城市轨道交通线网应急安全保障关键技术研究

发布时间：2018-04-15 15:49

本文选题：城市轨道交通 + 线网　；参考：《兰州交通大学》2014年博士论文

【摘要】：随着城轨交通系统大发展，城轨线网规模不断扩大，线网安全运营对线网的安全保障和线网大数据处理能力提出了新的要求。论文从线网运营应急安全角度出发，分析了线网指挥平台在应急安全保障方面的不足。针对这些不足，本文做了以下研究：针对线网运营安全问题，论文首先从线网组成元素的安全性入手，从三层结构不同粒度出发，对影响线网运营安全的安全因素进行了分析，完善了安全因素的安全模型,接着论文又对线网组成元素之间的相互影响和关联进行分析,提出了多粒度与关联度相结合的线网安全评估模型，为线网综合安全评估方法提供理论支持。最后运用了层次分析和模糊决策相结合的评估方法，对城轨线网安全进行动态评价，并使用马尔可夫原理对城轨线网突发事件进行分类演化和分级预测，采用闭环反馈的思想动态修改知识库，使安全评价结果和突发事件的分类分级更加准确合理，为应急预案的自动生成提供依据。针对线网安全评估结果中的应急处置，论文首先以ABC本体为原语本体，在城轨线网领域进行词汇扩充，完善了城轨线网应急预案领域本体，其次使用OWL本体描述语言和RDF对线网应急预案进行形式化描述，最后给出了线网应急事件处置方案自动生成模型。针对线网安全评估中应急安全信息的及时性要求，论文首先分析了线网指挥平台上各类数据信息的一致性和及时性，然后以Pull和Push为信息分发元机制，提出了将Push和Pull相结合的信息分发机制，实验结果表明，，Push和Pull相结合的信息分发机制一方面提高了系统柔性，另一方面通过动态调节参数，使该机制够满足城轨线网中各类数据信息分发的及时性需求。针对线网安全评估、应急决策和应急处置自动生成过程中会发生大数据任务处理，论文首先分析了线网指挥平台上处理的数据特征，搭建了线网云接口和线网云平台架构，然后对云环境下城轨线网大数据任务调度进行建模，采用PSO算法尝试使用云环境对城轨线网大数据任务调度进行优化，优化结果表明本文算法在调度长度上优于两步法，同时也证明了云计算适合处理线网大数据任务。论文最后对线网应急安全保障系统进行了设计和部分功能的实现。
[Abstract]:With the rapid development of urban rail transit system, the scale of urban rail network is expanding, and the safety of line network and the ability of big data to deal with it put forward new requirements for the safe operation of line network.From the point of view of emergency security of network operation, this paper analyzes the shortage of network command platform in emergency security.In view of these shortcomings, this paper has done the following research:Aiming at the problem of network operation security, this paper starts with the security of network elements, starting with the different granularity of three-layer structure, analyzes the security factors that affect the network operation safety, and perfects the security model of the security factors.Then, the paper analyzes the interaction and correlation between the elements of the network, and puts forward a network security evaluation model combining multi-granularity and correlation degree, which provides the theoretical support for the comprehensive security assessment method of the network.Finally, the method of AHP and fuzzy decision is used to evaluate the safety of urban rail network dynamically, and the Markov principle is used to classify and predict the unexpected events of urban rail network.The idea of closed-loop feedback is used to dynamically modify the knowledge base, which makes the result of safety evaluation and the classification and grading of emergency more accurate and reasonable, and provides the basis for the automatic generation of emergency plan.In view of the emergency disposal in the result of network security evaluation, the paper firstly takes ABC ontology as the original language ontology, expands the vocabulary in the field of urban rail network, and consummates the domain ontology of emergency plan of urban rail network.Secondly, the OWL ontology description language and RDF are used to formalize the emergency plan of wire network. Finally, the automatic generation model of emergency response scheme is given.In view of the timeliness requirement of emergency security information in network security assessment, this paper first analyzes the consistency and timeliness of all kinds of data information on the network command platform, and then takes Pull and Push as the information distribution meta-mechanism.An information distribution mechanism combining Push and Pull is proposed. The experimental results show that the information distribution mechanism combined with push and Pull improves the flexibility of the system on the one hand, and adjusts the parameters dynamically on the other.This mechanism can meet the timely demand of distribution of all kinds of data and information in urban rail network.Big data task processing will occur in the process of network security evaluation, emergency decision and emergency disposal automatic generation. Firstly, this paper analyzes the data characteristics of the network command platform, builds the network cloud interface and the network cloud platform architecture.Then, the task scheduling of big data in urban rail network under cloud environment is modeled, and the PSO algorithm is used to try to optimize task scheduling in urban rail network. The optimization results show that the algorithm in this paper is better than two-step method in scheduling length.At the same time, it also proved that cloud computing is suitable for big data task.At the end of the paper, the emergency safety system of wire network is designed and some functions are realized.
【学位授予单位】：兰州交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U298

【参考文献】