【作者】 郭召艳；

【导师】 陈庆国；

【作者基本信息】 哈尔滨理工大学 ， 高电压与绝缘技术， 2008， 硕士

【摘要】 大型发电机是电力系统中重要的电力设备之一,其绝缘发生局部放电会造成绝缘的劣化,最终导致绝缘击穿和短路事故。传统的离线试验方法不能完全反映电机运行条件下的状态,因而并不能确保发电机的安全运行,事故仍时有发生。因此,对发电机进行实时在线监测,可以有效了解发电机绝缘的状态,及时发现绝缘隐患,有利于提高机组运行的可靠性。本文在阅读国内外大量文献的基础上,主要论述了发电机定子绕组中三种不同局部放电的起因和特点,并对局部放电信号在电机定子绕组中的传播规律进行了分析。在此基础上,根据脉冲电流法在发电机局部放电中的应用,研发了一套大型发电机定子绕组局部放电在线监测系统,包括硬件和软件的研制。硬件系统包括宽频信号传感器,信号预处理电路和信号采集单元。软件系统基于VC++的开发平台,完成对数据采集卡的采集控制和上位机与下位机之间的通信。在现场中,局部放电信号往往被强烈的干扰信号所覆盖,所以局部放电在线监测的首要问题是如何将局部放电信号从强的电磁干扰中提取出来。本文针对监测现场可能存在的各种干扰提出了一种分层综合处理的方法,对于不同的干扰采用不同的信号处理方法分步骤地去除。对于严重的周期性窄带干扰,由于其在频域中频带窄而幅值大的特点,提出了FFT变换和小波分析联合消噪的方法;对于和局放信号一样遍历整个时域的白噪声,则利用小波包变换的处理方法抑制。仿真结果表明FFT变换和小波分析联合消噪的方法对含有强的连续周期性窄带干扰的局部放电信号消噪十分有效;采用小波包变换的方法可以有效的从白噪声的干扰信号中提取出局部放电信号,比小波变换更优越。利用本文设计的监测系统对高压定子绕组进行试验,并利用本文提出的抗干扰方法分层处理信号。结果证明这种分层综合处理的方法能较好地抑制局放信号中的干扰信号。随着信号处理技术不断发展,抑制不同干扰的方法将更加合理和完善,这种方法有较好的应用前景。更多还原

【Abstract】 The large generator is one of the important devices in power system. Partial discharge (PD) accured in its insulation can lead to the deterioration of insulation, resulting in insulation breakdown and short circuit. Traditional off-line test can not present the real status of generators in operation station, so it can not ensure safe-operation of generators. Insulation faults were still reported now and then. As a result, it’s necessary to develop a complete and practicable on-line partial discharge monitoring system for the large generator stator winding to find out the real insulation status effectively and to improve the operation reliability of generator units.On the basis of knowledge from many documents, the generator partial discharge generation reasons and characteristics of three different types of PD in the large generator stator winding are mainly elaborated in this paper. And the propagation law of PD pulse in the large generator stator winding is also analyzed. According to these theories and the application of the electrical research association in the generator partial discharge detection, a new type of on-line PD monitoring system for the large generator stator winding is developed,including hardware and software development. The hardware system is composed of the wide band sensor, the signal prepoccessing circuit and the PD signal collecting unit. The software system is developed based on the VC++ software development platform,realizing function of controlling the data acquisition card and communication between the upper computer and the lower computer.Considering different types of interferences which always submerge PD signals possibly encounter in the weak partial discharge signals, the first important problem of on-lining monitoring is how to pick up the nonstationary random signals from the interferences. Due to the variety of interference exsiting in the field, a measurement aimed to the hierarchically processing of signals is given in this paper, a method using different processing means to suppress different interferences. Firstly,considering the strong periodic narrow-band interference of PD signals has distinct difference from PD pulse singals in the frequency domain, a method is put forward to eliminate the interferences of partial discharge based on FFT and wavelet theory. Secondly, because of the similarity of white noises and PD pulse signals, wavelet packet transform is apt to eliminate white noises. Simulation proves that the wavelet transform combine with FFT method can be more effectively to eliminate the strong narrow-band interference of partial discharge signals and preserve the original signals very well. Meanwhile, the wavelet packet transform can pick up the partial discharge signals from the white noises, and it is much better than the wavelet transform.Further more, PD tests have been made on the high voltage generator stator winding using the system designed in this paper and the signals were processed using the interference suppression methods proposed in this paper, the result shows that the measurement aimed to the hierarchically processing of signals can pick up the PD signals effectively from the interferences. With the development of the signal processing technology, methods to process signals will be more perfect,so there is a practical future with this measurement.更多还原