基于Q-learning的不确定环境BDI Agent最优策略规划研究

BDI模型能够很好地解决在特定环境下的Agent的推理和决策问题,但在动态和不确定环境下缺少决策和学习的能力。强化学习解决了Agent在未知环境下的决策问题,却缺少BDI模型中的规则描述和逻辑推理。针对BDI在未知和动态环境下的策略规划问题,提出基于强化学习Q-learning算法来实现BDI Agent学习和规划的方法,并针对BDI的实现模型ASL的决策机制做出了改进,最后在ASL的仿真平台Jason上建立了迷宫的仿真,仿真实验表明,在加入Q-learning学习机制后的新的ASL系统中,Agent在不确定环境下依然可以完成任务。     

基于知识情感等级BDI主体的决策行为模型研究

为了将知识、信念、愿望和意图以及害怕、焦虑和自信对决策行为的影响的强烈程度进行形式化,把无穷值的卢卡斯维茨逻辑的真值取值范围从[0,1]扩展到[-1,1]。知识情感等级BDI主体的决策行为通过添加了具体条件的不同背景的不同测度来决定;知识情感等级BDI主体模型清晰地表示了主体的知识状态、心智状态和情感状态的不确定性;此模型可以轻易地向其他类型的主体进行推广。在给出了知识情感等级BDI模型的语言、语义以及此模型的不同背景之间的相互关系之后,对知识情感等级BDI主体的军事决策行为进行了实例分析。此研究旨在为分布式人工智能和军事仿真提供形式支持。     

基于JADE的Multi—Agent合作仿真的设计与实现

Multi—Agent合作是分布式人工智能领域的一个研究热点,多种合作算法被提出,需要一个平台进行比较验证,而从底层开发出一个Multi-Agent合作系统是比较复杂和困难的,该文在研究JADE仿真平台技术的基础上,构建了一个Multi—Agent合作系统,为研究Multi—Agent合作算法提供了验证平台。     

Collaborative assignment using belief-desire-intention agent modeling and negotiation with speedup strategies

In this paper, we propose a distributed agent model that applies belief-desire-intention (BDI) reasoning and negotiation for addressing the linear assignment problem (LAP) collaboratively. In resource allocation, LAP is viewed as seeking a concurrent allocation of one different resource for every task to optimize a linear sum objective function. The proposed model provides a basic agent-based foundation needed for efficient resource allocation in a distributed environment. A distributed agent algorithm that has been developed based on the BDI negotiation model is examined both analytically and experimentally. To improve performance in terms of average negotiation speed and solution quality, two initialization heuristics and two different reasoning control strategies are applied, with the latter yielding different variants of the basic algorithm. Extensive simulations suggest that all the heuristic-algorithm combinations can produce a near optimal solution soon enough in some specific sense. The significance and applicability of the research work are also discussed.     

Concurrent consideration of evacuation safety and productivity in manufacturing facility planning using multi-paradigm simulations

Manufacturing facilities are expected to maintain a high level of production and at the same time, employ strict safety standards to ensure the safe evacuation of the people in the event of emergencies (fire is considered in this paper). These two goals are often conflicting. This paper presents a methodology to evaluate evacuation safety versus productivity concurrently for various, widely known manufacturing layouts. While the safety performance indicators such as evacuation times are inferred from the crowd (agent based) simulation, the productivity performance indicators (e.g. throughput) are analyzed using the discrete event simulation. To this end, this research focuses on creating innovative techniques for developing accurate crowd simulations, where Belief-Desire-Intention (BDI) agent framework is employed to build each person’s individual actions and the interactions between them. The data model and rule based action algorithms for each agent are reverse-engineered from the human-in-the-loop experiments in the immersive virtual reality environments. Finally, experiments are conducted using the constructed simulations to compare safety and productivity for different layouts. To demonstrate the proposed methodology, an automotive power-train (engine and transmission) manufacturing plant was used. Initial results look quite promising.     

Agent的意图模型

意图是Agent的一个不可缺少的意识属性,在决定理性Agent的行为时起着重要的作用.已经有了若干种基于正规模态逻辑的意图模型,但它们存在着严重的“逻辑全知”问题.该文阐明意图不是正规模态算子,并提出了另一种意图模型,它不存在“逻辑全知”问题和其他相关问题(例如,副作用问题等).这种意图模型与Konolige和Pollack的意图模型相比,比较简单、自然,且满足K公理和联合一致性原理,实际上,为非正规模态算子基于正规可能世界的语义表示提供了一种新的方法.     

基于BDI的Agent合同网实现

合同网是一个用于分布式问题求解的高层通信协议,本文给出了一个扩充的合同网协议的会话策略和一个BDI精神状态形成模型,并在此基础上将BDI模型用于合同网协议,给出了实现合同同网协议的一系列BDI行为驱动规则。本文在合同网协议的招投标过程中引入了信任度、稳健度和积极程度等精神状态参数,这类参数有助于改善协作质量,从而使系统具有较好的适应性和灵活性。     

AUV心智逻辑*

为了表示和处理AUV（Autonomous Underwater Vehicle,自治水下机器人）的心智状态及心智活动,对Rao和Georgeff的BDI逻辑进行了拓展,定义了一种含模态词（信念、目标、能力、意图、上一次、下一次、曾经、将会、总有、总会）的一阶模态时态混合逻辑系统——AML（AUV心智逻辑）;基于Kripke可能世界语义模型给出了AML语义,构造了AML公理系统及演绎规则,并证明了AML的可靠性与完备性。AML为建造AUV心智模型提供了有力的表示和处理工具,弥补了Agent理论在机器人应用     

基于BDI模型构建信息Agent

本文提出一种用于分布式多Ａｇｅｎｔ环境的以ＢＤＩ模型为基础构建信息Ａｇｅｎｔ的方法,给出了ＰＥＲＦＯＲＭ（ＰＥＲｓｏｎａｌ ｉｎＦＯＲｍａｔｉｏｎ Ｍａｎａｇｅｒ）系统的模型和算法。     

Design and development of a fuzzy agent-based model to measure interest rate expectations

The financial system, which has been investigated by various researchers, is a rather complicated environment. Most research has only been concerned with quantitative factors (technical indexes), though qualitative factors (e.g., political situation, social conditions, international events, government policies, among others) play a critical role in the financial system environment, determining the regulatory policies within an economy. This paper presents a fuzzy knowledge-based model to measure the qualitative aspects related to one of the most important financial instruments used to regulate an economy, the base interest rate. The development and assessment of the proposed model was based on the Brazilian economy. Evaluation of the results obtained indicates that our approach gives good results when compared with real data and statistical-based forecasting tools. The main advantage of our approach is its capability to forecast long term interest rate expectations when combined with a powerful econometric model.