【作者】 单子丹；

【导师】 高长元；

【作者基本信息】 哈尔滨理工大学 ， 管理科学与工程， 2009， 博士

【摘要】 日益激烈的全球竞争、迅速发展的知识经济使得知识型战略合作成为衡量以技术创新为主导的高技术企业综合竞争优势与持续发展能力的核心。高技术虚拟企业(High-tech Virtual Enterprise,HTVE)是虚拟企业组织形式在高技术领域中的升级,它是高技术企业整合核心优势,分散技术研发风险,应对市场个性需求的有效途径。由于成员企业之间信息占有与知识学习的不完备,参与目标与信任程度的不相同,能力贡献与利益回报的动态变化导致联盟运营过程中积聚大量的冲突问题,因此,研究与建立HTVE多边动态协商机制对协调成员合作关系、发挥联盟协同效应以及优化联盟合作绩效具有重要的理论意义和实践价值。本文在对国内外虚拟企业相关理论、方法以及多边协商研究脉络进行系统分析的基础上,给出HTVE的内涵特征、形成路径与网络演化规律,揭示HTVE冲突产生的动因与层次类型,构建了合作冲突对联盟整体收益的SMART影响模型;通过对HTVE协商决策组织模式的深入研究,提出协商模式的选择依据及其管理规则,结合成员企业的任务角色探讨HTVE谈判力变迁的决定因素和过程特性,建立了HTVE协商成员谈判能力的测度模型;并从实际管理与应用角度设计出IITVE多边协商机制的总体框架、功能结构和协作方式。形式化描述HTVE中Multi-Agent协商决策过程,着重研究协商个体隐式效用曲线的空间分布特征和多重冲突议题的递阶层级结构,在对Multi-Agent协商联合状态逻辑分析的基础上,引入效用算子和概率算子作为协商个体形成动态提议区间方案的衡量依据和评价方法。据此建立了多边多议题交互协商的粗规划模型和细规划模型。从Multi-Agent协商策略的影响因素出发,以交互时间依赖策略、协商资源依赖策略、参与行为模仿策略和多议题互补策略作为策略联结的单项变量,提出HTVE组合协商策略的实施方案、运行原则和调整方法。通过分析各类协商策略选择机制间的互动关系,得出个体谈判对策对协商模型运行效果的作用与影响方式,形成HTVE中Multi-Agent协商策略的配置体系。建立了HTVE中Multi-Agent合作协议的管理机制,基于自适应改进粒子群算法生成Multi-Agent交互协商结果,分析比较各类自然启发式算法间的优化性能。提出HTVE协商内部信任的全过程管理结构,并通过协商直接信任度、间接信任度以及合作友好度构建了协商可信度计量模型,为成员企业信誉等级评定提供参考性依据;建立了HTVE合作成员的动态奖惩机制,确定协商奖惩的分析对象,提出奖励再评估和惩罚可申辩的管理体制,为HTVE技术标准的形成、发展与确立提供行为规范与制度保证。利用领域本体的思想,搭建了HTVE中Multi-Agent协商协议平台,为群体合作规则的共享与合成提供语义处理方法和通信手段。基于智能谈判支持系统(Intelligence Negotiation Support System,INSS)双驱动集成的问题处理机制,形成HTVE智能谈判支持系统(HTVE-INSS)的总体结构与模块功能,结合HTVE协商决策的组织方式,建立了HTVE-INSS的匹配模式。分别从协商决策问题的识别与理解、合作求解模型的调用架构、INSS综合推理的运行机制入手,给出冲突议题特征抽取、变换与规约的处理方法,提出概念-元模型、逻辑-结构模型和应用-业务模型多维模型的构成体系,利用在线合作学习与离线经验学习构建了INSS综合推理的执行结构,基于向量夹角余弦的证据合成理论,建立了HTVE-INSS知识推理与组织的运作机理,并针对INSS推理知识的存储器-动态知识库进行全面的分析与设计。在上述研究的基础上,以大庆石科科技开发有限公司发起的DSR-16BT项目作为实证研究的应用背景,采集HTVE成员企业关于互补冲突议题的谈判数据,结合参与成员协商策略的选择倾向,比较分析实施HTVE多边协商机制前后议题谈判的交互效果,并实现DSR-INSS部分模块的关键过程,验证HTVE动态协商机制及其支持系统的可操作性和有效性,提出DSR-16BT项目运行的有关对策与建议。本文形成了HTVE在不完备信息条件下多边多议题多阶段的交互式协商机制和冲突消解理论、方法与策略体系,既可为HTVE项目任务的稳定发展提供科学化、规范化和制度化的理论指导和方法支持,也可为高技术产业整体竞争力的强化与提升提供解决方案。HTVE-INSS的建立既可以提高HTVE协商机制的运行效率与效力,也可为网络合作联盟提供异地远程智能谈判的信息处理方法与辅助决策工具。更多还原

【Abstract】 Increasing ungovernable global competition and vertiginous knowledge economy make knowledge-type strategy collaboration become the core to measure High-tech enterprise synthesis competition advantage and persistent development capability with the dominant for technology innovation. High-tech Virtual Enterprise(HTVE) is called the virtual enterprise organization form upgrade in the high technology domain, which is the effective approach for core advantage conformity, technology R&D risk decentralization and market individuation demand of High-tech enterprise. For information possession and knowledge learning imperfection among members, for participating target and trust degree disaffinity and for capability contribution and benefit dynamic change, those cause coalition business accumulate large numbers of conflict problem. Therefore, research and establish HTVE Multi-Agent dynamic negotiation mechanism has the great academic significance and practical meaning to coordinating member cooperative relation, promoting coalition collaborative effect and optimizing coalition cooperative performance.Based on the systematic analysis of domestic and foreign theory and method related to virtual enterprise as well as the research aspects of Multi-Agent negotiation, this paper presents the connotation, characteristic, forming route and evolvement rule for HTVE, expounds conflict forming causation and type level and establishes SMART influence model on coalition benefit, Meanwhile HTVE negotiation decision organization mode is thorough researched, selection standard and management rule for negotiation mode is also put forward, according to participants assignment role it discusses determinative factors and process specialty for negotiation power transition, as well as establishing HTVE negotiation participants negotiation capability measure model; from practical managing and applying viewpoint, overall framework, function structure and collaborative ways for HTVE Multi-Agent negotiation mechanism is also designed.After formalized describing HTVE Multi-Agent negotiation decision process, it emphasizes space distribution characteristic for individual obscure utility curve and hierarchical structure for Multi-issue conflict, based on logic analysis for Multi-Agent negotiation united state, utility arithmetic operators and probability arithmetic operators is imported as measure reference and evaluation method for dynamic proposal interval scheme. Accordingly, Multi-Agent and Multi-issue alternative negotiation general programming model and particular programming model are established.From the influence factors for Multi-Agent negotiation strategy, it takes alternative time depending strategy, negotiation resource depending strategy, participating behavior imitating strategy and Multi-issue compensation strategy as monomial variables of coupling strategy, provides the executing scheme, operating principal and adjusting method for HTVE combination negotiation. Based on interactive relation analysis for various negotiation strategies selection mechanism, individual negotiation strategy effect and influencing ways to negotiation decision model is gained, and finally the configuration system of HTVE Multi-Agent negotiation strategies is formed.It constructs HTVE Multi-Agent collaboration protocol management mechanism. Multi-Agent iterative negotiation results are created based on self-adaptive improvement Particle Swarm Optimization as well as comparing optimizing capability on several of native heuristic arithmetic. Comprehensive process management structure for HTVE negotiation inner trust is brought forward, from negotiation direct trust degree, indirect trust degree and collaborative amity degree, it establishes negotiation reliability measure model, which provides the rank references assessed for participants’ credit; HTVE participants dynamic rewards and punishment mechanism is constituted, confirming correlative analysis objects and putting forwards management system for rewards upper evaluation and punishment upper averment, which forms the behavior criterion and system pledge for forming, developing and establishing HTVE technology standard. Using domain ontology thought, it constructs HTVE Multi-Agent negotiation protocol platform, providing semantic processing method and communication means for colony collaborative rule share and composition.According to problem process mechanism by double and integrated drive for Intelligence Negotiation Support System(INSS), detailed configuration and module function of HTVE Intelligence Negotiation Support System(HTVE-INSS) is constructed, and then this paper forms HTVE-INSS matching mode by the analysis of HTVE negotiation decision organization ways. Meanwhile, from the aspects of negotiation decision problem reorganization and comprehension, collaborative solution model transferring framework, and INSS synthesis reasoning mechanism, conflict issue processing method for character extraction, transformation and stipulation is presented, multidimensional model composing system including concept-meta model, logic-structure model and application-business model is provided; INSS synthesis reasoning executing structure using on-line collaborative learning and off-line experience learning, based on argument combination theory with vector angle cosine, HTVE-INSS knowledge reasoning and organization operating mechanism is established, and then dynamic knowledge base is thorough analyzed and designed as reasoning knowledge memory implement for INSS.Based on the above research, this paper takes DSR-16BT project as application background of empirical research initiated by Daqing Shike ltd. Collects negotiation data on compensatory conflict issue from HTVE participants, according to the selection preference for participants negotiation strategy, issue negotiation iterative effects using and non-using HTVE multi-agent negotiation mechanism are compared and analyzed. In order to certify the feasibility and validity of HTVE negotiation dynamic mechanism and its support system, it realizes HTVE-INSS important process for part module, and brings forward concerned countermeasures and suggestions on DSR-16BT project.The HTVE Multi-Agent, Multi-issue and Multi-stage iterative negotiation mechanism and conflict resolution theory, method and strategy system in the condition of imperfection information, would offer advantageous theory guidance and scientific method support for scientific and institutionalized and standardized HTVE project fine development as well as settlement scheme for High-tech industry whole competition aggrandizement and improvement. HTVE-INSS would enhance operation efficiency and effect of HTVE negotiation mechanism, as well as providing the information processing method and aid decision tool for distributed and long-distance intelligence negotiation in network collaborative coalition.更多还原