【作者】 陆斌；

【导师】 徐志科； 曹胜华；

【作者基本信息】 东南大学 ， 电气工程（专业学位）， 2022， 硕士

【摘要】 随着全世界现代工业技术的不断创新发展,电力行业也随之逐步进入了大规模建设智能电网的阶段,而智能变电站的建设是智能电网建设中的重要工作环节。作为目前智能变电站中最重要的组成部分,智能变电站二次系统发展迅速,二次系统网络发生了巨大的变化。由于电力系统对电网的安全性和稳定性要求极高,各种新技术和设备在智能变电站中的应用受到很大的限制。因此,如何在现有的技术条件下,在工程实践中合理优化智能变电站的二次系统网络就成为了一个必须深入研究的问题。本文首先简要概述了智能电网和智能变电站的发展,分析了国内外智能变电站及其二次系统网络优化技术的研究现状,并研究了存在的一些问题,阐述了本文的研究背景和必要性。然后介绍了智能变电站最重要的基础性标准,IEC61850系列标准体系,深入研究了智能变电站二次设备的选型标准和智能变电站二次系统网络结构的设计方法,介绍了现有的智能变电站的系统设备配置、网络架构等情况,重点分析了当前的组网方案。在此基础上分析了智能变电站二次网络存在的问题,包括设备功能单一以及网络架构不合理,研究了二次网络优化目标及优化原则,研究了间隔层、过程层网络设备优化的过程,并对优化效果进行了分析。对二次系统网络架构方面存在的问题,研究了基于SDN技术的过程层及间隔层网络架构,并在此基础上研究了朴素贝叶斯算法、KNN算法以及MACCA算法,并对这三种算法在流量分类时间、训练时间以及分类准确率进行了基本性能验证,发现KNN算法分类准确度高,但是分类时间长,朴素贝叶斯算法分类时间短,但是对网络数据分类准确度低,MACCA算法则充分结合了两者的优点,在分类速度、训练时间、数据分类准确度方面实现了基本平衡。最后在Mininet环境下对智能变电站二次网络下的网络流量传输进行了模拟,验证了MACCA算法在网络流量分类上的性能。同时研究了SDN架构下SDN控制器的流量优化策略,采用K条最短路径算法KSP(k-shortest path),在此基础上,本文提出了遗传算法的K条最短路径流量均衡算法,结果显示,该算法实现了流量均衡,从而增加了网络中链路的剩余带宽。最后,分析了目前智能变电站二次设备和二次系统网络存在的不足,给出了智能变电站二次系统今后的改进和发展方向。更多还原

【Abstract】 With the continuous development of modern industry,smart grid is constructed with large-scale in the power industry,and the construction of intelligent substations is an important link of the smart grid construction.The secondary system network of the intelligent substation has developed rapidly,and the secondary system network has also undergone great changes.Because the power system has extremely high requirements for the safety and stability,the application of various new technologies and new equipment in intelligent substations has been greatly restricted.Therefore,how to reasonably optimize the secondary system network of intelligent substations in engineering practice has become a problem that must be studied in depth.This thesis first briefly summarizes the development history of smart grid and intelligent substations,analyzes the current research status of intelligent substations and its domestic and overseas optimization technologies for secondary system network,and studies some of their existing problems,and elaborates the research background of this thesis and the necessity of research.IEC61850 standard system as the most important basic standards for the intelligent substation was introduced,then the intelligent substation secondary equipment selection and design of intelligent substation secondary system network structure was further studied.The system equipment configuration and network architecture of the existing intelligent substation is introduced,and current networking scheme is analyzed the emphatically.On this basis the problems existing in the secondary network of smart substation are analyzed,including single equipment function and unreasonable network architecture.The optimization objectives and principles of the secondary network are put forward,and the optimization process of network equipment in the interval layer and the process layer is proposed,and the optimization effect is analyzed.Problems existing with the network architecture of secondary systems,The process layer and interval layer network architecture based on SDN technology is studied,Based on this study,naive Bayes,KNN and MACCA algorithms,The basic performance of the three algorithms in the traffic classification time,training time and classification accuracy is verified,Results show that the KNN algorithm has a high classification accuracy,But the classification time is long,The Naive Bayes algorithm with a short classification time,However,the classification accuracy of network data is low,The MACCA algorithm fully combines the advantages of the both,A basic balance is achieved in terms of classification speed,training time,and data classification accuracy.Finally,the network traffic transmission under the secondary network of the smart substation is simulated in the Mininet environment to verify the performance of the MACCA algorithm on the network traffic classification.At the same time,the traffic optimization strategy of SDN controller under SDN architecture is studied,and the K shortest path algorithm KSP(k-shortest path)is adopted.On this basis,the K most short-circuit path equalization algorithm of genetic algorithm is proposed.The results show that the algorithm realizes the traffic equilibrium,thus increasing the remaining bandwidth of the link in the network.Finally,the current deficiencies in the secondary equipment and secondary system networks of intelligent substations are analyzed,and the direction for the improvement and development of the secondary systems of intelligent substations in the future is proposed.更多还原