traditional consensus algorithms. In the continuous-time case, we first show that the outdated state information combined with the current state information does not necessarily jeopardise the stability of a single system, but may improve the convergence speed without increasing the maximal control effort. Then this idea is extended from the single-agent case to the multi-agent case. When the directed communication graph is fixed, the corresponding Laplacian matrix and the outdated state information satisfy certain conditions, we show that the consensus algorithm with both the current and outdated states can achieve a faster convergence speed than the standard one. We also consider the case of a switching directed communication graph and derive corresponding conditions. In the discrete-time case, we propose a discrete-time consensus algorithm with both the current and outdated states under an undirected fixed communication graph. We then derive conditions on the communication graph, the sampling period and the outdated state information such that the proposed algorithm can achieve a faster convergence speed than that using the standard one. In both the continuous-time and discrete-time settings, we show that the maximum control efforts for the proposed consensus algorithms are identical to those for the standard ones. Several simulation examples are presented as a proof of concept.