CN108008244A - A kind of small current grounding fault progressive classifying identification method at many levels - Google Patents
A kind of small current grounding fault progressive classifying identification method at many levels Download PDFInfo
- Publication number
- CN108008244A CN108008244A CN201711006521.8A CN201711006521A CN108008244A CN 108008244 A CN108008244 A CN 108008244A CN 201711006521 A CN201711006521 A CN 201711006521A CN 108008244 A CN108008244 A CN 108008244A
- Authority
- CN
- China
- Prior art keywords
- fault
- faults
- zero
- linear
- resistance
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
- Locating Faults (AREA)
Abstract
A kind of small current grounding fault progressive classifying identification method at many levels, there is provided including fault signature point is pressed time domain specification, property of trouble point etc., top-down four layers of progressive failure modes data module, it is imperfect to efficiently solve small current grounding fault classification, the problems such as identification types are single, 5 steps of recognition methods are proposed accordingly, residual voltage is extracted in different levels, the fault waveform feature such as zero-sequence current, recognized using temporal signatures, heuristic segmentation algorithm, fft analysis etc. successively identifies fault type, it is final to draw the fault type for representing fault signature comprehensively.The method can effectively avoid the problem that concept obfuscation during fault identification, on the one hand profound can grasp failure cause and its regularity of distribution, promote power distribution network fortune to check horizontal persistently perfect of management lean；On the other hand operation maintenance personnel can be aided in find power distribution network " latency hidden danger " in time, further shortens trouble shooting time, promote the continuous lifting of distribution operational reliability.
Description
Technical Field
The invention belongs to the field of power distribution network fault diagnosis, and particularly relates to a method for identifying the type of a ground fault of a power distribution network in a low-current grounding mode.
Background
The fault recording data of the power distribution terminal, the transient recording type fault indicator and the like record the electric quantity change information on the lines before and after the fault, can truly reflect the fault condition of the power grid, and becomes an important research and application direction by utilizing the fault recording data to accurately and effectively diagnose the fault. The fault type identification is used as a precondition and a theoretical basis of a fault diagnosis technology, and has great significance for deeply mastering fault reasons and distribution rules thereof, timely finding potential hidden dangers of the power distribution network, further shortening fault finding time and promoting the improvement of the operation reliability of the power distribution network and the continuous improvement of the lean level of operation and maintenance management.
In recent years, attention has been paid to type identification of ground faults of low-current grounding systems, and relevant research has played a positive role in fault type identification. The identification method can be classified into a feature analysis method (identifying intermittent faults according to the distribution features of zero-sequence voltage amplitude values on a multi-fractal spectrum, identifying arc faults according to nonlinear features of transition resistance, identifying high-resistance faults according to wavelet energy moment features of high-frequency components, etc.) and an intelligent method (artificial neural network, fuzzy logic, decision tree, etc.) according to different application ways of extracting feature quantities. The existing classification method for the small current grounding fault is classified according to certain characteristics of the fault, such as instantaneous grounding and permanent grounding according to whether the fault needs manual treatment or not; dividing the grounding into low-resistance grounding and high-resistance grounding according to the size of the transition resistor; the method is divided into stable grounding, unstable grounding and the like according to the arcing condition of the grounding point. The fault classification method can cause the fault classification to be incomplete, and the fault types have characteristic cross, so that the problems of single identification type, fuzzy concept and the like exist in the fault identification process. Most of the existing identification methods adopt simulation data for verification, strong support of field data is lacked, and effectiveness and feasibility of an algorithm are reduced.
Disclosure of Invention
In order to solve the problems of single identification type, fuzziness and the like in the existing fault identification technology, the invention provides a multi-level progressive classification identification method for small-current ground faults.
The multilevel progressive type classification and identification method for the small current ground fault comprises the steps of setting four layers of progressive type fault classification data modules according to the characteristics of the small current ground fault and providing basis for further establishing a fault identification model, specifically, the method comprises a 1 st module layer 1, wherein the 1 st module layer 1 divides the fault into a permanent fault, an instantaneous fault and an intermittent fault according to the time domain characteristics of the fault, and the permanent fault refers to a ground state which is formed after single-phase grounding and continuously exists until manual processing; the transient fault refers to a grounding state that the fault disappears after lasting for a period of time and the system automatically recovers to be normal; the intermittent grounding fault refers to a grounding state that the fault disappears after lasting for a period of time and the fault occurs again in the system recovery process; the 2 nd module layer 2, the 2 nd module layer 2 divides the fault into a single fault and a developmental fault according to the complexity of the fault, the single fault refers to a grounding state in which the fault property is kept unchanged in the fault process, and the developmental fault refers to a grounding state in which the fault property is changed in the fault process; the 3 rd module layer 3, the 3 rd module layer 3 divides the fault into a linear fault and a nonlinear fault according to the property of the grounding point, the linear fault indicates that the grounding point has no arcing condition and the transition resistance is in a linear grounding state, and the nonlinear fault indicates that the grounding point has the arcing condition; the 4 th module layer 4, the 4 th module layer 4 divides the linear faults into metallic faults, low-resistance faults and high-resistance faults according to the size of the transition resistance; dividing nonlinear faults into arc light low resistance faults and arc light high resistance faults, wherein the high resistance faults refer to linear faults with the transition resistance larger than 1k omega, the low resistance faults refer to linear faults with the transition resistance smaller than 1k omega, the metallic faults refer to linear faults with the transition resistance of 0 or close to 0, the arc light high resistance faults refer to nonlinear faults with the minimum value of the transition resistance larger than 1k omega, and the arc light low resistance faults refer to nonlinear faults with the minimum value of the transition resistance smaller than 1k omega;
according to the progressive fault classification data module, extracting features of different levels in a fault waveform to identify faults according to the following multi-level progressive identification method of the low-current ground fault type, and specifically comprising the following steps of:
step 1: collecting bus zero-sequence voltage and zero-sequence current signals of outgoing lines;
step 2: identifying permanent faults, transient faults and intermittent faults by combining manual processing and zero sequence voltage attenuation degrees;
and step 3: processing the zero-sequence current by adopting a heuristic segmentation algorithm, identifying single faults and developmental faults, and segmenting fault waveforms;
and 4, step 4: performing FFT analysis on the zero sequence current, and identifying linear faults and nonlinear faults by comparing amplitude-frequency characteristics of the zero sequence current before and after the faults;
and 5: identifying high-resistance faults and low-resistance faults according to the magnitude of the zero-sequence voltage;
and 6: and combining the recognition results of different levels to obtain the final fault type.
The specific steps of the step 2 for identifying the permanent fault, the transient fault and the intermittent fault by combining manual processing and zero sequence voltage attenuation degree are as follows:
a. detecting the amplitude of the zero-sequence voltage of each period, and calculating the attenuation degree a of the zero-sequence voltage waveform of each period from left to right; let attenuation a of zero sequence voltage of T-th cycle T The ratio of the zero sequence voltage amplitude of the T-th period to the zero sequence voltage amplitude of the T-1 th period is as follows:
b. definition counter M 1 、M 2 To identify the progress and completion of the recovery process of the system.
Counter M 1 Initial value of 0, when U 0 >U op.set And alpha is&0.85, the counter M is arranged at intervals of one cycle 1 Adding 1, otherwise M 1 Zero clearing, wherein, U 0 Is zero sequence voltage amplitude, U op.set Alpha is the attenuation degree for the zero sequence voltage starting value of the protection device; then when M is 1 &2, the system is considered to be in a recovery process;
counter M 2 Initial value of 0, when U 0 <U cl.set At intervals of one cycle, a counter M 2 Adding 1, otherwise M 2 Zero clearing, wherein, U cl.set And returning the zero sequence voltage of the protection device. Then when M is 2 &2, the system recovery process is considered to be finished;
c. defining a flag F, setting the initial value to be 0, if the fault is manually processed, assigning the value to be 1, and if not, keeping the value unchanged;
if F =0 and α in the process<, 1, final M 2 &2, judging the fault as a transient fault;
if F =0 in the process, M is present 1 >, 2 and alpha>, 1, final M 2 &Judging the fault as an intermittent fault;
if F =0 and M in the process 2 < 2, final F =1, judgedThis failure is a permanent failure.
The step 3 of processing the zero sequence current by adopting a heuristic segmentation algorithm, identifying the single fault and the developmental fault, and segmenting the fault waveform comprises the following specific steps:
a. calculating the number N of points of the left and right partial sequences of any point i in the sampling sequence X (t) l 、N r Mean value μ l (i)、μ r (i) And standard deviation s l (i)、s r (i) Then, the combined deviation s of the sequences at both sides of the point i can be obtained D (i) Comprises the following steps:
b. and calculating a T hypothesis test statistic value T (i) of the significance difference of the mean values on the left side and the right side of the point i as follows:
repeating the calculation process for each point in X (T) from left to right to obtain a test statistic value sequence T (T), wherein T is assumed to be a mathematical proper noun;
c. calculating the maximum value T of T (T) max Statistical significance of P max ，P max Can be generally expressed as:
in the formula, the content of the active carbon is shown in the specification,in the case of an incomplete beta function,for the integral upper bound of the incomplete beta function, δ v and δ are two parameters of the incomplete beta function, v is the degree of freedom of the t-test, where the rootThe Monte Carlo simulation shows that gamma =4.19lnN-11.54, delta =0.40 and v = N-2,N are the total point number of the sampling sequence;
d. setting a threshold value P 0 =0.95, maximum value T among T (T) max Statistical significance of P max >P 0 When the fault is detected, the X (T) mutation is obvious, and the corresponding wave recording data has a trend mutation, namely the fault is a developmental fault, wherein T is max The corresponding point i is a catastrophe point, and the waveform is divided into two parts from the point i.
The step 4 of performing FFT analysis on the zero sequence current, and identifying a linear fault and a nonlinear fault by comparing the amplitude-frequency characteristics of the zero sequence current before and after the fault specifically includes the steps of:
a. performing FFT analysis on the zero sequence current waveform obtained in the step 3;
b. taking the previous period of the fault as a reference T 0 Each period after the fault is T in sequence 1 、T 2 …T n Calculating T i (i =1,2, …, n) each frequency amplitude as a percentage of the fundamental frequency;
c. calculating T i Relative to T 0 The percentage change quantity of each frequency amplitude is sorted from large to small to obtain T i Mean value mu of amplitude percentage change of the first 5 frequency points i ；
d. Defining the initial value of the counter N as 0 when mu i &When the content is 10 percent, adding 1 to N; otherwise, N is cleared;
e. when M is 1 &2, the system is not in the recovery process; if N is present at the same time&Judging that the fault is a linear fault if the high-frequency component of the waveform does not change obviously, and judging that the fault is a nonlinear fault if the high-frequency component of the waveform does not change obviously;
in the step 4, by using the characteristic of complete classification in the low-current ground fault multi-level classification method, firstly, linear fault identification is performed to ensure that transient non-linear ground arcs are not ignored.
The specific steps of identifying the high resistance fault and the low resistance fault through the magnitude of the zero sequence voltage in the step 5 are as follows:
a. according to the fault-leading phase voltage U m And zero sequence current 3I of fault outgoing line in metallic fault 0 Calculating the zero sequence impedance Z of the system S0 ：
b. Judging the size of the transition resistance according to the zero sequence voltage at the bus:
the beneficial effects of the invention are as follows:
the invention provides a multilevel progressive classification method for small-current ground faults, which effectively solves the problems of incomplete classification, single identification type and the like. According to the method, the fault type is identified layer by extracting fault waveform characteristics such as zero sequence voltage, zero sequence current and the like. Compared with various fault identification methods based on single characteristics, the method provided by the invention can effectively avoid the problem of concept ambiguity in the fault identification process, and is clear in structure, easy to program and realize in a modularized mode. The method provided by the invention can effectively avoid the problem of concept ambiguity in the fault identification process, has a clear structure, is easy to realize in a programmed manner, can be used as a theoretical basis and an implementation guide of a fault type identification method based on recording data, and has very important significance: on one hand, the method can support the establishment of a power distribution network fault information system and a characteristic fingerprint database, deeply master fault reasons and distribution rules thereof and promote the continuous improvement of the lean level of operation inspection management of the power distribution network; on the other hand, operation and maintenance personnel can be assisted to find potential hidden dangers of the power distribution network in time, fault finding time is further shortened, and the running reliability of the power distribution network is promoted to be continuously improved.
Drawings
FIG. 1 is a multi-level progressive classification method for small current ground faults;
FIG. 2 is a flow chart of a multi-level progressive identification method for a low-current fault type;
FIG. 3 is a graph of intermittent low current ground fault recording;
FIG. 4 is a chart of nonlinear fault progression to linear fault oscillography;
FIG. 5 is a chart of the progression of a high resistance fault to a low resistance fault;
FIG. 6 is a zero sequence current T test sequence for nonlinear fault development into linear fault recording;
FIG. 7 is a zero sequence current T test sequence for high resistance fault development to low resistance fault recording;
FIG. 8 is an amplitude-frequency characteristic of a nonlinear fault developing into a linear fault;
Detailed Description
In order to achieve the above purpose, the technical solutions adopted by the present invention are as follows, referring to fig. 1 and 2:
the fault characteristics are decomposed into 4 different factors, and the 4 different factors are analyzed on different levels to form a multi-level progressive fault classification and identification method, so that a basis is provided for further establishing a fault identification model.
The multi-level progressive classified identification method for the small-current ground fault comprises the steps of setting four layers of progressive fault classified data modules according to the characteristics of the small-current ground fault, providing basis for further establishing a fault identification model, and specifically comprises a 1 st module layer 1, wherein the 1 st module layer 1 divides the fault into a permanent fault, an instantaneous fault and an intermittent fault according to the time domain characteristics of the fault, and the permanent fault refers to a grounding state which is formed after the single-phase grounding and continuously exists until manual processing; the transient fault refers to a grounding state that the fault disappears after lasting for a period of time and the system automatically recovers to be normal; the intermittent grounding fault refers to a grounding state that the fault disappears after the fault lasts for a period of time and the fault occurs again in the system recovery process; the 2 nd module layer 2, the 2 nd module layer 2 divides the fault into a single fault and a developmental fault according to the complexity of the fault, the single fault refers to a grounding state in which the fault property is kept unchanged in the fault process, and the developmental fault refers to a grounding state in which the fault property is changed in the fault process; the 3 rd module layer 3, the 3 rd module layer 3 divides the fault into a linear fault and a nonlinear fault according to the property of the grounding point, the linear fault indicates that the grounding point has no arcing condition and the transition resistance is in a linear grounding state, and the nonlinear fault indicates that the grounding point has an arcing condition; the 4 th module layer 4, the 4 th module layer 4 divides the linear faults into metallic faults, low-resistance faults and high-resistance faults according to the size of the transition resistance; dividing nonlinear faults into arc light low resistance faults and arc light high resistance faults, wherein the high resistance faults refer to linear faults with the transition resistance larger than 1k omega, the low resistance faults refer to linear faults with the transition resistance smaller than 1k omega, the metallic faults refer to linear faults with the transition resistance of 0 or close to 0, the arc light high resistance faults refer to nonlinear faults with the minimum value of the transition resistance larger than 1k omega, and the arc light low resistance faults refer to nonlinear faults with the minimum value of the transition resistance smaller than 1k omega;
according to the progressive fault classification data module structure, the method for identifying the fault by extracting the characteristics of different levels in the fault waveform according to the following multi-level progressive identification method of the low-current ground fault type specifically comprises the following steps:
step 1: collecting bus zero-sequence voltage and zero-sequence current signals of outgoing lines;
and 2, step: identifying permanent faults, transient faults and intermittent faults by combining manual processing and zero sequence voltage attenuation degrees;
the specific process of the step 2 is as follows:
a. detecting the amplitude of the zero-sequence voltage of each period, and calculating the attenuation degree alpha of the zero-sequence voltage waveform of each period from left to right T (ii) a Setting the attenuation degree alpha of the zero sequence voltage of the Tth period T The ratio of the zero sequence voltage amplitude of the T-th period to the zero sequence voltage amplitude of the T-1 th period is as follows:
b. definition counter M 1 、M 2 To identify the progress of the recovery process of the systemAnd finishing.
Counter M 1 Initial value of 0, when U 0 >U op.set And alpha is&0.85, the counter M is arranged at intervals of one period 1 Adding 1, otherwise M 1 Zero clearing, wherein, U 0 Is zero sequence voltage amplitude, U op.set Alpha is the attenuation degree for the zero sequence voltage starting value of the protection device; then when M is 1 &2, the system is considered to be in a recovery process;
counter M 2 Initial value of 0, when U 0 <U cl.set At intervals of one cycle, a counter M 2 Adding 1, otherwise M 2 Zero clearing, wherein, U cl.set And returning the zero sequence voltage of the protection device. Then when M is 2 &2, the system recovery process is considered to be finished;
c. and defining a flag F, setting the initial value to be 0, assigning the value to be 1 if the fault is manually processed, and otherwise, keeping the value unchanged.
If F =0 in the process, and α<, 1, final M 2 &2, judging the fault as a transient fault;
if F =0 in the process, M is present 1 >, 2 and alpha>, 1, final M 2 &2, judging the fault as an intermittent fault;
if F =0 and M is in process 2 And (5) less than 2, finally F =1, and judging the fault as a permanent fault.
And step 3: processing the zero-sequence current by adopting a heuristic segmentation algorithm (BG algorithm), identifying single faults and developmental faults, and segmenting fault waveforms;
the specific process of the step 3 is as follows:
a. calculating the number N of points of the left and right partial sequences of any point i in the sampling sequence X (t) l 、N r Mean value μ l (i)、μ r (i) And standard deviation s l (i)、s r (i) Then, the combined deviation s of the sequences at both sides of the point i can be obtained D (i) Comprises the following steps:
b. and calculating a T hypothesis test statistic value T (i) of the significance difference of the mean values on the left side and the right side of the point i as follows:
repeating the calculation process for each point in X (T) from left to right to obtain an inspection statistic value sequence T (T);
c. calculating the maximum value T of T (T) max Statistical significance of P max ，P max Can be generally expressed as:
in the formula, the content of the active carbon is shown in the specification,in the case of an incomplete beta function,for the upper integral limit of the incomplete β function, δ v and δ are two parameters of the incomplete β function, and v is the degree of freedom of the t-test. Wherein gamma =4.19lnN-11.54, delta =0.40 and v = N-2,N are obtained as the total point number of the sampling sequence according to Monte Carlo simulation.
d. Setting a threshold value P 0 =0.95, maximum value T among T (T) max Statistical significance of P max >P 0 When the fault is detected, the X (T) mutation is obvious, and the corresponding wave recording data has a trend mutation, namely the fault is a developmental fault, wherein T is max The corresponding point i is a mutation point, and the waveform is divided into two parts from the point i.
And 4, step 4: performing FFT analysis on the zero sequence current, and identifying linear faults and nonlinear faults by comparing amplitude-frequency characteristics of the zero sequence current before and after the faults;
the specific process of the step 4 is as follows:
a. carrying out FFT analysis on the zero sequence current waveform obtained in the step 3;
b. taking the previous period of the fault as a reference T 0 Each period after the fault is T in sequence 1 、T 2 …T n Calculating T i (i =1,2, …, n) each frequency amplitude as a percentage of the fundamental frequency;
c. calculating T i Relative to T 0 The percentage change quantity of each frequency amplitude is sorted from large to small to obtain T i Mean value mu of percentage change in amplitude of the first 5 frequency points i ；
d. Defining the initial value of the counter N as 0 when mu i &When the content is 10 percent, adding 1 to N; otherwise, N is reset;
e. when M is 1 &2, namely the system is not in a recovery process; if N is present at the same time&Judging that the fault is a linear fault if the high-frequency component of the waveform does not change obviously, and judging that the fault is a nonlinear fault if the high-frequency component of the waveform does not change obviously;
in the step 4, by using the characteristic of complete classification in the low-current ground fault multi-level classification method, firstly, linear fault identification is performed to ensure that transient non-linear ground arcs are not ignored.
And 5: identifying high-resistance faults and low-resistance faults according to the magnitude of the zero-sequence voltage;
the specific process of the step 5 is as follows:
a. according to the fault-leading phase voltage U m And zero sequence current 3I of fault outlet line in metallic fault 0 Calculating the zero sequence impedance Z of the system S0 ：
b. Judging the size of the transition resistance according to the zero sequence voltage at the bus:
step 6: obtaining a final fault type by combining the recognition results of different levels;
the method of the present invention is described in further detail below with reference to the figures and several exemplary fault type recording data.
Taking the intermittent ground fault shown in fig. 3 as an example, the attenuation threshold value α is set in step 2 of the fault identification process set The description is as follows:
the distribution network voltage grade is lower, and the ground point is unstable, and the trouble disappears the back, and the system has three phase voltage's recovery process, especially in resonant grounding system, because arc suppression coil's effect, voltage recovery probably needs tens of power frequency cycle. Therefore, the recovery process of the system can be identified through the attenuation process of the zero sequence voltage, and the intermittent fault can be identified by judging whether the fault occurs again in the system recovery process. In the recovery process, the zero sequence voltage u 0 Is a damping component of free oscillation, which can be expressed as
In the formula, u 0 Is zero sequence voltage in the recovery process; u shape pm Is the magnitude of the phase voltage; alpha is an attenuation coefficient, and the magnitude of alpha is related to the damping rate of the power grid; e is a natural constant; omega 0 Is the free oscillation angular frequency;the initial phase angle of the voltage and current for the system recovery process. Therefore, the zero sequence voltage amplitude is attenuated to 1/e of the original value every period 2πα . The duration of the system recovery process is considered to be 20 power frequency cycles at most, when the system is attenuated to be less than 5% of the amplitude during fault, the system is considered to be recovered to be normal, and the system is attenuated to be at least 85% of the previous cycle per cycle. So when alpha is&And lt, 0.85, the system is considered to be in a recovery process.
The BG algorithm described in step 3 is described by taking two different types of developmental ground faults shown in fig. 4 and 5 as examples:
the BG algorithm processing is respectively carried out on the zero sequence currents of two different types of faults to obtain a check statistic value sequence T (T), as shown in fig. 6 and 7, the fact that the T check sequence obtains the maximum value at a mutation point can be seen, and the BG algorithm can effectively identify the trend change situation in fault signals.
Taking fig. 4 as an example, the FFT analysis in step 4 is explained:
the fault recording diagram shown in fig. 4 has three stages: before fault, non-linear fault and linear fault respectively corresponding to T 0 、T 1 、T 2 . FFT analysis is respectively carried out on the two signals, the amplitude-frequency characteristic is shown in figure 8, and T is calculated 1 、T 2 Relative to T 0 The average value of the percentage change of the amplitude is mu 10 ＝95.72％、μ 20 =9.663%. Therefore, the linear fault and the nonlinear fault can be effectively identified by detecting the change of the frequency component of the zero sequence current.
The transition resistance measurement in the step 5 is related to the capacitance current of the system, the accuracy of the field data cannot be verified, 170 groups of field fault recording are selected to test the fault identification method in the steps 2-4 of the invention, and the test results are shown in the table 1:
TABLE 1 test results of the identification method of the type of the low-current ground fault
The comprehensive test accuracy rate is more than 92%, and the fault identification method in the step 2-4 is proved to be effective.
And (3) testing the effectiveness of the high-resistance fault identification method in the step (5) by adopting an artificial grounding test, wherein the test data records are shown in table 1:
table 1 artificial grounding test data
The zero sequence voltage values measured by the test are smaller than theoretical calculated values, and the possible reasons are that complete metallic grounding is difficult to achieve during artificial grounding, the fault phase voltage is not 0, the measured capacitance current is smaller, the calculated value of the zero sequence impedance of the system is larger, and the finally obtained zero sequence voltage is larger. Within the allowable error range, the test results under the two grounding modes are consistent with the theoretical calculation results, and the zero sequence voltage high resistance identification method in the step 5 is proved to be effective.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (5)
1. A multi-level progressive classification and identification method for small current ground faults is characterized by comprising the following steps: the method comprises the steps that four layers of progressive fault classification data modules are arranged according to the characteristics of the low-current ground fault to provide basis for further establishing a fault identification model, and specifically comprises a 1 st module layer (1), wherein the 1 st module layer (1) divides the fault into a permanent fault, a transient fault and an intermittent fault according to the time domain characteristics of the fault, and the permanent fault refers to a grounding state in which the fault continuously exists after single-phase grounding till manual processing; the transient fault refers to a grounding state that the fault disappears after lasting for a period of time and the system automatically recovers to be normal; the intermittent grounding fault refers to a grounding state that the fault disappears after lasting for a period of time and the fault occurs again in the system recovery process; the 2 nd module layer (2), the 2 nd module layer (2) divides the fault into a single fault and a developed fault according to the complexity of the fault, the single fault refers to a grounding state that the fault property is kept unchanged in the fault process, and the developed fault refers to a grounding state that the fault property is changed in the fault process; the 3 rd module layer (3), the 3 rd module layer (3) divides the fault into a linear fault and a nonlinear fault according to the property of the grounding point, the linear fault indicates that the grounding point has no arcing condition and the transition resistance is in a linear grounding state, and the nonlinear fault indicates that the grounding point has an arcing condition; the 4 th module layer 4, the 4 th module layer (4) divides the linear faults into metallic faults, low-resistance faults and high-resistance faults according to the size of the transition resistance; dividing nonlinear faults into arc light low resistance faults and arc light high resistance faults, wherein the high resistance faults refer to linear faults with the transition resistance larger than 1k omega, the low resistance faults refer to linear faults with the transition resistance smaller than 1k omega, the metallic faults refer to linear faults with the transition resistance of 0 or close to 0, the arc light high resistance faults refer to nonlinear faults with the minimum value of the transition resistance larger than 1k omega, and the arc light low resistance faults refer to nonlinear faults with the minimum value of the transition resistance smaller than 1k omega;
according to the progressive fault classification data module, extracting features of different levels in a fault waveform to identify faults according to the following multi-level progressive identification method of the low-current ground fault type, and specifically comprising the following steps of:
step 1: collecting bus zero-sequence voltage and zero-sequence current signals of outgoing lines;
step 2: identifying permanent faults, transient faults and intermittent faults by combining manual processing and zero sequence voltage attenuation degrees;
and step 3: processing the zero-sequence current by adopting a heuristic segmentation algorithm, identifying a single fault and a developing fault, and segmenting a fault waveform;
and 4, step 4: performing FFT analysis on the zero sequence current, and identifying a linear fault and a nonlinear fault by comparing amplitude-frequency characteristics of the zero sequence current before and after the fault;
and 5: identifying high-resistance faults and low-resistance faults according to the magnitude of the zero-sequence voltage;
step 6: and combining the recognition results of different levels to obtain the final fault type.
2. The multi-level progressive classification and identification method for the small-current ground fault as claimed in claim 1, wherein: the step 2 specifically comprises the following steps:
a. detecting the amplitude of the zero-sequence voltage of each period, and calculating the attenuation degree alpha of the zero-sequence voltage waveform of each period from left to right; let attenuation degree alpha of zero sequence voltage of T-th period T Is the ratio of the zero sequence voltage amplitude of the T-th cycle to the zero sequence voltage amplitude of the T-1 cycle, namely:
b. definition counter M 1 、M 2 To identify the progress and completion of the recovery process of the system;
counter M 1 Initial value of 0, when U 0 >U op.set And alpha is&0.85, the counter M is arranged at intervals of one period 1 Adding 1, otherwise M 1 Zero clearing, wherein, U 0 Is zero sequence voltage amplitude, U op.set Alpha is the attenuation degree for protecting the zero sequence voltage starting value of the device. Then when M is 1 &2, the system is considered to be in a recovery process;
counter M 2 Initial value of 0, when U 0 <U cl.set At intervals of one cycle, a counter M 2 Adding 1, otherwise M 2 Zero clearing, wherein, U cl.set And returning the zero sequence voltage of the protection device. Then when M is 2 &2, the system recovery process is considered to be finished;
c. defining a flag F, setting the initial value to be 0, if the fault is manually processed, assigning the value to be 1, and if not, keeping the value unchanged;
if F =0 in the process, and α<, 1, final M 2 &2, judging the fault as a transient fault;
if F =0 in the process, M is present 1 >, 2 and alpha>, 1, final M 2 &Judging the fault as an intermittent fault;
if F =0 in the process, and M 2 And (5) less than 2, finally F =1, and judging the fault as a permanent fault.
3. The multi-level progressive classification and identification method for the small-current ground fault as claimed in claim 1, wherein step 3 specifically comprises:
a. calculating the number N of points of the left and right partial sequences of any point i in the sampling sequence X (t) l 、N r Mean value μ l (i)、μ r (i) And standard deviation s l (i)、s r (i) Then, the combined deviation s of the sequences at both sides of the point i can be obtained D (i) Comprises the following steps:
b. and calculating a T hypothesis test statistic value T (i) of the significance difference of the mean values on the left side and the right side of the point i as follows:
repeating the calculation process for each point in X (T) from left to right to obtain an inspection statistic value sequence T (T);
c. calculating the maximum value T of T (T) max Statistical significance of P max ，P max Can be generally expressed as:
in the formula, the first step is that,in the case of an incomplete beta function,for the integral upper limit of the incomplete beta function, δ v and δ are two parameters of the incomplete beta function, v is the degree of freedom of the t test, wherein, the integral upper limit of the incomplete beta function is obtained according to Monte Carlo simulation, γ =4.19lnN-11.54, δ =0.40, ν = N-2,N is the total point number of the sampling sequence；
d. Setting a threshold value P 0 =0.95, maximum value T among T (T) max Statistical significance of P max >P 0 When the fault is detected, the X (T) mutation is obvious, and the corresponding wave recording data has a trend mutation, namely the fault is a developmental fault, wherein T is max The corresponding point i is a catastrophe point, and the waveform is divided into two parts from the point i.
4. The multi-level progressive classification and identification method for low-current ground faults as claimed in claim, wherein the step 4 is specifically:
a. performing FFT analysis on the zero sequence current waveform obtained in the step 3;
b. taking the previous period of the fault as a reference T 0 Each period after the fault is T in sequence 1 、T 2 …T n Calculating T i (i =1,2, …, n) each frequency amplitude as a percentage of the fundamental frequency;
c. calculating T i Relative to T 0 The percentage change quantity of each frequency amplitude is sorted from large to small to obtain T i Mean value mu of amplitude percentage change of the first 5 frequency points i ；
d. Defining the initial value of the counter N as 0 when mu i &When the content is 10 percent, adding 1 to N; otherwise, N is reset;
e. when M is 1 &2, the system is not in the recovery process; if N is present at the same time&Judging that the fault is a linear fault if the high-frequency component of the waveform does not change obviously, and judging that the fault is a nonlinear fault if the high-frequency component of the waveform does not change obviously;
in the step 4, by using the characteristic of complete classification in the low-current ground fault multi-level classification method, firstly, linear fault identification is performed to ensure that transient non-linear ground arcs are not ignored.
5. The multi-level progressive classification and identification method for the small-current ground fault as claimed in claim 1, wherein the step 5 specifically comprises:
a. according to the fault-leading phase voltage U m And metallic soZero sequence current 3I of fault time fault outgoing line 0 Calculating the zero sequence impedance Z of the system S0 ：
b. Judging the size of the transition resistance according to the zero sequence voltage at the bus:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711006521.8A CN108008244B (en) | 2017-10-25 | 2017-10-25 | A kind of small current grounding fault progressive classifying identification method at many levels |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711006521.8A CN108008244B (en) | 2017-10-25 | 2017-10-25 | A kind of small current grounding fault progressive classifying identification method at many levels |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108008244A true CN108008244A (en) | 2018-05-08 |
| CN108008244B CN108008244B (en) | 2019-07-26 |
Family
ID=62051810
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711006521.8A Active CN108008244B (en) | 2017-10-25 | 2017-10-25 | A kind of small current grounding fault progressive classifying identification method at many levels |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108008244B (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108802573A (en) * | 2018-06-28 | 2018-11-13 | 西门子电力自动化有限公司 | Detect the method and apparatus that earth fault disappears in electric system compensation network |
| CN108983042A (en) * | 2018-07-25 | 2018-12-11 | 国网湖北省电力有限公司电力科学研究院 | A kind of distribution ground fault reason recognition methods based on KNN |
| CN109494703A (en) * | 2018-11-29 | 2019-03-19 | 李景禄 | A kind of pro-active intervention type security personnel's arc extinguishing method based on compensation damping mode |
| CN109901007A (en) * | 2019-03-21 | 2019-06-18 | 国家电网有限公司 | Low current method for locating single-phase ground fault based on Hierarchical Agent intelligent decision |
| CN111220879A (en) * | 2020-01-19 | 2020-06-02 | 重庆大学 | Method for positioning single-phase earth fault section of small current grounding system |
| EP3745146A1 (en) * | 2019-05-29 | 2020-12-02 | Siemens Aktiengesellschaft | Method, device and system for determining a characteristic of a power transmission line |
| CN112364941A (en) * | 2020-12-07 | 2021-02-12 | 西安德纳检验检测有限公司 | New energy station frequency characteristic detection method, device and system |
| CN112557948A (en) * | 2020-11-27 | 2021-03-26 | 广东电网有限责任公司肇庆供电局 | Power distribution network single-phase earth fault identification method based on fault multi-feature extraction |
| CN113379279A (en) * | 2021-06-23 | 2021-09-10 | 黑龙江大学 | Deep reinforcement learning short-term voltage stability evaluation method based on incomplete data |
| CN113406455A (en) * | 2021-07-07 | 2021-09-17 | 福州大学 | Macroscopic characterization method for microscopic state of low-voltage alternating-current arc |
| CN113762412A (en) * | 2021-09-26 | 2021-12-07 | 国网四川省电力公司电力科学研究院 | Power distribution network single-phase earth fault identification method, system, terminal and medium |
| CN114156831A (en) * | 2021-11-22 | 2022-03-08 | 昆明理工大学 | Photoelectric combined instantaneous fault discrimination method |
| CN117233471A (en) * | 2023-11-09 | 2023-12-15 | 四川大学 | Medium-voltage distribution network branch line-collision fault detection method based on contact impedance gradual change characteristic |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009081215A2 (en) * | 2007-12-21 | 2009-07-02 | Dan Andras | Equipment and procedure to determine fault location and fault resistance during phase to ground faults on a live network |
| CN103323742A (en) * | 2013-05-31 | 2013-09-25 | 华北电力大学 | System and method for grid fault identification based on random point distribution PMU |
| CN103424668A (en) * | 2013-08-05 | 2013-12-04 | 昆明理工大学 | Arc light ground fault continuous route selection method utilizing principal component analysis of zero-sequence current of feeder line and evidence theoretical integration |
| CN103606909A (en) * | 2013-12-04 | 2014-02-26 | 国家电网公司 | Protecting system and method for power distribution network circuit |
| CN105044560A (en) * | 2015-08-18 | 2015-11-11 | 海南电网有限责任公司三亚供电局 | Fault-adaptive-technology-based fault judgment method of power distribution network |
| EP3136526A1 (en) * | 2015-08-25 | 2017-03-01 | GE Energy Power Conversion Technology Ltd | Ground fault protection methods |
| CN106655120A (en) * | 2016-11-30 | 2017-05-10 | 孙嘉宁 | Intelligent ground protection method and system for neutral point of power distribution network |
-
2017
- 2017-10-25 CN CN201711006521.8A patent/CN108008244B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009081215A2 (en) * | 2007-12-21 | 2009-07-02 | Dan Andras | Equipment and procedure to determine fault location and fault resistance during phase to ground faults on a live network |
| CN103323742A (en) * | 2013-05-31 | 2013-09-25 | 华北电力大学 | System and method for grid fault identification based on random point distribution PMU |
| CN103424668A (en) * | 2013-08-05 | 2013-12-04 | 昆明理工大学 | Arc light ground fault continuous route selection method utilizing principal component analysis of zero-sequence current of feeder line and evidence theoretical integration |
| CN103606909A (en) * | 2013-12-04 | 2014-02-26 | 国家电网公司 | Protecting system and method for power distribution network circuit |
| CN105044560A (en) * | 2015-08-18 | 2015-11-11 | 海南电网有限责任公司三亚供电局 | Fault-adaptive-technology-based fault judgment method of power distribution network |
| EP3136526A1 (en) * | 2015-08-25 | 2017-03-01 | GE Energy Power Conversion Technology Ltd | Ground fault protection methods |
| CN106655120A (en) * | 2016-11-30 | 2017-05-10 | 孙嘉宁 | Intelligent ground protection method and system for neutral point of power distribution network |
Non-Patent Citations (2)
| Title |
|---|
| 王谦等: "一种基于录波数据的故障事件检测的混合方法", 《华东电力》 * |
| 薛永端等: "基于零序电压比率制动的小电阻接地系统接地保护", 《电力系统自动化》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108802573A (en) * | 2018-06-28 | 2018-11-13 | 西门子电力自动化有限公司 | Detect the method and apparatus that earth fault disappears in electric system compensation network |
| CN108983042A (en) * | 2018-07-25 | 2018-12-11 | 国网湖北省电力有限公司电力科学研究院 | A kind of distribution ground fault reason recognition methods based on KNN |
| CN109494703A (en) * | 2018-11-29 | 2019-03-19 | 李景禄 | A kind of pro-active intervention type security personnel's arc extinguishing method based on compensation damping mode |
| CN109901007A (en) * | 2019-03-21 | 2019-06-18 | 国家电网有限公司 | Low current method for locating single-phase ground fault based on Hierarchical Agent intelligent decision |
| CN109901007B (en) * | 2019-03-21 | 2020-12-22 | 国家电网有限公司 | Small current single-phase earth fault positioning method based on layered agent intelligent decision |
| EP3745146A1 (en) * | 2019-05-29 | 2020-12-02 | Siemens Aktiengesellschaft | Method, device and system for determining a characteristic of a power transmission line |
| US11360156B2 (en) | 2019-05-29 | 2022-06-14 | Siemens Aktiengesellschaft | Method, device and system for determining an arc on a power transmission line |
| CN111220879A (en) * | 2020-01-19 | 2020-06-02 | 重庆大学 | Method for positioning single-phase earth fault section of small current grounding system |
| CN112557948A (en) * | 2020-11-27 | 2021-03-26 | 广东电网有限责任公司肇庆供电局 | Power distribution network single-phase earth fault identification method based on fault multi-feature extraction |
| CN112364941A (en) * | 2020-12-07 | 2021-02-12 | 西安德纳检验检测有限公司 | New energy station frequency characteristic detection method, device and system |
| CN113379279A (en) * | 2021-06-23 | 2021-09-10 | 黑龙江大学 | Deep reinforcement learning short-term voltage stability evaluation method based on incomplete data |
| CN113406455A (en) * | 2021-07-07 | 2021-09-17 | 福州大学 | Macroscopic characterization method for microscopic state of low-voltage alternating-current arc |
| CN113406455B (en) * | 2021-07-07 | 2022-03-15 | 福州大学 | Macroscopic characterization method for microscopic state of low-voltage alternating-current arc |
| CN113762412A (en) * | 2021-09-26 | 2021-12-07 | 国网四川省电力公司电力科学研究院 | Power distribution network single-phase earth fault identification method, system, terminal and medium |
| CN113762412B (en) * | 2021-09-26 | 2023-04-18 | 国网四川省电力公司电力科学研究院 | Power distribution network single-phase earth fault identification method, system, terminal and medium |
| CN114156831A (en) * | 2021-11-22 | 2022-03-08 | 昆明理工大学 | Photoelectric combined instantaneous fault discrimination method |
| CN117233471A (en) * | 2023-11-09 | 2023-12-15 | 四川大学 | Medium-voltage distribution network branch line-collision fault detection method based on contact impedance gradual change characteristic |
| CN117233471B (en) * | 2023-11-09 | 2024-01-23 | 四川大学 | Medium-voltage distribution network branch line-collision fault detection method based on contact impedance gradual change characteristic |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108008244B (en) | 2019-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108008244A (en) | A kind of small current grounding fault progressive classifying identification method at many levels | |
| Manohar et al. | Spatio-temporal information based protection scheme for PV integrated microgrid under solar irradiance intermittency using deep convolutional neural network | |
| CN102016607B (en) | Method and apparatus for analyzing waveform signals of a power system | |
| CN101027565B (en) | Method and device for detecting electric arc phenomenon on at least one electric cable | |
| CN109800660B (en) | Voltage sag source identification method and system based on big data clustering | |
| CN111160241B (en) | Power distribution network fault classification method, system and medium based on deep learning | |
| Lopes et al. | Analysis of signal processing techniques for high impedance fault detection in distribution systems | |
| CN110261746B (en) | Cable defect detection method based on periodic attenuation characteristics of oscillating wave voltage | |
| CN115796708B (en) | Big data intelligent quality inspection method, system and medium for engineering construction | |
| CN114942386A (en) | Power battery fault online detection method and system | |
| CN116540025A (en) | Fault detection method based on transfer learning and residual error network | |
| CN116298735A (en) | AC arc fault detection method and related device for low-voltage distribution network | |
| Asman et al. | Identification of transient overvoltage using discrete wavelet transform with minimised border distortion effect and support vector machine | |
| CN116047164A (en) | Detection method and detection device for insulation resistance abnormality of electric automobile | |
| Gadanayak et al. | High impedance fault detection in distribution networks using randomness of zero-sequence current signal: A detrended fluctuation analysis approach | |
| CN114089218A (en) | Power distribution network high-resistance grounding fault identification method, device, terminal and medium | |
| CN115186772A (en) | Method, device and equipment for detecting partial discharge of power equipment | |
| CN114755529A (en) | Multi-feature fusion single-phase earth fault type identification method based on deep learning | |
| CN116973695A (en) | Partial discharge mode identification method, device and equipment for power cable and storage medium | |
| CN113361729A (en) | Risk assessment method, device, terminal and storage medium based on maintenance plan | |
| CN108983042B (en) | KNN-based distribution network ground fault reason identification method | |
| Ashrafian et al. | Detection of internal incipient faults in transformers during impulse test using hyperbolic S‐transform | |
| CN113985733B (en) | Power distribution network fault identification method based on self-adaptive probability learning | |
| CN116756660B (en) | Single-phase wire contact vegetation ignition prediction method, system and medium | |
| CN113671293B (en) | Distribution line lightning stroke fault property distinguishing method, system, equipment and terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |