电气自动化专业外文翻译.doc

《电气自动化专业外文翻译.doc》由会员分享，可在线阅读，更多相关《电气自动化专业外文翻译.doc（19页珍藏版）》请在三一办公上搜索。

1、毕业设计(论文)外文资料翻译题 目: 使用小波变换在中性非有效接地的分配系统中识别单相接地故障馈线 附 件： 1.外文资料翻译译文；2.外文原文。 使用小波变换在中性非有效接地的分配系统中识别单相接地故障馈线 AbstractA scheme of single-phase-to-ground fault feede摘要：本文介绍小波变换技术的在配电网单相接地故障馈线识别的一种应用方案。该计划使用零序当前行波识别的故障馈线和母线的残余电压，以确定故障或开关操作引起的事件Thescheme uses zero-sequence current traveling waves to identfa

2、ulted feeder, and the busbar residual voltage to determievent caused by fault or 。The current travelin由零序电流传感器测量当前行波分解提出利用小波多分辨分析。The local modulus当地小波变换的模的maxima of the wavelet transform are extracted to determine the极大值确定了time of the initial traveling wave.初始行波的时间。所有馈线上的小波变换在时间上相互比较大小和极性找出故障馈线The

3、wavelet transforms on all所有所有 。The feeder identification馈线标识is independent of the network neutral-point grounding mode.独立于电网中性点接地方式。The proposed scheme was implemented and verified using Elec-tromagnetic Transients Program (EMTP)-generated signals.拟议的计划通过电磁暂态程式（EMTP）已经被实施和验证。在正常情况下Thescheme proved t

4、o be robust against transients generated during对如馈线充电和放电以及as capacitor bank switching.电容器组切换的瞬态响应来说，这个计划被证明是具有鲁棒性的。关键词：配电保护 、小波变换、残压、零序电流Index TermsPower distribution protection, residual voltage,关键词 看关键词I. I NTRODUCTION一、引言 带有如不接地，高电阻接地，and neutralizer grounded, are often called the neutral nonef-f

5、ectively grounded distribution systems.和中性点接地的小接地故障电流的分布式系统DISTRIBUTION systems with low levels of ground-fault带，通常被称为的非有效中性点接地的分布系统。A distinguishing fea-ture of such systems is the ability to continue to operate the这种系统的特点是在单线接地的情况下能够继续运行一段时间。Since the system neu-tral is not solidly grounded, there

6、 is little current flow in the由于系统的中性点没有完全接地，因此接地故障时电流很小，同时系统的工作状态没有达到非正常的电流水平。At the same在同一time, because of the very limited ground fault current, it is fre-quently considered impractical to try to provide ground-fault时间，因为在非常有限的接地故障电流，试图用这个系统去保护接地故障通常被认为是不切实际的。Typically, the only relaying provide

7、d 通常情况下，该系统唯一提供的继电保护是零序过电压保护(with no indication of which feeder is faulted).（没有指示是哪根馈线故障）。The alarm only仅报警indicates that a ground fault exists somewhere on the system.表示系统接地故障存在于某个地方。Typical operating practice is to manually run tests to determine典型工作的做法是手动运行测试确定the faulted feeder.故障馈线。 In the past

8、 decade many techniques 1 have been develop在过去的十年中，已经开发出许多技术to improve the identification of faulted feeder in neutral non-effectively grounded distribution system.以提高非中性点有效接地分布式系统馈线故障的识别。They can be roughly他们大致可以divided into two groups: fundamental system frequency com-ponents based algorithms and

9、transient signal based algorithms.分为两组：基于基本系统频率元件算法和基于暂态信号算法。Fundamental system frequency components based algorithms基本系统频率元件的算法 such as magnitude comparison relay, phase comparison relay如幅度比较继电器，相位比较继电器， 瓦特计式wattmetric detection relay 2, 3瓦特 检测继电器 are not fully successful,没有完全成功 ,either due to thei

10、r inability to identify faulted feeder with low由于他们无法确定在低电流下的故障馈线current or to their maloperation by the maximum load unbal-ance expected on the circuit. 另一方面线路上预期的最大不平衡负荷会引起他们的误动作。In fact, the signals which would事实上，零序设备检测的信号包含着显著的暂态成分。 Transients are generally much greater than瞬变一般远远大于the power fr

11、equency component.电源频率分量。Moreover, single phase fault此外，单相故障往往由连续的短时间瞬态自熄故障形成往往 。In the paper 4, Pronys method i在文献4，Prony的方法是applied to earth fault current analysis in Petersen-coil-protected在保护消弧线圈系统中应用接地故障电流分析systems.。The papers 5, 6 use artificial neural networks文献5 6利用人工神经网络在谐振接地配电系统中检测单相接地故障。A

12、nother study 7 began to apply另一项研究7开始在谐振接地配电系统应用wavelets to analyze transient earth fault signals in resonant小波分析暂态接地故障。Papers 8, 9 present文献8 9，提出single-phase-to-ground fault feeder selection based on wavelet基于小波变换模的极大值的单相接地故障馈线选择transform modulus maxima.。The paper 10 uses transient文献10 在所选择的频段采用瞬态

13、信号signals in the chosen frequency band to identify faulted feede以确定故障馈线。 In this paper, the idea of identifying faulted feeder by com-paring magnitude and polarity of wavelet coefficients, suggested 在本文中，通过比较小波系数的大小和极性确定故障馈线的想法，建议in 8 without deep practical verification is thoroughly studied. 8不深入的实

14、际验证和研究。这里给出的对在正常产生的瞬变events such as feeder energizing and de-energizing as well as事件的扩展计划Its extended scheme against transients generated during normal它给，如馈线通电和断电功能，以及capacitor bank switching is presented here and the results of ex-tensive simulative testing with Electromagnetic Transients Pro-gram

15、(EMTP) signals are described.电容器组切换产生的瞬变和广泛的电磁暂态程序（EMTP）信号的模拟测试结果描述。 The following sections of the paper include general introduc-tion of the wavelet technique (Section II), basic principle of the本文以下各节包括小波技术（第二部分），目前的行波计划和实施（第三部分）的故障馈线鉴定的基本原则（第四部分）以及采用EMTP的产生的信号测试结果的总体介绍（第五部分）。关闭本文的Conclusions and

16、references close this paper.结论和参考。 I二、 W AVELET TRANSFORM小波变换 The wavelet transform is a powerful tool in the analysis of 小波变换是一个功能强大的工具，由于它提取时间和能力frequency information from the transient signal 1116.暂态信号的频率信息能力在分析transient phenomena because of its ability to extract time and短暂的现象中 11 - 16。对于一个给定的连续

17、小波变换（CWT）定义为母线signal with respect to a mother wavelet is母线小波信号Fig图 1. 1。 Shape of the derivative function of cubic B-spline function. 三次B样条函数的导函数的形状。 (1) （1） where is the scale factor and is the translation factor. 其中a是比例因数和b是转换因数。 In fact, the signal in engineering areas we should analyze is 事实上，我们

18、应该分析的在工程领域的信号usually discrete sequence通常是离散序列。在Wavelet transforms of discrete se-quence are based on multiresolution analysis, and we have 离散序列小波x（n）变换多分辨率分析的基础上，我们有的j a nd k are integers, representing the scale factor and the 和k是整数，即比例因数和translation factor, respectively; is a discrete approx-imatio

19、n of at t he resolution ; is the discrete转换转换因数，是离散信号在x（n）的决议2j; 是离散detail signal (wavelet transform) at the resolution ; and细节信号（小波变换）的决议; 2jare coefficients and they are determined only by the是系数，gk仅由他们确定mother wavelet.母小波。 (2) （2） There are many types of mother wavelets, such as Harr and 母小波的种类有很

20、多，如哈尔和多贝西Daubichies wavelets.多小波。The choice of mother wavelet plays a sig-ni ficant role in detecting and localizing different types of fault母小波的选择在不同类型的故障检测和定位transients瞬变起到了重要的作用。In addition, the choice also depends on a particular ap-plication.此外，母小波的选择还依赖于一个特定的应用程序。Because the derivative functio

21、n of the cubic B-spline由于三次B样条函数function requires the least support in all of the polynomial spline功能要求至少支持在所有的多项式样条functions and can obtain a good approximation similar to th，可以得到类似的一个很好的近似Gauss function, a derivative of the cubic B-spline function was高斯函数的三次B样条函数的导数selected as the base wavelet fun

22、ction for the present study.作为本研究的基本小波函数选择。 The（function and its shape are as (3) and shown in Fig.3）如下它的形状和系数。 1, respectively, (3) （3） For the aforementioned wavelets, the coef ficients are 对于上述小波，它的系数是 In order to obtain the characteristics of abrupt signal simply 为了获得突变信号的特征，只需and clearly, the m

23、odulus maxima of wavelet transforms is in-troduced.并明确小波变换的模的极大值。They are de fined as the local maxima of the wavelet 他们被定义为本地的小波变换模的极大值，当且仅当满足以下条件， 15, 17: 15，17： (4) （4） It has been verified that any signal could be reconstructed by 它已被证实的信号可重构its modulus maxima 17, so the modulus maxima of wavele

24、t其模的极大值17，因此小波转换模的极大值transforms contains all the useful information of the original包含了原有的所有有用的信息signal, and the modulus maxima of wavelet transforms can fully信号，小波变换模的极大值可以充分represent the original signal, including the signal s waveform,代表原始信号，包括信号的波形，properties, and slope of the variations.属性和坡度的变化

25、。Therefore, the modulus因此，在本中利用小波变换模得极大值数来了实施这个计划。maxima of wavelet transforms is used in the paper to implement III.三、 基本原理B ASICPRINCIPLE基When a single-phase-to-ground fault occurs in one feeder, 当单相接地故障发生在一个馈线是，产生的fault generated traveling waves will propagate from fault point故障行波点将从故障点沿馈线传播到alon

26、g the feeder to the system.系统。 Arriving at the point where wave到达行波impedance changes, for example, busbar in the substation, the阻抗的变化的，例如，母线变电站，traveling wave will be re flected and refracted.行波将反射和折射。 At the instant在当前行波抵达瞬间when current traveling waves arrived at the point, they satisfy时，他们满足Kirchh

27、off s current law基尔霍夫电流定律 (5) （5） where n is the number of branches connected to the point, and 其中n是连接点的分行数目，ikis the current through number K branch.是通过K个分支电流。That is, the current in flow to the point is equal to the current也就是说，当前电流在流入是等于电流的outflow.流出。As far as the busbar is concerned, the current

28、 value in至于有关母线， faulted feeder is equal to the sum of current in healthy feeders故障馈线电流值等于正常的馈线电流母线连接的总和connected to the busbar.。Suppose the current reference direc-tion is de fined as from the busbar to the feeder, the currents in假设当前的参考方向是从母线到馈线，在all healthy feeders have the same direction.所有健康的馈线

29、电流有相同的方向。 Therefore current因此目前traveling wave of faulted feeder is reverse to, and much higher故障馈线电流的反向， than the wave of any healthy feeder, which is the basic principle比任何健康馈线的电流要高得多，这是to identify single-phase-to-ground fault feeder in neutral nonef-fectively grounded distribution system 18.确定单相接地

30、故障馈线中性非有效接地的配电系统基本原则18。 IV.四、PROPOSED SCHEME 建议计划 Because current traveling waves generated during normal 因为在正常事件中，如馈线通电以及电容器组切换产生的行波会导致继电器故障馈线误懂，母线残压引入计划to determine a disturbance as a fault or not.以确定故障或没用的干扰。So the proposed因此，拟议scheme uses the busbar residual voltage to determine whether计划使用母线的残余

31、电压，用来确定是否a phase-to-ground fault occurs, and zero-sequence current相接地故障时，零序电流traveling waves to identify which feeder is faulted.行波出现故障馈线。When a当single-phase-to-ground fault occurs in one feeder, the residual单相接地故障发生在一个馈线中，残余power frequency voltage in the busbar will rise to three times在母线电源频率电压将上升到

32、三倍the phase-to-phase voltage.相电压。When a switch operates such as当开关操作如feeder energizing, the residual power frequency voltage in the馈线充电，在剩余busbar is about zero.母线的工频电压是零。In the scheme, the busbar residual voltage在这项计划中，母线的残余电压threshold is set as phase-to-phase voltage.阈值设置为相电压。 Fig. 图 2. 2。 Hardware

33、 arrangement of the proposed scheme. 该方案的硬件安排。 Fig. 图 3. 3。 Flowchart of the proposed scheme. 拟议计划的流程图。 On the other hand the scheme uses initial traveling waves to 另一方面，该计划使用初始行波来identify faulted feeder.确定故障馈线。The feeder length in distribution system在配电系统馈线长度is usually as short as several kilometer

34、s, even hundreds of me-ters, while the traveling wave speed can reach 300 .通常只要短短几公里，甚至几百米，而行波的速度可以达到300m/s。There-fore, to acquire the initial traveling waves correctly, high fre-quency sampling is necessary.因此，正确获取初始行波，高频率的采样是必要的。And after analysis and compar-ison, the appropriate sampling rate is

35、selected as 1 MHz in the经过分析和比较，选择适当的采样率1兆赫在本文中paper.在。To meet different sampling frequency requirements, the为了满足不同的采样频率的要求，scheme needs special design 19.计划需要特殊的设计19。 Hardware arrangements and 硬件安排和flowchart of the proposed scheme are shown in Figs.拟议计划的流程图，在图所示 2 and 3,2和3respectively. 分别。 The har

36、dware arrangement is mainly composed of three units: 硬件的安排主要是三个单位组成：电源频率电压处理单元power frequency voltage processing unit (PFU), current trav-eling waves processing unit (TWU), and logic determination电源平（PFU），行波处理单元（TWU）的和逻辑处理单元（LDU）。 在unit (LDU).在 In PFU, residual voltage from potential transformerPFU中

37、，残余电压电压互感器(PT) is first transferred to 5 V voltage signals by V/V trans-former.（PT），首先通过V / V变压器转移到5 V的电压信号，。Then, 500 Hz low pass filter is used to remove high fre-quency noise, and the signal is sampled at 1 kHz and converted然后，500赫兹的低通滤波器用于去除高频率的噪音，并在1 kHz的采样信号转换to digital quantities.数字量。At last,

38、 CPU1 extracts the residual voltage最后，CPU1中提取的残余电压characteristics to determine whether the disturbance is fault or特性，以确定是否干扰故障或not没有。 In TWU, each feeder current from current transformers is first 在TWU中，首先是每个馈线电流从电流互感器前通过I / V变压器transferred to 5 V voltage signals by I/V transformers before高速采样转移至5 V

39、的电压信号high-speed sampling.。Then low-pass filters (LP) and band-pass然后低通滤波器（LP）和带通filters (BP) are used to remove noise and generate triggering过滤器（BP）用于消除噪音并产生触发signals for relay starting.为继电器信号开始。The upper limit frequency for the LP上限频率为LPfilters is set as 500 kHz.过滤器设置为500千赫。The frequency band of th

40、e BP filtersBP滤波器的频带is selected as 3 10 kHz, which ensures that the scheme starts被选定为3 10千赫，以确保该计划开始reliably in most fault situations.可靠的大多数故障情况。 Fault starting circuit compares 故障起动电路比较the signal output from BP filter with a certain threshold.从BP滤波器的输出信号具有一定的阈值。When何时setting the threshold, it is ne

41、cessary to consider fault inception设置阈值，有必要考虑继电器的灵敏度和可靠性。 In order to avoid frequent starting problem in nonfault situations, together 避免频繁启动问题非错误情况，连同with consideration of the practical situation, the threshold is考虑到实际情况，门槛set as 10% maximum value of the phase voltage in the paper.设置相电压的最大值为10。In t

42、his case, faults, whose inception angle is greater than 6 ,在这种情况下，故障，其成立以来的角度大于6,could all be detected.都可以被检测出来。The sampling/hold (S/H) circuit samples采样/保持（S / H）电路样品analog signals and holds them at a sampling rate of 1 MHz.模拟信号，并持有他们在1 MHz的采样率。The sampling signal voltage ranges from 5Vto 5 V. A/D采

43、样信号的电压范围从5Vto 5五的A / D converters convert analog voltage to digital quantities. 转换器转换成数字量的模拟电压。Thesynchronous data conversion produces data streams, and the 同步数据转换产生的数据流，streams are stored in a cycle double-port memory.流被储存在一个周期内双端口存储器。 The CPLD在CPLDchip is used to generate a controlling impulse and

44、 controls the芯片用于生成一个控制脉冲和控制converters, ensuring that they work in order转换器，确保他们为了工作。At last, CPU2在过去，CPU2的 executes wavelet algorithm and outputs the disturbance source.执行小波算法和输出的干扰源。 The CPU3 synthesizes fault determination result from PFU CPU3综合故障PFU测定结果 and disturbance source result from TWU to

45、output the faulted 从TWU的干扰源的结果输出的故障feeder.接驳。 The flowchart of the proposed scheme is shown in Fig. 拟议计划的流程图如图。 3.3 When a disturbance in the system is detected, current traveling当检测到系统中的干扰，当前wave signals of all feeders sampled at 1 MHz are transferred在1 MHz采样所有馈线波信号传输to data memory.数据存储器。A fixed da

46、ta window containing 128 samples 一个固定的数据窗口，其中包含128个样本is used.被使用。For each data window, four scales wavelet transforms对于每一个数据窗口，四个尺度小波变换are processed.处理。Because capacitor banks are always installed for因为总是安装电容器组power-factor correction in the substation, they will filter high-frequency components of c

47、urrent traveling waves在变电站功率因数校正，他们将过滤电流波的高频分量。And high-fre-quency components of traveling waves attenuate more than rel-atively lower frequency components even along the same path和高频率的行波组件衰减甚至超过低频率的部分沿相同的路径from the disturbance source to the measured point.从干扰源的测量点。Thereby,因此，scale wavelet transform

48、 signal, which corresponds to rela-tively lower frequency components (31.125 kHz 62.25 kHz)尺度小波变换信号，它对应关系较低的频率成分（62.25千赫31.125千赫）of the signal, is chosen as the base signal to identify the faulted的信号，被选为基信号来识别故障feeder.。Then the local modulus maxima of scale wavelet那么当地的尺度小波模极大值transform signals are

49、extracted.变换信号提取。Because fault generated cur-rent traveling wave in faulted feeder is much higher than the由于故障产生在故障馈线电流旅行波远远高于waves in healthy feeders, and they share the same threshold, the健康馈线波，和他们分享相同的阈值，time of the first local modulus maxima in each feeder may be在每个支线的第一个本地的模极大值的时间可能是different.不同。In fact,