一种带概率与效用算子的Agent形式化模型

形式化方法描述Agent时需要考虑信念的不确定性与决策的效用性要素.在经典Agent的BDI形式化模型基础上,定义了Agent形式化语言,引入概率算子与效用算子,提出了Agent形式化模型,在此模型中利用概率算子与效用算子对Agent的信念、愿望、意图与规划等意识属性进行了定义.该模型能满足Agent对逻辑理性、信念的不确定性与决策理性的要求.     

避免逻辑全知的BDI语义

BDI(belief, desire, intention)是基于Agent计算的理论模型,BDI语义关系着Agent计算的发展.通过把相信划分为主观相信和客观相信,把可能世界理解为认知的不同阶段,给出具有进化特征的BDI语义.该语义既能描述Agent,又避免了"逻辑全知"问题.     

基于描述逻辑的Agent组织

描述逻辑是基于对象的知识表示形式．已有的描述逻辑没有考虑组织目标,不适合描述多Agent组织模型．针对这个问题,首先给出多Agent组织的模型,结合组织目标考虑组织、子组织之间的关系．基于此模型,提出描述逻辑ALCA框架,用以描述结合目标任务的多Agent组织,证明了ALCA是可判定的,给出一个ALCA的可满足性判定算法．最后通过一个例子说明ALCA的描述性．     

Multi-Agent系统中Agent知识获取的合作模型

Agent的知识是Agent计算的前提.在动态、不确定的Multi-Agent系统中,Agent必须具备及时有效地获取所需知识的能力以求解问题.现有的知识获取模型不能有效地支持在动态、不确定的Multi-Agent系统中Agent对知识获取的要求,Agent的知识获取能力比较有限.提出一个系统的、用于Agent知识获取的合作模型KACM(knowledge-acquiringcooperationmodel)系列,包括被动模型、主动终止模型和主动非终止模型.基于言语行为理论和以分枝时序逻辑为基础的形式化框架,讨论了KACM所涉及的Agent通信行为,分析了Agent如何响应这些通信行为以完成知识交互,定义了各通信行为以及KACM的满足语义,最后讨论了研究工作的意义.     

Agent在多Agent系统中计算的意愿理论*

提出了Agent在多Agent系统中计算的意愿理论,以支持Agent计算的理论研究.区分了两种意愿：实现型意愿和维护型意愿.基于多Agent系统计算的逻辑框架,给出了两种意愿新的语义定义,获取和描述了它们的一些重要逻辑属性.     

一种结合效用的Agent思维状态模型

建立Agent思维状态模型是Agent理论 研究的一个重要课题.结合效用提出一种Agent思维状态模型BDICU(belief desire intentio n combined with utility),使Agent的行为具有逻辑理性和决策理性,为副作用问题提供了 解决方法.同时,给出目标和意图的生成和更新规则.BDICU模型改进和扩充了Rao和Georgeff 的信念-期望-意图理论,为逻辑和效用理性Agent系统提供了实现支持.     

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.     

生态化MAS的认知模型

生态化MAS是由异构的Agent所组成的一种具有协同进化的多Agent系统。提出了生态化MAS的理念和认知模型,它是形成Agent知识和信念的基本框架;设计了一个模型检验的程序结构,并通过典型应用表明,该模型能够使Agent对环境的认知达到三个层次——概念图、模糊词计算和关联性框架。该模型能够为基于Agent的问题求解提供依据。     

Agent-BDI逻辑

阐述了Agent的形式化描述应该采用含有正规和非正规模态算子的混合模态逻辑为逻辑工具 的观点.建立了Agent-BDI逻辑的代表系统A-BI,讨论了它的语法和语义.特别是给出了非正 规模态算子基于Kripke标准可能世界的新的语义解释,证明了A-BI逻辑系统不但是可靠的, 而且是完备的.A-BI逻辑系统恰当地刻画了信念与意图的本质与内在联系,可作为Agent形式 化研究的逻辑工具.     

Agent意图的双子集语义改进模型

意图是Agent的一个关键的意识属性,在决定理性Agent的行为中起着重要作用.为了克服现有意图逻辑中存在的缺陷,建立了适用于意图的语义表示.讨论了理性Agent性态的形式化中对意图语义的要求以及现有意图逻辑中存在的问题.介绍了在前期工作--真假子集语义基础上开发的双子集语义改进模型及其在Agent意图形式化中的应用,并且证明通过对模型的代数结构施加一定的约束,能获得许多希望得到的性质.在二值逻辑中,真和假是同等重要的.当然,对一个命题,描述了真值也就知道了假值;但对于一类命题却不是这样,对假值的刻画与对真值的刻画具有同等重要的意义.而对意图的描述是对一类命题(Agent意图实现的命题)的刻画.经典的正规模态算子的可能世界语义只重视真,用RI(w)来描述,可看成是单子集语义.而改进的双子集语义真假并重,用R_IT(w)来描述真,并用R_IF(w)来描述假,从而能更全面地描述二值逻辑中的模态算子.经典的正规模态算子的可能世界语义可以看成是改进的双子集语义当R_IF(w)=()时的退化情形.改进的双子集语义不仅避免了基于正规模态逻辑表示的"逻辑全知"问题以及由此带来的副作用等问题,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,而且克服了前期工作真假子集语义和双子集语义表示的缺陷.实际上,改进的双子集语义为非正规模态算子的语义表示提供了一种新的方法,可应用于建立新的合适的Agent逻辑系统.     

认识逻辑(2):多认识主体的认识逻辑MEL

认识逻辑ＥＬ给出了知识和信念在单一认识主体情形下的逻辑框架。本文将认识逻辑ＥＬ推广到多认识主体的情形，得到了一个可靠并且完备的系统ＭＥＬ。ＭＥＬ继承了ＥＬ的全部性质。ＭＥＬ的一个重要特点是各个认识主体的知识是可以共享的。ＭＥＬ可以作为分布式多ａｇｅｎｔ系统的逻辑基础。     

Logical omniscience, semantics, and models of belief

Logical omniscience may be described (roughly) as the state of affairs in which an agent explicitly believes anything which is logically entailed by that agent's beliefs. It is widely agreed that humans are not logically omniscient, and that an adequate formal model of belief, coupled with a correct semantic theory, would not entail logical omniscience. Recently, two prominent models of belief have emerged which purport both to avoid logical omniscience and to provide an intuitively appealing semantics. The first of these models is due to Levesque (1984 b ); the second to Fagin and Halpem (1985). It is argued herein that each of these models faces serious difficulties. Detailed criticisms are presented for each model, and a computationally oriented theory of intensions is presented which provides the foundation for a new formal model of belief. This formal model is presented in a decidable subset of first-order logic and is shown to provide a solution to the general problem of logical omniscience. The model provides for the possibility of belief revision and places no a priori restrictions upon an agent's representation language.     

Logical organization of knowledge with inconsistent and undecidablealgorithms using imaginary and transfinite exponential number forms in anon-Boolean field-I. Basic principles

In order to consider the organization of knowledge using inconsistent algorithms, a mathematical set-theoretic definition of axioms and undecidability is discussed. Ways in which imaginary numbers, exponentials, and transfinite ordinals can be given logical meanings that result in a new way to definite axioms are presented. This presentation is based on a proposed logical definition for axioms that includes an axiom and its negation as parts of an undecidable statement which is forced to the tautological truth value: true. The logical algebraic expression for this is shown to be isomorphic to the algebraic expression defining the imaginary numbers. This supports a progressive and Hegelian view of theory development, which means that thesis and antithesis axioms that exist in quantum mechanics (QM) and the special theory of relativity (STR) can be carried along at present and might be replaced by a synthesis of a deeper theory prompted by subsequently discovered experimental concept     

Reasoning tractably about explicit belief: A model‐theoretic approach

Existing epistemic logics such as the logic of implicit and explicit belief and the logic of awareness adopt a deductive‐theoretic approach for characterizing belief. In this approach, an agent represents the state of the world with a conjunction of axioms in its knowledge base (KB) and evaluates queries by trying to prove or disprove that they follow from KB. This paper presents a multivalued epistemic logic (MEL) that allows agents to reason both deductively and model theoretically about implicit and explicit belief. By characterizing an agent's KB with a class of finite models, the set of formulas that an agent believes can be determined by checking their validity in all these models. This rests on the fact that MEL has a complete axiomatization (sentences that are true in all these models will also be provable). In this paper, the soundness, completeness, and decidability of MEL are proven. Furthermore, a polynomial time model‐checking algorithm for determining the satisfiability of a sentence at a particular state in a given model of MEL is also presented. © 2000 John Wiley & Sons, Inc.     

情感等级BDI主体模型研究

信念、愿望和意图(BDI)模型是近年来影响最为深远的主体技术之一。文中把命题动态逻辑和无穷值的ukasiewicz逻辑进行融合后对情感等级BDI主体模型进行了形式化。为通过信念度、愿望度、意图度、害怕度、焦虑度和自信度对不确定性行为进行表示和推理,把相应的公理添加到ukasiewicz逻辑中。文中的情感等级BDI主体模型的行为是通过添加具体条件的每种背景的不同测度来决定,清晰地表示主体的心理状态和情感状态的不确定性。文中对情感等级BDI模型进行公理化,并说明它们对主体行为的影响。此模型可较轻易地向包括其它心理状态和情感状态的主体进行推广。文中在给出情感等级BDI模型的语言、语义及公理和演绎规则后,证明此逻辑系统的可靠性和完全性。随后给出情感等级BDI主体模型的不同背景之间的相互关系,并对该主体的买房行动进行实例分析。本研究立足于不确定性的表示和推理,旨在为分布式人工智能提供形式支持。     

Using Abstract Resources to Control Reasoning

Many formalisms for reasoning about knowing commit an agent to be logically omniscient. Logical omniscience is an unrealistic principle for us to use to build a real-world agent, since it commits the agent to knowing infinitely many things. A number of formalizations of knowledge have been developed that do not ascribe logical omniscience to agents. With few exceptions, these approaches are modifications of the possible-worlds semantics. In this paper we use a combination of several general techniques for building non-omniscient reasoners. First we provide for the explicit representation of notions such as problems, solutions, and problem solving activities, notions which are usually left implicit in the discussions of autonomous agents. A second technique is to take explicitly into account the notion of resource when we formalize reasoning principles. We use the notion of resource to describe interesting principles of reasoning that are used for ascribing knowledge to agents. For us, resources are abstract objects. We make extensive use of ordering and inaccessibility relations on resources, but we do not find it necessary to define a metric. Using principles about resources without using a metric is one of the strengths of our approach.We describe the architecture of a reasoner, built from a finite number of components, who solves a puzzle, involving reasoning about knowing, by explicitly using the notion of resource. Our approach allows the use of axioms about belief ordinarily used in problem solving – such as axiom K of modal logic – without being forced to attribute logical omniscience to any agent. In particular we address the issue of how we can use resource-unbounded (e.g., logically omniscient) reasoning to attribute knowledge to others without introducing contradictions. We do this by showing how omniscient reasoning can be introduced as a conservative extension over resource-bounded reasoning.     

对称协调博弈问题的多智能体强化学习

针对多机器人协调问题,利用协调博弈中智能体策略相似性,提出智能体的高阶信念修正模型和学习方法PEL,使智能体站在对手角度进行换位推理,进而根据信念修正将客观观察行为和主观信念推理结合起来。证明了信念修正模型的推理置信度只在0和1两个值上调整即可协调成功。以多机器人避碰为实验背景进行仿真,表明算法比现有方法能够取得更好的协调性能。     

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