基于因果网络的配电网故障诊断研究

发布时间：2018-06-21 23:00

本文选题：配电网 + 故障诊断　；参考：《青岛大学》2017年硕士论文

【摘要】：配电网故障诊断是电力系统研究的重要问题,配电网故障的快速诊断对减少配电网断电区域和缩短配电网断电时间具有重大意义,同时也起到了改善供电质量和提高配电网运行效率的作用。随着配电网技术的不断进步,其智能化程度越来越高,配电网正趋向越来越自动化的方向发展。在配电网发生故障时电网监视系统和数据采集系统(SCADA)会将各种信息传到管理中心来帮助我们进行分析,但故障处理人员需要从大量信息中快速判断和处理故障,这在故障判定时间上提出了很高的要求,因此在配电网监控系统中加入故障诊断模块有重要的实际价值。本文对目前常用的配电网故障诊断的人工智能方法进行了整理,总结了配电网故障诊断各种方法的工作原理及优缺点。并构建了基于因果网络的配电网故障诊断模型,根据配电网继电保护配置,充分利用因果网络快速的逆向推理能力,运用故障发生时配电网系统中产生的各种告警信息,将配电网各设备的关联抽象成三类基本因果关系,运用因果网络做逻辑逆向推理,实现配电网故障诊断。通过实际算例的分析验证了该模型能取得较为理想的诊断效果,证明该研究具有一定的实用价值。随着配电网的发展,大量的分布式电源(DG)的接入势必会对常规配电网的运行、控制以及保护配置造成影响,从而对配电网的故障诊断提出新的要求,本文对含DG的配电网系统做了简要介绍,介绍了分布式发电的基本知识以及目前主要的分布式发电技术及其运行方式。在这种运行方式下为了适应DG接入对保护配置带来的影响,引入虚拟节点的概念构造了含DG的配电网因果网络诊断模型。此外当配电网出现故障时,若故障信息受到干扰而丢失或畸变,因果网络纠错能力弱的短板就会突显出来。为了降低误判率,在原来的基础上首先对各种信息进行预处理,提出了基于信息预处理的改进因果网络对含DG的配电网故障诊断方法,实现了在保护和断路器存在误动或者拒动等情况下对含DG的智能配电网的故障诊断。并用实际算例验证了改进因果网络在含DG配电网故障诊断中的优越性,证明了该模型的实用性。
[Abstract]:Distribution network fault diagnosis is an important problem in the research of power system. The rapid diagnosis of distribution network fault is of great significance to reduce the distribution network power failure area and shorten the distribution network power failure time. At the same time, it also plays the role of improving the quality of power supply and improving the efficiency of distribution network operation. With the development of distribution network technology, its intelligence is becoming higher and higher, and the distribution network is becoming more and more automatic. Power network monitoring system and data acquisition system (SCADA) will transmit all kinds of information to the management center to help us analyze when the distribution network fails, but the fault handler needs to quickly judge and deal with the fault from a large amount of information. Therefore, it is of great practical value to add fault diagnosis module to the monitoring system of distribution network. In this paper, the common artificial intelligence methods of distribution network fault diagnosis are summarized, and the working principles, advantages and disadvantages of the methods are summarized. The fault diagnosis model of distribution network based on causality network is constructed. According to the configuration of relay protection in distribution network, the rapid reverse reasoning ability of causal network is fully utilized, and all kinds of alarm information generated in distribution network system are used when the fault occurs. The connection of each equipment in distribution network is abstracted into three kinds of basic causality, and the fault diagnosis of distribution network is realized by using causality network to do logic reverse reasoning. Through the analysis of practical examples, it is proved that the model can obtain more ideal diagnostic effect, and that the research has certain practical value. With the development of the distribution network, the access of a large number of distributed generation (DG) will inevitably affect the operation, control and protection configuration of the conventional distribution network, thus putting forward new requirements for the fault diagnosis of the distribution network. This paper briefly introduces the distribution network system with DG, introduces the basic knowledge of distributed generation, the main distributed generation technology and its operation mode. In order to adapt to the influence of DG access on protection configuration, a causal network diagnosis model with DG is constructed by introducing the concept of virtual node. In addition, when the fault occurs in the distribution network, if the fault information is lost or distorted due to interference, the weak error correction ability of the causal network will be highlighted. In order to reduce the error rate, the paper first preprocesses all kinds of information on the basis of the original information, and proposes an improved causality network based on information preprocessing for fault diagnosis of distribution network with DG. The fault diagnosis of the intelligent distribution network with DG is realized under the condition of the protection and the maloperation or the failure of the circuit breaker. The advantages of improved causality network in fault diagnosis of distribution network with DG are verified by a practical example, and the practicability of the model is proved.
【学位授予单位】：青岛大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM711

【参考文献】