多Agent系统的协同

近年来,基于Agent技术给软件的发展带来了日新月异的变化,这种技术提供了软件系统概念化、设计与实现的全新模式,特别是分布与开放环境中运行的软件系统。文章着重阐述研究多Agent系统协同的原因、一般问题、协同关键方法以及协同协议和策略、几个代表观点,并简单描述了在当前主要应用领域内的经典工作,文中最后还提出了当前存在问题和研究领域。     

多Agent系统中信任的动态性处理

信任是多Agent系统中进行决策和交互的重要内容。收集必要的信息确定信任关系,动态地管理、维护信任关系,以及监控和重估已有的信任关系是多Agent系统中信任管理的关键问题。虽然研究者对上述关键问题提出了一系列解决方案,但依然存在一些问题有待进一步解决。本文针对信任的动态性处理这一问题,在分析现有的典型信任模型基础之上,提出一个具有动态性的Confidence-Reputation信任模型。模型中,我们不仅考虑了Agent的直接交互历史(Confidence)和信誉(Reputation),同时也考虑了信任的本体性。     

基于中间观测器的多智能体系统分布式故障估计

针对发生执行器故障的多智能体系统, 论文提出一种新型分布式中间观测器的设计方法, 可以同时估计系统的状态和故障.本文设计的观测器可以应用于严格正实条件和观测器匹配条件不满足的系统.针对多智能体系统的通讯拓扑是有向图和无向图的情况, 分别获得估计误差系统稳定的条件.观测器的参数矩阵可以通过求解线性矩阵不等式(Linear matrix inequality, LMI)计算.针对具有有向拓扑的多智能体系统, 本文方法所需求解的LMI的维数, 等于对单个智能体系统设计观测器所需求解的LMI的维数.这表明应用本文方法进行故障估计时, 计算量不会随着系统中智能体数目的增加而增加.针对多智能体系统通讯拓扑是无向图的情况, 利用Laplacian矩阵的对称性, 可以得到保守性更小的结论.最后, 仿真算例验证了本文方法的有效性.     

Cooperation Protocols in Multi-Agent Robotic Systems

Multi-agent robotic systems are useful in many practical applications. For some tasks, such as holding a conference, cooperation among agents are necessary. For other tasks, such as cleaning a room, multiple agents can work in parallel for better performance. This paper provides help-based (HCP) and coordination-based (CCP) protocols for controlling agents to accomplish multi-agent tasks. The HCP utilizes efficient negotiation to coordinate agents into groups. The CCP improves overall performance by exchanging local knowledge among agents and making decisions in parallel. A reactive and modularized agent architecture was employed to implement the protocols. Since each protocol is embedded into the architecture, it is efficient and effective. In addition, the protocols are deadlock-free. The protocols were utilized to solve the Object-Sorting Task, which abstracts two characteristics of tasks: parallelism and cooperation. The experimental results showed that 1) both HCP and CCP are stable under different workload; 2) the protocols can effectively utilize the agent-power to achieve super-linear improvement; 3) The CCP is better than the HCP in both performance and speedup.     

含通信延时及输入延时的线性多智能体的一致性

针对通信延时及输入延时并存的情况下,研究具有一般线性动态特性的高阶多智能体在固定且无向的网络拓扑下的一致性.通过将多智能体的特征方程分解为一系列低阶的因子,一致性问题转换为一系列低阶因子的稳定性分析问题,降低了问题的分析复杂度.进一步地,分析无延时情况下多智能体达到一致性的控制增益的取值范围,即增益的稳定区间.当从稳定区间选取一个固定的控制增益时,应用频率扫描高级聚类法分析多个因子在两延时参数空间中的稳定性,并由此获得多智能体延时无关及延时相关的一致性结论.一个说明性的例子验证了分析结论.     

Intra-Organizational Logistics Management Through Multi-Agent Systems

Compared to inter-organizational logistics management, the goal of intra-organizational logistics management is to maximize the profits of the whole company sometimes at the cost of its individual units' profits. While many agent-based models have been proposed for logistics management, most of these models use an auction approach. Thus, they are not suitable for intra-organizational logistics management. In this paper, we first formulate logistics management as a distributed resource allocation problem. Then, we present our ongoing work on developing a multi-agent model for intra-organizational logistics management and using Lagrangian relaxation to decompose the problem into a set of subproblems. Our initial experimental results are very promising. We provide a detailed analysis of our results.     

多智能体系统在分布式采样控制下的动力学行为

本文运用图论、矩阵分析和现代控制理论等工具, 研究信息网络传输下多智能体系统的协调控制和动力学行为. 假设每个智能体通过数字化网络传感器获得其邻近智能体的位置状态,而且多智能体系统采取分布式线性控制协议. 每个智能体被描述为一个简单的采样系统,多智能体系统转化为混合动力学系统. 研究结果表明,多智能体系统所呈现的渐近聚集、周期振荡和发散动力学行为不仅和网络结构的代数特征有关, 而且和每个智能体的动力学方程、采样周期有关. 本文给出了具体精确的代数判据. 仿真例子进一步验证了本文结果的有效性.     

多Agent系统中实时消息传递机制

1 引言 Agent是具有信念、能力、决策和承诺等精神状态的实体,它的提出是以计算的社会学模型为背景,通过多个简单Agent协同工作,可以完成比较复杂的任务。分布式人工智能(DAI)的主要目标就是解决如何组织多Agent系统中的Agent,使它们能够协同工作以完成复杂问题求解。多Agent系统的关键问题是Agent间的协同,而根据计算社会学模型的观点,消息     

多Agent系统问题求解的广义粒子模型方法

提出一种多Agent系统分布式问题求解的新的广义粒子模型,将复杂环境下多Agent系统资源分配和任务规划的优化问题转变为广义粒子模型中的粒子运动学和动力学问题.广义粒子模型可以描述和处理的复杂环境包括多Agent系统中的Agent之间存在的随机、并发、多类型的交互行为.各Agent有不同的个性、自治性、生命周期、拥塞程度和故障几率等.本文讨论了广义粒子模型和多Agent系统分布式问题求解的关系,提出了广义粒子模型的数学物理模型和多Agent系统分布式问题求解算法,并且证明了它们的正确性、收敛性、稳定平衡性等基本性质.通过复杂环境下多Agent系统资源分配和任务规划问题的实验和比较,证实了广义粒子模型方法的有效性及其特点.     

The Origins of Ontologies and Communication Conventions in Multi-Agent Systems

The paper proposes a complex adaptive systems approach to the formation of an ontology and a shared lexicon in a group of distributed agents with only local interactions and no central control authority. The underlying mechanisms are explained in some detail and results of some experiments with robotic agents are briefly reported.     

A Continuous Leader-following Consensus Control Strategy for a Class of Uncertain Multi-agent Systems

In this paper, we study the robust leader-following consensus problem for a class of multi-agent systems with unknown nonlinear dynamics and unknown but bounded disturbances. The control input of the leader agent is nonzero and not available to any follower agent. We first consider a class of high order chain integrator-type multi-agent systems. By employing the robust integral of the sign of the error technique, a continuous distributed control law is constructed using local information obtained from neighboring agents. Using Lyapunov analysis theory, we show that under a connected undirected information communication topology, the proposed protocol achieves semiglobal leader-following consensus. We then extend the approach to a class of more general uncertain multiagent systems. A numerical example is given to verify our proposed protocol.      

复杂网络中选举免疫策略研究

受现代社会选举制度启发,提出了一种新的节点免疫策略,提出的选举免疫策略按节点得票数量多少选择节点并免疫。基于消息传递方法,给出了大规模配置模型网络上SIR传染病模型的选举免疫策略的数学解析,并比较了选举免疫、随机免疫、目标免疫和熟人免疫策略的效果。数学解析结果和仿真结果表明,选举免疫能够有效控制疾病在无标度网络上的传播。     

有向图中网络Euler-Lagrange系统的自适应协调跟踪

基于一致性理论, 在有向图中研究网络 Euler-Lagrange 系统的协调跟踪控制. 所有跟随智能体的动力学模型均为 Euler-Lagrange 方程. 在仅有部分跟随智能体能获取领航智能体信息的情形下, 同时考虑系统模型的参数不确定性, 设计分布式自适应控制律实现所有跟随智能体对领航智能体的跟踪. 针对领航智能体的运动状态, 考虑以下两种情形: 1) 领航智能体为固定点; 2) 领航智能体为动态点. 对第一种情形, 设计的控制律使得所有跟随智能体渐近交会于固定点; 对第二种情形, 首先对每个跟随智能体设计分布式连续估计器, 然后提出了分布式自适应控制律. 当每个跟随智能体均能获取领航智能体的加速度信息时, 设计的控制律能实现对领航智能体的渐近跟踪, 当跟随智能体不能获取领航智能体的加速度信息时, 跟踪误差是有界的. 最后通过仿真分析验证设计的控制算法是合理有效的.     

基于近似模拟的多智能体系统分布式层次控制设计

本文主要探讨基于近似模拟的多智能体系统分布式层次控制的设计问题。为降低问题复杂度,首先建立了一个简单的抽象系统指导各智能体,继而讨论这一抽象系统和多智能体之间的模拟关系。借助于该抽象系统,给出多智能体系统完成一类协调任务的分布式层次控制器。同时利用模拟函数和共同Lyapunov函数,进一步分析了多个体系统在切换拓扑下的集体行为。     

多智能主体协作规划理论及其在RoboCup中的应用

机器人足球赛是近年人工智能迅速发展起来的一个重要的研究领域。基于此平台,研究了将规划、竞争、约束、协调纳入一个协作规划框架之中的多智能体协作规划理论,并将其分为5个阶段①产生协作目标;②协作规划;③寻求协作伙伴;④选择协作方案;⑤实现所确定的目标。最后探讨了应用到机器人足球赛的一些技术问题。在比赛中,决策速度也较好。     

基于FOO投票协议的无收据电子投票方案

安全且实用的电子投票协议是信息安全领域的热点问题之一。引入了盲签名、投票编号、申诉标识等工具,提出了一种新的无收据的电子投票方案,该方案进一步完善了FOO投票协议,可保证选票的匿名性、可验证性和无收据性,并且允许投票者中途弃权。该方案不仅保持了原方案的各种优点,而且增强了系统的安全性和灵活性。因此,与其他类似方案相比较,该方案具有更好的通用性和实用性。     

Human-in-the-Loop Consensus Control for Nonlinear Multi-Agent Systems With Actuator Faults

This paper considers the human-in-the-loop leader-following consensus control problem of multi-agent systems(MASs)with unknown matched nonlinear functions and actuator faults.It is assumed that a human operator controls the MASs via sending the command signal to a non-autonomous leader which generates the desired trajectory.Moreover,the leader’s input is nonzero and not available to all followers.By using neural networks and fault estimators to approximate unknown nonlinear dynamics and identify the actuator faults,respectively,the neighborhood observer-based neural fault-tolerant controller with dynamic coupling gains is designed.It is proved that the state of each follower can synchronize with the leader’s state under a directed graph and all signals in the closed-loop system are guaranteed to be cooperatively uniformly ultimately bounded.Finally,simulation results are presented for verifying the effectiveness of the proposed control method.     

网络管理的多agent范型

网络管理系统是大规模的分布式软件系统 ,伸缩性和灵活性是今天的网络管理技术面临的两大难题 .软件 a-gent为这些难题提供了出色的解决方法 .本文首先指出当前网络管理中存在的问题 ;在此基础上 ,提出基于多 agent的网络管理结构 ;为了证明其技术可行性 ,专门建立了基于 agent的网络流量管理实验系统 .对该系统的评估证明了软件agent非常适合于开发分布式网络管理系统 ,具有实时性高、可扩充性好、自治能力强的特点 ,而且与传统系统相比 ,这种系统的建立、修改和升级更加经济     

Multi-Candidate Voting Model Based on Blockchain

Electronic voting has partially solved the problems of poor anonymity and low efficiency associated with traditional voting. However, the difficulties it introduces into the supervision of the vote counting, as well as its need for a concurrent guaranteed trusted third party, should not be overlooked. With the advent of blockchain technology in recent years, its features such as decentralization, anonymity, and non-tampering have made it a good candidate in solving the problems that electronic voting faces. In this study, we propose a multi-candidate voting model based on the blockchain technology. With the introduction of an asymmetric encryption and an anonymity-preserving voting algorithm, votes can be counted without relying on a third party, and the voting results can be displayed in real time in a manner that satisfies various levels of voting security and privacy requirements. Experimental results show that the proposed model solves the aforementioned problems of electronic voting without significant negative impact from an increasing number of voters or candidates.