• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    PURPOSE: A method for detecting a partial discharge of a gas insulation switchgear by using a method for decomposing the unique value from an attractor used in a chaos theory is provided to apply an algorism easily realized in the hardware. CONSTITUTION: A method for detecting a partial discharge of a gas insulation switchgear by using a method for decomposing the unique value from an attractor used in a chaos theory includes the steps of: a first step(S2) of removing and normalizing the noise for the partial discharge; a second step(S3) of converting into a surrogate data so as to have a non-linear characteristics by mixing the internal structure with maintaining the linear characteristics; a third step(S3) of reconstructing the attractor to geometrically analyzing the data by using a differential after the time delay and the buried dimension are determined by the surrogate data; a fourth step(S5) of calculating the distribution of the orthogonal axis by using the unique value in the attractor reconstruction and determining the status of the gas insulation switchgear among a normal, alarm and failure.
    公开号:KR20040070366A
    申请号:KR1020030006523
    申请日:2003-02-03
    公开日:2004-08-09
    发明作者:김철환；강진수
    申请人:학교법인 성균관대학；
    IPC主号:
    专利说明:

    Method for detecting partly spark of gas insulator switchgear by single value decomposition from chaos theory of attract}
    [10] The present invention relates to a method of detecting a partial discharge of a gas insulated switchgear, and in particular, it is possible to quickly detect a partial discharge in a gas insulated switchgear and determine a degree of failure, thereby enabling rapid response and emergency recovery, thereby improving reliability of power supply. A partial discharge detection method of a gas insulated switchgear using singular value decomposition in a chaotic theory attractor.
    [11] In general, the gas insulator switchgear (GIS) is used for substations, and it is a device that properly protects the system by opening and closing the line safely under abnormal conditions such as accidents and short circuits as well as the normal opening and closing under operating conditions. This gas insulated switchgear incorporates a switchgear and busbar such as a disconnector (DS), a ground switch (ES), and a breaker (CB) in a sealed steel container filled with SF 6 gas. As such, the gas insulation switchgear uses SF 6 gas, which has excellent insulation performance and arc-extinguishing function, as an insulating medium in a metal sealed container (tank) to accommodate conductors and various protective devices to improve reliability. As a result, the whole system is constructed by combining various devices and connecting tubes into units.
    [12] As such, the gas insulated switchgear is composed of many devices. Therefore, the signal to be measured contains a lot of noise and is nonlinear. However, the existing gas insulation switchgear state monitoring methods have a disadvantage in that partial discharge detection is difficult by approaching linear analysis of signals generated from the gas insulation switchgear.
    [13] Accordingly, an object of the present invention is to monitor the abnormal symptoms of the gas insulated switchgear on-line at all times and predict the situation to occur in advance in order to provide more economical maintenance and reliable power supply. The present invention provides a method of detecting a partial discharge of a gas insulated switchgear using a singular value decomposition method in a chaos theory attractor that can prevent fatal accidents caused by partial discharge of a gas.
    [1] 1 is a block diagram of chaos theory,
    [2] 2 is an attraction reconstruction using Takeens landfill,
    [3] 3 is a signal obtained by removing noise and normalizing the signal;
    [4] 4 shows surrogate data generation;
    [5] 5 is a time delay determination,
    [6] 6 is an attractor reconstruction of a signal measured in a real system gas insulated switchgear when it is normal,
    [7] 7 is an attractor reconstruction of a signal measured in a real system gas insulated switchgear in the case of a warning;
    [8] 8 is an attractor reconstruction of a signal measured in a real gas insulated switchgear in case of failure,
    [9] 9 is a partial discharge detection algorithm of the gas insulation switchgear using the singular value decomposition method.
    [14] Partial discharge detection method of the gas insulation switchgear using a singular value decomposition method in the chaos theory attractor of the present invention for achieving the object of the present invention comprises the steps of removing noise and normalizing the generated partial discharge; Converting the surrogate data into non-linear characteristics by mixing internal structures while maintaining the linear characteristics of the normalized signal; determining the time delay and the buried dimension based on the surrogate data; Reconstructing the attractor geometrically interpreting the time series data using the Takeens embedding theorem, which reconstructs the trajectory on the attractor using the difference of each time delay from the time series signal when there is a signal; and Calculate Orthogonal Axis Variance Using Singular Value Decomposition in Reconstruction And, is made to a specific value to determine the area of the second set comprises the step of determining the status of any of the gas-insulated switchgear apparatus in the normal, warning, and failure.
    [15] In this case, the first step uses wavelet transform. In addition, in the second step, the surrogate data is a data set generated by removing the deterministic properties of the data through phase noise or amplitude noise, and the surrogate data is phased at each frequency after Fourier transform of the measurement time series. Is the result of performing an Inverse Fourier Transform to make an irregular and return back to the time domain. In the third step, the correlation integral is used to determine the time delay. In the fourth step, the singular value may be set to any range within the range of 0.2 to 0.9.
    [16] Hereinafter, with reference to the accompanying drawings will be described a preferred embodiment of the present invention.
    [17] 1 is a block diagram of chaos theory. Referring to Figure 1, the signal measured in the field as shown in (a) can be divided into four types. Types include a signal that becomes constant over time, a periodic signal, a signal having two independent frequencies, and an irregular signal. In order to apply chaos theory to such signals, attractor reconstruction, which is a method of geometrically interpreting time series data, is essential. (b) is the Tarkens Landfill Theorem, which determines the appropriate time delay and landfill dimensions to reconstruct the time series signal into the attractor. The attractor reconstruction of each time series signal is shown as (c), and the four types of time series characteristics of (a) are each performing fixed point motion, periodic motion, quasi-periodic motion, and chaos motion due to the attraction reconstruction. It can be seen. The measured time series signal can be geometrically defined as the movement of the measured time series signal through the attraction reconstruction. To determine whether the attractor is a strange attractor representing chaos movement, the cross section of the Poang Curry in (d) is used. Because this Puan Curry cross section allows the attraction of complex structures to be viewed on a two-dimensional plane, as in (e), the structure determines its self-similarity as a fractal structure, and has a chaotic characteristic with a Liapunov index and a correlation dimension. It can be determined whether or not the signal having a. In general, the chaos signal has a fractal structure, and the Liafunop exponent is greater than zero and the correlation dimension has a real property.
    [18] 2 shows time delay in time series data It is shown that the attractor is reconstructed in the phase space by determining the and the buried dimension m and creating the data in the m-dimensional reconstruction state space by applying the Tokens landfill theorem. Attractor reconstruction of time series data uses the Tarkens Landfill Theorem. This way When you have a time series signal like From the time series data, the difference is reconstructed into the trajectory on the attractor using the difference of each time delay. Specifically, time series data From the time delay In the m-dimensional reconstructed state space, m-dimensional vectors as shown in [Equation 1] are created.
    [19]
    [20]
    [21]
    [22]
    [23] ㆍ ······ [Equation 1]
    [24] here, Is the time delay and m is the landfill dimension. Time series data Is represented in one dimension. Therefore, the application of the time delay and the embedding dimension is a way to correctly represent the phase of the measured system signal. Therefore, the m-dimensional vector connects these points to create a trajectory in phase space and to generate respective coordinates.
    [25] 3 is a signal obtained by removing noise and normalizing wavelets using the measured partial discharge. If the tractor is reconstructed using this signal, it becomes difficult to grasp the location and characteristics of the partial discharge. As such, the signal normalized using the wavelet has no internal structure, making it difficult to reassemble the attractor. Thus, Surrogate data is generated, which is well illustrated in FIG. 4.
    [26] FIG. 4 illustrates surrogate data through phase shifting of an internal structure while maintaining the linear characteristics of the signal in FIG. 3. Surrogate data is data created by breaking the internal causal relationship while maintaining the linear characteristics of the measured signal, ie, average, variance, distribution, frequency component, and autocorrelation. Surrogate data generation allows mixing of phases without spectral changes, and correcting the chaotic theory's attractor reconstruction with respect to time series data. Determination within the data through phase randomization or amplitude adjustment. Create a dataset with no theoretical properties to examine the significance of the time series. Significance is defined as in [Equation 2].
    [27] ㆍ ······ [Equation 2]
    [28] here, Is the nonlinear exponent value of the measurement time series Is the mean of the nonlinear exponential values for the replacement data, Represents the standard deviation of the nonlinear exponential values for the alternative dataset, respectively. If the nonlinear exponent value obtained from the measured time series exceeds some significance in comparison with the distribution of nonlinear indices obtained from the alternative data, the time series is internally nonlinear and deterministic rather than linear statistical time series. Surrogate data generation involves the Fourier transform of the measured time series, followed by an inverse Fourier transform to make the phase irregular at each frequency and back to the time domain.
    [29] 5 is a simulation result for determining the time delay. As shown in FIG. 5, C (r) is a coefficient value and is a method of determining, as a time delay of a given time series data, a point that is the first local minimum point in a value calculated according to a change in a time delay. Since the integral is used, the time delay is determined as 1.
    [30] FIG. 6 is an attractor reconstruction of a signal measured in a gas insulated switchgear in a normal state in which partial discharge does not occur (normal case). Referring to FIG. 6, in this normal case, the shape is densely formed around the center axis of the attractor at an angle of 45 ° on the phase plane, and the orthogonal axis dispersion of the attractor is very small. That is, it is a straight form.
    [31] 7 is an attractor reconstruction of a partial discharge having 5 partial discharges and a very large magnitude (in case of warning). As shown in FIG. 7, it can be seen that the orthogonal shaft dispersion of the attractor becomes larger as compared with the normal case.
    [32] 8 is a simulation result when the partial discharge is very severe (in case of failure). As shown in FIG. 8, the orthogonal axis variance orthogonal to the central axis is greatly changed in the attractor reconstruction, and this shows a spherical shape in which the orthogonal axis variance is increased as compared with the previous simulation. When the partial discharge is reconfigured, the orthogonal axis of the attractor changes as the number of occurrences and the magnitude of the partial discharge change the form of the attractor. It can be seen that this increases. Therefore, it is necessary to calculate the orthogonal axis variance in the attractor reconstruction of chaos theory, and the value of failure can be determined. That is, it is possible to determine whether a partial discharge has occurred.
    [33] 9 is a partial discharge detection algorithm of the gas insulation switchgear using the singular value decomposition method. Since the partial discharge measurement signal S1 measured in the field is difficult to distinguish due to many external noises, noise is removed and normalized with respect to the measured signal (S2). This signal is converted to surrogate data for attractor reconstruction (S3). This surrogate surrogate data is reconstructed with the Atkens theorem after determining the appropriate time delay and embedding dimensions (S4). The attractor increases the dispersion of the orthogonal axis along the central axis according to the size and number of partial discharges. The second singular value when the second singular value is obtained by using singular value decomposition for the dispersion of orthogonal axes in the attractor reconstruction. You can set the range. Calculated Second Singular Value Is less than or equal to 0.4 (Yes), the gas insulation switchgear is continuously monitored (S5), otherwise the second singular value Determine whether is less than or equal to 0.7 (S6) Is less than or equal to 0.7, a warning signal is sent (S7), otherwise it is determined as a failure and informs the driver (S8). Therefore, it is easy to determine the failure when calculating the orthogonal axis variance of the attractor using the singular value decomposition method. Such an algorithm is highly reliable because it sets a second singular value range by simulating a lot of real system data.
    [34] Table 1 shows the simulation results of the singular value decomposition method.
    [35]
    [36]
    [37] Table 1 shows the simulation results of the gas insulation switchgear state monitoring method using the singular value decomposition method described in FIG. 9 using various real system data. In the case where the partial discharge is not normal, the second singular value is small, and when the partial discharge is large and the size is large, the second singular value is large. However, since the second singular value has information on the number and size of partial discharges at the same time, the second singular value is relatively large when the magnitude of the second discharge is large. In addition, since the normalized signal is used, a failure is determined by setting a second singular value range.
    [38] Here, a brief description will be given of the singular value decomposition method in the attraction of chaos theory.
    [39] It is a method of calculating the orthogonal axis variance in the attraction of chaos theory through singular value decomposition. The reason for using singular value decomposition is that the measured data is not square. Measured Partial Discharge Signal By using singular value decomposition as in [Equation 3] Can be represented.
    [40] ㆍ ······ [Equation 3]
    [41] here, , Is an orthogonal matrix, Is a diagonal matrix. Diagonal matrix Can be divided into two groups as shown in [Equation 4].
    [42] ㆍ ······ [Equation 4]
    [43] The diagonal value of is the singular value, and this singular value contains the information about the system phase and the bad condition. Diagonal matrix Eigenvalues of on Is the singular value, and the first singular value Represents the central axis of the attractor, and the second singular value Represents the orthogonal axis of the attractor.
    [44] In the present invention, partial discharge data is acquired using a LEMKE device in a 154 GHz GIS, and noise cancellation using a wavelet and a normalized signal are used. The lack of non-linear characteristics for attractor reconstruction of this signal produces surrogate data that undergoes phase-out. For the surrogate data, the time delay is set to 1 and the embedding dimension is set to 2, and the attractor reconstruction is performed using the Tarkens Landfill Theorem, and the second value is obtained by using the singular value decomposition method for the orthogonal axis dispersion of the attractor. Outlier Set the range to monitor the condition of the gas insulation switchgear.
    [45] In the verification of the present invention, the data acquired using the LEMKE equipment is stored in the form of ASCII code in the magnitude of the charge amount over time, and the stored data is analyzed using Matlab.
    [46] As described above, the partial discharge detection method of the gas insulated switchgear using singular value decomposition in the chaos theory attractor according to the present invention has the advantage that the state of the gas insulated switchgear can be continuously and uninterrupted compared to the conventional method. In addition, by using an algorithm using a singular value decomposition method, it is possible to geometrically analyze a partial discharge signal having a nonlinear characteristic and to monitor the state of the gas insulated switchgear by accurate detection of the partial discharge. In particular, the application of an algorithm that is easy to implement hardware is expected to be effective in terms of practical value. The present invention is not limited to the above-described embodiment, and it will be apparent that many modifications are possible by those skilled in the art within the technical spirit of the present invention.
    权利要求:
    Claims (7)
    [1" claim-type="Currently amended] A first step of removing noise and normalizing the generated partial discharge;
    A second step of mixing internal structures to surrogate data to have nonlinear characteristics while maintaining the linear characteristics of the normalized signal;
    After the time delay and the embedding dimension are determined based on the surrogate data, time series data is obtained by using the Tarkens Landfill Theorem which reconstructs the time series signal from the time series signal to the trajectory on the attractor using the difference of each time delay from the time series signal. A third step of reconstructing the attractor geometrically interpreted; And
    A fourth step of calculating the variance of the orthogonal shaft using singular value decomposition in the attractor reconstruction and determining the state of any one of the gas insulation switchgear among normal, warning, and fault by determining the set second singular value region;
    Partial discharge detection method of the gas insulated switchgear using a singular value decomposition method in the attraction of chaos theory characterized in that it comprises a.
    [2" claim-type="Currently amended] The method of claim 1, wherein the first step comprises using a wavelet transform.
    [3" claim-type="Currently amended] The method of claim 1, wherein the surrogate data in the second step is to generate a data set from which deterministic properties of the data are removed through phase noise or amplitude noise. Partial discharge detection method of gas insulated switchgear used.
    [4" claim-type="Currently amended] 4. The method of claim 1 or 3, wherein the surrogate data is a result of performing Fourier transform of the measurement time series, irregular phase at each frequency, and inverse Fourier transform to return to the time domain. Partial discharge detection method of gas insulated switchgear using singular value decomposition method in attractor of chaos theory.
    [5" claim-type="Currently amended] The method of claim 1, wherein a correlation integral is used to determine the time delay in the third step.
    [6" claim-type="Currently amended] The method of claim 1, wherein the singular value in the fourth step is 0.2 Outlier It is set in the range of 0.9, the singular value is less than 0.2 is normal, the warning is when the singular value is 0.2 to 0.9, the singular value is greater than 0.9 is characterized in that the attraction of the attractor of the chaos theory Partial discharge detection method of gas insulated switchgear using value decomposition method.
    [7" claim-type="Currently amended] The method of claim 1, wherein the singular value in the fourth step is 0.4 Outlier It is set in the range of 0.7, and if the singular value is less than 0.4, it is normal, the warning is when the singular value is between 0.4 and 0.7, and when the singular value is more than 0.7, it is determined as a failure in the attraction of chaos theory Partial discharge detection method of gas insulated switchgear using value decomposition method.
    类似技术:
    公开号 | 公开日 | 专利标题
    Li et al.2014|Fault detection and classification in medium voltage DC shipboard power systems with wavelets and artificial neural networks
    Zhou et al.2005|An improved methodology for application of wavelet transform to partial discharge measurement denoising
    Copi et al.2004|Multipole vectors: A new representation of the CMB sky and evidence for statistical anisotropy or non-Gaussianity at 2⩽ l⩽ 8
    Han et al.2005|Wavelet packet based damage identification of beam structures
    Butler-Purry et al.2003|Characterization of transients in transformers using discrete wavelet transforms
    Dubey et al.2013|Wavelet singular entropy-based symmetrical fault-detection and out-of-step protection during power swing
    Johns et al.1976|Digital simulation of faulted ehv transmission lines with particular reference to very-high-speed protection
    US5600248A|1997-02-04|Fault distance locator for underground cable circuits
    Ghaderi et al.2017|High impedance fault detection: A review
    Reddy et al.2014|Smart fault location for smart grid operation using RTUs and computational intelligence techniques
    EP1714366B1|2017-04-05|High impedance fault detection
    Lee et al.2003|New fault diagnosis of circuit breakers
    Gilany et al.2006|Traveling-wave-based fault-location scheme for multiend-aged underground cable system
    James et al.1995|Development of computer-based measurements and their application to PD pattern analysis
    Chaari et al.1995|Prony's method: an efficient tool for the analysis of earth fault currents in Petersen-coil-protected networks
    US6483316B2|2002-11-19|Method of diagnosing partial discharge in gas-insulated apparatus and partial discharge diagnosing system for carrying out the same
    Song et al.2007|Second generation wavelet transform for data denoising in PD measurement
    Borghetti et al.2006|On the use of continuous-wavelet transform for fault location in distribution power systems
    Chung et al.2005|Application of phase detection frequency domain reflectometry for locating faults in an F-18 flight control harness
    Wu et al.2003|Ultra-high-speed directional protection of transmission lines using mathematical morphology
    CN102103183B|2014-11-05|Partial discharge measurement device under impulse voltage on site and signal processing method thereof
    Pandey et al.1998|Multiresolution signal decomposition: A new tool for fault detection in power transformers during impulse tests
    Paś et al.2013|Determination of the impact indicators of electromagnetic interferences on computer information systems
    Taniguchi et al.2009|Observation results of lightning shielding for large-scale transmission lines
    Patsch et al.2001|Pulse Sequence Analysis-a diagnostic tool based on the physics behind partial discharges
    同族专利:
    公开号 | 公开日
    KR100449285B1|2004-09-22|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    2003-02-03|Application filed by 학교법인 성균관대학
    2003-02-03|Priority to KR10-2003-0006523A
    2004-08-09|Publication of KR20040070366A
    2004-09-22|Application granted
    2004-09-22|Publication of KR100449285B1
    优先权:
    申请号 | 申请日 | 专利标题
    KR10-2003-0006523A|KR100449285B1|2003-02-03|2003-02-03|Method for detecting partly spark of gas insulator switchgear by single value decomposition from chaos theory of attract|
    [返回顶部]