【作者】 鄂越；

【导师】 曹永生； 诸叶平；

【作者基本信息】 中国农业科学院 ， 作物信息科学， 2012， 博士

【摘要】 我国蜂产业组织比较松散，人为因素作用较大，还缺乏信息化、智能化技术作为质量安全控制的支撑手段，进而导致蜂产品从“蜂场到餐桌”所经历的各个环节的质量不易控制。因此，如何对蜂产品质量安全进行有效管理与控制是近年来我国蜂业研究领域的热点问题。本论文运用现代信息技术，一方面，建立蜂产品质量安全控制系统，为蜂产品质量控制提供一种技术支持和通用性工具；另一方面，作为一种关键技术的研究和探索，采用Agent技术，提出一种具有智能性、协同性的蜂产品质量控制方法，为蜂产品质量安全控制研究提供新的研究思路。这对于我国养蜂业向信息化、智能化管理模式转变，提高我国蜂产品质量安全管理与控制水平，具有十分必要的现实意义和实用价值本论文主要的研究内容：（1）在分析研究我国蜂产品质量控制的基础上，采用Agent技术识别与定义了参与蜂产品质量控制的实体，提出了基于质量控制动态联盟形式的蜂产品质量控制系统体系结构参考模型，详述了参考模型的层次结构，为系统进一步实现奠定了基础。（2）针对蜂产品质量控制，提出了一个层次化的多维的Agent模型构建方法，确定了Agent模型的功能和结构，给出了Agent模型中一些关键行为的驱动规则，保证了Agent模型具有较好的适应性和灵活性。（3）提出了蜂产品质量控制链的概念。研究了基于Agent的蜂产品质量监控与应急处置的方法，给出了详尽的基于Agent逻辑架构的信息交互模式和流程图；定义了参与蜂产品质量控制的对象和属性与Agent的映射关系，并给出了基于Agent蜂产品质量控制的实现策略。（4）从Agent协同工作的角度，研究和探讨了蜂产品质量安全应急决策机制，提出了蜂产品质量应急决策模型，基于任务分解、多目标的规划实现了蜂产品应急辅助决策任务的制定。基于黑板通讯机制，给出了应急决策任务存储的层次结构、访问方法，为建立基于Agent的蜂产品质量应急辅助决策提供了数据信息层支持。本论文主要创新点：（1）提出了基于Agent的蜂产品质量控制框架，开展了基于动态联盟形式的质量控制方法研究。（2）设计了基于Agent的蜂产品质量控制链，提出了蜂产品质量控制中Agent模型的构建方法，实现了系统中各个功能Agent模块的设计；提出了基于Agent的蜂产品质量安全应急机制。（3）设计开发了基于Agent的蜂产品质量控制系统平台，验证了论文所提方法的有效性。更多还原

【Abstract】 In China, the bee farms are only organized loosely, so human factor has a great impact on beeproduct quality (BPQ). At the same time, the quality control methods related to information andintelligence technology did not give a useful support on BPQ. As a result, it is not easy control BPQduring production process of bee products. Therefore, how to control BPQ safety effectively always is ahot topic in BPQ control research field.In this study, using modern information technology, BPQ control system was developed to a kindof provide technical support and a common tool for BPQ control. Using a key technology based onAgent, a BPQ control method with intelligent and coordination was built to give a new researchapproach for BPQ control. This effort can promote apiculture management with informationization andintellectualization and upgrade BPQ management level, so it is very necessary and can be a tool withpractical value.The main research work as follow:(1) Based on the analysis of BPQ control in China, the BPQ control entity is identified by theAgent analysis, and then the reference model of BPQ control is proposed, which is dynamic union(DU)-based. The hierarchical structure of the reference model is also described at length, which is basefor model use.(2) For BPQ control, a hierarchical multidimensional Agent model is developed in this study. Theframework of agent model is presented and some issues in implementation are discussed. Also severalkey status-based action rules are presented to ensure the higher adaptability and flexibility.(3)The BPQ control chain is proposed. BPQ control and emergency mechanism is designed basedon Agent technology, and the Agent logical architecture, the data mapping and conversion model, workdata flow are detailed. In the architecture, the mapping relationship of Agent and BPQ control objectsand attributes is defined. The strategies of control BPQ are also proposed.(4) From Agent coordination, the BPQ safety emergency decision-making mechanism is studied.The BPQ emergency decision model was designed. In the architecture, the BPQ emergency decisiontask is implemented based on task decomposition and the multi-objective programming. Based onblackboard communication mechanism, the level of structure diagram, access methods, storagescalability and availability are detailed, which can provide data information layer support for BPQdecision task.This study incorporates several innovations as follow:(1)Agent-based BPQ control framework is proposed, and the quality control method based ondynamic union is studied.(2) Base on Agent technology, the bee product quality control chain was designed. The Agentmodel construction method for BPQ control is proposed, and then each function module in the modelcan be implemented. Agent-based BPQ safety emergency mechanism is also proposed. (3) The Agent-based BPQ control platform was built. By model evaluation, some of ideas anddesigns are validated for BQP control更多还原