铁道供电调度拥塞控制机制与海量信息集群压缩算法研究

Information flow of high-speed railway power supply dispatching is fast-changed for multipoint and burst characteristics, which is easy to cause fault delay or omission when excessive amount of information need to handle in the event of an emergency, massive information of long term operation will make access response too slow, or even crash, which will directly threaten the safe operation of power supply dispatching. Suddenly information and massive information for railway power supply system are as research object in this project, which will have a basic scientific research using newsboy model, publish subscribe mode, intelligent interaction method and cluster compression algorithm. Contents are as follows: (1)Establishing newsboy publish subscrible model, to control the quantitative relationship between information publish and system subscrible delay; (2)Using information intelligent interaction method, with intelligent alliance and news-event trigger communication mode, to explore new anti-congestion mechanism for emergency information; (3)Establishing cluster compression access component that massive storage information can be high ratio compression, which will fuse cloud platform and decomposition aggregation engine, and research new lossless cluster compression algorithm. The aims of this project are to reveal congested conditions and law of unexpected information in railway power supply system,to enhance the ability with sudden fault, and improve compression ratio of massive monitoring information, that will lay theoretical and technical foundation for safety dispatching of railway traction power supply and signal power supply.

高速铁道供电调度的信息流，具有点多、变化快和突发性特点，突发故障处理的信息量过大时，容易造成故障迟报、误报甚至漏报，长期运行产生海量信息导致存取反应慢，甚至死机，直接威胁供电调度的安全运行。本项目以铁道供电突发信息和海量信息为研究对象，利用报童模型、发布订阅模式、智能交互方法和集群压缩算法进行基础科学研究。内容包括：(1)建立报童发布订阅模型，利用报童模型调控信息发布量，研究对订阅信息延时的影响；(2)运用信息智能交互方法，借助智能联盟和消息事件触发器通信模式，探索突发信息的自主抗拥塞交互新机制；(3)建立海量存储信息大比率压缩存取的集群压缩组件，融合云平台和分解聚合引擎，研究无损集群压缩新算法。本项目旨在揭示铁道供电突发信息拥塞发生的条件和规律，提升应对突发故障的能力，提高海量监测信息的无损压缩率，为保障铁道牵引供电和信号供电的安全调度奠定理论与技术基础。

高速电气化铁道供电系统的调度监控信息流，具有点多、变化快和突发性特点，故障处理的信息量过大时会造成故障迟报或误报，长期运行的海量信息导致存取反应慢，有时甚至死机。本项目以铁道供电监测信息为研究对象，利用报童收益模型、发布订阅模式、智能交互方法和集群压缩等手段进行了基础科学研究。内容包括：(1)利用发布订阅消息交互技术，建立发布订阅的报童优化模型，结合观察者设计模式，提出基于发布订阅和报童收益的调度监控系统应用信息处理方法。(2)通过推导网络结构、层次结构、联盟结构智能体间信息的交互代价，比较得到分布式实时监控的信息联盟结构，利用消息触发技术封装批量实时监测信息，实现由关系模型、持久化对象模型到监控信息智能体交互的映射。(3)针对铁道供电调度信息压缩存储处理的研究，采用Hadoop云计算及Hive数据仓库框架，对海量信息进行分布式云存取和集群压缩处理，获得了铁路供电调度信息存储处理的集群无损压缩处理方法。

内容获取失败，请点击重试