CN110244219B - Breaker fault identification method based on closing coil current time domain statistical characteristics - Google Patents
Breaker fault identification method based on closing coil current time domain statistical characteristics Download PDFInfo
- Publication number
- CN110244219B CN110244219B CN201910618553.6A CN201910618553A CN110244219B CN 110244219 B CN110244219 B CN 110244219B CN 201910618553 A CN201910618553 A CN 201910618553A CN 110244219 B CN110244219 B CN 110244219B
- Authority
- CN
- China
- Prior art keywords
- circuit breaker
- closing coil
- current
- value
- signals
- 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.)
- Active
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
- G01R31/327—Testing of circuit interrupters, switches or circuit-breakers
- G01R31/3271—Testing of circuit interrupters, switches or circuit-breakers of high voltage or medium voltage devices
- G01R31/3275—Fault detection or status indication
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention discloses a circuit breaker fault identification method based on current time domain statistical characteristics of a closing coil, which comprises the steps of collecting current data of the closing coil during closing operation of a circuit breaker and extracting characteristic points of the current data; and calculating Euclidean distances between the characteristic points of the signals to be identified and the characteristic points of the known class signals, finding out a plurality of signals closest to the characteristic points of the signals to be identified, and judging the state of the circuit breaker where the signals to be identified are located according to the state of the circuit breaker of the signals, thereby realizing fault identification of the circuit breaker. The invention can diagnose common mechanical faults of the circuit breaker operating mechanism and accurately position fault elements and fault parts when the faults occur, thereby helping maintenance personnel to process the faults, improving the maintenance efficiency and ensuring the safe and reliable operation of the circuit breaker.
Description
Technical Field
The invention relates to the technical field of breaker fault diagnosis, in particular to a breaker fault identification method based on closing coil current time domain statistical characteristics.
Background
Circuit breakers carry the critical tasks of control and protection in electrical power systems. If the breaker breaks down and can not be switched on and off in time, the fault range in the power system is necessarily enlarged, and the safe and reliable operation of the power system is seriously threatened. According to investigation, most of the faults of the circuit breaker are mechanical faults of an operating mechanism of the circuit breaker; in addition, the circuit breaker operating mechanism is complex in structure, and when the circuit breaker operating mechanism has mechanical failure, the failure part or failure element is often difficult to determine, so that the difficulty of maintenance work is improved.
The closing coil is one of the key elements in the circuit breaker operating mechanism, and when current flows in the coil, magnetic flux is generated, so that an iron core in the coil starts to act under the action of electromagnetic attraction, and the circuit breaker is controlled to realize the closing function. On the contrary, when the current in the coil is cut off, the magnetic flux in the coil disappears, so that the iron core loses the electromagnetic attraction force and is restored to the original position under the action of the spring. Therefore, the working conditions of different parts in the operating mechanisms such as the iron core, the tripping mechanism and the auxiliary switch can be reflected through the current of the closing coils in different stages, and the current signal can carry out fault diagnosis on the mechanical fault of the operating mechanism. At present, for the selection of current characteristic points of the switching-on and switching-off coil, the essential characteristics of different distribution of the current characteristic points of the coil in time domains under different states are not considered, and the selected characteristic points cannot comprehensively reflect the change condition of the current of the coil; most of methods for diagnosing by adopting the current of the closing coil adopt artificial intelligence algorithms such as a support vector machine, a neural network and the like, the training speed is low, problems such as overfitting, local optimization and the like are easily caused, and the accuracy of fault diagnosis is influenced, so that a method for diagnosing the open circuit fault, which has more reasonable characteristic point selection and more accurate and reliable fault identification, is necessarily introduced.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide a method for identifying a circuit breaker fault based on closing coil current time domain statistical characteristics, which can quickly diagnose the element, part and type of the mechanical fault of a circuit breaker operating mechanism. The technical scheme is as follows:
a breaker fault identification method based on closing coil current time domain statistical characteristics comprises the following steps:
step A: respectively acquiring experimental data of current signals of a closing coil of a circuit breaker operating mechanism in a normal state and each fault state;
and B: extracting signals to be analyzed of the current of a closing coil of the circuit breaker operating mechanism in a normal state and in each fault state,
extracting time domain statistical characteristic points of the current to-be-analyzed signal of the closing coil;
and C: collecting current signals of a closing coil of a circuit breaker operating mechanism in actual engineering, extracting signals to be identified of the current of the closing coil from the current signals, and extracting time domain statistical characteristic points of the signals to be identified of the current of the closing coil;
step D: and calculating Euclidean distances between the time domain statistical characteristic points of the signals to be identified and the time domain statistical characteristic points of the signals to be analyzed, finding out a plurality of signals which are closest to the time domain statistical characteristic points of the signals to be identified, and judging the state of the circuit breaker where the signals to be identified are located according to the state of the circuit breaker of the signals, thereby realizing fault identification of the circuit breaker.
Further, the step a specifically includes:
carry out N times switching-on operation experiments to normal condition and operating device bite, not hard up, the iron core bite three kinds of fault state's of base screw high voltage circuit breaker respectively to sampling frequency f gathers respectively:
current signal i of closing coil of circuit breaker operating mechanism in n-th experiment under normal state0 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under jammed state of operating mechanism1 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under loose state of base screw2 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under iron core jamming state3 n;
Current signal i of the switching-on coil0 n、i1 n、i2 n、i3 nT × f sampling points are included, where T is the time of each closing operation experiment, and N is 1, 2, …, N.
Further, the step B specifically includes:
step B1: denoising current signals of a closing coil in the nth experiment of each state of the circuit breaker by utilizing wavelet transformation, comparing the denoised current signals of the closing coil with a judgment threshold value point by point, and recording the sequence number of a sampling point corresponding to the current signal value of the closing coil with the 1 st value larger than the judgment threshold value as 1 when 5 continuous points are larger than the judgment threshold value; until 5 are continuedWhen all the points are smaller than the judgment threshold value, recording the sampling point serial number corresponding to the current signal value of the 1 st closing coil smaller than the judgment threshold value as kb n(ii) a Selecting a current signal of a closing coil in the time period as a signal i to be analyzedb *n(k) (ii) a Wherein b is a breaker state, b is 0 to represent a normal state, b is 1 to represent an operating mechanism jamming fault, b is 2 to represent a base screw loosening fault, and b is 3 to represent an iron core jamming fault; the sampling point sequence number takes the value as: 1, 2, …, kb n;
Step B2: to-be-analyzed signal i of current of closing coil of circuit breaker operating mechanism for n-th experiment under each state of circuit breaker is extracted respectivelyb *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a signal i to be analyzed of the current of a closing coil of a circuit breaker operating mechanism in the nth experiment in the normal state of the circuit breaker respectivelyb *n(k) Middle 0.2kb n、0.4kb n、0.6kb n、0.8kb nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of the signal to be analyzed of the nth experiment of the switching-on coil current in each state of the circuit breakerb(n)=[tb1 n,tb2 n,tb3 n,tb4 n,tb5 n,tb6 n,tb7 n,tb8 n,ib1 n,ib2 n,ib3 n,ib5 n,ib6 n,ib7 n,ib8 n];tb1 n、tb2 n、tb3 nCurrent signals i representing the switching-on coils, respectivelyb *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is tb4 n、tb5 n、tb6 n、tb7 n、tb8 nAre each kb n、0.2kb n、0.4kb n、0.6kb n、0.8kb n;ib1 n、ib2 n、ib3 nSignals i to be analyzed respectively representing current of closing coilsb *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. ib5 n、ib6 n、ib7 n、ib8 nSignals i to be analyzed respectively representing current of closing coilsb *n(k) Middle 0.2kb n、0.4kb n、0.6kb n、0.8kb nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
Further, the step C specifically includes:
step C1: collecting a switching-on coil current signal of an operating mechanism during switching-on operation of a high-voltage circuit breaker in actual engineering in real time according to a sampling frequency f, and carrying out denoising processing on the collected switching-on coil current signal by utilizing wavelet transformation; comparing the denoised closing coil current signal point by point with a judgment threshold, and recording the sequence number of a sampling point corresponding to the closing coil current signal value greater than 1 as 1 when 5 continuous points are greater than the judgment threshold; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as km(ii) a Selecting a closing coil current signal in the time period as a signal i to be identifiedm(k) And the sampling point sequence number is: 1, 2, …, km;
Step C2: extracted and collected current to-be-identified signal i of closing coil of circuit breaker operating mechanismm(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; meanwhile, extracting a signal i to be identified of the current of the closing coilm(k) Middle 0.2km、0.4km、0.6km、0.8kmThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of the actually measured current signal of the closing coil of the circuit breaker to be identifiedm=[t1,t2,t3,t4,t5,t6,t7,t8,i1,i2,i3,i5,i6,i7,i8]Wherein t is1、t2、t3Signals i to be identified respectively representing current of closing coilm(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t4、t5、t6、t7、t8Are each km、0.2km、0.4km、0.6km、0.8km;i1、i2、i3Signals i to be identified respectively representing current of closing coilm(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i5、i6、i7、i8Signals i to be identified respectively representing current of closing coilm(k) Middle 0.2km、0.4km、0.6km、0.8kmAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
Further, the step D specifically includes:
step D1: the characteristic quantities of all experimental sample data in four states of normal state, jamming of an operating mechanism, loosening of a base screw and jamming of an iron core and the actually measured time domain statistical characteristic quantity X of a signal to be recognized of the current of a closing coil of the circuit breaker to be recognized are measuredmForming a 4N +1 row 15 column high-voltage circuit breaker operating mechanism closing coil current signal time domain statistical characteristic quantity set matrix X, X ═ X0(1);X0(2);…;X0(N);X1(1);X1(2);…;X1(N);X2(1);X2(2);…;X2(N);X3(1);X3(2);…;X3(N);Xm](ii) a Solving the maximum value X in each column element of the characteristic quantity set matrix Xmax(j) And the minimum value xmin(j) Wherein j is 1, 2, …, 15;
step D2: for the element X of the p row and the j column in the characteristic quantity set matrix Xp,jNormalization was performed as follows:
wherein p is 1, 2, …, 4N + 1; x'p,jIs the element X of the p-th row and the j-th column in Xp,jA normalized value of (d);
step D3: the normalized values X 'of the last row of elements in X are calculated separately as follows'4N+1,jEuclidean distances to the normalized values of the other row elements in X:
step D4: selecting the row number p corresponding to the minimum s Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D31、p2、…、psIf p is1、p2、…、p10If the number of the breaker distributed in the range of 1-N is the maximum, the breaker is identified to be in a normal state; if p is1、p2、…、psThe maximum number of the circuit breakers distributed in the range of N + 1-2N is determined, and the jamming fault of the operating mechanism of the circuit breaker is identified; if p is1、p2、…、psThe number of the base screws distributed in the range of 2N + 1-3N is the largest, and then the base screw loosening fault of the circuit breaker is identified; if p is1、p2、…、psThe number of the iron core blocking faults distributed in the range of 3N + 1-4N is the largest, and the iron core blocking faults of the circuit breaker are identified;
step D5: if p is in step D41、p2、…、psSelecting the row number p corresponding to the minimum s +1 Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D3 when the maximum values of the numbers in different ranges are the same respectively1、p2、…、ps、ps+1And repeating the step D4 until the distribution number is the only one.
Furthermore, the number N of the switching-on experiments of the high-voltage circuit breaker in each state is not less than 200.
Furthermore, the value range of the judgment threshold is 2 mA-10 mA.
Furthermore, the value range of s in the minimum s Euclidean distance values is 5% -10% of the switching-on experiment frequency N of the high-voltage circuit breaker.
The invention has the beneficial effects that: according to the invention, not only are peak and valley points of the coil current taken as characteristic points, but also the time and current values corresponding to the 0.2k, 0.4k, 0.6k, 0.8k and k sampling points in the k sampling points of the closing coil current are selected as the characteristic points, so that an improved coil current characteristic point set is obtained, and the change condition of the closing coil current of the circuit breaker can be more comprehensively reflected. The method and the device find out s pieces of experimental data which are closest to the characteristic point distance of the signal to be identified, judge the state of the circuit breaker of the signal to be identified according to the state of the circuit breaker of the experimental data, and compared with the prior art, the method and the device can more quickly and reliably realize the fault identification of the circuit breaker and have higher engineering application value.
Detailed Description
The present invention will be described in further detail with reference to specific examples. The invention relates to a breaker fault identification method based on closing coil current time domain statistical characteristics, which comprises the following steps of: firstly, respectively acquiring experimental data of current signals of a closing coil of a circuit breaker operating mechanism in a normal state and in each fault state; extracting signals to be analyzed of the current of the closing coil of the circuit breaker operating mechanism in a normal state and each fault state, and extracting time domain statistical characteristic points of the signals to be analyzed of the current of the closing coil; then collecting current signals of a closing coil of a circuit breaker operating mechanism in actual engineering, extracting signals to be identified of the current of the closing coil from the current signals, and extracting time domain statistical characteristic points of the signals to be identified of the current of the closing coil; and finally, calculating the Euclidean distance between the time domain statistical characteristic points of the signals to be identified and the time domain statistical characteristic points of the signals to be analyzed, finding out a plurality of signals which are closest to the time domain statistical characteristic points of the signals to be identified, and judging the state of the circuit breaker where the signals to be identified are located according to the state of the circuit breaker of the signals, thereby realizing the fault identification of the circuit breaker. The method comprises the following specific steps:
A. acquisition of failure data of circuit breaker
Carry out N times switching-on operation experiments to normal condition and operating device bite, not hard up, the iron core bite three kinds of fault state's of base screw high voltage circuit breaker respectively to sampling frequency f gathers respectively:
current signal i of closing coil of circuit breaker operating mechanism in n-th experiment under normal state0 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under jammed state of operating mechanism1 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under loose state of base screw2 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under iron core jamming state3 n;
Current signal i of closing coil0 n、i1 n、i2 n、i3 nT × f sampling points are included, where T is the time of each switching-on operation experiment, and N is 1, 2, …, N.
B. Extraction of time domain statistical characteristics of experimental data of circuit breaker
B1, using wavelet transformation to carry out current signal i of closing coil of operating mechanism in nth experiment when breaker is normal0 nDenoising, namely comparing the denoised current signals of the closing coil point by point with a judgment threshold value; when the continuous 5 points are all larger than the judgment threshold value, recording the serial number of a sampling point corresponding to the current signal value of the 1 st closing coil larger than the judgment threshold value as 1; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as k0 nStopping comparison; selecting a current signal of a closing coil in the time period as a signal i to be analyzed0 *n(k) And the sampling point sequence number is: 1, 2, …, k0 n。
Operating mechanism closing coil current signal i of nth experiment under state of jamming of circuit breaker operating mechanism by utilizing wavelet transformation1 nDenoising, namely comparing the denoised current signals of the closing coil point by point with a judgment threshold value; when the continuous 5 points are all larger than the judgment threshold value, recording the serial number of a sampling point corresponding to the current signal value of the 1 st closing coil larger than the judgment threshold value as 1; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as k1 nStopping comparison; selecting a current signal of a closing coil in the time period as a signal i to be analyzed1 *n(k) And the sampling point sequence number is: 1, 2, …, k1 n。
Operating mechanism closing coil current signal i for nth experiment under loose state of circuit breaker base screw by utilizing wavelet transformation2 nDenoising, namely comparing the denoised current signals of the closing coil point by point with a judgment threshold value; when the continuous 5 points are all larger than the judgment threshold value, recording the serial number of a sampling point corresponding to the current signal value of the 1 st closing coil larger than the judgment threshold value as 1; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as k2 nStopping comparison; selecting a current signal of a closing coil in the time period as a signal i to be analyzed2 *n(k) And the sampling point sequence number is: 1, 2, …, k2 n。
Operating mechanism closing coil current signal i for nth experiment under state of jamming of circuit breaker iron core by utilizing wavelet transformation3 nDenoising, namely comparing denoised closing coil current signals with a judgment threshold value point by point, and recording the sequence number of a sampling point corresponding to the 1 st closing coil current signal value larger than the judgment threshold value as 1 when 5 continuous points are larger than the judgment threshold value; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as k3 nRatio of stopComparing; selecting a current signal of a closing coil in the time period as a signal i to be analyzed3 *n(k) And the sampling point sequence number is: 1, 2, …, k3 n。
In this embodiment, the threshold value is set to 2mA to 10 mA.
B2, extracting signals i to be analyzed of current of closing coil of circuit breaker operating mechanism in nth experiment under normal state of circuit breaker respectively0 *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a signal i to be analyzed of the current of a closing coil of a circuit breaker operating mechanism in the nth experiment in the normal state of the circuit breaker respectively0 *n(k) Middle 0.2k0 n、0.4k0 n、0.6k0 n、0.8k0 nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of current sample data of the switching-on coil in the normal state of the circuit breaker0(n)=[t01 n,t02 n,t03 n,t04 n,t05 n,t06 n,t07 n,t08 n,i01 n,i02 n,i03 n,i05 n,i06 n,i07 n,i08 n]Wherein t is01 n、t02 n、t03 nSignals i to be analyzed respectively representing current of closing coils0 *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t04 n、t05 n、t06 n、t07 n、t18 nAre each k0 n、0.2k0 n、0.4k0 n、0.6k0 n、0.8k0 n;i01 n、i02 n、i03 nRespectively indicating closing coilsCurrent of i to be analyzed0 *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i05 n、i06 n、i07 n、i08 nSignals i to be analyzed respectively representing current of closing coils0 *n(k) Middle 0.2k0 n、0.4k0 n、0.6k0 n、0.8k0 nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
Circuit breaker operating mechanism closing coil current to-be-analyzed signal i for n-th experiment under jammed state of closing circuit breaker operating mechanism is extracted respectively1 *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a signal i to be analyzed of the current of a closing coil of the circuit breaker operating mechanism in the nth experiment under the jammed state of the circuit breaker operating mechanism respectively1 *n(k1) Middle 0.2k1 n、0.4k1 n、0.6k1 n、0.8k1 nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining time domain statistical characteristic quantity X of current sample data of a closing coil in a jammed state of a circuit breaker operating mechanism1(n)=[t11 n,t12 n,t13 n,t14 n,t15 n,t16 n,t17 n,t18 n,i11 n,i12 n,i13 n,i15 n,i16 n,i17 n,i18 n]Wherein t is11 n、t12 n、t13 nSignals i to be analyzed respectively representing current of closing coils1 *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t14 n、t15 n、t16 n、t17 n、t18 nAre each k1 n、0.2k1 n、0.4k1 n、0.6k1 n、0.8k1 n;i11 n、i12 n、i13 nSignals i to be analyzed respectively representing current of closing coils1 *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i15 n、i16 n、i17 n、i18 nSignals i to be analyzed respectively representing current of closing coils1 *n(k) Middle 0.2k1 n、0.4k1 n、0.6k1 n、0.8k1 nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
Circuit breaker operating mechanism closing coil current to-be-analyzed signal i for n-th experiment under loose state of circuit breaker base screw is extracted respectively2 *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a signal i to be analyzed of current of a closing coil of a circuit breaker operating mechanism in the nth experiment under the condition that a screw of a circuit breaker base is loosened2 *n(k) Middle 0.2k2 n、0.4k2 n、0.6k2 n、0.8k2 nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining time domain statistical characteristic quantity X of current sample data of a closing coil in a state that a base screw of the circuit breaker is loosened2(n)=[t21 n,t22 n,t23 n,t24 n,t25 n,t26 n,t27 n,t28 n,i21 n,i22 n,i23 n,i25 n,i26 n,i27 n,i28 n]Wherein t is21 n、t22 n、t23 nRespectively indicating the current of the closing coil to be dividedAnalysis signal i2 *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t24 n、t25 n、t26 n、t27 n、t28 nAre each k2 n、0.2k2 n、0.4k2 n、0.6k2 n、0.8k2 n;i21 n、i22 n、i23 nSignals i to be analyzed respectively representing current of closing coils2 *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i25 n、i26 n、i27 n、i28 nSignals i to be analyzed respectively representing current of closing coils2 *n(k) Middle 0.2k2 n、0.4k2 n、0.6k2 n、0.8k2 nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
Circuit breaker operating mechanism closing coil current to-be-analyzed signal i for n-th experiment under blocking state of circuit breaker iron core is extracted respectively3 *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a current to-be-analyzed signal i of a closing coil of a circuit breaker operating mechanism of an nth experiment under the state of jamming of a circuit breaker iron core respectively3 *n(k) Middle 0.2k3 n、0.4k3 n、0.6k3 n、0.8k3 nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining time domain statistical characteristic quantity X of current sample data of closing coil in circuit breaker iron core jamming state3(n)=[t31 n,t32 n,t33 n,t34 n,t35 n,t36 n,t37 n,t38 n,i31 n,i32 n,i33 n,i35 n,i36 n,i37 n,i38 n]Wherein t is31 n、t32 n、t33 nSignals i to be analyzed respectively representing current of closing coils3 *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t34 n、t35 n、t36 n、t37 n、t38 nAre each k3 n、0.2k3 n、0.4k3 n、0.6k3 n、0.8k3 n;i31 n、i32 n、i33 nSignals i to be analyzed respectively representing current of closing coils3 *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i35 n、i36 n、i37 n、i38 nSignals i to be analyzed respectively representing current of closing coils3 *n(k) Middle 0.2k3 n、0.4k3 n、0.6k3 n、0.8k3 nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
C. Acquisition of time domain statistical characteristic quantity of current signal of closing coil of circuit breaker operating mechanism
C1, collecting the current signals of the closing coil of the operating mechanism in real time when the high-voltage circuit breaker operates in cooperation in the actual engineering according to the sampling frequency f, and carrying out denoising processing on the collected current signals of the closing coil by utilizing wavelet transformation; comparing the denoised closing coil current signal point by point with a judgment threshold value, and recording the sampling point serial number corresponding to the closing coil current signal value greater than 1 as 1 when 5 continuous points are greater than the judgment threshold value; then, when 5 continuous points are all smaller than the preset value, recording the serial number of the sampling point corresponding to the current signal value of the 1 st smaller closing coil as km(ii) a Selecting a closing coil current signal in the time period as a signal i to be identifiedm(k) Where k is 1, 2, …, km。
C2, extracting collected current signal i of closing coil of circuit breaker operating mechanismm(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; at the same time, a current signal i of a closing coil is extractedm(k) Middle 0.2km、0.4km、0.6km、0.8kmThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of the actually measured current signal of the closing coil of the circuit breaker to be identifiedm=[t1,t2,t3,t4,t5,t6,t7,t8,i1,i2,i3,i5,i6,i7,i8]Wherein t is1、t2、t3Current signals i representing the switching-on coils, respectivelym(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t4、t5、t6、t7、t8Are each km、0.2km、0.4km、0.6km、0.8km;i1、i2、i3Current signals i representing the switching-on coils, respectivelym(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i5、i6、i7、i8Current signals i representing the switching-on coils, respectivelym(k) Middle 0.2km、0.4km、0.6km、0.8kmAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
D. Identification of breaker fault type
D1, carrying out time domain statistical characteristic quantity X of characteristic quantities of all experimental sample data and actually measured current signals of the circuit breaker closing coil to be identified in four states of normal state, clamping of the operating mechanism, loosening of base screws and clamping of iron coresmForming a 4N +1 row 15 column high-voltage circuit breaker operating mechanism closing coil current signal time domain statistical characteristic quantity set matrix X, X ═ X0(1);X0(2);…;X0(N);X1(1);X1(2);…;X1(N);X2(1);X2(2);…;X2(N);X3(1);X3(2);…;X3(N);Xm](ii) a Solving the maximum value X in each column element of the characteristic quantity set matrix Xmax(j) And the minimum value xmin(j) Wherein j is 1, 2, …, 15.
D2, element X of p row and j column in characteristic quantity set matrix Xp,jNormalization was performed as follows:
wherein p is 1, 2, …, 4N + 1; x'p,jIs the element X of the p-th row and the j-th column in Xp,jThe normalized value of (a).
D3, calculating the normalized value X 'of the last line element in X according to the formula'4N+1,jEuclidean distances to the normalized values of the other row elements in X.
D4, selecting the row number p corresponding to the minimum s Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D31、p2、…、psIf p is1、p2、…、p10If the number of the breaker distributed in the range of 1-N is the maximum, the breaker is identified to be in a normal state; if p is1、p2、…、psThe maximum number of the circuit breakers distributed in the range of N + 1-2N is determined, and the jamming fault of the operating mechanism of the circuit breaker is identified; if p is1、p2、…、psThe number of the base screws distributed in the range of 2N + 1-3N is the largest, and then the base screw loosening fault of the circuit breaker is identified; if p is1、p2、…、psAnd the number of the iron cores distributed in the range of 3N + 1-4N is the largest, and then the iron core jamming fault of the circuit breaker is identified.
In the embodiment, the value range of s is 5-10% of the switching-on experiment frequency N of the high-voltage circuit breaker.
D5, if p in step D41、p2、…、psSelecting the row number p corresponding to the minimum s +1 Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D3 when the maximum values of the numbers in different ranges are the same respectively1、p2、…、ps、ps+1And repeating the step D4 until the distribution number is the only one.
The principle of the fault identification method of the invention is as follows:
the current of the closing coil of the circuit breaker can show the action condition of the closing coil and the running condition of a controlled operating mechanism in the working process, the coil electrifying time, the operating mechanism starting time, the iron core moving time and the like, and can be obtained in the detection of the current of the closing coil. And whether the mechanical fault occurs to the circuit breaker operating mechanism in the working process can be judged according to the circuit breaker parameters and the historical information. The invention firstly extracts the wave crest and the wave trough as characteristic points, and then selects the time and the current value corresponding to the sampling points of 0.2k, 0.4k, 0.6k, 0.8k and k in k sampling points of the current of the closing coil as the characteristic points to reflect the change condition when the current does not reach the wave crest and the wave trough, thereby more comprehensively reflecting different characteristics of different currents of the closing coil.
In order to distinguish the characteristic points under different breaker states and realize breaker fault identification, the invention firstly carries out unified normalization processing on the characteristic points of the test data and the characteristic points of the signals to be identified. And the Euclidean distance between the characteristic points is shorter according to the test data which is the same as the state of the circuit breaker to which the signal to be identified belongs. According to the principle, the distance between the characteristic point of the signal to be identified and other test data characteristic points is calculated, s samples with the shortest distance to the characteristic point are found, the category of the signal to be identified can be determined by observing the categories of the samples, and therefore fault identification of the circuit breaker is achieved.
Claims (6)
1. A breaker fault identification method based on closing coil current time domain statistical characteristics is characterized by comprising the following steps:
step A: respectively acquiring experimental data of current signals of a closing coil of a circuit breaker operating mechanism in a normal state and each fault state;
and B: extracting signals to be analyzed of closing coil current of a circuit breaker operating mechanism in a normal state and each fault state, and extracting time domain statistical characteristic points of the signals to be analyzed of the closing coil current;
and C: collecting current signals of a closing coil of a circuit breaker operating mechanism in actual engineering, extracting signals to be identified of the current of the closing coil from the current signals, and extracting time domain statistical characteristic points of the signals to be identified of the current of the closing coil;
step D: calculating Euclidean distance between a time domain statistical characteristic point of a signal to be identified and a time domain statistical characteristic point of a signal to be analyzed, finding out a plurality of signals with the closest distance to the time domain statistical characteristic point of the signal to be identified, and judging the state of a circuit breaker where the signal to be identified is located according to the state of the circuit breaker of the signals, thereby realizing fault identification of the circuit breaker;
the step A specifically comprises the following steps:
carry out N times switching-on operation experiments to normal condition and operating device bite, not hard up, the iron core bite three kinds of fault state's of base screw high voltage circuit breaker respectively to sampling frequency f gathers respectively:
current signal i of closing coil of circuit breaker operating mechanism in n-th experiment under normal state0 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under jammed state of operating mechanism1 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under loose state of base screw2 n;
Current signal i of closing coil of circuit breaker operating mechanism in nth experiment under iron core jamming state3 n;
Current signal i of the switching-on coil0 n、i1 n、i2 n、i3 nEach of which contains T × f sampling points, whereinT is the time of each switching-on operation experiment, and N is 1, 2, …, N;
the step B specifically comprises the following steps:
step B1: denoising current signals of a closing coil in the nth experiment of each state of the circuit breaker by utilizing wavelet transformation, comparing the denoised current signals of the closing coil with a judgment threshold value point by point, and recording the sequence number of a sampling point corresponding to the current signal value of the closing coil with the 1 st value larger than the judgment threshold value as 1 when 5 continuous points are larger than the judgment threshold value; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as kb n(ii) a Selecting a current signal of a closing coil in the time period as a signal i to be analyzedb *n(k) (ii) a Wherein b is a breaker state, b is 0 to represent a normal state, b is 1 to represent an operating mechanism jamming fault, b is 2 to represent a base screw loosening fault, and b is 3 to represent an iron core jamming fault; the sampling point sequence number takes the value as: 1, 2, …, kb n;
Step B2: to-be-analyzed signal i of current of closing coil of circuit breaker operating mechanism for n-th experiment under each state of circuit breaker is extracted respectivelyb *n(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; simultaneously, extracting a signal i to be analyzed of the current of a closing coil of a circuit breaker operating mechanism in the nth experiment in the normal state of the circuit breaker respectivelyb *n(k) Middle 0.2kb n、0.4kb n、0.6kb n、0.8kb nThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of the signal to be analyzed of the nth experiment of the switching-on coil current in each state of the circuit breakerb(n)=[tb1 n,tb2 n,tb3 n,tb4 n,tb5 n,tb6 n,tb7 n,tb8 n,ib1 n,ib2 n,ib3 n,ib5 n,ib6 n,ib7 n,ib8 n];tb1 n、tb2 n、tb3 nCurrent signals i representing the switching-on coils, respectivelyb *n(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is tb4 n、tb5 n、tb6 n、tb7 n、tb8 nAre each kb n、0.2kb n、0.4kb n、0.6kb n、0.8kb n;ib1 n、ib2 n、ib3 nSignals i to be analyzed respectively representing current of closing coilsb *n(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. ib5 n、ib6 n、ib7 n、ib8 nSignals i to be analyzed respectively representing current of closing coilsb *n(k) Middle 0.2kb n、0.4kb n、0.6kb n、0.8kb nAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
2. The circuit breaker fault identification method based on closing coil current time domain statistical characteristics according to claim 1, wherein the step C specifically comprises:
step C1: collecting a switching-on coil current signal of an operating mechanism during switching-on operation of a high-voltage circuit breaker in actual engineering in real time according to a sampling frequency f, and carrying out denoising processing on the collected switching-on coil current signal by utilizing wavelet transformation; comparing the denoised closing coil current signal point by point with a judgment threshold, and recording the sequence number of a sampling point corresponding to the closing coil current signal value greater than 1 as 1 when 5 continuous points are greater than the judgment threshold; until 5 continuous points are smaller than the judgment threshold value, recording the sequence number of the sampling point corresponding to the current signal value of the switching-on coil of which the 1 st point is smaller than the judgment threshold value as km(ii) a Selecting theTaking a current signal of a closing coil in a time period as a signal i to be identifiedm(k) And the sampling point sequence number is: 1, 2, …, km;
Step C2: extracted and collected current to-be-identified signal i of closing coil of circuit breaker operating mechanismm(k) The first wave peak value, the second wave peak value, the wave trough value and the corresponding sampling serial number; meanwhile, extracting a signal i to be identified of the current of the closing coilm(k) Middle 0.2km、0.4km、0.6km、0.8kmThe current signal value of the closing coil corresponding to the serial number of each sampling point; obtaining the time domain statistical characteristic quantity X of the actually measured current signal of the closing coil of the circuit breaker to be identifiedm=[t1,t2,t3,t4,t5,t6,t7,t8,i1,i2,i3,i5,i6,i7,i8]Wherein t is1、t2、t3Signals i to be identified respectively representing current of closing coilm(k) Sampling serial numbers corresponding to a first wave peak value, a second wave peak value and a wave valley value; t is t4、t5、t6、t7、t8Are each km、0.2km、0.4km、0.6km、0.8km;i1、i2、i3Signals i to be identified respectively representing current of closing coilm(k) The first wave peak value, the second wave peak value and the wave trough value; i.e. i5、i6、i7、i8Signals i to be identified respectively representing current of closing coilm(k) Middle 0.2km、0.4km、0.6km、0.8kmAnd the current signal value of the closing coil corresponding to the serial number of each sampling point.
3. The circuit breaker fault identification method based on closing coil current time domain statistical characteristics according to claim 2, wherein the step D specifically includes:
step D1: the normal state, the jamming of the operating mechanism, the loosening of the base screw and the jamming of the iron core are all four kindsCharacteristic quantities of all experimental sample data under the state and actually measured time domain statistical characteristic quantity X of signals to be identified of current of closing coil of circuit breaker to be identifiedmForming a 4N +1 row 15 column high-voltage circuit breaker operating mechanism closing coil current signal time domain statistical characteristic quantity set matrix X, X ═ X0(1);X0(2);…;X0(N);X1(1);X1(2);…;X1(N);X2(1);X2(2);…;X2(N);X3(1);X3(2);…;X3(N);Xm](ii) a Solving the maximum value X in each column element of the characteristic quantity set matrix Xmax(j) And the minimum value xmin(j) Wherein j is 1, 2, …, 15;
step D2: for the element X of the p row and the j column in the characteristic quantity set matrix Xp,jNormalization was performed as follows:
wherein p is 1, 2, …, 4N + 1; x'p,jIs the element X of the p-th row and the j-th column in Xp,jA normalized value of (d);
step D3: the normalized values X 'of the last row of elements in X are calculated separately as follows'4N+1,jEuclidean distances to the normalized values of the other row elements in X:
step D4: selecting the row number p corresponding to the minimum s Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D31、p2、…、psIf p is1、p2、…、p10If the number of the breaker distributed in the range of 1-N is the maximum, the breaker is identified to be in a normal state; if p is1、p2、…、psThe maximum number of the circuit breakers distributed in the range of N + 1-2N is determined, and the jamming fault of the operating mechanism of the circuit breaker is identified; if p is1、p2、…、psThe number of the base screws distributed in the range of 2N + 1-3N is the largest, and then the base screw loosening fault of the circuit breaker is identified; if p is1、p2、…、psThe number of the iron core blocking faults distributed in the range of 3N + 1-4N is the largest, and the iron core blocking faults of the circuit breaker are identified;
step D5: if p is in step D41、p2、…、psSelecting the row number p corresponding to the minimum s +1 Euclidean distance values in the Euclidean distances { D (1), D (2), …, D (4N) } calculated in the step D3 when the maximum values of the numbers in different ranges are the same respectively1、p2、…、ps、ps+1And repeating the step D4 until the distribution number is the only one.
4. The method for identifying the fault of the circuit breaker based on the current time domain statistical characteristics of the closing coil according to claim 1, wherein the number N of closing experiments of the high-voltage circuit breaker in each state is not less than 200.
5. The circuit breaker fault identification method based on closing coil current time domain statistical characteristics according to claim 1 or 2, characterized in that the value range of the determination threshold is 2 mA-10 mA.
6. The method for identifying the fault of the circuit breaker based on the closing coil current time domain statistical characteristics as claimed in claim 3, wherein s in the minimum s Euclidean distance values has a value range of 5% -10% of the number N of closing experiments of the high-voltage circuit breaker.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910618553.6A CN110244219B (en) | 2019-07-10 | 2019-07-10 | Breaker fault identification method based on closing coil current time domain statistical characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910618553.6A CN110244219B (en) | 2019-07-10 | 2019-07-10 | Breaker fault identification method based on closing coil current time domain statistical characteristics |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110244219A CN110244219A (en) | 2019-09-17 |
CN110244219B true CN110244219B (en) | 2020-07-17 |
Family
ID=67891659
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910618553.6A Active CN110244219B (en) | 2019-07-10 | 2019-07-10 | Breaker fault identification method based on closing coil current time domain statistical characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110244219B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008082896A2 (en) * | 2006-12-28 | 2008-07-10 | General Electric Company | Measurement of analog coil voltage and coil current |
CN103336243A (en) * | 2013-07-01 | 2013-10-02 | 东南大学 | Breaker fault diagnosis method based on separating/closing coil current signals |
CN105866669A (en) * | 2016-05-19 | 2016-08-17 | 华南理工大学 | Circuit breaker breaking-closing control loop fault diagnosis method |
CN108919104A (en) * | 2018-05-21 | 2018-11-30 | 国网江苏省电力有限公司检修分公司 | A kind of circuit breaker failure diagnostic method based on Fisher identification and classification method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8441768B2 (en) * | 2010-09-08 | 2013-05-14 | Schweitzer Engineering Laboratories Inc | Systems and methods for independent self-monitoring |
-
2019
- 2019-07-10 CN CN201910618553.6A patent/CN110244219B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008082896A2 (en) * | 2006-12-28 | 2008-07-10 | General Electric Company | Measurement of analog coil voltage and coil current |
CN103336243A (en) * | 2013-07-01 | 2013-10-02 | 东南大学 | Breaker fault diagnosis method based on separating/closing coil current signals |
CN105866669A (en) * | 2016-05-19 | 2016-08-17 | 华南理工大学 | Circuit breaker breaking-closing control loop fault diagnosis method |
CN108919104A (en) * | 2018-05-21 | 2018-11-30 | 国网江苏省电力有限公司检修分公司 | A kind of circuit breaker failure diagnostic method based on Fisher identification and classification method |
Non-Patent Citations (3)
Title |
---|
High-Voltage Circuit Breaker Fault Diagnosis Model Based on Coil Current and KNN;Tong Li;《2018 Prognostics and System Health Management Conference (PHM-Chongqing)》;20181028;全文 * |
基于多参量的高压断路器分_合闸线圈的故障诊断;靳文娟;《高压电器》;20190316;第55卷(第3期);全文 * |
高压断路器合_分_闸线圈电流在线监测***的研制;陈志英;《厦门理工学院学报》;20161031;第24卷(第5期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN110244219A (en) | 2019-09-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109188258A (en) | The high-voltage circuitbreaker feature extraction and classification method being electrically coupled based on vibration | |
Xiong et al. | Incipient fault identification in power distribution systems via human-level concept learning | |
CN104181460B (en) | A kind of permanent magnetic actuator vacuum circuit breaker method for diagnosing faults based on Multi-information acquisition | |
Gu et al. | High impedance fault detection in overhead distribution feeders using a DSP-based feeder terminal unit | |
CN106526468B (en) | Circuit-breaker status detection method based on wave character identification | |
CN104852346A (en) | Circuit testing closer apparatus | |
CN111458599A (en) | Series arc fault detection method based on one-dimensional convolutional neural network | |
Lu et al. | A morphological scheme for inrush identification in transformer protection | |
CN111382809B (en) | Isolating switch mechanical fault diagnosis method based on motor output power | |
Jiang et al. | Machine learning approach to detect arc faults based on regular coupling features | |
CN109828181A (en) | A kind of transformer winding minor failure detection method based on MODWT | |
CN109406949A (en) | Power distribution network incipient fault detection method and device based on support vector machines | |
CN108919104B (en) | Breaker fault diagnosis method based on Fisher discriminant classification method | |
CN107765065B (en) | Fundamental wave attenuation factor-based power distribution network magnetizing inrush current identification method | |
CN112083327A (en) | Mechanical fault diagnosis method and system for high-voltage vacuum circuit breaker | |
CN114757110B (en) | Circuit breaker fault diagnosis method based on sliding window detection and current extraction signals | |
CN110737996A (en) | high-voltage circuit breaker opening and closing coil current identification method | |
CN113376474A (en) | Neural network fault arc identification system and method based on generalized S transformation | |
CN113917294B (en) | Intelligent self-adaptive arc detection method based on wavelet decomposition and application device thereof | |
CN110244219B (en) | Breaker fault identification method based on closing coil current time domain statistical characteristics | |
CN113297786B (en) | Low-voltage fault arc sensing method based on semi-supervised machine learning | |
CN111079647A (en) | Circuit breaker defect identification method | |
CN115113038B (en) | Circuit breaker fault diagnosis method based on current signal phase space reconstruction | |
De et al. | A fuzzy ARTMAP fault classifier for impulse testing of power transformers | |
CN115728628A (en) | On-line monitoring and early warning method and device for circuit breaker control loop fault |
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 |