摘要
突发事件应急管理中的协同式群决策问题是一种复杂系统在复杂条件下协同决策的科学问题。此问题不但要以应急预案体系和各协作成员的知识库与数据库为基础，同时还要考虑不同协作成员的协作方式、推理方式以及组织规范等因素。突发事件处置过程中的协同式群决策目标是基于应急预案中各种过程、组织与协作要求根据各个成员的局部可选方案快速生成针对突发事件的总体应急处置方案。目前由于突发事件的处置过程具有复杂巨系统的特点，在进行协同式群决策过程中，会产生决策方式各异、各个协作成员自律性差、组合方案庞杂和系统稳定性与有效性差等一系列问题。因此本文针对上述问题进行以下具体研究：
1、根据突发事件应急管理的协同处置过程，构建了基于规范的应急处置过程云计算服务框架。通过在应急多核决策系统中引入组织规范和约束要求，增强了各个成员对任务的自律性和协商能力，使得系统的协作方式与应急环境中各部门的协作形式更加接近，以充分调动云计算系统的服务资源。
2、建立了基于组织规范的多核心任务规划和协同逻辑推理模型。为能够提高系统决策的适应性和决策效率，设计了基于突发事件处置流程的多层次推理方式。处置过程中每个部门中都具有不同形式的推理行为，尤其是在物联网条件下，每个部门的专业领域系统都具有高度的独立性和智能性。因此这使得多部门联动的不同的推理方式之间的衔接成为成功进行决策推理的关键。
3、根据部门内部和部门间的应急协作方式，讨论了构架内核中心内部成员与核中心成员之间的协同关系，基于此建立了基于情景的核逻辑推理模式和基于情景的多核应急协同决策推理模型。使系统在做决策时将突发事件情景与各部门基于应急处理方式的逻辑推理模式同时考虑，以确定具体应急情景下的行动方案。
4、建立了基于情景和语义的逻辑推理模式匹配方法。系统进行决策时，情景确定不但要与真实的突发事件环境相一致，还要与相互协作推理部门的请求语义相一致。通过建立一种环境-情景-语义的逻辑推理模式匹配方法，能够为各个agent在进行决策时确定与环境和协作成员适应度高的情景。
5、从系统角度引入了多目标鲁棒优化方法实现应急处置过程中的推理鲁棒性和有效性。突发事件决策过程是一种分布式系统决策过程。因此从系统角度看，系统的稳定性、有效性以及应急管理所要求的时效性，必须满足突发事件处置过程的要求，这就需要从系统角度对协同决策进行优化，实现其在真实环境中的鲁棒能力。由此通过建立相应的多目标鲁棒优化模型和设计非支配排序遗传算法实现了系统的稳定性。
The Collaborative group decision making problem is a scientific problems undercomplicated conditions of coordinated decision in emergency management. Theproblems not only based on the emergency plan system and each collaborationmembers of the knowledge base and database, also want to consider the differentcooperation members of the collaboration way, reasoning method and Organizationstandard and other factors. The target is to product a general emergency disposal plan,based on the emergency plan of all kinds of process, organization and coordinationrequirements, according to various members of the local alternative. At present, it willproduce a series of problems it includes different ways of decision-making, eachcollaboration members self-discipline poor, confused combination scheme and theworse of the system stability and effectiveness, considerring the disposal process ofcomplex giant system characteristics,in Collaborative group decision making process.In this thesis, much research is deployed aiming at resolving above problems. Thedetails of results are as follows:
1) A multi-agent mission planning and decision making framework based onorganization standard has been established. The framework added self-discipline andnegotiation ability about members, made multi-agent cooperation way and thecooperation form of different departments in emergency environment more closer,thus, improved the system decision adaptability and decision-making efficiency byintroducing the organizational standard and constraint requirements in emergencymulti-agent decision-making system.
2) The emergency disposal process of multi-level inference method is designed.The disposal process in each department has different forms of reasoning behavior.Especially in internet of things, each department of professional fields had highlyindependence and intelligent. This made the join which come from different ways ofreasoning in multi sectoral linkage become critical in decision making reasoning.
3) According to the department of a variety of alternative emergency disposal,based on scenarios of logical reasoning mode determining method has beenestablished. Based on scenarios of logical reasoning mode determining method hasbeen put forward, in the emergency disposal process for scenarios alternativespartition problem. In system making decisions, emergency scenarios and departmentalemergency treatment were considered at the same time, the target is to determine the programme of action in some specific emergency scenarios.
4) Base on the scenarios and semantic reasoning pattern matching method hasbeen established. When the system made a decision, it made the emergency situationand emergency treatment as a factor, at the same time. The scenario was identified notonly with the real accident environment consistent, but also with collaborativereasoning department requests a semantic consistency. It can be determined for eachagent high fitness scene, which can match the environment and the collaborativemembers, when making a decision.
5) From the angle of system introducing the multi-objective robust optimizationmethod, it can realize the robust and effective of reasoning in the emergency disposalprocess. From the system point of view, the system stability, validity and emergencymanagement requirements of timeliness must meet the requirements of emergencydisposal process. This required to optimizate multi-agent collaborative decisionmaking from the system point and achieved its robust ability in real environment.Thus, it established the corresponding multi-objective robust optimization model anddesigned non dominated sorting genetic algorithm to achieve the stability of thesystem.
引文
[1]钟开斌.国家应急管理体系建设战略转变：以制度建设为中心.经济体制改革，2006(5):7-11.
[2]陈祖义.应急响应管理系统的研究与实现.东南大学硕士学位论文，2005.
[3]刘红芹，沙勇忠，刘强.应急管理协调联动机制构建：三种视角的分析.情报杂志，2011(4):18-23.
[4]贾学琼，高恩新.应急管理多元参与的动力与协调机制.中国行政管理，2011(1):70-73.
[5]Han Zhengqiang，Deng Jingyi.EmergencyProcessCapability Assessment Based onStochastic Petri Nets.Proceedings2010IEEE International Conference on EmergencyManagement and Management Sciences，2010:367-371.
[6]刘天虎，许维胜，吴启迪.大规模突发事件下医疗资源配置建模及算法.计算机工程与应用，2010,46(29):13-17.
[7]张海龙，李雄飞，王仁彪.启发式遗传算法求解应急资源调度.吉林大学学报，2010,40(3):758-762.
[8]王晓，庄亚明.基于案例推理的非常规突发事件资源需求预测.西安电子科技大学学报（社会科学版），2010,20(4):22-26.
[9]黄卫东.基于环境感知的应急管理知识需求研究.四川大学学报（工程科学版），2007,39(5):182-186.
[10]杜磊，王文俊，董存祥，李力雄，高珊.基于多Agent的应急协同Petri网建模及协同检测.计算机应用[J],2010,30(10):2567-2571.
[11]李藐，陈建国，陈涛，袁宏永.突发事件的事件链概率模型.清华大学学报（自然科学版），2010(8):1173-1177.
[12]季伟学，翁文国，赵前胜.突发事件链的定量风险分析方法.清华大学学报（自然科学版），2009(11):1749-1756.
[13]裘江南，师花艳，王延章.基于事件链的知识导航模型研究.中国管理科学，2009(1):138-143.
[14]陈安，马建华，李季梅.现代应急管理应用与实践.北京：科学出版社,2010.
[15]元菁晶，陈安.突发事件与应急管理的机理体系.科技与社会[J],2009,24(5):496-503.
[16]唐珺珺.城市应急管理系统框架模型的研究.硕士学位论文.上海：同济大学，2006.
[17]齐二石，王嵩.城市应急管理系统的构建及系统分析研究.现代管理科学[J],2008,(7):3-7.
[18]张明媛，袁永博，周晶，冯士森.基于灰色系统模型的城市承灾经济协调性分析.系统工程理论与实践[J],2008,(3):171-177.
[19]安金朝，付跃强，刘卫东.基于工作流的应急系统响应可靠性研究.机电产品的研发与创新[J],2007,(1):52-56.
[20]陈秀万，杨凯欣，方裕等.基于空间信息技术的城市应急救援联动系统研究.地理与地理信息科学[J],2003,19(4):52-55.
[21]姜立新.地震现场应急指挥技术系统的结构与设计.地震[J],2004,(3):94-99.
[22]邵全等.突发事件模型库中模型的层次网络表示方法.清华大学学报（自然科学版）[J],2009,(49):5-10.
[23]王文俊等.应急事件Ontology语义模型及其应用.计算机工程[J],2005,31(10):10-12.
[24]王旭坪，李小龙，梁阿密.基于SOA非常规突发事件资源协调决策系统研究.情报杂志[J],2010,29(6):164-169.
[25]吴建宏，翁文国，倪顺江.基于GIS和Multi-Agent的城市应急疏散.清华大学学报（自然科学版）[J],2010,50(8):1168-1172.
[26]姚杰，计雷，池宏.突发事件应急管理中的动态博弈分析.管理评论[J],2005,(3):46-50.
[27]林莉，郭建忠，王强.浅谈城市应急管理信息系统的可重构性.地理信息世界[J],2007,(2):32-36.
[28]付跃强.应急系统响应可靠性理论及在火灾应急中的应用研究.[博士学位论文],南昌:南昌大学，2008.
[29]赵淑红.应急管理中的动态博弈模型及其研究.[硕士学位论文],2008.
[30]付跃强，刘卫东，安金朝.突发公共事件应急系统的组织结构分析.江西社会科学[J],2007,(8):167-170.
[31]汪季玉,王金桃.基于案例推理的应急决策支持系统研究.管理科学[J],2003,22(6):46-51.
[32]Jackson H. First responders problems and solutions: office of emergencymanagement.Technology in Society,2003,25(4):539-541.
[33]Kevany M J. GIS in the World Trade Center Attack-trial by fire. Computers,Environment and Urban System.2003.27(6):571-583.
[34]李娟莉,闫克裕.基于本体的Apriori算法在消防应急决策系统中的应用研究.科学之友[J],2007,28(11):128-131.
[35]凌学武.公共危机管理中的协调联动机制研究.前沿[J],2007,28(9):135-138.
[36]Atoji Y, Koiso T, Nishida S. Information filtering for emergency management.Robot and Human Interactive Communication [J],2000,27(29):96-100.
[37]周芳，韩立岩.多传感器信息融合技术综述.遥感遥测[J],2006,27(3):1-7.
[38]韩崇昭，朱洪艳，段战胜.多源信息融合.北京：清华大学出版社,2006.
[39]Barron A,Rissanen J,Yu B.The minimum description length principle in codingand modeling.IEEE Transactions on Information Theory,1998,44(6):2743-2760.
[40]Zhou Y.F, Leung H. Minimum entropy approach for multi-sensor data fusion.Proceeding of the1997.IEEE Signal Processing Workshop on Higher-OrderStatistics.Los Alamitos, CA, USA: IEEE.1997.336-339.
[41]Shafer G. A mathematical theory of evidence. Princeton: Princeton UniversityPress,1976.
[42]孙怀江，胡钟山，杨静宇.基于证据理论的多分类器融合方法研究.计算机学报,2001,24(3):231-235.
[43]Smets P.The combination of evidence in the transferable belief model.IEEETransactions on Pattern Analysis and Machine Intelligence,1990,12(5):447-458.
[44]Beynon M. Curry B, Morgan P. The Dempster-Shafer theory of evidence: Analternative approach to multicriteria decision modeling. Omega (The InternationalJournal of Management Science),2000,28(1):37-50.
[45]Vapnik V.N. Statistical learning theory.New York: Wiley,1998.
[46]张学工.关于统计学习理论与支持向量机.自动化学报,2000,26(1):32-42.
[47]Vapnik V.N.统计学习理论的本质.张学工译,北京:清华大学出版社,2000.
[48]Vo B.N, Singh S, Doucet A.Sequential montecarlo methods for multitargetfiltering with rando, finite sets. IEEE Transactions on Aerospace and ElectronicSystems,2005,41(4):1224-1245.
[49]邓勇，施文康.随机集理论在数据融合中的应用研究.红外与激光工程,2002,31(6):545-549.
[50]Shu H.C, Sun X.F, Si D.J.A new method for locating faults on transmission linesbased on rough set and FNN.Proceedings of IEEE International Conference on PowerSystem Technology,2002,2584-2588.
[51]张晓祥,姚静,李满春.模糊集方法在空间数据处理中的应用综述.遥感信息,2005,38(2):47-51.
[52]邓聚龙.灰色理论教程.武汉:华中理工大学出版社,1990.
[53]Bryant I, Gonfalini M. Moffat J, et al. Understanding uncertainty. Oilfield Review,2002,14(3):2-15.
[54]Pear J.Baysian networks:A model of self-activated memory for evidencereasoning. Proceedings of the7th Conference of Cognitive Science Society,University of California, Irvine, CA.1985,329-334.
[55]Bouquet P,Giunchiglia F,Harmelen FV,et al.Contextualizing ontologies.Journal ofWeb Semantics,2004,1(4):325-343.
[56]Pierra G.Context representation in domain ontologies and its use for semanticintegration of data.Journal on Data Semantics X, LNCS4900,2008,174-211.
[57]Kaneiwa K, Mizoguchi R.Distributed reasoning with ontologies and rules inorder-sorted logic programming.Web Semantics: science, Services and Agents on theWWW,2009(7):252-270.
[58]Lutz C, Wolter F, Zakharyashev M.Temporal description logics:a survey.Proc ofTIME.2008.3-14.
[59]常亮.基于逻辑的行为理论综述.桂林电子科技大学学报,2010,30(5):440-446.
[60]单承戈.基于部门协调的分布式群决策支持系统的研究.西南交通大学博士学位论文,2005.
[61]辛明军,李伟华,沈钧毅.面向协同设计系统的分布式群决策支持技术研究.计算机工程,2005,31(6):30-33.
[62]Kuwata Y,Noda I,Ohta M, et al.Evaluation of Decision Support Systems forEmergency Management. Proceedings of the41stSICE Annual Conference,2002,2(5):860-864.
[63]Mendonca D, et al.Assessing group decision support systems for emergencyresponse using gaming simulation. Safty Science,2006,44(6):523-535.
[64]Chen M, Liou Y C, Wang C W, et al.TeamSpirit: design, implementation, andevaluation of a web-based group decision support system. Decision Support System,2007,43(4):1186-1202.
[65]Wang K J, Chien C F.Designing an Internet-based group decision supportsystem.Robotics and Computer Integrated Manufacturing,2003,9(1):65-77.
[66]Haseman D W, Nazareth D L, Paul S.Implementation of a group decision supportsystem utilizing collective memory.Information&Management.2005.42(4)591-605.
[67]Rees J, Barkhi R.The problem of highly constrained tasks in group decisionsupport system.European Journal of Operational Research,2001,135(1):220-229.
[68]徐作宁.战略信息系统规划群决策支持系统框架分析.杭州电子科技大学学报,2007,3(3):14-19.
[69]Wang H Q.Knowledge level analysis of group decision support system.ExpertSystems with Applications.1997,13(2):155-162.
[70]马应龙,金蓓弘,冯玉琳.基于进化分布式本体的语义Web服务动态发现.计算机学报,2005.28(4):603-614.
[71]M.Paolucci, K.RSycara, T.Nishimura, and N.Srinivasan. Using DAML-S for P2Pdiscovery. In Proceedings of the Intemational Conference on Web Services.2003.203-207.
[72] Sivashanmugam, K.Verma, K.Sheth, A.P.Miller, J.A. Adding Semantics to WebServices Standards.Proeeedings of the Int.Conf.on Web Services, ICWS.2003.395-401.
[73]周超,张毅.论转型期我国城市突发事件应急决策系统之构建[J].重庆城市管理职业学院学报，2006，6(4):434-437.
[74]Kwan M P，Lee J Y. Emergence response after9/11: the potential of real-time3DGIS for quick emergency response in micro-spatial environments. Computers,Environment and Urban System.2005.29(2):93-113.
[75]K.Arabsian, H.Schulzrinne, D.Trossen, andD.pavel,” Glosery: Global servicediscovery using the OWL web onto1ogy language”,in IEEE International Workshopon Intelligent Environments(IEDS).IEE,2005.
[76]N.Srinivasan,M.Paolucci, and K.Sycara,”An efficient algorithm for OWL-S basedsemantic search in UDDI, in Lecture Notes in Computer Science,vol.3387,2005.
[77]B.SaPkota,L.Vasiliu,I.Toma.D.Roman. Peer-to-Peer Technology Usage in webservice Discovery and Matchlnaking. Proceedings of the6th International Conferenceon Web Information Systems Engineering(WISE),2005.
[78] A.W.Van, M.Dumas, H.A.Ter. Web Service Composition Languages: Old Winein New Bottles. Proceedings of Euromiero Conferenee.2003. P298-305.
[79] P.F.Pires, M.Benevides, M.Mattoso. Building Reliable Web ServicesCompositions. Lecture Notes in Computer Seienee,2002.
[80] Chakraborty, F.Perieh, A.Joshi, T.W.Finin. A Reactive Service CompositionArchitecture for Pervasive Computing Environments. Proceedings of theIFIPTC6/WG6.8Working Conference on Personal Wireless Communieations.2002.53-62.
[81] S.Mellraith, T.Son. Adapting Golog for Composition of Semantic Web Services.Proceedings of the Eighth International Conference on Knowledge Representationand Reasoning.2002.482-493.
[82]S.Narayanan, S.A.Mellraith.Simulation, Verification and Automated Compositionof Web Services. Proceedings of the eleventh international conference on World WideWeb.2002.77-88.
[83] S.Mellraith, T.Son. Adapting Golog for Composition of Semantic Web Services.Proceedings of the Eighth International Conference on Knowledge Representationand Reasoning.2002.482-493.
[84] J.Rao. Semantie Web Serviee Composition via Logic-based Program synthesis.Norwegian University of Science and Technology,2004.
[85]刘峰,谭庆平,杨艳萍.基于图论的web服务合成算法.华中科技大学学报(自然科学版),2005,33(z1):202-204.
[86]刘家茂,顾宁,施伯乐.基于Mediator的Web services无回溯反向链动态合成.计算机研究与发展,2005,42(7):1153-1158.
[87] S.Narayanan, S.Mellraith. Analysis and Simulation of Web Services ComputerNetworks,2003,42(5):675-693.
[88]刘大有,杨鲲,陈建中. Agent研究现状与发展趋势[J].软件学报，2000，11(3):315-321.
[89]黄小兵,唐文胜.基于Agent系统的概念、方法和应用[J].计算机与现代化,2000(4):6-11.
[90]Agotnes, T.，Hoek, W. V. D., Rodriguez-Aguilar, J. A., et al. On the logic ofnormative systems[C]. Proceedings of the Twentieth International JointConference on Artificial Intelligence (IJCAI2007),2007:1175-1180.
[91]Conte, C. Understanding the Functions of Norms in Social Groups throughSimulation Artificial Societies [D]. The Computer Simulation of Social Life， UCLPress,1995.
[92]Foster, Kesselman, C. The grid: blueprint for a new computing infrastructure [D].Morgan-Kaufman,1999.
[93]Hatfield, U. K. Normative multi-agent systems[C].Proeeedings of theSymposium on Normative Multi-Agent Systems,2005: l-8.
[94]Jennings, N.R., Wooldridge, M. J. Software agents [J]. IEEE Review,1996,42(l):17-20.
[95]Tuomela, B. T. Norms and Agreement. European Journal of Law [J]. Philosophyand Computer Science,1995:41-46.
[96]Boella, G., Torre, L., Verhagen, H. Introduction to the special issue onnormative multiagent systems [J]. Autonomous Agents and Multi-Agent Systems，2008,17(l): l-10.
[97]Fox, M. S. An organizational view of distributed systems [J]. IEEE Transactionson Systems, Man and Cybernetics,1981,11(l):70-80.
[98]Dignum, F. Autonomous agents with norms [J]. Artificial Intelligence and Law，1999,1(7):69-79.
[99] STUDER R，et a1．Knowledge Engineering：Principles and Methods[J]． Dataand Knowledge Engineering，1998，25(1-2)：161-197．
[100]Boella, G., Torre, L. V. D. Permission and authorization in normative multiagentsystems[C].Proeeedings of the10thinternational conference on Artificialintelligence and law,2005:236-237.
[101]Grossi, D., Dignum, Frank, Meyer, John-Jules C. A formal road frominstitutional norms to organizational structures[C]. Proceedings of the6thInternational Joint Conference on Autonomous Agents and Multi-agent Systems(AAMAS07),2007:616-623.
[102]Brian, H. Cloud computing [J]. Communications of the ACM,2008,7(51):9-11
[103]Boella, G, Torre, V, Verhagen, H. Introduction to Normative MultiagentSystems [J]. Computational and Mathematical Organization Theory,2006,12(2-3):71-79.
[104]Boella, G., Torre, L. V. D. Norm negotiation in multiagent systems [J].International Journal of Cooperative Information Systems (IJCIS)Special Issue:Emergent Agent Societies,2007,16(2):97-122.
[105]曾大军，王飞跃，曹志东.开源信息在突发事件应急管理中的应用.科技导报[J],2008,26(16):27-33.
[106]Di Noia T,Di Sciacio E,Donini F.M,et al.Semantic matchmaking in a P2Pelectronic marketplace.In: Proceedings of the18th Annual ACM Symposium onApplied Computing (SAC),Melbourne.2003.582-586.
[107] Uszok A,Bradshaw J.M,Johnson M,et al.KAoS policy management for semanticweb services.IEEE Intell.Syst.2004,19(4):32-41.
[108] Haas, H.and A.Brown.Web Services Glossary.W3C Working Group Note,2004.
[109]M.Stollberg, U.Keller, D.Fensel: Partner and Service Discovery forCollaboration Establishment with Semantic Web Services.Proceedings of IEEEInternational Conference on Web Services.2005.480.