【作者】 刘卫东；

【导师】 张申；

【作者基本信息】 中国矿业大学 ， 通信与信息系统， 2016， 博士

【摘要】 信息物理系统是一个综合计算、网络与物理环境的多维复杂系统,通过计算、通信与控制技术的有机融合与深度协作,实现大型工程系统的实时感知、动态控制和信息服务。兴隆庄煤矿选煤厂在前期建设过程中已实现自动化控制,但工序之间的信息交互还比较弱。在此基础上,通过引入新的计算、通信和控制技术实现信息系统和控制系统更紧密的融合与协作是选煤厂信息化建设进一步发展的关键。本文基于CPS概念并结合选煤厂控制系统的特点,构建选煤厂CPS模型并对关键技术进行研究,为CPS技术在选煤厂的实施以及智能化、有自主行为的现代化选煤厂实现提供理论指导。论文首先根据工艺流程设计了基于偏序知识共享的选煤厂CPS模型,通过对每一生产单元分配部分存储用于知识库的构建,以及设定知识的异步共享与传递方式,实现了生产单元的松耦合关联。结合信息物理融合系统与复杂网络的特点,通过将随机行走机制以及基于小世界特性的RLOC拓扑优化算法引入到建立选煤厂CPS拓扑结构的过程中,使得所建立的选煤厂CPS拓扑结构具有无标度网络和小世界网络的特性,同时具有抗毁性特点。任务调度是连接选煤厂CPS任务计划与实际生产的关键环节,结合兴隆庄选煤厂PLC控制系统,论文将动态联盟思想引入选煤厂信息物理系统任务调度中,提出了基于交货期窗口的CPS静态调度模型,将TS引入到PSO搜索的后期提高了调度性能;同时基于选煤厂生产环境的动态性,提出了设备异常及紧急任务到达时的选煤厂CPS动态任务调度方法,仿真结果表明算法用于选煤厂CPS任务调度的可行性。结合选煤厂实时数据管理要求,论文提出了一种反馈控制架构来实现实时嵌入式数据库系统中事务的期望延迟,与以往实时数据库的反馈控制不同,该方法可同时控制相互作用的多种资源,更好的权衡事务的时间性和数据的新鲜度之间的关系,从而提供更多的有效性和鲁棒性。生产部件对信息输入源有着不同的可信赖度和可靠性需求,论文最后构建了选煤厂CPS网络信任度模型及评价方法,结合影响可信度的因素,利用D-S证据理论,通过引入环境函数,设计了选煤厂CPS可信度计算流程,并给出每一部分合理的计算方法,实现了选煤厂CPS可信度计算。更多还原

【Abstract】 Cyber physical system(CPS) is a kind of multi-dimension complex system that integrates computing, network and physical environment together. Through the organic integration and deep cooperation of computing, communication and control techniques, CPS realizes the real time sensing, dynamic control and information service of large engineering system. During earlier building stage of Xinlong Zhuang coal preparation plant the automation control has been realized, but the interactions among working procedures were rather weak. In such base, it is the key for the informatization of coal preparation plant to introduce new computing, communication and control techniques to realize the more closely fusion of information system and control system. In the thesis based on the conception of CPS combined with characteristic of coal preparation plant control system, the CPS model of coal preparation plant is built. Some key techniques in the CPS model are researched. The results can provide theoretic guidelines for applying CPS in coal preparation plant and building modern coal preparation plant with autonomic behavior.Firstly, according to the technological process a coal preparation plant CPS model based on partial order knowledge sharing is designed. Through dispatching partial storage to build knowledge base for each production unit, and setting the methods of knowledge asynchronous sharing and transmitting, the loose correlation of production units is realized.Combining the characteristics of CPS with complex network, and introducing random walk mechanism and small world based RLOC(Removing superfluous Links to Optimize the Cluster) topological optimization algorithm to the procedure of building topological structure of coal preparation plant CPS, the coal preparation plant CPS topological structure has the characteristic of scale-free and small world network, and at the same time has the feature of strong survivability.Task scheduling is the key point to link coal preparation plant CPS task plan with actual production. Combining Xinlong Zhuang coal preparation plant PLC(Programmable Logic Controller) control system, the thought of dynamic alliance is adopted to coal preparation plant CPS task scheduling. A CPS static scheduling model based on due date window is proposed. By introducing TS(Tabu Search) into the PSO(Particle Swarm Optimization) searching late period, the scheduling performance is enhanced. At the same time based on the dynamic characteristic of production environment of coal preparation plant a dynamic task scheduling method is also put forward for the situation of equipment exception and urgent task. Simulating results indicate the feasibility of the algorithm used for coal preparation plant CPS task scheduling.On the demand of real time data management of coal preparation plant a kind of feedback control framework is presented to realize the expected delay of transaction in real time embedding database system. Comparing with traditional real time database the method can control multiple interacted resources, and better balance the relationship of transaction time and data freshness to provide more effectiveness and robustness.Production parts have different demands on information inputs for trustworthiness and reliability. Therefore at last of the thesis a trust degree model and its evaluating method for coal preparation plant CPS network are built. Combining the factors of affecting trust, D-S(Dempster-Shafer) evidence theory is used. Through introducing environment parameters a computing flow is designed for coal preparation plant CPS trustworthiness. And suitable computing method for each part is given. The trust degree calculation is realized for coal preparation plant CPS.更多还原