ѧ
JOURNAL OF SOFTWARE
1999 4 10 Vol.10 No.4 1999



Agent˼ά״̬ģ
ΰ͡ʯһ
ժҪ2090AgentϵͳAgent˼ά״̬ģ͵о.ֱ۷ʽģ͡ṹģ͡оӦ÷Bratmanѧ۵㡢CohenLevesqueͼģ͡RaoGeorgeffBDIģ͵Ҫɹ,ڿչĹ.
ؼʡAgent,MAS(multi-agent system),˼ά״̬,,Ը,Ŀ,ͼ,ŵ,,ͼ.
ͼšTP18
About the Mental State Model for Agent
MA Guang-wei XU Jin-hui SHI Chun-yi
(Department of Computer Science and Technology Tsinghua University Beijing 100084)
Abstract This paper summarizes the research of mental state model for agent, which is a major subject in the research on multi-agent system. Most references are published after 1990. From different perspectives, covering informal analysis, formal models, architectures, cooperation researches, and applications, many important results are introduced in this paper, such as Bratman's analysis, Cohen and Levesque's intention model, Rao and Georgeff's BDI model, etc. Finally, some prosperous works are introduced as well. We hope that this presentation can help readers deepen their understanding of agent.
Key wordsAgent, multi-agent system, mental state, belief, desire, goal, intention, commitment, cooperation, joint-intention.
ڶAgentϵͳ(multi-agent system,MAS),Agent˼ά״̬ģоAgent˼άԺ֮Ĺ,Լ֪滮ΪЭȻĹϵ.ʶ,һ(belief)Ը(desire)ͼ(intention)˼ά(BDI).AgentBDIͬϤ֪ʶʾ,Ƕ˼ά̵һ,ΪӦMASлĶ,ͻ¼,֤ԵġһµĺΪ.
Dennettǿϵͳ(1) ,ϵͳԺ͹ɣ(2) ,ϵͳĿꣻ(3) ʶ,ϵͳAgent,ͨԸʶԤAgentΪ.SinghΪAgentʶϵͳĺôǣ(1) ߺͷ˵,Ȼģ(2) ϵͳΪṩ˼ıʾ,ͽͣ(3) ʵ־ͿԵõAgentΪĹģʽ(4) ɱAgentл1.ʵ,ิϵͳ,ʹ꾡Ľṹ͹,ҲѴԤͽΪ,ʺϲʶ.ʶһֳ󹤾,ΪͺԤ⸴ϵͳΪṩһַķʽ2.
,йAgentģ͵оĶ(1) Ʒֲʽĸԣ(2) ˷˻ľ.ֲʽĹ㷺ʹ,Agent֮以ĸ.ͼʹ˼άӦĸ,ڹ滮ṩһڶԵͲͨЭķװ.Ե,ֱӲݽʧ,ͼAgentʵһּӹ1.Ŀǰ,Agent˼ά״̬ģģ̬߼Ϊ,һ֪ʶ߼KTD45߼KD45ĿͼKD.Agent߼һЩ,磺(1) ߼ȫ֪⣺BEL(),ҪԴ޵Agentǲʵ;(2) Ϊ⣺AgentΪձȻΪ߱ȻΪ,ôΪĿͼ;(3) ⣺AgentΪզñȻΪ沢Ҳͼ,ôҲǿͼ.
1 ֱ۷
1.1 BDIֱ
AgentԵǰ״ԼΪﵽĳЧܲȡΪ·ߵĹ,˼ά״̬֪.ԸAgentδ״ԼܲȡΪ·ߵϲ,˼ά״̬ĸ鷽.AgentӵлݵԸ,ҲԸǿʵֵ.ĿAgentԸѡӼ,ûвȡжĳŵ.һAgentĿǿʵֵ.AgentԴ,һȥ׷еĿ,,AgentѡĳĿ(Ŀ꼯)׷ĳŵ,γͼ.Agentĵǰͼ(ͼṹ)ǱѡĿ꼯ʹ״̬,ͼ˼ά״̬,ලAgentĶ.ŵʾĿ굽ͼת,Agent׷ͼļ̶ֳȲƶͼ¿.MAS,һΪŵǶAgentȺĽϼ,Agent֮ĺЭΪͬʵŵ.
ԸĿͼĹϵ3ݣ
(1) ͼ-һԣAgentӦͼǿܵ,ŴﲻĿ,ȷ,ŻﵽĿ.
(2) ͼ-ȫԣAgentͼﵽĳ״,״һʵ,ҲAgentͼֲȫǵ.
(3) ⣺Agentͼ,Ҫ,ôҲҪͼ.
(4) ڲһԣAgentҪӵгͻͼ,ӵгͻԸ.
(5) ֶ-ĿͼҪAgentδĳʱҪ˼,ԸûбҪ.
(6) ٳɹʧܣͼɱΪԸжʵֵĳŵ,,ͼĳɹʧܽи,ʧʱ¹滮. 
1.2 BratmanƽΪͼ
ƽʹAgentΪϻ,ֿ͹,AgentԼڣ(1) ޵ļϢԴ,Agentκʱ̶м,Ҳڸʱļ,ҲᾭԤı仯,֪ʶҲȫ;(2) Э,AgentҪЭеĻͽĻ,ҪЭͬAgentĻ.BratmanΪʹͼƽ,ͼܹΪԸ,뿴һ˼άԣ4.
ͼΪͼ͵ǰͼ.ǰͼҪΪ,Ϊйϵ,ͼΪ,ͬʱͼ,滮ϵ.ΪͼAgentΪԼһAgentıԼеΪͼ,Ѿʵʩͼһ,Agentӻı仯,ֲʵʻҪͼ5.
1.3 ͼ
Tuomela6Ϊͼѧĸ,˲μߵжĲ.:(1) 뵥Agentͼһ,⡢AgentΪѡ(2) жĳʼָͼأ(3) Ⱥϵ(4) 淶Agent˼ж.AgentGĸAСWe-Intends to do XĸͼҽGеиִͬXĻУ(1) AִĲ,(2) Aжжǰ,(3) AGдڶжǰ໥,(4) ʹ(1)(2)(3).AСWe will do Xgroup-intentionͼ,ҽGеиִͬXĻУ(1) AСWe-Intends to do X,(2) AγһִXgroup-intention,ǡWe-Intends to do X.
ִжJoint Intentionҽ(1) ЩAgentִжgroup-intention;(2) (1),໥.
ѾγִXͼ(JIP)ҽ(1) ÿִXĹ滮,(2) ÿ彫ǵĽϢ֪ͨ,(3) (1)(2)ڸ֮໥,ϳŵȥִX,ÿŵ.
2BDIģ
2.1˼ά״̬ģ
CohenLevesqueͼģͣ7ݻģ̬߼Ŀģ.ģ̬ӴʶĿ,ݳĿ궨ͼ,һЩʵ.ʽΪ
(1) M,,n,v(BEL x ) iff дnͨϵBɴĿ綼ЦΪ.
(2) M,,n,v(GOAL x ) iff дnͨĿϵGɴĿ綼ЦΪ.
(3) (P-GOAL x p)=def(GOAL x (LATER p))(BEL x  p)(BEFORE (BEL x p)(BEL x p))(GOAL x (LATER p))).
(4) (INTEND1 x a)=def(P-GOAL xDONE x (BEL x (HAPPENS a))?;a),aʾһΪ.
(5) (INTEND2 x p)=def(P-GOAL x e(DONE x (BEL x e(HAPPENS x e;p?))ġ(GOAL x  (HAPPENS x e;p?))?;e;p?)),pһν.
CohenLevesqueͼģǱȽڵĹ,ΪԱȽϵĻ׼.,Ϊ,ͼΪԸڶAgentԵļ.
RaoGeorgeffBDIģͣ8ݵʽϵͳǶԼ߼CTL*,ʹþĿģ.ͼ1ʾt1ʱAgentɴϵ.ÿʱṹһʱ,Agentӵһ͵ʷͷ֧͵δ,֧AgentӦʱ̵ѡ.ڸģ,ʽ״̬ʽ·ʽ,ģ̬ӴʶĿͼ,һЩģԼ͹,˼ά֮Լϵ,رǿʵԶԵ,˲ͬĳŵAgentĲͬԸ.ںĹУкܶ඼ǻRaoGeorgeffĹ.

ͼ1ʱģ
KonoligePollackͼģͣ9ݻڷģ̬߼,ʹ֪ṹAgent˼ά״̬,ĹǼµĳ,ͼϵͼĸֱӵķʽʾͼĽṹ,means-endĽṹ.
BellԸıģͣ10ݽһ׳,Agentᱣ˼άֱиı,ʽʱĸ.ʽǶHintikkaȻչ,ģ̬ӴʶӰ졢Ըͼ,ʽؼʱ.
GasparCoelhoָ11ͼģ͵Ĳ㣺(1) ģ̬߼裻(2) ûиĿͼβ(3) û仯ʱ,Ŀͼ.ڴ,AgentĿͼģԼ,չǰAgent˼ά״̬еģ.ĿģͶģ,AgentڲĿ,ҲһЭĿ꼯ѡƫӼ.
LinderHoek12ݵģͺб,ϲáĿͳŵ,ڳŵ˾̬Ͷ̬.ָ̬Agentĳŵ,ָ̬AgentŵΪ.ģͶ߼ȫ֪бȽƵĴ.
CavedonPadgham13ָͼ߼ķ,Kripke롰硱,չ˷,ͼṩϸ,ģ̬߼ͼĲ߼.
GeorgeffRao14ݷͼ,ͨԭBDICTL*߼šԸͼonlyģ̬(OBEL,ODESOINTEND),ģԼ͹,̻ͼԸĹϵ.ַΪ̽˼ά״̬Ķ̬ݻָһ.
ף15ݷģ̬߼оͼľ,ˡͼάṹۡ,ڹϵ۴Ժʱϵά϶ͼп̻,̽ͼĲͶ̬仯.ͼܺͻBDIṹAgent.
Donghaĳŵģͣ4ݻʱ̬߼Ӽ,Ϊŵṩ,ɳŵԭָAgentͼȡ,ҲΪAgentιĿ滮ṩģ.
2.2 ʵ
Wooldridgeָ,ģ̬߼ӿAgentֻܽг,ܵ˵޷Գķʽʵ,ԭǣ(1) ʹõ߼ʵDAIϵͳṹ֮ȱĹϵ,ر,ģͶʵϵͳڳ.(2) Щ߼Agent˲ʵ趨16.
RaoΪAgentʵԭǣ߼Ķ֤ģͼĸ⻹ûг,ʵֵBDIϵͳڰ3Ҫ˼άԿݽṹģ̬.ʵʵϵͳּ򻯵ļȱʵۻ,߼ٶԽʵṩ.RaoΪ,BDIоһĿʹĳֺҺбģ֤ۡۺͳ߼һһӦϵ17.
оMASϵͳķʶ߼ʱ̬߼ͽ̴,ûһȫõ,Wooldridge׶η(1) ҪоMAS,ģͣ(2) (1)ģʵֿģ,ʽϵͳУ(3) (2)пģ͵ʷΪ߼,(1)еģͽƵ18.
ʵϵͳ(dMARS)RaoչAgentSpeak(L),ԼǿʵϵͳԵļ.AgentSpeak(L)һԲ¼ͶĴ.ԸͼûбʾΪģ̬ʽ,ֻǶӦAgentSpeak(L)еʵ,߸.ͨһ׻͹滮Agent.RaoAgentSpeak(L)ĲBDI㷨,ֽһظ֤,һЩ֤,ڲͬ¼BDIõת,֤Agentԣ17.
2.3 ˼ά״̬
ڵAgent˼ά״̬ģ,LevesqueCohen19ݸAgent/ӵͼʽ,ͼΪϳĿ,˼ʵĶ.Castelfranchi20Ϊ˶岻ǳֱҪ,Ϊ:(1) ȱ໥ε,Щ,϶,Ǻ,ûͼ(2) ĳЩ⻷,ʹȱͨźͶԹͬжĹ,Ҳں.
Castelfranchi21MASдڵĶֳŵʽ,˴ŵĸȨ.ڶAgent֮ſ͹۵ϵ,Agent֮ĸΪϵ.໥Ϊ,Ϊ.
GroszKraus22ݻShareplansģͺPollackĵAgent滮,չ˺滮ۣ(1) ʹøͼȥ߶ǵϻĳŵ;(2) Agentĺжĳŵ;(3) ںı½ͰΪ;(4) аа;(5) ͨŵҪǵ,ڶĳŵ;(6) ϵӹ滮֤.
Dunin-KepliczVerbrugge23ݻRaoGeorgeff18ݵBDIģ,˼ͼͳŵĶԼ3ֳŵ.
Haddadi3ݻRaoGeorgeffBDI߼,ϳŵ,Э̷µ̽.ں˵,ĿǱίɻɵĿ,ŵҪ˫ͬͬǶϣ.ֶ-ĿΪĿꡢͼ;״̬ת,Ӧһ,˵ʲô³ŵγɡά.ͨŻԼںϵ.
2.4 BDIṹ
PRSǻBDIģ͵Agentṹ24(ͼ2ʾ),RaoGeorgeff25BDI߼PRS߼,˳BDI.PRSֻ˵Agent.

ͼ2PRSṹ
GRATE*ǻBDIģ͵Ľṹ26(ͼ3ʾ),ʹͼεĸΪΪִ.GRATE*˶Agent.

ͼ3GRATE*Ĺܽṹ
2.5 Ӧ 
INTERRAP27ݴӹ̽ǶAgentĻʽֲṹ.Agent֪ʶ⡢3ȼĿƲ,˼ά״̬Щ˼ά״̬Ļܼ֪ж֮Ĺܹϵ.
Jennings28ݷ˸ӻAgentȱͼֶ֧»ҵ,ͼģ,ָGRATE*.ͼΪŵǰͳŵĸЭ,Э鶨˳ŵĸıӦĻ.
Tambe29ݻͼ,˶ģ,ɶ״̬Ͷ,״̬ǹڶӹ,ӱʾжϳŵ,ȱʾ˸öĿǰȡϻ,ָÿһԱݵĽɫ,ֶ֧ӹAgentṹSTEAM.
Shoham30AOP(agent-oriented programming),˼ά״̬ģʹϵͳʵֵĽǶAgentӦеĽṹΪ.Agent״̬˼ά,ԪBͳŵOBL,ǶԼĳŵ.˼άģ͸Ԫعϵļٶڲһ¡ʡ.AOPй涨Ĵڻ򲻴״̬ŵԶ.Shoham,ͨõAgent.
3 
ף79ݲӰϴ,ģеʽؿʱ,е˼ά.ĽۺʵѽģͺоҲ,ۻ̽.
о෽.һΪϺ߼Ե,Ϊ˸߼ϵ,˹ϵͳ,֤һЩضǷ,ΪΪǻڵǰϺ߼Ƶ.Ŀǰ˼ά״̬ģ͵о෽.һ෽ǲþ,ģһжᷢʲô,Ϊÿ.ԸģʵʾЩ״̬Ч.һжʹЧŻж,ҪԸʹøۼõ.Ӹ,߼ʵԵ,۷ͨŻЧöʵԵľ.Ӽ,ʹ÷߼޷ʹЧŻ,ʹֵľҲ.ڴڶ̬Դ޵Agent˵,Ҫ,ҲҪľ,ʹΪл,ȻҪںַо.
˼ά״̬оǣ(1) ڲۣо˼ά״̬ݻ,ǿ̬ԡԡʵԼݻ(2) Ϊ()Ŀǽ⡢ЧʺͿɿԵ,ΪĽ,Ϊ˼ά״̬Ƶ,о˼ά״̬βһ¡Ϊ,ʹõ.
Agent滮봫ͳ滮űʵĲ.MAS,AgentĸԻв,ϵͳϱǨ,ʹù滮ɹִ̺й̽֯һ,κʱ̵Ĺ滮ֻǲֹ滮,ʱ,˼ά״̬ͼලAgent滮𵽹ؼԵ.
ʶΪ߼Чۡڲ۵ⲿ۵㡢ۺʵɽԺͿɼԼ̬Ͷ̬,AgentΪǰصDAIогì,о۶ӦѵЩìΪŬķ.ĿǰڿչĹУ
(1) BDIģеĸȷ߼ȫ֪ȱ̬ϵȱ,Ըͼ˼ά״̬֪ĺ,AgentĶ̬BDIģ.
(2) Agent,AgentȺĺṹͺ˼ά״̬.
(3) BDIģص˼ά,淶Эо,AgentBDIģ.
(4) Ч÷ĳŵ.
(5) огAgent˼ά״̬Ķģ.
л ĵоõȻѧ,ĿΪ6977302669733020.MAS۰Ķ.μ۰ĻУСǿ㡢һ·֡﷼졢ͬ־,ӦҲμ۲ָ,ڴһʾл.
*оõȻѧ.
ͨѶϵ:ΰ,100084,廪ѧѧ뼼ϵ
߼顡ΰ,1969,ʿ,ҪоΪAgentϵͳ,Ϊ淶.
,1966,ʿ,ҪоΪAgentϵͳ,˻.ʯһ,1935,,ʿʦ,ҪоΪ˹Ӧû.
ߵλ廪ѧѧ뼼ϵ100084
E-mail: mgw@est4.cs.tsinghua.edu.cn
ο
1Bradshaw M. An introduction to software agents. In: Bradshaw M ed. Software Agents. Menlo Park, California: AAAI Press, 1997. 346
2Wooldridge M, Jennings N R. Intelligent agents: theory and practice. The Knowledge Engineering Review, 1995,10(2):115152
3Haddadi A S. Communication and Cooperation in Agent Systems. Berlin: Springer-Verlag KG, 1996. 1134
4Dongha P. Toward a formal model of commitment for resource bounded agents. In: Wooldridge M J, Jennings N R eds. Intelligent Agents, Proceedings of the ECAI'94 Workshop on Agent Theories, Architectures, and Languages. Berlin: Springer-Verlag KG, 1995. 86101
5Shi Chun-yi et al. The advances of distributed artificial intelligence. Pattern Recognition and Artificial Intelligence, 1995,8(supplement):7292
6Tuomela R. Philosophy and distributed artificial intelligence: the case of joint intention. In: O'Hare G, Jennings N R eds. Foundations of Distributed Artificial Intelligence. New York: John Wiley & Sons, Inc., 1996. 487504
7Cohen P R, Levesque H J. Intention is choice with commitment. Artificial Intelligence, 1990,42(23):213261
8Rao A S, Georgeff M P. Modeling rational agents within a BDI-architecture. In: Allen J, Fikes R, Sandewall E eds. Principles of knowledge representation and reasoning: Proceedings of the 2nd International Conference (KR91). San Mateo, CA: Morgan Kaufmann Publishers, Inc., 1991. 473484
9Konolige K, Pollack M E. A representationalist theory of intention. In: Bajcsy R ed. Proceedings of the 13th International Joint Conference on Artificial Intelligence. San Mateo, CA: Morgan Kaufmann Publishers, Inc., 1993. 390395
10Bell J. Changing attitudes. In: Wooldridge M J, Jennings N R eds. Intelligent Agents, Proceedings of the ECAI'94 Workshop on Agent Theories, Architectures, and Languages. Berlin: Springer-Verlag KG, 1995. 4055
11Gaspar G, Coelho H. Where do intentions come from?: a framework for goals and intentions adoption, derivation and evolution. In: Ferreira C P, Mamede N J eds. Progress in Artificial Intelligence, Proceedings of the 7th Portuguese Conference on Artificial Intelligence, EPIA'95. Berlin: Springer-Verlag KG, 1995. 115128
12Linder B van, Hoek W van der, Meyer J. Formalising motivational attitudes of agents: on preferences, goals and commitments. In: Wooldridge M, Mller J P, Tambe M eds. Intelligent Agents II: Agent Theories, Architectures, and Languages, Proceedings, 1995. Berlin: Springer-Verlag KG, 1996. 1732
13Cavedon L, Padgham L, Rao A et al. Revisiting rationality for agents with intentions. In: Xin Yao ed. Proceedings of the 8th Australian Joint Conference on Artificial Intelligence. Singapore: World Scientific Publishing Co. Pte. Ltd., 1995. 131138
14 Georgeff M P, Rao A S. The semantics of intention maintenance for rational agents. In: Mellish C S ed. Proceedings of the 14th International Joint Conference on Artificial Intelligence. San Mateo, CA: Morgan Kaufmann Publishers, Inc., 1995. 704710
15쳯.ͼάṹۼǵо۲ʿѧλģ.Ͼպѧ,1998
(Zhu Zhao-hui. Two-dimensional structure intention theory and nonmonotonic reasoning Ph.D Thesis. Nanjing University of Aeronautics and Astronautics, 1998)
16Wooldridge M. This is MYWORLD: the logic of an agent-oriented DAI testbed. In: Wooldridge M J, Jennings N R eds. Intelligent Agents, Proceedings of the ECAI'94 Workshop on Agent Theories, Architectures, and Languages. Berlin: Springer-Verlag KG, 1995. 160178
17Rao A S. AgentSpeak(L): BDI agents speak out in a logical computable language. In: Velde Walter Van de, Perram John W eds. Agents Breaking Away, Proceedings of the 7th European Workshop on Modelling Autonomous Agents in a Multi-Agent World, MAAMAW'96. Berlin: Springer-Verlag KG, 1996. 4255
18Wooldridge M. Temporal belief logics for modeling distributed artificial intelligence systems. In: O'Hare G, Jennings N R eds. Foundations of Distributed Artificial Intelligence. New York: John Wiley & Sons, Inc., 1996. 267286
19Levesque H J, Cohen P R, Nunes J H T. On acting together. In: Proceedings of the 8th National Conference on Artificial Intelligence. Menlo Park, CA: AAAI Press/MIT Press, 1990. 9499
20Castelfranchi C, Conte R. Distributed artificial intelligence and social science: critical issues. In: O'Hare G, Jennings N R eds. Foundations of Distributed Artificial Intelligence. New York: John Wiley & Sons, Inc., 1996. 527544
21Castelfranchi C. Commitments: from individual intentions to groups and organizations. In: Victor Lesser ed. Proceedings of the 1st International Conference on Multi-Agent Systems. Menlo Park, California: AAAI Press/The MIT Press, 1995. 4148
22Grosz B F, Kraus S. Collaborative plans for complex group action. Artificial Intelligence, 1996,86(2):269357
23Dunin-Keplicz B, Verbrugge R. Collective commitments. In: Durfee F ed. Proceedings of the 2nd International Conference on Multi-Agent Systems. Menlo Park, California: AAAI Press, 1996. 5663
24Georgeff M P, Ingrand F F. Decision-making in an embedded reasoning system. In: Sridharan N S ed. Proceedings of the 11th International Joint Conference on Artificial Intelligence. San Mateo, CA: Morgan Kaufmann Publishers, Inc., 1989. 972978
25Rao A S, Georgeff M P. An abstract architecture for rational agents, In: Nebel B, Rich C, Swartout W eds. Principles of Knowledge Representation and Reasoning: Proceedings of the 3rd International Conference (KR'92). San Mateo, CA: Morgan Kaufmann Publishers, Inc., 1992. 439449
26Jennings N R. Specification and implementation of a belief-desire-joint-intention architecture for collaborative problem solving. International Journal of Intelligent and Cooperative Information Systems, 1993,2(3):289318
27Mller J P. The design of intelligent agents: a layered approach. Berlin: Springer-Verlag KG, 1996. 1227
28Jennings N R. Controlling cooperative problem solving in industrial multi-agent systems using joint intentions. Artificial Intelligence, 1995,75(2):195240
29Tambe M. Agent architectures for flexible, practical teamwork. In: Proceedings of the 14th National Conference on Artificial Intelligence. Menlo Park, CA: AAAI Press, 1997. 2228
30Shoham Y. An overview of agent-oriented programming. In: Bradshaw M ed. Software Agents. Menlo Park, California: AAAI Press, 1997. 271289
1998-07-01յԭ,1998-10-23յ޸ĸ
