目标驱动—情境感知的柔性Web服务组合研究

发布时间：2019-01-07 10:40

【摘要】：随着新的信息技术和网络技术的发展,突发事件与信息技术耦合,变得更加复杂,呈现出跨界、交互和不确定的特性,其发生频率和影响范围都出现迅速增长的态势。这种情况下,农业应急管理信息化面临最大的考验就是如何针对突发事件的即时需求,及时、准确、快速地进行应急组织,协同各部门和各主体做出应急响应和处理。而Web服务组合技术,作为一种资源整合的模式,其核心思想是通过有效地结合不同功能的Web服务资源来实现更强大的功能,进而解决更复杂的问题。因此,Web服务组合技术的研究和发展恰好与农业应急管理信息化建设的需求相契合,能够为新环境下的农业应急管理提供技术支持。然而,现有的Web服务组合方法通常采用预先定义完整的服务组合方案方式,难以适应在应急环境下用户需求的不明确性、执行过程中的不断演化性的特征。因此,需要充分考虑服务组合中和用户相关的柔性问题、和资源相关的柔性以及和服务所处情境相关的柔性问题,以提高服务组合的效率和适应性。本论文围绕服务组合生命周期中各个阶段的柔性需求,对服务组合的描述、服务组合的自动生成以及服务执行过程中的动态调整进行了研究,主要研究内容包括以下几个方面： (1)提出了一种基于目标和情境的服务本体模型(Goal Context based Ontology Web Language of Services, GC-OWL-S)。该模型从目标-情境-服务三元问题出发,对OWL-S本体进行了扩展和改进,使其不仅包含服务的基本属性还包含服务的需求目标标注和服务所依赖情境的属性,从而更好地刻画了服务的领域特性、功能特性和非功能特性,为正确和完备地描述组合服务和实现柔性服务奠定了基础。 (2)针对用户需求的不确定性和情境的不断演变性,建立了服务组合的动态智能规划知识模型,并设计了相应的动态智能规划方法。该模型将服务生成问题转换为动态智能规划问题,采用两阶段的服务生成策略,按照用户目标逐步分解和情境信息匹配的方式构造一个满足用户偏好需求和情境约束的服务组合方案,实现服务组合自动化生成的同时也增强了服务组合的准确性和灵活性。 (3)在充分分析了服务组合方案执行过程可能出现的异常行为的基础上,提出一种服务重定位、重规划、重建模的三层次动态调整机制,实现了服务组合方案的动态调整和优化,以在最大限度上保证服务组合方案在运行中的灵活性。 (4)在以上研究基础上,设计了一个多层次应急决策系统框架,该框架涉及应急资源的服务化描述,应急行动方案的制定以及行动过程中的调整,同时也在应用层面上验证了上述方法的有效性和实用性。
[Abstract]:With the development of new information technology and network technology, the coupling of emergency and information technology becomes more complex, showing the characteristics of cross boundary, interaction and uncertainty, and its frequency and influence range are all increasing rapidly. In this case, agricultural emergency management information faces the greatest test is how to meet the immediate needs of emergencies, timely, accurate and rapid emergency organization, and cooperate with various departments and main bodies to make emergency response and processing. As a mode of resource integration, the core idea of Web services composition technology is to realize more powerful functions by effectively combining Web services resources with different functions, and then to solve more complex problems. Therefore, the research and development of Web service composition technology coincide with the needs of agricultural emergency management information construction, and can provide technical support for agricultural emergency management under the new environment. However, existing Web service composition methods usually adopt a predefined and complete service composition scheme, which is difficult to adapt to the uncertainty of user needs in emergency environment and the evolving characteristics of execution process. Therefore, in order to improve the efficiency and adaptability of service composition, it is necessary to fully consider the flexibility problems related to users, resources and situations in service composition. This paper focuses on the flexible requirements of each stage of the service composition life cycle, and studies the description of service composition, the automatic generation of service composition and the dynamic adjustment in the process of service execution. The main contents are as follows: (1) A service ontology model (Goal Context based Ontology Web Language of Services, GC-OWL-S) based on goal and context is proposed. The model extends and improves the OWL-S ontology from the perspective of the object-situation-service ternary problem, so that it includes not only the basic attributes of the service, but also the requirement target annotation of the service and the attributes of the situation on which the service depends. Thus, the domain, functional and non-functional characteristics of services are described better, which lays a foundation for the correct and complete description of composite services and the implementation of flexible services. (2) aiming at the uncertainty of user's demand and the evolving of situation, the dynamic intelligent planning knowledge model of service composition is established, and the corresponding dynamic intelligent planning method is designed. The model transforms the waiter problem into a dynamic intelligent planning problem and adopts a two-stage service generation strategy. According to the way of user goal decomposition and situation information matching, a service composition scheme is constructed to meet the user's preference requirements and situation constraints, which can automate the generation of service composition and enhance the accuracy and flexibility of service composition. (3) based on the analysis of the possible abnormal behavior in the execution process of the service composition scheme, a three-level dynamic adjustment mechanism of service repositioning, replanning and remodeling is proposed. The dynamic adjustment and optimization of the service composition scheme are realized to ensure the flexibility of the service composition scheme in operation to the maximum extent. (4) based on the above research, a framework of multi-level emergency decision-making system is designed, which involves the service description of emergency resources, the formulation of emergency action plan and the adjustment of action process. At the same time, the effectiveness and practicability of the above methods are verified in the application level.
【学位授予单位】：中国农业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TP393.09

【参考文献】