基于Agent的导弹兵力行动方案辅助决策的系统建模

自主规划、适应环境和学习功能等多项智能行为是现代战争的决策活动对辅助决策系统的基本要求.以常规导弹兵力行动方案的辅助决策为背景,提出了基于多Agent的系统总体设计,确立了系统构成的核心:机动方案Agent、测试方案Agent、运输方案Agent、保障方案Agent、,集成方案Agent以及任务管理Agent,建立了各类Agent的结构模型,并阐述了其功能.     

基于角色驱动的多Agent主体工作流模型的研究

在分析了传统的几种工作流建模方法的基础上，提出了基于角色驱动的多Agent体工作流模型，把多Agent的协作交互作为模型的主体，模型中引入了多Agent，把任务的执行者看作Agent，该Agent具有角色库，可以拥有一个或者多个角色，并用这些Agent之间的交流协作来描述工作流过程，使得执行者Agent成为模型的主体，改变了传统的模型中人的被动地位，使活动能更加顺利圆满地完成，同时也很好地解决了业务过程中可能存在的活动执行者的多角色和角色变化的问题。     

用遗传算法实现多智能体联盟的形成

在多Agent系统中，当单个Agent不能或不能有效完成特定的任务时，任务就必需分配给一组Agent，即Agent将形成联盟。本文将讨论Agent联盟的形成，并将遗传算法应用于Agent联盟形成中，希望可以快速、高效地找出合适的Agent联盟。     

一种优化的任务分配方法研究

在计算机领域中,由于数据量和待处理任务的复杂度迅速增加,多台计算机共同协作完成任务已是未来发展的趋势,而合理的任务分配方法是高效地执行任务的前提。目前,借助多Agent系统去研究任务分配的方法是目前研究的一个热点。本文基于多Agent系统提出一种基于执行代价和传输代价的任务分配的优化方法,建立相应的任务分配模型,推导出评价任务分配优化效果的目标函数,并通过实例说明了该分配机制对多Agent系统处理问题性能的影响。结果表明,该种优化方法能有效地分析多Agent系统处理问题时的性能。     

安全工作流任务分配策略

工作流系统访问控制为企业的工作流安全技术框架提供了可靠的保证。基于任务角色的访问控制模型通过任务将角色和权限联系起来．然后给用户指派合适的角色．用户通过其指派的角色获得可以执行的任务。然而在工作流多个任务实例同时执行过程中用户通过角色获得多个任务需要同时执行，如果用户执行的任务选择不当，将会导致工作流中某一任务实例中途挂起或夭折。本文根据工作流执行的特点．在基于任务角色访问控制模型的基础上，时工作流执行过程中任务实例出现的瓶颈问题进行了分析，提出了任务基本分配策略，事务任务，互斥任务，任务基数约束等分配方案，有效地提高了工．作流的执行效率。     

基于遗传算法的多agent联盟的生成

联盟生成是多Agent系统的一个关键问题,主要研究如何在多Agent系统中动态生成面向任务的最优Agent联盟.本文讨论了Agent联盟的生成,并将遗传算法应用于Agent联盟生成中,对比实验结果表明,遗传算法可以快速、高效地找出合适的Agent联盟,并在解的质量和收敛速度上均优于相关算法.     

基于Agent的CFD自主求解系统研究

在用Agent概念对复杂软件系统抽象的基础上,强调Agent在CFD求解系统中的自主性和智能性,实现具有推理能力的计算卖体。通过基于角色的系统分析和改进的BDI模型Agent构建方法,将求解系统构建为一个多Agent系统。求解过程中由Agent控制子区域独立计算,根据解的收敛变化情况在不同数值方法间切换,根据子域之间边界点数据的变化实时决定通信的时机。同时介绍了系统的结构和各个Agent的主要工作。     

合同网协议中的信用模型

针对经典的合同网协议CNP(Contractor Network Protocol),提出多Agent系统环境下自私Agent的信用模型。为追求利益最大化,自私Agent在自己能力不足的情况下仍有可能对宣布的任务进行投标。通过引入信用模型,任务管理者在发布标书和选择中标者时将充分考虑Agent投标及任务完成情况的历史记录,从而减少因随机选择Agent发标或仅考虑效用选择中标Agent而引发多次协商或任务无法完成的状况。实验结果表明,基于信用模型的Agent协商策略在保证任务完成质量的基础上,能显著提高系统性能,尤其是在任务数目较大的情况下。     

基于消息报告机制的多Agent调度系统研究

针对生产车间调度系统中因资源庞大以及动态变化而引起的监管困难问题,构建了一种 基于消息报告机制的多代理调度系统分层体系结构。根据Agent在调度系统中的作用和角色的不 同,将Agent划分为管理Agent、任务Agent和资源Agent,并通过一个具有消息报告机制的管理Agent 来有效管理和监控系统的运行。实例分析表明,消息报告机制不仅能保证系统的有效运行,而且能提 高系统的工作效率。     

基于Agent的自主式智能机器人体系结构及集成系统

自主式智能机器人体系结构的关键是寻求精确推理和快速反应之间的最佳折衷,分布式多Agent系统为体系结构的实现提供了新的途径.本文提出基于多Agent的面向任务集成的自组织体系结构IRASO,详细讨论该结构的计算机系统集成方案,介绍基于异质分布式计算的网络集成环境Robix,并通过仿真分析该系统的特性.     

基于强化学习的多Agent协作研究

强化学习为多Agent之间的协作提供了鲁棒的学习方法．本文首先介绍了强化学习的原理和组成要素，其次描述了多Agent马尔可夫决策过程MMDP，并给出了Agent强化学习模型．在此基础上，对多Agent协作过程中存在的两种强化学习方式：IL(独立学习)和JAL(联合动作学习)进行了比较．最后分析了在有多个最优策略存在的情况下，协作多Agent系统常用的几种协调机制．     

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.     

Distributing RePast agent‐based simulations with HLA

Large, experimental multi‐agent system (MAS) simulations are highly demanding tasks, both computationally and developmentally. Agent toolkits provide reliable templates for the design of even the largest MAS simulations, without offering a solution to computational limitations. Conversely, distributed simulation architectures offer performance benefits, but the introduction of parallel logic can complicate the design process significantly. The motivations of distribution are not limited to this question of processing power. True interoperation of sequential agent‐simulation platforms would allow agents designed using different toolkits to transparently interact in common abstract domains. This paper discusses the design and implementation of a system capable of harnessing the computational power of a distributed simulation infrastructure with the design efficiency of an agent toolkit. The system permits integration, through a higher‐level architecture (HLA) federation, of multiple instances of the Java‐based lightweight agent‐simulation toolkit RePast. This paper defines abstractly the engineering process necessary in creating such middleware, and reports on the experience in the specific case of the RePast toolkit. The paper also presents performance results that illustrate that significant speedup can be achieved through the integration of RePast with HLA. Copyright © 2008 John Wiley & Sons, Ltd.     

多Agent系统的上下文感知增强

如今,多agent系统(multi-agent system,简称MAS)被广泛用于开发各种应用系统.当前,开放、动态、难控的计算平台要求软件系统能够根据系统自身及其环境状态信息及其改变,动态地调节自身的行为,即具备一定的上下文感知能力.然而,现有工作并未就如何向既有的MAS应用系统引入上下文感知能力提出有效的解决方案.依照关注分离原则,结合面向上下文程序设计技术(context-oriented programming,简称COP)、反射技术(reflection)以及代码植入技术(code instrumentation),提出了一套MAS系统上下文感知增强框架和底层支撑技术.开发人员可以在既有应用源码不可得的情况下,自动地将指定agent类型转换为(扩展为)具有上下文感知能力的agent类型.此外,利用底层运行支撑环境,系统管理员可以在系统运行时刻动态地调整指定agent的上下文感知行为.     

基于MAS Builder开发多Agent系统的方法与实例

随着计算机领域面临的应用问题越来越呈现出分布、开放、动态的特征，多Agent成为近年来AI研究的热点之一。本文简要介绍开发和集成多Agent系统的环境－MAS Builder，并以实例说明了基于该系统开发多Agent系统的方法：问题域分解、Agent定义、任务模块的生成。     

An Evolutionary Dynamical Analysis of Multi-Agent Learning in Iterated Games

In this paper, we investigate Reinforcement learning (RL) in multi-agent systems (MAS) from an evolutionary dynamical perspective. Typical for a MAS is that the environment is not stationary and the Markov property is not valid. This requires agents to be adaptive. RL is a natural approach to model the learning of individual agents. These Learning algorithms are however known to be sensitive to the correct choice of parameter settings for single agent systems. This issue is more prevalent in the MAS case due to the changing interactions amongst the agents. It is largely an open question for a developer of MAS of how to design the individual agents such that, through learning, the agents as a collective arrive at good solutions. We will show that modeling RL in MAS, by taking an evolutionary game theoretic point of view, is a new and potentially successful way to guide learning agents to the most suitable solution for their task at hand. We show how evolutionary dynamics (ED) from Evolutionary Game Theory can help the developer of a MAS in good choices of parameter settings of the used RL algorithms. The ED essentially predict the equilibriums outcomes of the MAS where the agents use individual RL algorithms. More specifically, we show how the ED predict the learning trajectories of Q-Learners for iterated games. Moreover, we apply our results to (an extension of) the COllective INtelligence framework (COIN). COIN is a proved engineering approach for learning of cooperative tasks in MASs. The utilities of the agents are re-engineered to contribute to the global utility. We show how the improved results for MAS RL in COIN, and a developed extension, are predicted by the ED. Author funded by a doctoral grant of the institute for advancement of scientific technological research in Flanders (IWT).     

MULTI-AGENT INFRASTRUCTURES FOR OBJECTIVE AND SUBJECTIVE COORDINATION

Coordination in multi-agented systems (MAS) can be conceived as either an agent activity (the subjective viewpoint) or an activity over agents (the objective viewpoint). The two viewpoints have generated two diverging and often contrasting lines of research, as well as different and noncompatible technologies, however, their integration is mandatory for modeling and engineering complex MAS. In this paper, we explore the issue of integration at both the model and the technology levels.

First, by taking FIPA agents and coordination artifacts as reference notions for subjective and objective approaches, respectively, we sketch a framework where agent interactions with coordination artifacts are modeled as physical acts, deliberated and executed by agents analogously to communicative actions. Then, we show how the JADE infrastructure for FIPA-compliant agents, and the TuCSoN infrastructure providing agents with coordination artifacts can be integrated at the technology level, allowing JADE agents to access TuCSoN tuple centers through JADE services.     

Agents in object‐oriented software engineering

Software engineers of multi‐agent systems (MASs) are faced with different concerns such as autonomy, adaptation, interaction, collaboration, learning, and mobility, which are essentially different from classical concerns addressed in object‐oriented software engineering. MAS developers, however, have relied mostly on object‐oriented design techniques and programming languages, such as Java. This often leads to a poor separation of MAS concerns and in turn to the production of MASs that are difficult to maintain and reuse. This paper discusses software engineering approaches for MASs, and presents a new method for integrating agents into object‐oriented software engineering from an early stage of design. The proposed approach encourages the separate handling of MAS concerns, and provides a disciplined scheme for their composition. Our proposal explores the benefits of aspect‐oriented software development for the incorporation of agents into object‐oriented systems. We also illustrate our aspect‐oriented approach through the Portalware multi‐agent system, a Web‐based environment for the development of e‐commerce portals. Copyright © 2004 John Wiley & Sons, Ltd.     

An application perspective evaluation of multi-agent system in versatile environments

Multi-agent systems (MAS) based computing is the most appropriate paradigm for the problem domain, where data, control, expertise or resources are distributed and also it is interesting to the user only if the technologies address the issues of interest to the user. The MAS has the hypothesis that the agent based computing offers better approach to manage the complex systems and process. They are large-scale systems and collaborate with one another to achieve their functions in a highly modular and flexible way. In this point of view, the work presented in this paper is an enhanced attempt to validate the MAS based on application perspective. As a test-bed, a distributed MAS for software testing is constructed such that to provide a hybrid testing environment based on variety of agents, which possibly incorporate several testing techniques. The developed framework is validated on two perspectives namely, efficiency of the application domain, i.e. software testing using MAS and efficiency of the proposed framework. The validation of the later case has been carried out on two conditions: regular working environment and exceptional working environment. The second type of validation provided the guidelines for implementing proper exception handling mechanism in the enhanced MAS, which is being developed for software testing Purpose.     

A multi-agent system for the decentralized resource-constrained multi-project scheduling problem

A restart evolution strategy (RES) for the resource‐constrained project scheduling problem (RCPSP), as well as its integration in a multi‐agent system (MAS) for solving the decentralized resource‐constrained multi‐project scheduling problem (DRCMPSP) will be presented. To evaluate the developed approach, problem instances of the RCPSP taken from the literature with up to 300 activities are used, as well as 80 generated instances of the DRCMPSP, with up to 20 projects and with up to 120 activities each. For 73 instances of the RCPSP, the RES found better solutions than the best ones found so far. In addition, the MAS is suitable for solving large multi‐project instances decentrally. The results for the DRCMPSP instances show that the presented decentralized MAS is competitive with a central solution approach.