接触网污秽绝缘子泄漏电流分析研究

发布时间：2018-06-26 14:21

本文选题：接触网 + 绝缘子　；参考：《西南交通大学》2015年硕士论文

【摘要】：铁路作为我国的重要基础设施,是人们在日常生活中首选的交通工具之一,是综合交通运输体系中不可或缺的部分。电气化铁路在铁路事业中的比重越来越大,这也是铁路发展的趋势。接触网作为电气化铁路的重要部分,负责给电力机车提供电能。接触网一般都是裸露在空气中,这样,接触网的重要部件——绝缘子,就很容易被污染而降低绝缘特性。近年来,多地持续出现雾霾天气,加剧了接触网绝缘子的染污程度。这就很容易造成绝缘子表面放电,继而发生污秽闪络,甚至线路跳闸,威胁着电气化铁路牵引供电系统的安全正常运行。因此深入开展接触网污秽绝缘子沿表面放电过程及其影响因素的研究很有必要,通过绝缘子污秽在线监测装置持续监测绝缘子运行状态,深入研究绝缘子表面污秽积累发展过程,探索导致污闪的潜在因素,及时做出相应的防污方案,进一步强化绝缘管理,防止突发事故发生,尽可能避免污秽闪络,对电气化铁路运输具有十分重要的意义,为电气化铁路的可持续发展提供技术支持。本文针对接触网复合绝缘子和瓷质绝缘子开展了人工污秽试验,利用接触网绝缘子污秽在线监测装置,采集并储存泄漏电流信号和环境温湿度信息,从试验现象和试验数据可以观测出绝缘子泄漏电流随污秽度和湿度变化的特征。低湿度情况下,泄漏电流值很小,随污秽度变化特征不是很明显；高湿度情况下,泄漏电流幅值随污秽度的增加而明显变大,脉冲含量和谐波含量也随之增多。另外,由试验放电现象和泄漏电流数据说明复合绝缘子的抗污性能优于瓷质绝缘子。为了实现接触网绝缘子污秽在线监测,论文分析了铁路现场对泄漏电流影响的大因素——过车影响。在机车经过时,绝缘子泄漏电流信号中低频(0～10Hz)含量明显增加,利用这一特性,采用小波能量比的方法,可以有效识别过车时的泄漏电流波形,从而通过剔除这段数据来达到消除过车对泄漏电流影响的目的。绝缘子污秽在线监测装置中的电流传感器会对采集到的泄漏电流信号造成偏移,文中对此进行了分析,经大量数据验证,信号除去经验模态分解后的趋势项可以有效地改善这种偏移。文中阐述了两种针对泄漏电流信号的去噪方法,即奇异值分解去噪和小波阈值去噪。从去噪后的波形来看,奇异值分解去噪过多地滤除了细节特征,平滑了过多的噪声,使信号损失了一些细节信息；而小波阈值去噪则合理地滤除了噪声,使信号比较光滑,又很好地保留了信号本身的细节信息,因此小波阈值去噪方法比奇异值分解去噪方法更加适用于泄漏电流信号去噪。论文中分析了泄漏电流的时频域特性。给出了泄漏电流有效值、最大值、和时域熵三个时域参考量,泄漏电流三次谐波幅值与基波幅值之比、泄漏电流总谐波畸变率两个频域参考量,来评估绝缘子表面污秽状态。这五个参考量与绝缘子表面污秽度均呈现指数增长关系,拟合成指数曲线后拟合度都在0.95以上,满足工程应用要求,因此这五个参考量可以用来评估绝缘子表面污秽状态。最后,文中采用支持向量机来识别接触网绝缘子泄漏电流波形对应绝缘子污秽等级的分类方法。将环境湿度RH、泄漏电流有效值Ie、泄漏电流最大值Ih、泄漏电流三次谐波幅值与基波幅值之比K13和泄漏电流总谐波畸变率THD这五个参考量作为支持向量机模型的输入,污秽的四个等级作为支持向量机分类器的四类输出,选取了高斯径向基核函数,通过样本训练和测试,给出预测模型的结果,实验表明本文中建立的支持向量机预警模型的平均误差能够完全满足实际工程要求。
[Abstract]:As an important infrastructure of our country, railway is one of the preferred transportation tools in our daily life. It is an indispensable part of the comprehensive transportation system. The proportion of the electrified railway in the railway industry is increasing, which is also the trend of the railway development. As an important part of the electrified railway, the contact network is responsible for the power machine. The car provides electrical energy. The contact net is generally exposed to the air, so that the insulators, the important parts of the catenary, are easily polluted and reduce the insulation. In recent years, more and more haze weather has been sustained, which aggravates the pollution degree of the contact insulator. It is easy to cause the insulator surface discharge, and then the pollution flashover, Even line tripping threatens the safe and normal operation of the traction power supply system of electrified railway. Therefore, it is necessary to carry out the research on the discharge process along the surface of the polluted insulator of the contact network and its influencing factors. The insulator contamination on-line monitoring device is continuously monitored and the insulator surface contamination accumulation is deeply studied. The development process, exploring the potential factors leading to the pollution flashover, making the corresponding anti fouling scheme in time, further strengthening the insulation management, preventing the accident from happening, avoiding the pollution flashover as much as possible, is of great significance to the electrified railway transportation, and provides technical support for the sustainable development of the electrified railway. The edge and porcelain insulators have carried out artificial pollution test. Using the on-line insulator contamination monitoring device of the contact network, the leakage current signal and the ambient temperature and humidity information are collected and stored. From the test and test data, the leakage current of the insulator can be measured with the pollution and humidity. The leakage current is very low. In the case of high humidity, the amplitude of leakage current increased obviously with the increase of pollution degree and the content of pulse content and harmonic wave increased. In addition, the pollution resistance of composite insulators was superior to porcelain insulators by experimental discharge and leakage current data. In order to achieve contact network insulation. On line monitoring of subpollution, the paper analyzes the influence of the leakage current on the railway field - the effect of passing on the vehicle. When the locomotive passes, the low frequency (0 to 10Hz) of the leakage current signal in the insulator is obviously increased. Using this characteristic, the wavelet energy ratio method can be used to identify the leakage current waveform of the vehicle, and then pass through the filter. In addition to this data, the current sensor in the insulator contamination on-line monitoring device will offset the collected leakage current signal. In this paper, an analysis is made in this paper. It is verified by a large number of data that the signal can effectively improve the migration after removing the trend term after the empirical mode separation. In this paper, two denoising methods for leakage current signal are described, namely singular value decomposition de-noising and wavelet threshold denoising. From the wave de-noising waveform, singular value decomposition de-noising too much to filter out the details, smoothing too much noise and losing some details of the signal; and the wavelet threshold denoising is reasonable to filter the noise. In order to make the signal more smooth and retain the details of the signal itself, the wavelet threshold denoising method is more suitable for the leakage current signal denoising than the singular value decomposition denoising method. In this paper, the characteristics of the leakage current in the time and frequency domain are analyzed. The leakage current effective value, the maximum value, and the time domain entropy three time domain reference quantities are given. The ratio of the three harmonic current of leakage current and the amplitude of the fundamental wave, the total harmonic distortion rate of the leakage current of two frequency domain reference amounts to evaluate the surface contamination of insulators. These five reference quantities have an exponential growth relationship with the surface contamination of insulators, and the fitting degree of the quasi synthetic index curve is above 0.95, which satisfies the requirements of the engineering application, thus five A reference quantity can be used to evaluate the contamination status of the insulator surface. Finally, the support vector machine is used to identify the classification method of the insulator contamination level by the contact insulator leakage current waveform. The environmental humidity RH, the leakage current effective value Ie, the maximum leakage current Ih, the ratio of the leakage current to the amplitude of the three harmonic current and the amplitude of the fundamental wave are K1 3 and the total harmonic distortion rate of the leakage current THD these five reference quantities as the input of the support vector machine model, the four levels of pollution as the four class of the support vector machine classifier, select the Gauss radial basis kernel function, and give the result of the prediction model through the sample training and test. The experiment shows the support vector machine set up in this paper. The average error of the police model can fully meet the actual engineering requirements.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U225.43;TM216

【参考文献】