发电机组故障模型建立及诊断方法研究

发布时间：2018-05-13 09:41

本文选题：大型发电机 + 转子一点接地　；参考：《长春工业大学》2016年硕士论文

【摘要】：电力工业作为国民经济可持续发展的先行工业发展迅速,许多大型机组陆续投入运行。然而机组容量的增大使其结构和系统日趋复杂,如何保证这些机组能安全、可靠、经济和高效运行,对整个电力系统的安全稳定起着至关重要的作用。运行中的大型发电机转子由于处于高速旋转状态,承受着很大的机械应力和热负荷,同时还受到巨大的电磁力作用,所以对转子部分的故障诊断就显得尤为重要。转子绕组的一点接地和匝间短路是常见的发电机电气故障,本文从传统的试验方法和电气二次入手对发电机转子一点接地故障诊断作出了详细的探讨。也对转子绕组匝间短路后电气参数变化进行了分析,将传统的方法和智能化方法相结合,对转子绕组匝间短路的存在性和故障严重性进行了诊断。本文的主要工作包括四个部分:第一部分研究发电机转子一点接地故障及实测数据分析。介绍了发电机转子的结构,转子系统故障产生原因及转子接地故障危害。并以实际电厂中转子回路接地报警故障为例,详细阐述了一点接地故障的出现、监测、诊断及其处理情况,作了相应的电气试验,分析了引发此次故障的可能原因,作出相应处理,提出了防范措施。第二部分研究低频方波注入式的转子一点接地故障诊断及其应用。分析了方波电压注入式的转子一点接地故障诊断的特性,推导出判据,并用仿真实验来验证理论分析的正确性。第三部分研究发电机转子绕组匝间短路故障。介绍了发电机转子绕组匝间短路形式分类,各种传统检测方法的优缺点及适用情况。对匝间短路后的电磁特性进行分析,确定了典型故障相关参数。第四部分研究将人工神经网络应用到转子匝间短路故障诊断中。介绍了人工神经网络的基本概念及优缺点。根据发电机转子绕组的匝间短路会造成转子电流增大但无功却相对减少,获取神经网络学习样本,利用发电机监测的数据来建立基于LM算法的BP神经网络预测模型,并对转子绕组匝间短路的存在性和故障严重性进行了诊断。
[Abstract]:As the leading industry of national economy sustainable development, electric power industry develops rapidly, and many large units are put into operation one after another. However, the increase of unit capacity makes its structure and system more and more complex. How to ensure the safety, reliability, economy and high efficiency of these units plays an important role in the safety and stability of the whole power system. Since the rotor of large generator in operation is in the state of high speed rotation, it bears great mechanical stress and thermal load, and is also subjected to huge electromagnetic force, so the fault diagnosis of rotor part is particularly important. The point grounding and inter-turn short circuit of rotor winding are common electrical faults of generator. This paper discusses the diagnosis of one-point grounding fault of generator rotor in detail from the traditional test method and electrical secondary. The change of electrical parameters after inter-turn short circuit of rotor winding is also analyzed. The existence and fault severity of inter-turn short circuit of rotor winding are diagnosed by combining traditional method with intelligent method. The main work of this paper includes four parts: the first part studies the fault of one point grounding of generator rotor and the analysis of measured data. The structure of generator rotor, the cause of rotor system fault and the harm of rotor grounding fault are introduced. Taking the grounding alarm fault of rotor circuit in actual power plant as an example, the occurrence, monitoring, diagnosis and treatment of a point grounding fault are described in detail, and the corresponding electrical tests are made, and the possible causes of this fault are analyzed. Corresponding treatment is made and preventive measures are put forward. In the second part, the fault diagnosis and application of low frequency square wave injection are studied. The characteristics of one-point grounding fault diagnosis of rotor with square wave voltage injection are analyzed and the criterion is deduced. The correctness of the theoretical analysis is verified by simulation experiments. In the third part, the inter-turn short circuit fault of generator rotor winding is studied. This paper introduces the classification of inter-turn short circuit of generator rotor windings, the advantages and disadvantages of various traditional detection methods and their application. The electromagnetic characteristics of interturn short circuit are analyzed and the related parameters of typical faults are determined. In the fourth part, artificial neural network is applied to rotor interturn short circuit fault diagnosis. The basic concept, advantages and disadvantages of artificial neural network are introduced. According to the fact that the inter-turn short circuit of generator rotor windings will result in the increase of rotor current but the relative decrease of reactive power, a BP neural network prediction model based on LM algorithm is established by using the data of generator monitoring. The existence and fault severity of inter-turn short circuit in rotor winding are diagnosed.
【学位授予单位】：长春工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM31;TM307

【相似文献】