WO2016019593A1 - 一种电缆过电流原因辨识方法及装置 - Google Patents

一种电缆过电流原因辨识方法及装置 Download PDF

Info

Publication number
WO2016019593A1
WO2016019593A1 PCT/CN2014/084248 CN2014084248W WO2016019593A1 WO 2016019593 A1 WO2016019593 A1 WO 2016019593A1 CN 2014084248 W CN2014084248 W CN 2014084248W WO 2016019593 A1 WO2016019593 A1 WO 2016019593A1
Authority
WO
WIPO (PCT)
Prior art keywords
overcurrent
index
phase
sample
phase current
Prior art date
Application number
PCT/CN2014/084248
Other languages
English (en)
French (fr)
Inventor
刘理峰
张静
李题印
留毅
姚海燕
徐超
屠永伟
Original Assignee
浙江群力电气有限公司
国家电网公司
国网浙江杭州市余杭区供电公司
国网浙江省电力公司杭州供电公司
国网浙江省电力公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 浙江群力电气有限公司, 国家电网公司, 国网浙江杭州市余杭区供电公司, 国网浙江省电力公司杭州供电公司, 国网浙江省电力公司 filed Critical 浙江群力电气有限公司
Priority to EP14897888.5A priority Critical patent/EP3012643B1/en
Publication of WO2016019593A1 publication Critical patent/WO2016019593A1/zh

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08Locating faults in cables, transmission lines, or networks
    • G01R31/088Aspects of digital computing
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R19/00Arrangements for measuring currents or voltages or for indicating presence or sign thereof
    • G01R19/25Arrangements for measuring currents or voltages or for indicating presence or sign thereof using digital measurement techniques
    • G01R19/2513Arrangements for monitoring electric power systems, e.g. power lines or loads; Logging

Definitions

  • the present invention relates to the field of electrical systems, and in particular to a method and apparatus for identifying a cause of overcurrent of a cable. Background technique
  • the early detection method of power cable is mainly to detect whether there is a short overcurrent in the cable.
  • this method can detect the short-time single-phase overcurrent signal, it cannot accurately indicate the result detected by the above method.
  • Early failure of the cable was detected because of the many causes of short-time overcurrent in a single phase, such as the commissioning of transformers and capacitors, load input, and starting of the motor. Therefore, the specific cause of the short-time single-phase overcurrent signal cannot be determined by the above method.
  • the detection of early cable faults mainly focuses on detecting the detection of short-time single-phase overcurrent signals without further identifying whether the overcurrent is caused by an early failure of the cable. Therefore, how to find the cause of the current single-phase overcurrent from many causes of short-time overcurrent is a problem that those skilled in the art need to solve at present. Summary of the invention
  • the embodiment of the present invention provides the following technical solutions:
  • a cable overcurrent cause identification method includes:
  • the method for detecting the single-phase overcurrent of the three-phase current signal is specifically: performing wavelet transformation on the three-phase current signal of the cable, calculating a maximum value of the wavelet transform mode of each phase current, and detecting whether there is only a wavelet transform mode pole of the one-phase current Whether the large value has two consecutive mutations in the 0.5 to 5 cycle, and if so, it is determined that there is a short overcurrent in the phase.
  • the method for determining the index characteristic vector sample of the three-phase current of the overcurrent is specifically: selecting a wavelet transform by using the db4 wavelet for the three-phase current signal of the cable, and performing 5-layer decomposition to obtain 5 layers of high-frequency detail coefficients and 1 layer The sum of the low frequency approximation coefficient and the energy value of each layer detail coefficient, the current phase current rms value and the current duration, and these nine indicators are used as the index feature vector samples.
  • the calculation formula of the sum of the energy values of the detail coefficients of each layer is: Where J ⁇ is the sum of the energy values of the detail coefficients of each layer, which is the detail coefficient of the _/ ⁇ layer.
  • the calculation formula of the degree of association is: Wherein, ⁇ is the correlation degree between the index feature vector sample and the first indicator vector reference sample, > is the first
  • the weight coefficient of each indicator is the correlation coefficient between the index feature vector sample and the first indicator of the first indicator vector reference sample, and S is the number of indicators.
  • the embodiment of the invention further provides a cable overcurrent cause identification device, comprising:
  • a three-phase current processing unit for performing single-phase overcurrent detection on the three-phase current signal; an index feature vector sample determining unit configured to determine an overcurrent phase current when a single-phase overcurrent is detected in the three-phase current signal An indicator feature vector sample of the signal;
  • a first calculating unit configured to calculate a correlation coefficient between the index feature vector sample and each index of the predetermined index vector reference samples, and one index vector reference sample corresponds to a cause of single-phase overcurrent
  • a second calculating unit configured to calculate, according to the index weight vector and the correlation coefficient of each indicator, the degree of association between the index feature vector sample and the predetermined indicator vector reference samples;
  • the discriminating unit is configured to find a single-phase overcurrent cause corresponding to the index vector reference sample with the largest correlation value.
  • the three-phase current processing unit includes: a wavelet transformer and a discriminator,
  • the wavelet transformer is configured to perform wavelet transform on the three-phase current signal of the cable, and calculate a wavelet transform modulus maximum value of each phase current;
  • the discriminator is configured to detect whether the wavelet transform modulus maximum value of only one phase current has two consecutive mutations in the 0.5 to 5 cycle, and if so, it is determined that the phase has a short overcurrent.
  • the indicator feature vector sample determining unit includes:
  • the coefficient determining sub-unit is configured to perform wavelet transform on the three-phase current signal of the cable by using db4 wavelet, and perform five-layer decomposition to obtain a sum of five layers of high-frequency detail coefficients and one layer of low-frequency approximation coefficients and energy values of each layer of detail coefficients, current Phase current rms and current duration, and these nine indicators are used as index feature vector samples.
  • the coefficient determining subunit includes:
  • £ ⁇ is the sum of the energy values of the detail coefficients of each layer, and ⁇ is the detail coefficient of the _ / layer.
  • the second calculating unit includes:
  • the weight coefficient of each indicator is the index of the indicator feature vector sample and the first indicator vector reference sample.
  • Correlation coefficient of / indicator, S is the number of indicators.
  • the cable overcurrent cause identification method provided by the embodiment of the present invention performs single-phase overcurrent detection on the three-phase current signal; and determines overcurrent when detecting the single-phase overcurrent of the three-phase current signal
  • An index feature vector sample of the phase current signal; a correlation coefficient between the index vector sample and the predetermined index of each index vector reference sample, and an index vector reference sample corresponding to a cause of single-phase overcurrent combined with the index weight vector and
  • the correlation coefficient of each index calculates the degree of association between the index feature vector sample and the predetermined index vector reference samples; and finds the single-phase overcurrent cause corresponding to the index vector reference sample with the largest correlation value.
  • FIG. 1 is a flowchart of a method for identifying a cause of overcurrent of a cable according to an embodiment of the present invention
  • FIG. 2 is a flowchart of a method for detecting single-phase overcurrent of a three-phase current signal according to an embodiment of the present invention
  • FIG. 3 is a structural block diagram of a cable overcurrent cause identification device according to an embodiment of the present invention
  • FIG. 4 is a structural block diagram of a three-phase current processing unit according to an embodiment of the present invention.
  • FIG. 5 is a structural block diagram of an indicator feature vector sample determining unit according to an embodiment of the present invention
  • FIG. 6 is a structural block diagram of a coefficient determining subunit provided by an embodiment of the present invention
  • FIG. 7 is a structural block diagram of a first computing unit according to an embodiment of the present invention.
  • FIG. 8 is a structural block diagram of a second computing unit according to an embodiment of the present invention.
  • FIG. 9 is a structural block diagram of a discriminating unit according to an embodiment of the present invention. detailed description
  • FIG. 1 is a flowchart of a method for identifying a cause of overcurrent of a cable according to an embodiment of the present invention, where the method may include:
  • Step sl00 performing single-phase overcurrent detection on the three-phase current signal
  • the three-phase current signal of the cable is sampled, and the three-phase current signals of a sample window are respectively transformed to determine whether there is one phase and only one phase has a single-phase overcurrent. If it exists, the process proceeds to the next step, if it does not exist or is greater than The presence of an overcurrent in one phase transforms the three-phase current signals in a sample window.
  • Step sll0 when detecting that the three-phase current signal has a single-phase overcurrent, determining an index feature vector sample of the overcurrent phase current signal;
  • the required characteristic values of each index obtained during the conversion process are recorded and calculated, and the index feature vector of the overcurrent phase current signal is formed according to the order of each index of the reference vector reference sample. sample.
  • Step sl20 calculating a correlation coefficient between the indicator feature vector sample and each index of the predetermined index vector reference sample, and an indicator vector reference sample corresponding to a cause of a single phase overcurrent;
  • the index vector reference sample is a reference value of each index when a single-phase overcurrent is formed due to early failure of the cable, transformer commissioning, capacitor operation, load input, and motor starting, which are formed based on a large amount of measured data or empirical data.
  • the reference value of each indicator forms a corresponding reference sample of the indicator vector.
  • a correlation coefficient between the index feature vector sample and a predetermined reference vector reference sample formed by each cause is calculated.
  • an indicator vector reference sample corresponds to a cause of single-phase overcurrent.
  • Step sl30 combining the index weight vector and the correlation coefficient of each index to calculate an indicator feature vector sample The degree of association between the present and the predetermined indicator vector reference samples;
  • the predetermined index weight vector sample is an expert weighting according to an expert's analysis of a large amount of measured data or empirical data, and the obtained index is weighted in the cause of the single-phase overcurrent of the cable, thereby forming an index weight vector sample.
  • the degree of association between the correlation coefficient and the predetermined index weight vector sample is calculated.
  • Step sl40 Find the single-phase overcurrent cause corresponding to the indicator vector reference sample with the largest correlation value.
  • the correlation degree value For the obtained correlation degree value, find the one with the largest value, and find the correlation coefficient correspondingly. In the reading correlation coefficient, it is calculated from which index vector reference sample of the index feature vector sample, and each index vector reference sample will have a corresponding one.
  • the reason for the single-phase overcurrent of the cable is to find the cause of the current single-phase overcurrent of the cable.
  • the cable overcurrent cause identification method performs single-phase overcurrent detection on the three-phase current signal; and determines an overcurrent phase current signal indicator when detecting that the three-phase current signal has a single-phase overcurrent
  • the eigenvector sample calculating the correlation coefficient between the index vector sample and the predetermined index of each index vector reference sample, and one index vector reference sample corresponding to a cause of single-phase overcurrent; combining the index weight vector with the correlation coefficient of each index The degree of association between the index feature vector sample and the predetermined index vector reference samples; and the single-phase overcurrent cause corresponding to the index vector reference sample having the largest correlation value.
  • the specific cause of the current single-phase overcurrent of the cable is determined, and it can also be determined which single-phase overcurrent of the cable is caused by an early failure. Therefore, it is of great practical significance to be able to accurately identify the early failure of the cable, so that it can take measures to repair and replace it, and actively and effectively reduce the ability of the grid to operate safely and extend the operating life of the grid line.
  • FIG. 2 is a flowchart of a method for performing single-phase overcurrent detection on a three-phase current signal according to an embodiment of the present invention.
  • the method for performing single-phase overcurrent detection on the three-phase current signal may be Includes:
  • Step s200 performing wavelet transform on the three-phase current signal of the cable
  • the wavelet transform of the three-phase current signal of the cable can be performed by using the db4 wavelet, decomposing the 5-layer method for wavelet transform, and selecting the db4 wavelet 5-layer decomposition for the three-phase current signal of each sample window. The number is separately wavelet transformed.
  • Step s210 Calculate a wavelet transform modulus maximum value of each phase current
  • the wavelet transform is a convolutional form with filtering significance. If the scale is ⁇ 3 ⁇ 4, the convolutional wavelet transform of the current signal ⁇ ) is WT a 0 ,t.
  • WTi(a 0 , to) is ⁇ ) The local modulus maximal of wavelet transform near t 0 at scale ⁇ 3 ⁇ 4 value. If U ⁇ . , t) on the upper there is a maximum value, then iQ) is the current signal ⁇ wavelet transform ot
  • Step s220 detecting whether the wavelet transform mode maximum value of each phase current continuously occurs twice in the 0.5 ⁇ 5 cycle wave, if yes, step S230 is performed, if not, step S260 is performed;
  • the three-phase current signals are respectively decomposed by the db4 wavelet 5 layer for wavelet transform, it is detected whether there is an overcurrent in each phase.
  • the early faults of the cable are generally single-phase ground faults, and the overcurrent caused generally only lasts for 0.5 ⁇ 5 cycles. Therefore, it is necessary to detect whether a second time occurs within 0.5 ⁇ 5 cycles of detecting the sudden change of the first current signal of a phase. mutation.
  • the sudden change point of the current signal in the time domain corresponds to the maximum point of 73 ⁇ 4fl, 0 after the wavelet transform, and according to the nature of the derivative, the zero-crossing point of W7 (a, 0 takes the maximum value and ⁇ , ⁇ ) Correspondence. Therefore, you can use W7 (a, 0's maximum point and W7 (a, 0's zero-crossing point to detect the sudden change of the current signal. If yes, go to the next step, if not, the three-phase current signal for the next sample window respectively Perform wavelet transformation.
  • Step s230 determining that there is a short-time overcurrent in the phase
  • Step S240 determining whether there is only a short overcurrent in the phase, if yes, proceeding to step S250, otherwise, proceeding to step s260;
  • Step S250 determining that the three-phase current signal has a single-phase overcurrent
  • Step s260 Perform wavelet transformation on the three-phase current signal of the next sample window.
  • each indicator of the index feature vector sample and the index vector reference sample may be obtained by summing the five layers of high frequency detail coefficients, one layer of low frequency approximation coefficients, and high frequency detail coefficients of the wavelet signal after the wavelet signal is decomposed, and the overcurrent continues. Time and the RMS value of the overcurrent phase current.
  • the calculation method of the indicator feature vector sample is:
  • s is the number of indicators. The contribution of each index to the identification of overcurrent is different. According to the analysis of a large number of measured data or empirical data, the weights of the various indicators obtained in the cause of the single-phase overcurrent of the cable are obtained.
  • the weight of the overcurrent, the index weight vector is formed by the wavelet transform decomposition principle, the detail coefficient and the approximation coefficient obtained by decomposing the signal on the first layer are respectively (in the present invention, the corpse 1 , 2, . . . , 5 ):
  • each indicator of the five causes of single-phase grounding early failure, transformer commissioning, capacitor commissioning, load input, and motor starting can be used to form an index vector reference sample.
  • the reference samples for each indicator when single-phase grounding early failure, transformer commissioning, capacitor operation, load input, and motor starting occur are established.
  • the index coefficient, the approximation coefficient and other index values, and the index values are obtained from measured data or empirical data.
  • a single calculation may be performed for one reason or one reason, or may be calculated by using a matrix form.
  • the specific calculation method is:
  • p is the resolution coefficient
  • a single calculation may be performed for one reason or one reason, or may be calculated in the form of a matrix.
  • the short-time single-phase overcurrent is judged to be caused by the corresponding cause, so that short-time single-phase overcurrent caused by early failure of the cable can be identified from many short-time single-phase overcurrent causes. .
  • the cable overcurrent cause identification method provided by the embodiment of the present invention can analyze the transient characteristics of the overcurrent in the time domain and the frequency domain simultaneously by performing the wavelet transform on the current signal in the time-frequency domain analysis method.
  • the post-modulus maxima can effectively detect sudden changes in the current signal and can be reconstructed in the time domain to obtain the duration of the mutation in the time domain.
  • the beneficial effect of the invention is that after the calculation, the correlation degree between the index feature vector and the reference sample vector of each factor is obtained, and the item with the largest correlation degree is found to find the cause of the overcurrent, thereby determining the single-phase overcurrent causing the current cable.
  • the embodiment of the invention provides a method for identifying a cable overcurrent cause, and the cable overcurrent cause identification can be performed by the above method.
  • the cable overcurrent cause identification device provided by the embodiment of the present invention will be described below.
  • the cable overcurrent cause identification device described below and the cable overcurrent cause identification method described above can be referred to each other.
  • FIG. 3 is a structural block diagram of a cable overcurrent cause identification device according to an embodiment of the present invention.
  • the cable overcurrent cause identification device may include: a three-phase current processing unit 100 for performing single-phase overcurrent detection on the three-phase current signal; an index feature vector sample determining unit 200, configured to detect the When there is a single-phase overcurrent in the three-phase current signal, an indicator characteristic vector sample of the overcurrent phase current signal is determined;
  • the first calculating unit 300 is configured to calculate a correlation coefficient between the indicator feature vector sample and each indicator of the predetermined index vector reference samples, and one indicator vector reference sample corresponds to a cause of single-phase overcurrent;
  • a second calculating unit 400 configured to calculate, according to the index weight vector and the correlation coefficient of each indicator, the degree of association between the index feature vector sample and the predetermined indicator vector reference samples;
  • FIG. 4 is a structural block diagram of a three-phase current processing unit 100 according to an embodiment of the present invention.
  • the three-phase current processing unit 100 may include:
  • the wavelet transformer 110 is configured to perform wavelet transform on the three-phase current signal of the cable, and calculate each phase current Wavelet transform modulus maxima,
  • the discriminator 120 is configured to detect whether there is only one phase change of the wavelet transform mode maximum value of the one-phase current in the 0.5 ⁇ 5 cycle wave, and if so, it is determined that the phase has a short-time overcurrent.
  • FIG. 5 is a structural block diagram of an indicator feature vector sample determining unit 200 according to an embodiment of the present invention.
  • the target feature vector sample determining unit 200 may include:
  • the coefficient determining subunit 210 is configured to perform wavelet transform on the cable three-phase current signal by using db4 wavelet, and perform 5-layer decomposition to obtain a sum of five layers of high-frequency detail coefficients, one layer of low-frequency approximation coefficients, and energy values of each layer of detail coefficients.
  • Current phase current rms value and current duration, and these nine indicators are used as index feature vector samples;
  • FIG. 7 is a structural block diagram of a first computing unit 300 according to an embodiment of the present invention.
  • the first computing unit 300 may include:
  • the first calculation execution subunit 310 calculates the correlation coefficient using the following formula
  • FIG. 8 is a structural block diagram of a second computing unit 400 according to an embodiment of the present invention.
  • the second computing unit 400 may include:
  • the weighting coefficient of each indicator is the correlation coefficient between the index feature vector sample and the first indicator of the first indicator vector reference sample, and s is the number of indicators.
  • FIG. 9 is a structural block diagram of a discriminating unit 500 according to an embodiment of the present invention.
  • the unit 500 can include:
  • the comparator 510 is configured to compare the size of the correlation degree, find the correlation degree with the largest value, and correspondingly find the index vector reference sample corresponding to the maximum correlation degree, because one index vector reference sample corresponds to one cause of single-phase overcurrent, That is, the cause of the current single-phase overcurrent is found.
  • the embodiment of the invention provides a cable overcurrent cause identification device, and the cable overcurrent cause identification can be performed by the above device.
  • the steps of a method or algorithm described in connection with the embodiments disclosed herein can be implemented directly in hardware, a software module executed by a processor, or a combination of both.
  • the software module can be placed in random access memory (RAM), memory, read only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or technical field. Any other form of storage medium known.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Power Engineering (AREA)
  • Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
  • Testing Relating To Insulation (AREA)
  • Emergency Protection Circuit Devices (AREA)

Abstract

一种电缆过电流原因辨识方法及装置,该方法包括:对三相电流信号进行单相过电流检测(S100);在检测到所述三相电流信号存在单相过电流时,确定过电流相电流信号的指标特征向量样本(S110);计算指标特征向量样本与预定的各指标向量参考样本各指标的关联系数(S120),一个指标向量参考样本对应一个引起单相过电流的原因;结合指标权重向量和各指标的关联系数计算指标特征向量样本与预定的各指标向量参考样本的关联度(S130);找到关联度数值最大的指标向量参考样本所对应的单相过电流原因(S140),从而识别电缆早期故障。

Description

一种电缆过电流原因辨识方法及装置
本申请要求于 2014年 8月 6日提交中国专利局、申请号 201410383429.3, 发明名称为 "一种电缆过电流原因辨识方法及装置"的中国专利申请的优先权, 其全部内容通过引用结合在本申请中。 技术领域
本发明涉及电气系统领域,特别是涉及一种电缆过电流原因辨识方法及装 置。 背景技术
随着城市化建设的不断推进,目前城市配电网日益趋向于釆用电力电缆进 行电力传输。 由于电力电缆一般埋设于土壤中或敷设于室内、 沟道、 隧道中, 线间绝缘距离小, 不用杆塔占地少且基本不占地面上空间, 其受到电击的可能 性小,受气候条件和周围环境影响小,传输性能稳定且可靠性高,供电更可靠、 更安全。但是电力电缆发生的闪络、局放等情况可能会导致电力电缆永久性故 障, 且该故障查找困难。
电力电缆的早期故障通常是由于绝缘层内部老化而逐渐恶化导致,一般表 现为持续时长为 0.5 ~ 5个周波的短时单相过电流。 如果早期故障不能够及时、 有效的得到检测和识别并加以修复或更换,将会发展成为永久性故障。 因此对 电力电缆的早期故障提前预警, 就能够积极有效地减少发生严重故障的情况, 力, 延长电网线路的运行寿命, 都具有十分重要的实际意义。
目前电力电缆的早期故障的检测方法主要是通过检测电缆是否存在短时 过电流,但是该方法虽然能够检测到短时单相过电流信号,但是并不能以上述 方法检测到的结果来准确地表明检测到了电缆的早期故障,因为引起单相短时 过电流的原因比较多, 例如变压器和电容器的投运、 负荷投入以及电动机的起 动等原因。因此利用上述方法并不能够确定引起短时单相过电流信号的具体原 因。 目前对电缆早期故障的检测主要集中检测到短时单相过电流信号的检测, 而没有进一步辨识该过电流是否由电缆的早期故障引起。 因此如何从众多引起短时过电流的原因中找到引起当前单相过电流的原 因, 是本领域的技术人员目前需要解决的问题。 发明内容
本发明的目的是提供一种电缆过电流原因辨识方法,该方法能够完成对引 起电缆当前单相过电流的原因的辨识;本发明的另一目的是提供一种电缆过电 流原因辨识的装置。
为解决上述技术问题, 本发明实施例提供如下技术方案:
一种电缆过电流原因辨识方法, 包括:
1 )对三相电流信号进行单相过电流检测;
2 )在检测到所述三相电流信号存在单相过电流时, 确定过电流相电流信 号的指标特征向量样本;
3 )计算指标特征向量样本与预定的各指标向量参考样本的各指标的关联 系数, 一个指标向量参考样本对应一个引起单相过电流的原因;
4 )结合指标权重向量和各指标的关联系数计算指标特征向量样本与预定 的各指标向量参考样本的关联度;
5 )找到关联度数值最大的指标向量参考样本所对应的单相过电流原因。 其中, 所述三相电流信号进行单相过电流检测方法具体为: 对电缆三相电 流信号进行小波变换,计算各相电流小波变换模极大值,检测是否仅有一相电 流的小波变换模极大值在 0.5 ~ 5周波内是否连续发生两次突变, 若是, 则确 定该相存在短时过电流。
其中, 所述确定过电流的三相电流的指标特征向量样本的方法具体为:对 电缆三相电流信号选用 db4小波进行小波变换, 并进行 5层分解,得到 5层高 频细节系数和 1层低频近似系数及各层细节系数能量值之和,电流相电流有效 值和电流持续时间, 并将这九个指标作为指标特征向量样本。
其中, 所述各层细节系数能量值之和的计算公式为:
Figure imgf000004_0001
其中, J ^为各层细节系数能量值之和, 为第 _/·层的细节系数。
其中, 所述关联度的计算公式为: 其中 ^为指标特征向量样本与第 个指标向量参考样本关联度, >为第
/个指标的权重系数, 为指标特征向量样本与第 个指标向量参考样本的第 /个指标的关联系数, S为指标个数。 本发明实施例还提供一种电缆过电流原 因辨识装置, 包括:
三相电流处理单元, 用于对三相电流信号进行单相过电流检测; 指标特征向量样本确定单元,用于在检测到所述三相电流信号存在单相过 电流时, 确定过电流相电流信号的指标特征向量样本;
第一计算单元,用于计算指标特征向量样本与预定的各指标向量参考样本 的各指标的关联系数, 一个指标向量参考样本对应一个引起单相过电流的原 因;
第二计算单元,用于结合指标权重向量和各指标的关联系数计算指标特征 向量样本与预定的各指标向量参考样本的关联度;
判别单元,用于找到关联度数值最大的指标向量参考样本所对应的单相过 电流原因。 其中, 所述三相电流处理单元包括: 小波变换器和判别器,
小波变换器, 用于对电缆三相电流信号进行小波变换,计算各相电流小波 变换模极大值;
判别器, 用于检测是否仅有一相电流的小波变换模极大值在 0.5 ~ 5周波 内是否连续发生两次突变, 若是, 则确定该相存在短时过电流。
其中, 所述指标特征向量样本确定单元包括:
系数确定子单元, 用于对电缆三相电流信号选用 db4小波进行小波变换, 并进行 5层分解,得到 5层高频细节系数和 1层低频近似系数及各层细节系数 能量值之和, 电流相电流有效值和电流持续时间, 并将这九个指标作为指标特 征向量样本。
其中, 所述系数确定子单元包括:
能量值计算子单元,用于根据公式 =∑ ∑ 计算各层细节系数能量 值之和;
其中, £^为各层细节系数能量值之和, ^为第 _ /层的细节系数。
其中, 所述第二计算单元包括:
第二计算执行子单元, 用于根据公式 ^= ) ) 计算关联度; 其中 ^为指标特征向量样本与第 个指标向量参考样本关联度, >为第
/个指标的权重系数, 为指标特征向量样本与第 个指标向量参考样本的第
/个指标的关联系数, S为指标个数。 基于上述技术方案, 本发明实施例所提 供的电缆过电流原因辨识方法,对三相电流信号进行单相过电流检测; 在检测 到所述三相电流信号存在单相过电流时,确定过电流相电流信号的指标特征向 量样本;计算指标特征向量样本与预定的各指标向量参考样本的各指标的的关 联系数,一个指标向量参考样本对应一个引起单相过电流的原因; 结合指标权 重向量和各指标的关联系数计算指标特征向量样本与预定的各指标向量参考 样本的关联度;找到关联度数值最大的指标向量参考样本所对应的单相过电流 原因。从而确定出引起当前电缆单相过电流的具体原因,也可以确定哪些电缆 单相过电流是由于早期故障所引起的。 因此能够准确的辨别电缆早期故障,从 而就可以釆取措施进行修复和更换,就能够积极有效地减少发生严重故障的情 能力, 延长电网线路的运行寿命, 都具有十分重要的实际意义。 附图说明
图 1为本发明实施例提供的电缆过电流原因辨识方法的流程图; 图 2 为本发明实施例提供的对三相电流信号进行单相过电流检测的方法 流程图;
图 3为本发明实施例提供的电缆过电流原因辨识装置的结构框图; 图 4为本发明实施例提供的三相电流处理单元的结构框图;
图 5为本发明实施例提供的指标特征向量样本确定单元的结构框图; 图 6为本发明实施例提供的系数确定子单元的结构框图;
图 7为本发明实施例提供的第一计算单元的结构框图;
图 8为本发明实施例提供的第二计算单元的结构框图;
图 9为本发明实施例提供的判别单元的结构框图。 具体实施方式
本发明的核心是提供一种电缆过电流原因辨识方法,该方法能够完成对引 起电缆当前单相过电流的原因的辨识;本发明的另一目的是提供一种电缆过电 流原因辨识的装置。
为使本发明实施例的目的、技术方案和优点更加清楚, 下面将结合本发明 实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然, 所描述的实施例是本发明一部分实施例, 而不是全部的实施例。基于本发明中 的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其 他实施例, 都属于本发明保护的范围。
请参考图 1, 图 1为本发明实施例提供的电缆过电流原因辨识方法的流程 图, 该方法可以包括:
步骤 sl00、 对三相电流信号进行单相过电流检测;
这里对电缆三相电流信号进行釆样,对一个釆样窗口的三相电流信号分别 进行变换, 判断是否存在且仅有一相存在单相过电流, 若存在则进入下一步, 若不存在或者大于一相存在过电流则对在一个釆样窗口的三相电流信号分别 进行变换。
步骤 sll0、在检测到所述三相电流信号存在单相过电流时,确定过电流相 电流信号的指标特征向量样本;
在检测到三相电流信号存在单相过电流时,记录、计算变换过程中得到的 所需要的各指标特征值,按照与指标向量参考样本各指标的顺序形成过电流相 电流信号的指标特征向量样本。
步骤 sl20、 计算指标特征向量样本与预定的各指标向量参考样本的各指 标的关联系数, 一个指标向量参考样本对应一个 I起单相过电流的原因;
其中,指标向量参考样本是根据大量实测数据或经验数据形成的当电缆发 生早期故障、 变压器投运、 电容器投运、 负荷投入及电动机起动时等原因形成 单相过电流时的各项指标参考数值,各项指标参考数值形成对应的指标向量参 考样本。
计算指标特征向量样本与每一个原因形成的预定的各指标向量参考样本 的关联系数。当然在这里一个指标向量参考样本对应一个引起单相过电流的原 因。
步骤 sl30、 结合指标权重向量和各指标的关联系数计算指标特征向量样 本与预定的各指标向量参考样本的关联度;;
其中,预定指标权重向量样本是根据专家对大量实测数据或经验数据的分 析进行专家打分, 得到的各个指标在引起电缆单相过电流原因中所占的权重, 从而形成的指标权重向量样本。
计算得到关联系数与预定指标权重向量样本的关联度。
步骤 sl40、 找到关联度数值最大的指标向量参考样本所对应的单相过电 流原因。
对得到的关联度数值, 找到数值最大的一个, 对应的找到关联系数, 在看 次关联系数是由指标特征向量样本于哪个指标向量参考样本计算得到,每个指 标向量参考样本都会对应有一个形成电缆单相过电流的原因,这样就找到了引 起当前电缆单相过电流的原因。
本发明实施例所提供的电缆过电流原因辨识方法,对三相电流信号进行单 相过电流检测; 在检测到所述三相电流信号存在单相过电流时,确定过电流相 电流信号的指标特征向量样本;计算指标特征向量样本与预定的各指标向量参 考样本的各指标的关联系数,一个指标向量参考样本对应一个引起单相过电流 的原因;结合指标权重向量和各指标的关联系数计算指标特征向量样本与预定 的各指标向量参考样本的关联度;找到关联度数值最大的指标向量参考样本所 对应的单相过电流原因。从而确定出引起当前电缆单相过电流的具体原因,也 可以确定哪些电缆单相过电流是由于早期故障所引起的。因此能够准确的辨别 电缆早期故障,从而就可以釆取措施进行修复和更换, 就能够积极有效地减少 电网的安全运行的能力,延长电网线路的运行寿命,都具有十分重要的实际意 义。
可选的,图 2给出了本发明实施例提供的对三相电流信号进行单相过电流 检测的方法流程图, 参照图 2, 该对三相电流信号进行单相过电流检测的方法 可以包括:
步骤 s200、 对电缆三相电流信号进行小波变换;
可选的, 对电缆三相电流信号进行小波变换可以利用 db4 小波, 分解 5 层的方法进行小波变换,选用 db4小波 5层分解对每一釆样窗口的三相电流信 号分别进行小波变换。
步骤 s210、 计算各相电流小波变换模极大值;
可选的,在定义信号的小波变换模极大值时, 小波变换是釆用具有滤波意 义的卷积形式。若在尺度为 ί¾时,电流信号 ί)的卷积形小波变换为 WT a0,t。
如果在 t0的领域内满足条件:
\WTi(a0, t)\<\WTi(a0, t0)\ (1) 则称 WTi(a0, to)为 ί)在尺度 ί¾时在 t0附近的小波变换局部模极大值。 若 U^。,t)在 上有一过极大值, 则称 iQ)为电流信号^的小波变换 ot
模极大值点。
步骤 s220、 检测各相电流小波变换模极大值在 0.5 ~ 5周波内是否连续发 生两次突变, 若是, 执行步骤 S230, 若否, 执行步骤 S260;
可选的, 由于将三相电流信号分别用 db4小波 5层分解进行小波变换,检 测各相是否存在过电流。 电缆早期故障一般为单相接地故障,且引起的过电流 一般仅持续 0.5 ~ 5 个周波, 因此需要在检测到某相第一次电流信号突变的 0.5 ~ 5个周波内检测是否发生第二次突变。
可选的, 突变的计算为:
设函数 6>(0满足小波函数的基本性质, 即
Figure imgf000009_0001
(2) 且 0为具有低通平滑的滤波函数。 若 0二阶可导, 一阶导数和二阶导数分 别为 it)和 "( ), 由傅里叶变换的微分性质可知 和 ?/(χ)仍然满足容许条件, 即可以作为小波母函数。 对 ί)分别用 it)和?/ (t)进行小波变换, 得
WTi, (a, t) =
Figure imgf000009_0002
WTin (a, t) =丄厂 i ^ h = a2^ [i(t) * θα (t)] 由式 (3)和 (4)可知 WT (a, ί)可以等效为 0通过与 6>fl(t)卷积小波变换后, 再求一阶导数得到的, 而 7)ηΟ,ί)是 ί)通过与 (t)卷积小波变换后, 再求二 阶导数得到的。电流信号在时域的突变点对应于小波变换后 7¾fl,0的极大值 点, 而根据导数的性质可知 W7 (a, 0取得极大值的点与 ΨΤφ, ί)的过零点相 对应。 因此可以用 W7 (a, 0的极大值点和 W7 (a, 0的过零点来检测电流信号 的突变点。是则进入下一步, 不是则对下一个釆样窗口的三相电流信号分别进 行小波变换。
步骤 s230、 确定该相存在短时过电流;
若检测各相电流小波变换模极大值在 0.5 ~ 5周波内连续发生两次突变; 则该相存在短时过电流。
步骤 S240、 判断是否只有该相存在短时过电流, 是则, 执行步骤 S250, 否则, 进行步骤 s260;
步骤 S250、 确定所述三相电流信号存在单相过电流;
步骤 s260、 对下一釆样窗口三相电流信号进行小波变换。
可选的,指标特征向量样本和指标向量参考样本的各指标可以由小波信号 分解后的 5层高频细节系数、 1层低频近似系数、 以及高频细节系数的能量值 之和、 过电流持续时间以及过电流相电流有效值构成。
可选的, 指标特征向量样本的计算方法为:
利用上述小波变换分析,计算过电流相小波信号分解后的 5层高频细节系
$i[d d2, d3, d4, d5]、 1层低频近似系数 以及高频细节系数的能量值之和 Ed 并结合过电流持续时间 T以及过电流相电流有效值 构成电缆早期故障辨识 指标的特征向量
Figure imgf000010_0001
, s为指 标的个数。各个指标对辨识过电流的贡献不同,根据预定指标权重向量样本是 根据对大量实测数据或经验数据的分析,得到的各个指标在引起电缆单相过电 流原因中所占的权重得到各指标在辨识过电流时的权重, 形成指标权重向量 由小波变换分解原理可知,信号在第 层上分解得到的细节系数 和近似 系数 分别为 (本发明中取尸1 , 2, . . . , 5 ):
' (5) 其中 g和 分别为低通和高通滤波器, 上式表明, 第 尺度上的细节系数和逼 近系数可有第 -1尺度上的逼近系数分别与高、 低通滤波器进行卷积再进行二 抽取后得到; ,为窗口中的当前釆样点, 对 I的求和表示对当前窗口中所有釆 样点的求和(例如, 当釆用频率为 3.2kHZ时, 每窗口中含有 8个釆样点, 则 z=l, 2, ... , 8 ); «为小波函数的离散化程度, "e Z
第 层分解信号的细节系数 的能量值可表示为
^ =∑ (6) 那么当前窗口 5层高频细节系数的能量值之和
(7) 1 1 1 可选的, 可以将单相接地早期故障、 变压器投运、 电容器投运、 负 荷投入及电动机起动时的 5种原因的各指标参考样本形成指标向量参考 样本。
可选的, 按照 ¾中各指标的顺序建立当发生单相接地早期故障、 变压器 投运、 电容器投运、 负荷投入及电动机起动时的各指标参考样本
Figure imgf000011_0001
其中 ^=1,2,... ,5)为上述五种原因引起的短时单相过电流特征向量,与;^维 数相同; (1)到 )为上述电缆短时单相过电流的细节系数、 近似系数等各 指标值, 各指标值由实测数据或经验数据得到。
可选的, 计算关联系数时可以一个原因一个原因进行单个计算, 也 可以利用矩阵的形式进行计算。 可选的, 当利用矩阵形式计算式, 具体计算方法为:
计算比较样本 与各参考特征向量 的关联系数, 比较样本 与参考 序列 X( ^第 1(1=1, 2,... 个指标的关联系数为, min | i (0 - xok ( | + max
l≤I≤s l≤I≤s l i ( - ( |
| i (0 ~ xok ( | + max | i (0 ~ ·½ ( |
\<l≤s
\<k<5
式中 p为分辨系数, 通过选取合适的分辨率能够提高关联分析的抗 干扰能力, 一般取 >=0.5。 可选的, 计算关联度时可以一个原因一个原因进行单个计算, 也可 以利用矩阵的形式进行计算。 可选的, 当利用矩阵形式计算式, 具体计算方法为: 由关联系数矩阵 <f结合指标的权重向量 ωχ, 即可得到比较样本 与参考 序列;^!^的关联度]^ =∑ω^ (10) 式中 )为第 /个指标的权重系数。 由于权重系数 和关联系数^都是 [0, 1] 之间的数,所以比较样本 与各参考序列 的关联度] ¾的范围也为 [0,1]。 当 }¾=0时, 表明 与 完全不相关; 当 ] ¾=1时, 表明 越大则 与 的相关程度越大。
找到 最大的一项,则本次短时单相过电流判定为由 对应的原因引起的, 从而可以从众多短时单相过电流原因中辨识出由电缆早期故障引起的短时单 相过电流。
基于上述技术方案, 本发明实施例所提供的电缆过电流原因辨识方法,在 时频域分析方法可以同时在时域和频域对过电流的暂态特征进行分析,通过对 电流信号进行小波变换后的模极大值能有效地检测出电流信号的突变,并能够 在时域进行重构以获得时域内突变持续的时间。利用在进行小波变换的时频域 分析过程中得到的各个指标、在时域内突变持续的时间以及电流信号进行小波 变换后的模极大值构成指标特征向量,利用该指标特征向量与引起电缆早期故 障的各因素所建立的参考样本向量中的各个指标进行关联系数的计算,利用关 联系数矩阵与预先设置好的指标权重向量进行关联度的计算, 得到各关联度, 本发明的有益效果就是经过计算,得到指标特征向量与各因素的参考样本向量 的关联度,找到关联度最大的项从而找到了引起过电流的原因,从而确定出引 起当前电缆单相过电流的具体原因,也可以确定哪些电缆单相过电流是由于早 期故障所引起的。 因此能够准确的辨别电缆早期故障,从而就可以釆取措施进 行修复和更换, 就能够积极有效地减少发生严重故障的情况,对于减少因电力 电缆发生故障而导致的停电损失,提高电网的安全运行的能力,延长电网线路 的运行寿命, 都具有十分重要的实际意义。
本发明实施例提供了电缆过电流原因辨识方法, 可以通过上述方法 进行电缆过电流原因辨识。
下面对本发明实施例提供的电缆过电流原因辨识装置进行介绍,下文描述 的电缆过电流原因辨识装置与上文描述的电缆过电流原因辨识方法可相互对 应参照。
图 3为本发明实施例提供的电缆过电流原因辨识装置的结构框图。 参照图 3, 该电缆过电流原因辨识装置可以包括: 三相电流处理单元 100, 用于对三相电流信号进行单相过电流检测; 指标特征向量样本确定单元 200, 用于在检测到所述三相电流信号存在单 相过电流时, 确定过电流相电流信号的指标特征向量样本;
第一计算单元 300, 用于计算指标特征向量样本与预定的各指标向量参考 样本的各指标的关联系数,一个指标向量参考样本对应一个引起单相过电流的 原因;
第二计算单元 400, 用于结合指标权重向量和各指标的关联系数计算指标 特征向量样本与预定的各指标向量参考样本的关联度;
判别单元 500, 用于找到关联度数值最大的指标向量参考样本所对应的单 相过电流原因。 可选的, 图 4 示出了本发明实施例提供的三相电流处理单元 100的结构框图, 三相电流处理单元 100可以包括:
小波变换器 110, 用于对电缆三相电流信号进行小波变换, 计算各相电流 小波变换模极大值,
判别器 120, 用于检测是否仅有一相电流的小波变换模极大值在 0.5 ~ 5 周波内是否连续发生两次突变, 若是, 则确定该相存在短时过电流。
可选的, 图 5示出了本发明实施例提供的指标特征向量样本确定单元 200 的结构框图, 标特征向量样本确定单元 200可以包括:
系数确定子单元 210, 用于对电缆三相电流信号选用 db4小波进行小波变 换, 并进行 5层分解,得到 5层高频细节系数和 1层低频近似系数及各层细节 系数能量值之和, 电流相电流有效值和电流持续时间, 并将这九个指标作为指 标特征向量样本;
可选的,图 6示出了本发明实施例提供的系数确定子单元 210的结构框图, 系数确定子单元 210 可以包括能量值计算子单元 220, 用于根据公式 Ed =∑ ∑ d计算各层细节系数能量值之和; 其中, £^为各层细节系数能量值之和, 4为第 ·/·层的细节系数。
可选的, 图 7示出了本发明实施例提供的第一计算单元 300的结构框图, 第一计算单元 300可以包括:
第一计算执行子单元 310, 利用下面公式计算关联系数;
min | ι (0 ~ xok ( | + max | ι ( - ·½ ( |
ξ l≤l≤s l≤I≤s
(1) _ \<k<5 \≤k<5
\x\ (0 ~ xok ( | + max l i ( - ·½ ( |
l≤l≤s
l≤k≤5
其中, 式中 p为分辨系数, 通过选取合适的分辨率能够提高关联分 析的抗干扰能力, 一般取 p=0.5。 可选的, 图 8示出了本发明实施例提供的第二计算单元 400的结构框图, 第二计算单元 400可以包括:
第二计算执行子单元 410, 用于根据公式 ^ = ; '^) 计算关联度; 其中 ^为指标特征向量样本与第 个指标向 ΐ1参考样本关联度, )为第
/个指标的权重系数, 为指标特征向量样本与第 个指标向量参考样本的第 /个指标的关联系数, s为指标个数。
可选的, 图 9示出了本发明实施例提供的判别单元 500的结构框图, 判 别单元 500可以包括:
比较器 510, 用于比较关联度的大小, 找到数值最大的关联度, 并对应找 到此最大关联度所对应的指标向量参考样本,由于一个指标向量参考样本对应 一个引起单相过电流的原因, 即找到了引起当前单相过电流的原因。
本发明实施例提供了电缆过电流原因辨识装置, 可以通过上述装置进行 电缆过电流原因辨识。
本说明书中各个实施例釆用递进的方式描述,每个实施例重点说明的都是 与其他实施例的不同之处,各个实施例之间相同相似部分互相参见即可。对于 实施例公开的装置而言, 由于其与实施例公开的方法相对应, 所以描述的比较 简单, 相关之处参见方法部分说明即可。
专业人员还可以进一步意识到,结合本文中所公开的实施例描述的各示例 的单元及算法步骤, 能够以电子硬件、 计算机软件或者二者的结合来实现, 为 了清楚地说明硬件和软件的可互换性,在上述说明中已经按照功能一般性地描 述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行,取决于 技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来 使用不同方法来实现所描述的功能, 但是这种实现不应认为超出本发明的范 围。
结合本文中所公开的实施例描述的方法或算法的步骤可以直接用硬件、 处理器执行的软件模块, 或者二者的结合来实施。软件模块可以置于随机存储 器 (RAM )、 内存、 只读存储器 (ROM )、 电可编程 ROM、 电可擦除可编程 ROM, 寄存器、 硬盘、 可移动磁盘、 CD-ROM、 或技术领域内所公知的任意 其它形式的存储介质中。
以上对本发明所提供的电缆过电流原因辨识方法及装置进行了详细介绍。 说明只是用于帮助理解本发明的方法及其核心思想。应当指出,对于本技术领 域的普通技术人员来说,在不脱离本发明原理的前提下,还可以对本发明进行 若干改进和修饰, 这些改进和修饰也落入本发明权利要求的保护范围内。
+

Claims

权 利 要 求
1、 一种电缆过电流原因辨识方法, 其特征在于, 该方法包括:
1 )对三相电流信号进行单相过电流检测;
2 )在检测到所述三相电流信号存在单相过电流时, 确定过电流相电流信 号的指标特征向量样本;
3 )计算指标特征向量样本与预定的各指标向量参考样本的各指标的关联 系数, 一个指标向量参考样本对应一个引起单相过电流的原因;
4 )结合预定指标权重向量和各指标的关联系数计算指标特征向量样本与 预定的各指标向量参考样本的关联度;
5 )找到与关联度数值最大的指标向量参考样本所对应的单相过电流原因。
2、 如权利要求 1所述的方法, 其特征在于, 所述步骤 1 ) 中对三相电流 信号进行单相过电流检测方法具体为: 对电缆三相电流信号进行小波变换,计 算各相电流小波变换模极大值,检测是否仅有一相电流的小波变换模极大值在 0.5 ~ 5周波内是否连续发生两次突变, 若是, 则确定该相存在短时过电流。
3、 如权利要求 1所述的方法, 其特征在于, 所述步骤 2 ) 中确定过电流 的三相电流的指标特征向量样本的方法具体为:对电缆三相电流信号选用 db4 小波进行小波变换, 并进行 5层分解,得到 5层高频细节系数和 1层低频近似 系数及各层细节系数能量值之和, 电流相电流有效值和电流持续时间, 并将这 九个指标作为指标特征向量样本。
4、 如权利要求 3所述的方法, 其特征在于, 所述各层细节系数能量值之 和的计算公式为: Ed = U j ; 其中, ^为各层细节系数能量值之和, 为第 ·层的细节系数。
5、 如权利要求 1所述的方法, 其特征在于, 所述步骤 4 ) 中关联度的计 算公式为:
= M、 , 其中 ^为指标特征向量样笨与第 个指标向量参考样本的关联度, )为 第 /个指标的权重系数, ^ 为指标特征向量样本与第 个指标向量参考样本的 第 /个指标的关联系数, S为指标个数。
6、 一种电缆过电流原因辨识装置, 其特征在于, 包括: 三相电流处理单元, 用于对三相电流信号进行单相过电流检测; 指标特征向量样本确定单元,用于在检测到所述三相电流信号存在单相过 电流时, 确定过电流相电流信号的指标特征向量样本;
第一计算单元,用于计算指标特征向量样本与预定的各指标向量参考样本 的各指标的关联系数, 一个指标向量参考样本对应一个引起单相过电流的原 因;
第二计算单元,用于结合指标权重向量和各指标的关联系数计算指标特征 向量样本与预定的各指标向量参考样本的关联度;
判别单元,用于找到关联度数值最大的指标向量参考样本所对应的单相过 电流原因。
7、 如权利要求 6所述的电缆过电流原因辨识装置, 其特征在于, 所述三 相电流处理单元包括: 小波变换器和判别器, 其中,
小波变换器, 用于对电缆三相电流信号进行小波变换,计算各相电流小波 变换模极大值;
判别器, 用于检测是否仅有一相电流的小波变换模极大值在 0.5 ~ 5周波 内是否连续发生两次突变, 若是, 则确定该相存在短时过电流。
8、 如权利要求 6所述的电缆过电流原因辨识装置, 其特征在于, 所述指 标特征向量样本确定单元包括:
系数确定子单元, 用于对电缆三相电流信号选用 db4小波进行小波变换, 并进行 5层分解,得到 5层高频细节系数和 1层低频近似系数及各层细节系数 能量值之和, 电流相电流有效值和电流持续时间, 并将这九个指标作为指标特 征向量样本。
9、 如权利要求 8所述的电缆过电流原因辨识装置, 其特征在于, 所述系 数确定子单元包括:
能量值计算子单元,用于根据公式 Ed = 计算各层细节系数能量 值之和;
其中, ^为各层细节系数能量值之和, 为第 ·层的细节系数。
10、 如权利要求 6所述的电缆过电流原因辨识装置, 其特征在于, 所述第 二计算单元包括: 第二计算执行子单元, 用于根据公式 ^ =t^^) 计算关联度; 其中 ^为指标特征向量样本与第 个指 rf 量参考样本关联度, 为第
/个指标的权重系数, 为指标特征向量样本与第 个指标向量参考样本的第 /个指标的关联系数, S为指标个数。
PCT/CN2014/084248 2014-08-06 2014-08-13 一种电缆过电流原因辨识方法及装置 WO2016019593A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP14897888.5A EP3012643B1 (en) 2014-08-06 2014-08-13 Method and apparatus for identifying causes for cable overcurrent

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410383429.3 2014-08-06
CN201410383429.3A CN104122487B (zh) 2014-08-06 2014-08-06 一种电缆过电流原因辨识方法及装置

Publications (1)

Publication Number Publication Date
WO2016019593A1 true WO2016019593A1 (zh) 2016-02-11

Family

ID=51767977

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/084248 WO2016019593A1 (zh) 2014-08-06 2014-08-13 一种电缆过电流原因辨识方法及装置

Country Status (3)

Country Link
EP (1) EP3012643B1 (zh)
CN (1) CN104122487B (zh)
WO (1) WO2016019593A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104501873B (zh) * 2014-12-29 2017-02-22 浙江群力电气有限公司 一种环网柜的集成信息监控系统
CN108896870B (zh) * 2018-06-28 2019-12-24 四川大学 工频和冲击联合作用下的输电线路故障识别方法
CN112505484B (zh) * 2020-11-27 2023-03-21 广东电网有限责任公司佛山供电局 一种小电阻接地系统中压配电电缆潜伏性故障识别方法
TWI781006B (zh) * 2021-12-10 2022-10-11 國立勤益科技大學 遠端電力電纜瑕疵故障檢測方法
CN115754598B (zh) * 2022-11-10 2024-04-16 海南电网有限责任公司乐东供电局 基于相关滤波的电缆故障查找方法和装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101907437A (zh) * 2010-07-23 2010-12-08 西安科技大学 一种基于小波差分算法的电缆故障测距方法
CN102253327A (zh) * 2011-06-16 2011-11-23 长沙河野电气科技有限公司 开关电流电路故障诊断方法
CN102567783A (zh) * 2012-02-10 2012-07-11 广西电网公司电力科学研究院 并联混合型电能质量调节器专家故障分析与诊断方法
CN102570392A (zh) * 2012-01-17 2012-07-11 上海电力学院 基于改进概率神经网络的变压器励磁涌流鉴别方法
CN103713237A (zh) * 2013-12-30 2014-04-09 华北电力大学 一种电力系统输电线路短路故障诊断方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1244819C (zh) * 2004-03-22 2006-03-08 西安交通大学 交联聚乙烯电力电缆外护套绝缘在线检测装置
CN201156070Y (zh) * 2007-11-30 2008-11-26 北京市电力公司 电缆故障指示器
CN102435908B (zh) * 2011-09-07 2013-12-04 兖州煤业股份有限公司 基于零模和相模电流行波单相接地故障在线测距方法
CN102680854B (zh) * 2012-05-08 2014-08-20 山东康威通信技术股份有限公司 一种电力电缆隐性故障及接头工艺缺陷在线监测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101907437A (zh) * 2010-07-23 2010-12-08 西安科技大学 一种基于小波差分算法的电缆故障测距方法
CN102253327A (zh) * 2011-06-16 2011-11-23 长沙河野电气科技有限公司 开关电流电路故障诊断方法
CN102570392A (zh) * 2012-01-17 2012-07-11 上海电力学院 基于改进概率神经网络的变压器励磁涌流鉴别方法
CN102567783A (zh) * 2012-02-10 2012-07-11 广西电网公司电力科学研究院 并联混合型电能质量调节器专家故障分析与诊断方法
CN103713237A (zh) * 2013-12-30 2014-04-09 华北电力大学 一种电力系统输电线路短路故障诊断方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3012643A4 *

Also Published As

Publication number Publication date
EP3012643A4 (en) 2017-04-26
CN104122487B (zh) 2017-02-08
EP3012643B1 (en) 2020-09-30
EP3012643A1 (en) 2016-04-27
CN104122487A (zh) 2014-10-29

Similar Documents

Publication Publication Date Title
WO2016019593A1 (zh) 一种电缆过电流原因辨识方法及装置
CN110068759B (zh) 一种故障类型获得方法及装置
Ananthan et al. Real‐time fault analysis of transmission lines using wavelet multi‐resolution analysis based frequency‐domain approach
CN110320434B (zh) 基于零序电流波形区间斜率曲线的高阻故障辨识方法及系统
WO2019105486A1 (zh) 差动保护的判断方法
CN105510760B (zh) 一种基于小波分析的短路故障数据检测方法
JP2021148778A (ja) 非接地配電システムのための過渡状態に基づいた故障位置特定方法
CN112255492A (zh) 一种强噪声背景下的配电网单相接地高阻故障识别方法
Cao et al. Speeded‐up robust features based single‐ended travelling wave fault location: a practical case study in Yunnan power grid of China
CN116679161A (zh) 一种电网线路故障诊断方法、设备和介质
Taheri et al. A fault‐location algorithm for parallel line based on the long short‐term memory model using the distributed parameter line model
CN112162172A (zh) 一种基于有限采样数据的串联故障电弧检测方法及系统
CN112946515A (zh) 一种用于换流变网侧套管的在线监测方法和装置
Lei et al. A faulty pole detection method for the VSC-HVDC system based on the aperiodic component energy
JP2011237182A (ja) 部分放電判別装置及び部分放電判別方法
CN114236212A (zh) 电压测量方法、电压测量电路及设备
CN111044865B (zh) 一种直流气体绝缘电气设备局部放电小波去噪方法
CN109324268B (zh) 基于贝叶斯推理的配电网早期故障检测方法及装置
CN104849530B (zh) 一种moa阻性电流基波的测量方法
Ahmadi-khaneghahi et al. A fast and robust local-based protection algorithm based on the high-frequency transient for HVDC grid
CN107561400B (zh) 一种基于实时小波变换的配网短路故障快速检测方法
CN114201991A (zh) 一种基于超声传感器阵列的局放信号检测方法和系统
CN114884027A (zh) 一种电流差动保护方法和系统
CN111649819A (zh) 一种利用迭代软阈值的变压器状态振声检测信号滤波方法和系统
JP2005283489A (ja) 電路の部分放電検出方法

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 2014897888

Country of ref document: EP

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14897888

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE