CN113514784A - Method for detecting and judging mechanical state of transformer winding under no-load closing - Google Patents

Method for detecting and judging mechanical state of transformer winding under no-load closing Download PDF

Info

Publication number
CN113514784A
CN113514784A CN202110282781.8A CN202110282781A CN113514784A CN 113514784 A CN113514784 A CN 113514784A CN 202110282781 A CN202110282781 A CN 202110282781A CN 113514784 A CN113514784 A CN 113514784A
Authority
CN
China
Prior art keywords
transformer
whale
state
imf1
imf1 component
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
Application number
CN202110282781.8A
Other languages
Chinese (zh)
Other versions
CN113514784B (en
Inventor
李勇
许洪华
张勇
马宏忠
陈冰冰
刘宝稳
陈寿龙
朱昊
朱雷
王凌燕
王徐延
刘金品
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hohai University HHU
Nanjing Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
Hohai University HHU
Nanjing Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hohai University HHU, Nanjing Power Supply Co of State Grid Jiangsu Electric Power Co Ltd filed Critical Hohai University HHU
Priority to CN202110282781.8A priority Critical patent/CN113514784B/en
Publication of CN113514784A publication Critical patent/CN113514784A/en
Application granted granted Critical
Publication of CN113514784B publication Critical patent/CN113514784B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
    • G01R31/72—Testing of electric windings
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
    • G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
    • 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
    • G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
    • G01R31/62—Testing of transformers

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

The invention relates to a method for detecting and judging the mechanical state of a transformer winding under no-load closing, belonging to the technical field of safety monitoring of power transformers. Testing time domain vibration acceleration signals generated by transformers in four states in the same model under no-load switching-on, and decomposing the time domain vibration acceleration signals through WOA-VMD to obtain IMF1 component frequency spectrum peak values respectively at three frequencies of 100Hz, 200Hz and 400Hz and obtain four conditions of the transformers in the four states in the same model; and then, acquiring a time domain vibration acceleration signal of the transformer to be detected when the transformer is switched on in an idle state, decomposing the time domain vibration acceleration signal by WOA-VMD to obtain IMF1 component frequency spectrum peak values at three frequencies of 100Hz, 200Hz and 400Hz, and comparing the IMF1 component frequency spectrum peak values at the three frequencies of the transformer to be detected by taking the obtained four conditions as judgment conditions so as to judge that the mechanical state of a winding of the transformer to be detected belongs to one of the four states. The detection and judgment method provided by the invention has the advantages of timely, accurate, reliable, simple and feasible diagnosis.

Description

Method for detecting and judging mechanical state of transformer winding under no-load closing
Technical Field
The invention relates to a method for detecting and judging the mechanical state of a transformer winding under the condition of no-load closing. Belongs to the technical field of power transformer safety monitoring.
Background
The power transformer is used as an important device in a power system and plays roles of voltage transformation, electric energy distribution and electric energy transmission, and the normal operation of the power transformer is an important guarantee for the safe, reliable, high-quality and economic operation of the power system. Statistical analysis shows that a significant number of transformer faults result from winding loosening deformations. The loose deformation of the winding is used as the latent defect of the transformer winding, and early warning is an effective means for preventing the fault from being enlarged. At present, the diagnosis transformer winding has on-line methods such as a short-circuit reactance method and a steady-state current vibration method. The early warning capability of the characteristic off-line diagnosis method based on short-circuit fault disturbance in the operation of the transformer is poor, and the diagnosis is delayed; the transformer winding fault diagnosis method based on the steady-state vibration signals is weak in anti-interference capacity, and low in sensitivity and accuracy of fault diagnosis and fault degree identification.
The vibration signal of the transformer in operation is mainly the superposition of the iron core vibration and the winding vibration. When the transformer is impacted by excitation inrush current, because the amplitude of the impact current is far higher than the rated current, the vibration of the winding is far larger than the vibration of the iron core, and the vibration of the transformer is considered to be mainly caused by the vibration of the winding during the impact of the inrush current. The existing vibration analysis method detects the winding state of the transformer by measuring the vibration signal transmitted to the wall of the oil tank, but there is a lack of a method for judging the mechanical state of the winding by analyzing the vibration signal.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: and judging the mechanical state of the transformer winding by using the vibration signal of the transformer.
The technical scheme provided by the invention for solving the technical problems is as follows: a method for detecting and judging the mechanical state of a transformer winding under no-load closing executes the following steps:
1) the following four states of transformers of the same type are selected for testing,
firstly, 100% standard pressing force of the winding is in a first state of a normal mechanical state without looseness,
② the 90 percent standard pressing force of the winding is in a second state of 10 percent slight loose mechanical state,
third, the 70% standard pressing force of the winding is in a third state of a medium loose mechanical state of 30%,
fourthly, the 50 percent standard pressing force of the winding is in a fourth state of a serious loose mechanical state of 50 percent;
respectively arranging vibration acceleration sensors on the same type of transformer in the four states, and respectively collecting time domain vibration acceleration signals of the same type of transformer in the four states when the transformer is switched on in a no-load mode;
2) performing WOA-VMD decomposition on the time domain vibration acceleration signal acquired in the step 1) to obtain IMF1 component frequency spectrum peaks respectively at three frequencies of 100Hz, 200Hz and 400Hz, and obtaining four conditions of the transformer with the same model in the four states as follows:
A. in the first state, the 100Hz IMF1 component spectrum peak is more than 3 times of the 400Hz IMF1 component spectrum peak, or the 200Hz IMF1 component spectrum peak is more than 7 times of the 400Hz IMF1 component spectrum peak;
B. in the second state, the 100Hz IMF1 component spectrum peak is more than 5 times of the 400Hz IMF1 component spectrum peak, or the 200Hz IMF1 component spectrum peak is more than 7 times of the 400Hz IMF1 component spectrum peak; (ii) a
C. In the third state, the 400Hz IMF1 component spectral peak is greater than 4 times the 100Hz IMF1 component spectral peak or greater than 7 times the 200Hz IMF1 component spectral peak; (ii) a
D. In the fourth state, the 400Hz IMF1 component spectral peak is more than 25 times the 100Hz IMF1 component spectral peak or more than 14 times the 200Hz IMF1 component spectral peak;
3) for a transformer to be detected which is actually prepared to be put into operation, arranging a vibration acceleration sensor on the transformer to be detected, and collecting a vibration acceleration signal of the transformer to be detected when the transformer to be detected is switched on in a no-load state;
4) decomposing the vibration acceleration signal acquired in the step 3) by WOA-VMD to obtain IMF1 component frequency spectrum peak values at three frequencies of 100Hz, 200Hz and 400Hz, and comparing the IMF1 component frequency spectrum peak values at the three frequencies of the transformer to be detected by taking the four conditions obtained in the step 2) as judgment conditions to judge whether the mechanical state of the winding of the transformer to be detected belongs to one of the four states.
Further, the WOA-VMD decomposition specifically comprises the following steps:
optimizing the VMD parameter decomposition layer number k and the penalty factor alpha by using a whale algorithm, and setting initial values of k and alpha as 5 and 2000 respectively.
② initializing whale optimization algorithm population scale, iterationThe times and the self-adaptive weight value, IMF energy entropy are as the following formula (1), and a fitness function F is takenfitThe reciprocal of the mean value of the energy entropy of each IMF is shown in the following formula (2),
Figure BDA0002979247960000021
Figure BDA0002979247960000022
in the formula, EkIs the energy of the IMF component of the k-th order,
Figure BDA0002979247960000031
the energy of the k-th order IMF component accounts for the total energy proportion of the signal;
calculating the fitness value of each whale with a fixed head, comparing the fitness values and determining the whale with the optimal current fitness;
fourthly, entering an algorithm main loop, and updating the position according to the values of p and G;
evaluating the whole population and determining the globally optimal whale position;
sixthly, repeating the steps from the third step to the fifth step until the maximum iteration times is reached, and outputting the optimal k and alpha combination;
and (c) initializing the VMD by using the optimal k and alpha parameters, and decomposing the vibration signals into IMF1 component frequency spectrum peaks of three frequencies of 100Hz, 200Hz and 400 Hz.
Still further, the whale optimization algorithm is as follows:
1) assuming that the whale population scale is N, the position of the ith whale in the solution space of the D-dimension optimization problem to be solved is
Figure BDA0002979247960000032
The algorithm assumes the problem variable to be optimized and the position of the optimal whale (prey) solved by the problem variable;
2) in the prey surrounding stage, assuming that the current optimal post-selection solution is close to the optimal whale position, then other whales will automatically update the self position, and the position updating equation is as follows (3):
Figure BDA0002979247960000033
in the formula (3), D is the optimal solution candidate position, G and C are coefficients, and n is the element (1, n)max) Is the current iteration number, X*(n) is the optimal solution for the nth iteration, X (n) is the position of whale in the nth iteration, r ∈ [0,1 ∈]Is a random number, a decreases linearly from 2 to 0 as the number of iterations increases;
3) in the local search stage, a bounding mechanism or a spiral bubble mode is selected, and the position update equation of the local search is the following formula (4):
Figure BDA0002979247960000034
in the formula (4), p is [0,1 ]]Is a random number, D' ═ X*(n) -X (n) is the distance between the ith whale and the optimal whale, and is ∈ [ -1,1 [ ]]Is a random number, b is a spiral constant, e is the base of a natural logarithmic function;
4) in the global search phase, the algorithm is allowed to perform global search by using the convergence factor | G |, and when | G | > 1, the position update equation for performing global search is as follows (5):
D＝|CXrand(n)-X(n)|
X(n+1)＝Xrand(n)-AD (5)；
in the formula (5), XrandIs the position of a random whale in the current iteration population.
In the above retrieval scheme, the vibration acceleration sensors are arranged in three of A, B, C three phases of the transformer and are distributed at equal intervals between the high-voltage insulated terminals and the low-voltage insulated terminals of the transformer.
The invention has the beneficial effects that: for the collected time domain vibration acceleration signals of the transformer winding under no-load closing, IMF1 component frequency spectrum peaks respectively at three frequencies of 100Hz, 200Hz and 400Hz are obtained after decomposition through WOA-VMD (variation mode decomposition optimized by whale optimization algorithm), and the relation that the IMF1 component frequency spectrum peaks of the three frequencies respectively occupy the ratio is taken as the condition to be met or not, so that the judgment of the mechanical state of the transformer winding is realized. Compared with the existing detection and analysis method for the mechanical state of the transformer winding, the detection and judgment method provided by the invention has the advantages of timely diagnosis, accuracy, reliability, simplicity, convenience and feasibility.
Furthermore, it should be noted that: the invention judges the mechanical state of the transformer winding by measuring a time domain vibration acceleration signal of the transformer winding under no-load closing and decomposing the time domain vibration acceleration signal by a WOA-VMD algorithm to obtain an IMF1 component frequency spectrum peak value, and is based on the discovery of the inventor that the axial acceleration of the transformer winding under short circuit impact has internal relation with the power frequency. The inventor obtains an expression of the axial acceleration of the transformer winding under the short circuit impact with respect to the time t through theoretical derivation as the following formula (6),
Figure BDA0002979247960000041
in the formula (6), ayIs the axial acceleration of the winding, Ay、B1、B3、GyIs a dimensionless number, C'yIs damping coefficient, M is total mass of winding wire cake, gamma is initial phase angle of voltage during closing, omega0Is the power angular frequency at the time of closing, beta is the discontinuous angle of the magnetizing inrush current in one cycle, omega is the power angular frequency, T is the circuit time constant, theta is arctan [ (1-omega)2T2)/2ωT]And
Figure BDA0002979247960000042
is a function of ω, K'yIs the stiffness coefficient and e is the base of the natural logarithmic function.
From the equation (6), the vibration signal characteristic quantity (a) of the winding vibration characteristic (axial acceleration) when the transformer receives the inrush current impacty) Is an even harmonic of the supply frequency (ω). Thus, changes in the mechanical dynamics of the transformer winding caused by loosening or deformation of the transformer winding can be accelerated by vibration of the transformer winding under short-circuit impactThe degree signal is reflected.
The above formula (6) reflects a scientific theory discovery of the present inventor, and the formula derivation process is complex, and does not belong to the scope of patent, so the present invention only indicates the relationship between the winding vibration acceleration signal and the power supply frequency by the formula (6), and the detailed formula derivation process is not described herein.
Drawings
The following describes the method for detecting and determining the mechanical state of the transformer winding under no-load closing with reference to the accompanying drawings.
Fig. 1 is a schematic structural diagram of a vibration acceleration sensor mounted on a transformer in an embodiment.
Fig. 2 is a graph of vibration acceleration signals collected from phase 1 of a transformer a in the embodiment.
FIG. 3 is a graph of a winding vibration time-domain signal measured for phase 1 of transformer A and its IMF1 component time-domain signals of each order after WOA-VMD decomposition.
Fig. 4 is a frequency domain signal graph of the vibration acceleration signal in the first state after fast fourier transform.
Fig. 5 is a frequency domain signal graph of the vibration acceleration signal in the second state after fast fourier transform.
Fig. 6 is a frequency domain signal graph of the vibration acceleration signal in the third state after the fast fourier transform.
Fig. 7 is a frequency domain signal graph of the vibration acceleration signal in the fourth state after the fast fourier transform.
Detailed Description
Examples
The method for detecting and judging the mechanical state of the transformer winding under no-load closing comprises the following steps:
1) the following four states of transformers of the same type are selected for testing,
firstly, 100% standard pressing force of the winding is in a first state of a normal mechanical state without looseness,
② the 90 percent standard pressing force of the winding is in a second state of 10 percent slight loose mechanical state,
third, the 70% standard pressing force of the winding is in a third state of a medium loose mechanical state of 30%,
and fourthly, the standard pressing force of 50 percent of the winding is in a fourth state of a serious loose mechanical state of 50 percent.
Respectively arranging vibration acceleration sensors, and respectively collecting vibration acceleration signals of transformers in four states in the same model when the transformers are switched on in a no-load mode;
as shown in fig. 1, the vibration acceleration sensors 10 are respectively disposed on the same type of transformer in four states, and three vibration acceleration sensors, nine in total, respectively, reference numerals 1, 2, 3, 4, 5, 6, 7, 8 and 9 in the drawing are disposed on each phase of the transformer A, B, C, and the positions of the nine vibration acceleration sensors are equidistantly distributed between the high-voltage insulated terminal 11 and the low-voltage insulated terminal 12. After closing, respectively acquiring time domain vibration acceleration signals of the four transformers during no-load closing; for convenience of explanation, the vibration acceleration signal collected by one vibration acceleration sensor of phase a labeled 1 is taken as an example.
2) Decomposing the time domain vibration acceleration signals acquired in the step 1) by WOA-VMD to obtain IMF1 component frequency spectrum peaks respectively at three frequencies of 100Hz, 200Hz and 400Hz, decomposing the acquired vibration acceleration signals by WOA-VMD to obtain IMF1 component frequency spectrum peaks respectively at three frequencies of 100Hz, 200Hz and 400Hz as shown in figure 2, wherein WOA is a whale optimization algorithm, VMD is a variational modal decomposition, and WOA-VMD is a combination application of a whale optimization algorithm and a variational modal decomposition, and is concretely as follows:
optimizing the VMD parameter decomposition layer number k and the penalty factor alpha by using a whale algorithm, and setting initial values of k and alpha as 5 and 2000 respectively.
Secondly, initializing the population scale, the iteration times and the self-adaptive weight value of the whale optimization algorithm, wherein the IMF energy entropy is as the following formula (1), and taking a fitness function FfitThe reciprocal of the mean value of the energy entropy of each IMF is shown in the following formula (2),
HEk＝-pklgpk (1)
Figure BDA0002979247960000061
in the formula (1), EkIs the energy of the IMF component of the k-th order,
Figure BDA0002979247960000062
the energy of the k-th order IMF component accounts for the total energy proportion of the signal;
calculating the fitness value of each whale with a fixed head, comparing the fitness values and determining the whale with the optimal current fitness;
fourthly, entering an algorithm main loop, and updating the position according to the values of p and G;
evaluating the whole population and determining the globally optimal whale position;
sixthly, repeating the steps from the third step to the fifth step until the maximum iteration times is reached, and outputting the optimal k and alpha combination;
the VMD is initialized by the optimal k and alpha parameters, and the frequency spectrum peak values of IMF components of which the vibration signals are decomposed into 100Hz frequency, 200Hz frequency and 400Hz frequency are shown in the following table 1:
TABLE 1
Figure BDA0002979247960000071
From table 1, the following four cases can be derived:
A. for the first transformer, the peak value of 100Hz IMF1 component spectrum is more than 3 times of that of 400Hz IMF1 component spectrum, or the peak value of 200Hz IMF1 component spectrum is more than 7 times of that of 400Hz IMF1 component spectrum;
B. for the second transformer, the 100Hz IMF1 component spectrum peak value is more than 5 times of the 400Hz IMF1 component spectrum peak value, or the 200Hz IMF1 component spectrum peak value is more than 7 times of the 400Hz IMF1 component spectrum peak value; (ii) a
C. For the third transformer, the peak value of the IMF1 component spectrum at 400Hz is more than 4 times of the peak value of the IMF1 component spectrum at 100Hz or more than 7 times of the peak value of the IMF1 component spectrum at 200 Hz; (ii) a
D. For the fourth transformer, the 400Hz IMF1 component spectrum peak value is more than 25 times of the 100Hz IMF1 component spectrum peak value or more than 14 times of the 200Hz IMF1 component spectrum peak value.
3) For a to-be-detected transformer which is actually prepared to be put into operation, arranging a vibration acceleration sensor on the transformer, and collecting a vibration acceleration signal of the transformer when the transformer is switched on in a no-load state;
4) and (3) decomposing the vibration acceleration signal acquired in the step 3) by WOA-VMD to obtain IMF1 component frequency spectrum peak values at three frequencies of 100Hz, 200Hz and 400Hz, wherein the specific decomposition process is as above, and is not described herein again.
And taking the four conditions obtained in the step 2) as judgment conditions, and comparing the frequency spectrum peaks of the IMF1 components with three frequencies to judge which of the four transformers the mechanical state of the transformer winding to be detected belongs to. The peak value of the frequency spectrum of the IMF component for decomposing the vibration signal into 100Hz frequency, 200Hz frequency and 400Hz frequency in the embodiment is shown in the following table 2:
TABLE 2
Figure BDA0002979247960000081
As can be seen from comparison, the present embodiment meets the above-mentioned case B, and therefore, it can be determined that the mechanical state of the transformer winding detected by the present embodiment is slight loose.
FIG. 3 is a graph showing the vibration time domain signal of the transformer winding obtained from the phase-A1 measuring point of this embodiment and the time domain signal of each IMF component after WOA-VMD decomposition; fig. 4-7 are graphs showing frequency domain signals of time domain signals of IMF components of each order in four states after fast fourier transform.
The whale optimization algorithm in step 2) above is existing (refer to Mirjalili S, Lewis a. the hand optimization algorithm [ J ]. Advances in engineering software development 2016,95: 51-67). This embodiment gives the algorithm as follows:
1) assuming the whale population size is N, the ith in the solution space of the D-dimension optimization problem to be solvedThe whale is in the position of
Figure BDA0002979247960000082
The algorithm assumes the problem variable to be optimized and the position of the optimal whale (prey) solved by the problem variable;
2) in the prey surrounding stage, assuming that the current optimal post-selection solution is close to the optimal whale position, then other whales will automatically update the self position, and the position updating equation is as follows (3):
Figure BDA0002979247960000091
in the formula (3), D is the optimal solution candidate position, G and C are coefficients, and n is the element (1, n)max) Is the current iteration number, X*(n) is the optimal solution for the nth iteration, X (n) is the position of whale in the nth iteration, r ∈ [0,1 ∈]Is a random number, a decreases linearly from 2 to 0 as the number of iterations increases;
3) in the local search stage, a bounding mechanism or a spiral bubble mode is selected, and the position update equation of the local search is the following formula (4):
Figure BDA0002979247960000092
in the formula (4), p is [0,1 ]]Is a random number, D' ═ X*(n) -X (n) is the distance between the ith whale and the optimal whale, and is ∈ [ -1,1 [ ]]Is a random number, b is a spiral constant, e is the base of a natural logarithmic function;
4) in the global search phase, the algorithm is allowed to perform global search by using the convergence factor | G |, and when | G | > 1, the position update equation for performing global search is as follows (5):
Figure BDA0002979247960000093
in the formula (5), XrandIs the position of a random whale in the current iteration population.
Example two
The method for detecting and judging the mechanical state of the transformer winding under no-load closing in the embodiment is the same as the first embodiment, and is different from the first embodiment in that: the vibration acceleration signal obtained in the step 4) is decomposed by WOA-VMD to obtain IMF1 component frequency spectrum peaks at three frequencies of 100Hz, 200Hz and 400Hz as shown in the following Table 3:
TABLE 3
Figure BDA0002979247960000094
Figure BDA0002979247960000101
As can be seen from comparison, the present embodiment meets the above condition C, and therefore, it can be determined that the mechanical state of the transformer winding detected by the present embodiment is moderate loose.
The above description is only for the preferred embodiment of the present invention, but the present invention is not limited thereto, for example. All equivalents and modifications of the inventive concept and its technical solutions are intended to be included within the scope of the present invention.

Claims (4)

1. A method for detecting and judging the mechanical state of a transformer winding under no-load closing executes the following steps:
1) the following four states of transformers of the same type are selected for testing,
firstly, 100% standard pressing force of the winding is in a first state of a normal mechanical state without looseness,
② the 90 percent standard pressing force of the winding is in a second state of 10 percent slight loose mechanical state,
third, the 70% standard pressing force of the winding is in a third state of a medium loose mechanical state of 30%,
fourthly, the 50 percent standard pressing force of the winding is in a fourth state of a serious loose mechanical state of 50 percent;
respectively arranging vibration acceleration sensors on the same type of transformer in the four states, and respectively collecting time domain vibration acceleration signals of the same type of transformer in the four states when the transformer is switched on in a no-load mode;
2) performing WOA-VMD decomposition on the time domain vibration acceleration signal acquired in the step 1) to obtain IMF1 component frequency spectrum peaks respectively at three frequencies of 100Hz, 200Hz and 400Hz, and obtaining four conditions of the transformer with the same model in the four states as follows:
A. in the first state, the 100Hz IMF1 component spectrum peak is more than 3 times of the 400Hz IMF1 component spectrum peak, or the 200Hz IMF1 component spectrum peak is more than 7 times of the 400Hz IMF1 component spectrum peak;
B. in the second state, the 100Hz IMF1 component spectrum peak is more than 5 times of the 400Hz IMF1 component spectrum peak, or the 200Hz IMF1 component spectrum peak is more than 7 times of the 400Hz IMF1 component spectrum peak; (ii) a
C. In the third state, the 400Hz IMF1 component spectral peak is greater than 4 times the 100Hz IMF1 component spectral peak or greater than 7 times the 200Hz IMF1 component spectral peak; (ii) a
D. In the fourth state, the 400Hz IMF1 component spectral peak is more than 25 times the 100Hz IMF1 component spectral peak or more than 14 times the 200Hz IMF1 component spectral peak;
3) for a transformer to be detected which is actually prepared to be put into operation, arranging a vibration acceleration sensor on the transformer to be detected, and collecting a time domain vibration acceleration signal of the transformer to be detected when the transformer to be detected is switched on in a no-load state;
4) performing WOA-VMD decomposition on the time domain vibration acceleration signal acquired in the step 3) to obtain IMF1 component frequency spectrum peak values at three frequencies of 100Hz, 200Hz and 400Hz, and taking the four conditions obtained in the step 2) as judgment conditions to compare the IMF1 component frequency spectrum peak values at the three frequencies of the transformer to be detected so as to judge that the mechanical state of the winding of the transformer to be detected belongs to one of the four states.
2. The method of claim 1, further comprising: the WOA-VMD decomposition comprises the following specific steps:
optimizing the VMD parameter decomposition layer number k and the penalty factor alpha by using a whale algorithm, and setting initial values of k and alpha as 5 and 2000 respectively.
Secondly, initializing the whale optimization algorithm population scale, iteration times, adaptive weight values and IMF energy entropy
Figure FDA0002979247950000021
The following formula (1) and taking fitness function FfitFor each IMF energy entropy
Figure FDA0002979247950000022
The reciprocal of the average value is as follows (2),
Figure FDA0002979247950000023
Figure FDA0002979247950000024
in the formula (1), EkIs the energy of the k-th IMF component, pk＝Ek/E
Figure FDA0002979247950000025
The energy of the k-th order IMF component accounts for the total energy proportion of the signal;
calculating the fitness value of each whale with a fixed head, comparing the fitness values and determining the whale with the optimal current fitness;
fourthly, entering an algorithm main loop, and updating the position according to the values of p and G;
evaluating the whole population and determining the globally optimal whale position;
sixthly, repeating the steps from the third step to the fifth step until the maximum iteration times is reached, and outputting the optimal k and alpha combination;
and (c) initializing the VMD by using the optimal k and alpha parameters, and decomposing the vibration signals into IMF1 component frequency spectrum peaks of three frequencies of 100Hz, 200Hz and 400 Hz.
3. The method of claim 1, further comprising: the whale optimization algorithm is as follows:
1) assuming that the whale population scale is N, the position of the ith whale in the solution space of the D-dimension optimization problem to be solved is
Figure FDA0002979247950000026
The algorithm assumes the problem variable to be optimized and the position of the optimal whale (prey) solved by the problem variable;
2) in the prey surrounding stage, assuming that the current optimal post-selection solution is close to the optimal whale position, then other whales will automatically update the self position, and the position updating equation is as follows (3):
Figure FDA0002979247950000031
in the formula (3), D is the optimal solution candidate position, G and C are coefficients, and n is the element (1, n)max) Is the current iteration number, X*(n) is the optimal solution for the nth iteration, X (n) is the position of whale in the nth iteration, r ∈ [0,1 ∈]Is a random number, a decreases linearly from 2 to 0 as the number of iterations increases;
3) in the local search stage, a bounding mechanism or a spiral bubble mode is selected, and the position update equation of the local search is the following formula (4):
Figure FDA0002979247950000032
in the formula (4), p is [0,1 ]]Is a random number, D' ═ X*(n) -X (n) is the distance between the ith whale and the optimal whale, and is ∈ [ -1,1 [ ]]Is a random number, b is a spiral constant, e is the base of a natural logarithmic function;
4) in the global search phase, the algorithm is allowed to perform global search by using the convergence factor | G |, and when | G | > 1, the position update equation for performing global search is as follows (5):
D＝|CXrand(n)-X(n)|
X(n+1)＝Xrand(n)-AD (5)；
in the formula (5), XrandIs the position of a random whale in the current iteration population.
4. The method of claim 1, further comprising: the vibration acceleration sensors are three of A, B, C three phases arranged in the transformer and are distributed at equal intervals between high-voltage insulated terminals and low-voltage insulated terminals of the transformer.
CN202110282781.8A 2021-03-16 2021-03-16 Method for detecting and judging mechanical state of transformer winding under no-load closing Active CN113514784B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110282781.8A CN113514784B (en) 2021-03-16 2021-03-16 Method for detecting and judging mechanical state of transformer winding under no-load closing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110282781.8A CN113514784B (en) 2021-03-16 2021-03-16 Method for detecting and judging mechanical state of transformer winding under no-load closing

Publications (2)

Publication Number Publication Date
CN113514784A true CN113514784A (en) 2021-10-19
CN113514784B CN113514784B (en) 2023-09-15

Family

ID=78061763

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110282781.8A Active CN113514784B (en) 2021-03-16 2021-03-16 Method for detecting and judging mechanical state of transformer winding under no-load closing

Country Status (1)

Country Link
CN (1) CN113514784B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070078611A1 (en) * 2005-09-30 2007-04-05 Usa As Represented By The Administrator Of The National Aeronautics And Space Administration System and method of analyzing vibrations and identifying failure signatures in the vibrations
CN102721465A (en) * 2012-06-13 2012-10-10 江苏省电力公司南京供电公司 System and method for diagnosing and preliminarily positioning loosening faults of iron core of power transformer
CN102721464A (en) * 2012-06-13 2012-10-10 江苏省电力公司南京供电公司 Method and system for detecting winding deformation faults of power transformer

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070078611A1 (en) * 2005-09-30 2007-04-05 Usa As Represented By The Administrator Of The National Aeronautics And Space Administration System and method of analyzing vibrations and identifying failure signatures in the vibrations
CN102721465A (en) * 2012-06-13 2012-10-10 江苏省电力公司南京供电公司 System and method for diagnosing and preliminarily positioning loosening faults of iron core of power transformer
CN102721464A (en) * 2012-06-13 2012-10-10 江苏省电力公司南京供电公司 Method and system for detecting winding deformation faults of power transformer

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
王涛云;马宏忠;姜宁;李凯;许洪华;万达;崔杨柳;: "基于空载合闸振动信号的变压器绕组松动诊断", 中国电力, no. 05 *
陆启宇;李腾飞;王备;王丰华;: "基于自适应EEMD算法的变压器绕组状态检测仿真", 计算机仿真, no. 01 *

Also Published As

Publication number Publication date
CN113514784B (en) 2023-09-15

Similar Documents

Publication Publication Date Title
CN105067966B (en) The low-voltage alternating-current fault arc detection method of feature based modal components energy spectrometer
CN110221137B (en) Distribution transformer abnormal state detection method based on vibration-sound correlation
CN104597344A (en) Fault arc online detecting method based on wavelet first-layer high-frequency component correlation
CN112556828A (en) Transformer winding loosening fault detection method
CN111856166B (en) Detection method of box-type transformer and transformer
CN113805010A (en) Method and system for studying and judging single-phase earth fault of power distribution network
CN111487491A (en) Health state detection and evaluation system and method for oil immersed smoothing reactor
CN113391244B (en) VMD-based transformer closing vibration signal characteristic frequency calculation method
CN117233687B (en) CVT initial error assessment method, medium and terminal based on historical data
CN111856167A (en) Detection method of box-type transformer and transformer
CN108845074B (en) State evaluation method for oil-immersed transformer
CN113514784A (en) Method for detecting and judging mechanical state of transformer winding under no-load closing
CN113514783B (en) Method for detecting and judging mechanical state of transformer winding under no-load closing
Jiang et al. Characteristic Analysis and Optimization Strategy of Transformer Vibration Signal
CN112881839B (en) Transformer diagnosis method based on mutual information of frequency concentration and vibration stability
CN110286291B (en) Method and system for detecting vibration and sound of running state of transformer by using principal components
CN115308644A (en) Transformer winding fault detection method and system based on current offset ratio difference analysis
Tang et al. Series Arc Fault Identification Method Based on Lightweight Convolutional Neural Network
CN111929608A (en) Load switch event detection method and system
CN112098066A (en) High-voltage shunt reactor fault diagnosis method and system based on gate control circulation unit
CN106908735B (en) A kind of test method of charging pile energy-storage battery
CN116184141B (en) Gas insulation equipment fault diagnosis method and system
Xiao et al. Transformer Internal Mechanical Fault Diagnosis Method Based on Improved CEEMDAN and IMVO-HKELM
CN113608031B (en) Impact impedance monitoring method and device for substation lightning arrester
Liu et al. Fault simulation and identification of capacitive casings based on fault tree theory

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