有领航者的多智能体系统的稳定性分析

本文研究了一类有领航者的多智能体系统在平面和高斯型源下的稳定性分析.该系统由领航者和跟随者两种智能体组成,只有领航者具有环境的信息;同时,吸引与排斥函数也被推广到具有某些性质的一类函数.证明了在满足适当假设条件时,在领航者智能体的导航作用下,多智能体系统的跟随者在没有环境信息的情况下也能准确地到达目标或远离目标.当多智能体系统在不同源的条件下聚集时有界区域的大小也被估计.给出数值仿真以证明理论结果的正确性.     

基于有向网络的一致性跟踪算法

研究了leader有控制输入且followers未知该输入条件下的线性多智能 体一致性跟踪问题.提出两种一致性跟踪算法,证明两种算法在leader到followers存在一棵 有向生成树且follower间拓扑是有向条件下,网络就能跟踪leader的状态.对于第一种算法,节点根 据相邻节点或leader的状态来求解其控制输入,并基于代数Riccati不等式给出 连续情形下算法稳定性条件.第二种算法直接利用相邻节点或leader的状态,使followers在上述网络条件下跟踪leader的状态,同样基于代数Riccati不等式给出算法稳定性条件. 仿真结果验证了算法的有效性.     

双目标推动下群体行为的元胞自动机模拟

采用元胞自动机对自推动粒子改进模型进行模拟,在此基础上分别探讨无目标与双目标群体行为的演化特征。无目标条件下群体行为的元胞模拟结果显示低密度环境下大部分个体处于动态均衡,高密度环境下大部分个体处于静态均衡,中等密度下两种均衡均存在。基于双目标吸引力的群体行为模拟结果显示当智能体的理性程度较低时系统演化能够产生一定的集聚效应,处于吸引源以外的智能体与吸引源保持一定距离并呈湍流式运动,而当智能体具备小概率全局理性判断能力时,群体行为的演化结果为大部分智能体都聚集在吸引源附近。     

含有障碍物环境下多智能体系统的聚集行为

研究多动态智能体系统在含有障碍物环境下的聚集行为．首先建立一类多智能体系统在有障碍物的环境下的数学模型，以描述个体在遇到障碍物时的行为特点．其次从理论上证明了整个多智能体系统聚集行为的稳定性，指出在有障碍物环境下仍然能够涌现系统的聚集行为．最后，通过数值模拟验证了有关结论的有效性．     

CGF队形形成及避障行为研究

计算机生成兵力(CGF)在模拟战术级兵力仿真时,多个兵力实体要按照一定的队形行进。首先对基本队形的分布式控制方法进行分析;然后,将leader-follower算法思想应用其中,简化了算法模型;其次,为了满足底层仿真的需要,对兵力实体的避障行为进行了研究,引入了切线避障法。增加follower与leader之间的距离对leader速度的反馈,在follower避障过程中,动态地改变leader的速度,避免掉队现象。实验表明,该方法能很好地适用于底层战术级仿真模型。     

大规模移动智能群体的建模及联合行为分析

提出一种基于个体局部信息的移动智能群体的模型,该模型的建立仅依靠相互可检测到的个体之间的局部位置信息.在此模型基础上,研究了移动智能群体联合趋向目标行为的稳定性问题.仿真实验表明,基于局部位置信息的群体模型能够实现群体的联合跟踪全局目标的行为,同时对群体的规模具有较强的可扩展性.     

大规模智能群体的建模及稳定性分析

提出一种基于个体局部信息的智能群体模型．在此模型基础上,给出了个体的运动控制方程。并基于Lyapunov稳定性理论研究了大规模智能群体动态集聚行为的稳定性．该模型的建立仅依靠相互可检测到的个体之间的局部信息。结合所提出的个体局部控制算法。便能实现群体的稳定全局集聚行为,体现出了较强的适应性、鲁棒性和可扩展性．     

多智能体入侵杂草算法

针对标准入侵杂草算法缺乏信息共享机制的缺陷,将多智能体系统融入标准入侵杂草算法,提出了一种新的多智能体入侵杂草算法.该算法通过多智能体系统中改进的邻域竞争合作算子实现个体间信息的交流,提高收敛速率;利用多智能体系统中的自学习算子增强算法求解精度.五个基准函数测试对比分析结果表明,多智能体入侵杂草算法的求解精度、收敛速度和稳定性优于标准入侵杂草算法、粒子群算法和差分进化算法.     

多智能体系统输入约束下的一致性与轨迹规划研究

针对多智能体系统提出了一种分布式预测控制方法. 首先, 研究了有输入约束下的一致性问题. 其次, 对环境中有障碍物的多智能体轨迹规划进行了研究, 其中只有当障碍物进入智能体有限感知区域内时, 障碍物状态信息才能被获取. 基于预测控制方法, 设计了一种分布式控制算法来解决上面两个问题. 构造一个与每个智能体动力学相交互的代价函数, 设计相应最优控制问题, 从而实现优化控制算法. 智能体间交互信息是其邻居在上一时刻的最优控制状态. 系统稳定性可以通过构造代价函数中的一个终点状态控制器与最优控制问题中的一个终点状态区域来保证. 仿真研究表明所提方法的有效性.     

邻域交互结构优化的多智能体快速蜂拥控制算法

针对多智能体系统在动态演化过程中容易出现的"局部聚集"现象,融 合复杂网络中的拓扑结构优化理论与多智能体系统协调蜂拥控制研究,提出了一种基 于邻域交互结构优化的多智能体快速蜂拥控制算法.该算法首先从宏观上分析多智 能体的局部聚集现象,利用社团划分算法将局部相对密集的多个智能体聚类成一个 社团,整个多智能体系统可以划分成多个相对稀疏的社团,并为每个社团选择度 最大的个体作为信息智能体,该个体可以获知虚拟领导者信息;随后从多智能体 系统中不同社团相邻个体间的局部交互结构入手,取消社团间相邻个体的交 互作用,设计仅依赖于社团内部邻居个体交互作用的蜂拥控制律;理论分 析表明,只要每个社团存在一个信息智能体,在虚拟领导者的引导作用下,整个多 智能体系统就可以实现收敛的蜂拥控制行为;仿真实验也证实了对多智 能体系统进行邻域交互结构优化可以有效提高整个系统的收敛速度.     

Formation control for multi-robot systems with collision avoidance

In this paper, distributed formation is studied for a team of mobile robots including leaders and followers. Followers are able to sense the relative displacements to neighbouring followers and all of the leaders, and the leaders can be sensed by the followers. Based on such assumption of sensing, distributed formation control scheme is designed, under which both followers team and leaders team are fully independent. The followers and leaders have exchangeable roles within their own group. The leaders can have an arbitrary formation, and around the leaders, the followers need to reach a regular polygon formation with a suitable orientation. Distributed control laws and localised collision avoidance algorithms are designed for each follower, and they use only local displacements. Speed and acceleration sensors are avoided. As the leaders and the followers are independent and exchangeable, both robot teams are scalable and robust against member failures and system delays.     

Event-based broadcasting containment control for multi-agent systems under directed topology

The event-based broadcasting containment control problem for both first-order and second-order multi-agent systems under directed topology is investigated. Based on certain event, each agent decides when to transmit its current states to its neighbours and the agents’ distributed control algorithms are based on these sampled state measurements, which can significantly decrease the number of the controllers’ updates. All the agents are divided into two groups, namely, the leaders and the followers. The formation control is introduced. The leaders exchange their information to converge to a formation. The followers utilise the information from both their leader neighbours and their follower neighbours and are driven to the convex hull of the leaders using the proposed control algorithms. Numerical simulations are provided to illustrate the effectiveness of the obtained theoretical results.     

Leader–follower consensus over numerosity-constrained random networks

In this work, we study a discrete-time consensus protocol for a group of agents which communicate over a class of stochastically switching networks inspired by fish schooling. The network model incorporates the phenomenon of numerosity, that plays a prominent role in the collective behavior of animal groups, by defining the individuals’ perception of numbers. The agents comprise leaders, which share a common state, and followers, which update their states based on information exchange among neighboring agents. We establish a closed form expression for the asymptotic convergence factor of the protocol, that measures the decay rate of disagreement among the followers’ and the leaders’ states. Handleable forms of this expression are derived for the physically relevant cases of large networks whose agents are composed of primarily leaders or followers. Numerical simulations are conducted to validate analytical results and illustrate the consensus dynamics as a function of the number of leaders in the group, the agents’ persuasibility, and the agents’ numerosity. We find that the maximum speed of convergence for a given population can be enhanced by increasing the proportion of leaders in the group or the agents’ numerosity. On the other hand, we find that increasing the numerosity has also a negative effect as it reduces the range of agents’ persuasibility for which consensus is possible. Finally, we compare the main features of this leader–follower consensus protocol with its leaderless counterpart to elucidate the benefits and drawbacks of leadership in numerosity-constrained random networks.     

异质多智能体系统的合围控制

在固定有向拓扑下,研究由一阶与二阶动力学智能体组成的异质多智能体系统的合围控制问题.首先,分别针对系统内同时包含一阶动力学领航者和二阶动力学跟随者、二阶动力学领航者和一阶动力学跟随者的两种情况,共给出4种有效的控制协议;其次,综合利用代数图论、矩阵理论和稳定性分析工具,通过系统变换方法或构建误差向量法得到上述两类异质多智能体系统能够实现合围控制的判据条件;最后,通过仿真进一步验证所提出协议的有效性.     

Graph-theoretic characterisations of the steady states for containment control

Containment control is a consensus-like protocol whose objective is to drive all the followers into the convex hull formed by the leaders. The steady states of the followers have great effect on system performance. Two kinds of cell partitions from graph theory are employed to investigate the reflections of the information topology on the steady states of the followers. It is proved that agents that belong to the same cell have the same steady states.     

Distributed containment control with multiple stationary or dynamic leaders in fixed and switching directed networks

In this paper, we study the problem of distributed containment control of a group of mobile autonomous agents with multiple stationary or dynamic leaders under both fixed and switching directed network topologies. First, when the leaders are stationary and all followers share an inertial coordinate frame, we present necessary and sufficient conditions on the fixed or switching directed network topology such that all followers will ultimately converge to the stationary convex hull formed by the stationary leaders for arbitrary initial states in a space of any finite dimension. When the directed network topology is fixed, we partition the (nonsymmetric) Laplacian matrix and explore its properties to derive the convergence results. When the directed network topology is switching, the commonly adopted decoupling technique based on the Kronecker product in a high-dimensional space can no longer be applied and we hence present an important coordinate transformation technique to derive the convergence results. The proposed coordinate transformation technique also has potential applications in other high-dimensional distributed control scenarios and might be used to simplify the analysis of a high-dimensional system to that of a one-dimensional system when the decoupling technique based on the Kronecker product cannot be applied. Second, when the leaders are dynamic and all followers share an inertial coordinate frame, we propose a distributed tracking control algorithm without velocity measurements. When the directed network topology is fixed, we derive conditions on the network topology and the control gain to guarantee that all followers will ultimately converge to the dynamic convex hull formed by the dynamic leaders for arbitrary initial states in a space of any finite dimension. When the directed network topology is switching, we derive conditions on the network topology and the control gain such that all followers will ultimately converge to the minimal hyperrectangle that contains the dynamic leaders and each of its hyperplanes is normal to one axis of the inertial coordinate frame in any high-dimensional space. We also show via some counterexamples that it is, in general, impossible to find distribute containment control algorithms without velocity measurements to guarantee that all followers will ultimately converge to the convex hull formed by the dynamic leaders under a switching network topology in a high-dimensional space. Simulation results are presented as a proof of concept.     

Distributed Containment Control of Networked Nonlinear Second-order Systems With Unknown Parameters

In this paper, we study the containment control problem for nonlinear second-order systems with unknown parameters and multiple stationary/dynamic leaders. The topologies that characterize the interaction among the leaders and the followers are directed graphs. Necessary and sufficient criteria which guarantee the control objectives are established for both stationary leaders (regulation case) and dynamic leaders (dynamic tracking case) based protocols. The final states of all the followers are exclusively determined by the initial values of the leaders and the topology structures. In the regulation case, all the followers converge into the convex hull spanned by the leaders, while in the dynamic tracking case, not only the positions of the followers converge into the convex hull but also the velocities of the followers converge into the velocity convex hull of the leaders. Finally, all the theoretical results are illustrated by numerical simulations.      

Interconnection topologies for multi-agent coordination under leader–follower framework

In this paper, the formation control problem of the network of multiple agents is studied in terms of controllability, where the network is of the leader–follower structure with some agents taking leaders role and others being followers interconnected via the neighbor-based rule. It is shown that the controllability of a multi-agent system can be uniquely determined by the topology structure of interconnection graph, for which the investigation comes down to that for a multi-agent system with the interconnection graph being connected. Based on these observations, two kinds of uncontrollable interconnection topologies are characterized, and a necessary and sufficient eigenvector-based condition is presented. Our studies also touch upon the selection of leaders.     

Containment in leader–follower networks with switching communication topologies

We study bipartite, first-order networks where the nodes take on leader or follower roles. Specifically, we let the leaders’ positions be static and assume that leaders and followers communicate via an undirected switching graph topology. This assumption is inspired by the swarming behavior of silkworm moths, where female moths intermittently release pheromones to be detected by the males. The main result presented here states that if the followers execute the linear agreement protocol, they will converge to the convex hull spanned by the leaders’ positions as long as the time-varying undirected graph defining the communication among all agents is jointly connected. The novelty of this research is that we use LaSalle’s Invariance Principle for switched systems, and additionally, the result is shown to hold for arbitrary state dimensions.