光纤电流互感器故障诊断方法研究

论文总字数：35956字

摘 要

随着智能电网的发展，电力系统的测量逐渐朝着高效化、智能化和互联网化发展。传统的电磁式电流互感器由于体型大、灵敏度低、精确度低、适应性差、易受到外界环境的干扰等一系列缺点逐渐被市场淘汰。体积小、性能好、精确度高、适应性强、抗干扰能力强的光纤电流互感器占领的市场份额越来越大。而针对光纤电流互感器的研究，如何能够高效的判别互感器的工作状态以及正确判别光纤传感器的故障类型成为当下研究的热点。

本毕业设计针对现状，采用了基于数据特征提取和信息分类的光纤电流互感器的故障诊断方法，建立光纤电流互感器故障诊断模型，采用小波分析以及支持向量机算法对故障信息进行特征提取和分类，进行仿真实验，从而有效地识别出传感器工作状态，并且对发生故障时故障类别进行判断。

本毕业设计首先讨论了光纤电流互感器的研究和发展现状，以及其结构特点；其次使用故障检测领域的研究现状以及常用的研究方法作为参考，提出了使用小波分析和支持向量机算法进行光纤电流互感器故障检测的方法；然后建立了光纤电流互感器数学模型，并且对其固定偏差故障、漂移偏差故障、变比偏差故障、完全失效故障这几种故障类型进行建模分析；之后介绍了小波分析和支持向量机算法的理论基础，探讨了将二者用于光纤电流互感器故障检测领域的可行性；最后进行了仿真实验，实现了对故障信息进行特征提取和分类的目的，有效的判断出光纤电流互感器的工作状态以及故障类型，验证了将小波分析理论和支持向量机理论应用于光纤电流互感器故障诊断的可行性。

关键词：光纤电流互感器，小波变换，支持向量机，多分类，故障诊断

Abstract

With the development of smart grids, the measurement of power systems is gradually becoming more efficient, intelligent and Internet-based. Traditional electromagnetic current transformers have been gradually eliminated by the market due to their large size, low sensitivity, low accuracy, poor adaptability, and easy interference from the external environment. Fiber optic current transformers with small size, good performance, high precision, strong adaptability and strong anti-interference ability occupy an increasing market share. For the fault diagnosis research of fiber-optic current transformer, how to effectively determine the working state of the transformer and correctly determine the fault type of the fiber sensor has become a hot topic of current research. Based on the current situation, this graduation design adopts the fault diagnosis method of fiber-optic current transformer based on data feature extraction and information classification, establishes the fault diagnosis model of fiber-optic current transformer, and uses wavelet analysis and support vector machine algorithm to extract and classify fault information. The simulation experiment is carried out to effectively identify the working state of the sensor and judge the fault category at the time of the fault.

Based on the above ideas, This graduation design first discusses the research and development status of fiber-optic current transformers, as well as its structural characteristics. Secondly, using the research status of fault detection and the commonly used research methods as reference, the paper proposes the use of wavelet analysis and support vector machine algorithm for fiber current. The method of transformer fault detection; then the mathematical model of fiber-optic current transformer is established, and the fault types such as fixed deviation fault, drift deviation fault, variable ratio deviation fault and complete failure fault are modeled and analyzed; then wavelet is introduced. The theoretical basis of the analysis and support vector machine algorithm is discussed. The feasibility of using the two in the field of fiber-optic current transformer fault detection is discussed. Finally, the simulation experiment is carried out to realize the purpose of feature extraction and classification of fault information, and effective judgment. The working state of the fiber-optic current transformer and the type of fault are verified, and the feasibility of applying wavelet analysis theory and support vector machine theory to the fault diagnosis of fiber-optic current transformer is verified.

KEY WORDS: fiber optic current sensor，wavelet analysis，support vector machine，fault diagnosis

目 录

第一章 绪论 1

1.1引言 1

1.2光纤电流互感器研究背景 2

1.3光纤电流互感器介绍 3

1.4故障诊断领域研究现状 5

1.5论文内容安排 7

第二章 光纤电流互感器故障特征分析 8

2.1引言 8

2.2光纤电流互感器测量结果数学模型 8

2.2.1固定偏差故障 8

2.2.2漂移偏差故障 9

2.2.3变比偏差故障 10

2.2.4完全失效故障 11

2.3本章小结 12

第三章 光纤电流互感器故障识别算法 13

3.1引言 13

3.2小波变换理论基础 13

3.3小波分析的特点及应用领域 14

3.4支持向量机理论基础 15

3.5支持向量机解决多分类问题 18

3.6本章小结 20

第四章 仿真实验 21

4.1引言 21

4.2实验思路 21

4.3样本集的生成 22

4.4对光纤电流互感器的各类故障进行小波变换的结果分析 23

4.5提取特征值后使用支持向量机算法对故障进行分类 26

4.5.1对光纤电流互感器固定偏差故障进行诊断 26

4.5.2对光纤电流互感器漂移偏差故障进行诊断 27

4.5.3对光纤电流互感器变比偏差故障进行诊断 28

4.5.4对光纤电流互感器完全失效故障进行诊断 31

4.6本章小结 32

第五章 主要结论及未来展望 33

参考文献 34

致 谢 36

第一章 绪论

1.1引言

剩余内容已隐藏，请支付后下载全文，论文总字数：35956字