基于机器学习理论的电力通信网的安全性及可靠性研究

发布时间：2018-05-06 06:20

  本文选题：电力通信网 + 关键节点　； 参考：《南京信息工程大学》2016年硕士论文

【摘要】：电力通信网是用来服务电力系统的一种通信专网,电力通信网包括语音、数据、视频和多媒体这四大类业务。运行上述四大类电力通信业务需要电力通信网有着较高的可靠性和安全性。电力通信网在智能电网的应用中,起到了至关重要的作用。最近几年里,电力系统发生了巨大的变化,与此同时电力系统的规模也在日益的扩大,并且电力部门对电力系统的维护和管理水平也在不断的提高,然而电力通信网发生的故障次数和故障率却没有得到相应的减少和降低。其主要的是：大量的电力系统业务都加载在电力通信网上运行和传输,因此导致了电力通信网的压力剧增,而电力通信部门又没有有效的方法来提高通信网的可靠性和安全性。过去,电力系统生产运行部门采取的方法是单一的依靠更新设备并对通信网络进行改造、升级和扩容来减少通信中断次数和故障率,然而这些方法是远远不够的。因为只有在通信中断或是通信业务运行出现问题的时候才能够察觉,但那时损失己经造成。这是一种“后知后觉”的做法,并不能有效的避免电力公司造成的损失。本文针对以上问题而做出如下几个方面的研究工作,目的在于确实有效的提高电力通信网的安全性和可靠性并对电力通信网故障做到“防患于未然”：利用复杂网络识别电力通信网中网络拓扑结构中关键点。通过快速密度聚类算法进行了无监督学习分类,将节点的重要性进行等级划分,可以有效的应用于电力通信网节点的重要性评价中,为电力通信网的规划作为支撑。利用最优化理论实现电力通信网带宽的分配。在实际规划中,要实现对每个链路之间的带宽进行分配是一个很复杂的问题。本文利用最优化理论实现通信链路的最优化,并在最优化的基础上实现带宽的合理分配。利用大数据分析技术对电力通信网的带宽进行预测。电力通信网带宽的预测是电力通信网规划的一个重要方面,本文利用大数据分析技术,通过对历史实际使用带宽数据的分析,建立预测模型,实现电力通信网的带宽预测。
[Abstract]:Power communication network is a kind of special communication network used to serve power system. Power communication network includes four kinds of services: voice, data, video and multimedia. The operation of the four types of power communication services requires high reliability and security of the power communication network. Power communication network plays an important role in the application of smart grid. In recent years, great changes have taken place in the electric power system. At the same time, the scale of the power system is expanding day by day, and the level of maintenance and management of the power system has been continuously improved. However, the number of failures and the failure rate of power communication network have not been reduced. The main point is that a large number of power system services are loaded on the power communication network to run and transmit, which results in the pressure of the power communication network increased sharply, and there is no effective method to improve the reliability and security of the communication network in the power communication department. In the past, the methods adopted by the production and operation department of power system were to reduce the number of communication interruptions and the failure rate by simply updating the equipment and reforming, upgrading and expanding the communication network, but these methods were far from enough. Because it can only be detected when the communication is interrupted or if the communication service is not running properly, but the loss is already caused by that time. This is a "hindsight", can not effectively avoid the losses caused by power companies. In view of the above problems, this paper makes the following research work in several aspects, The purpose of this paper is to improve the security and reliability of electric power communication network effectively and to prevent the trouble in the power communication network. The key points of the network topology in the power communication network are identified by using the complex network. By using the fast density clustering algorithm to classify unsupervised learning and classify the importance of nodes, it can be effectively applied to the evaluation of the importance of nodes in electric power communication networks, and can be supported by the planning of power communication networks. The optimization theory is used to distribute the bandwidth of power communication network. In practical planning, it is a complex problem to allocate the bandwidth between each link. In this paper, the optimization theory is used to realize the optimization of communication link, and the reasonable allocation of bandwidth is realized on the basis of optimization. Using big data analysis technology to forecast the bandwidth of electric power communication network. The prediction of power communication network bandwidth is an important aspect of power communication network planning. This paper makes use of big data analysis technology, through the analysis of historical actual use of bandwidth data, establishes a prediction model to realize the power communication network bandwidth prediction.
【学位授予单位】：南京信息工程大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM73;TP181
，

本文编号：1851158