CN110927490A - Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) - Google Patents
Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) Download PDFInfo
- Publication number
- CN110927490A CN110927490A CN201911212783.9A CN201911212783A CN110927490A CN 110927490 A CN110927490 A CN 110927490A CN 201911212783 A CN201911212783 A CN 201911212783A CN 110927490 A CN110927490 A CN 110927490A
- Authority
- CN
- China
- Prior art keywords
- svm
- winding
- cwaa
- ivmd
- transformer winding
- 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.)
- Pending
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
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
A transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM comprises the following steps: collecting vibration signals of a transformer winding in normal and fault states; and optimizing VMD related parameters by adopting PSO, calculating permutation entropy of IMF to extract characteristic quantity, taking the characteristic quantity as CWAA-SVM input, comparing the characteristic quantity with the traditional SVM and WOA-SVM, and comparing the prediction accuracy. The transformer winding fault diagnosis method can effectively extract the mechanical fault characteristics of the transformer winding, accurately diagnose and predict the transformer winding fault diagnosis, has obvious fault characteristics, strong operability, no false component problem, low calculation amount, universality of characteristic quantity, less occupation of processor resources of a diagnosis and prediction system and low power consumption of a processor of the diagnosis and prediction system, and has the technical effect superior to that based on the traditional SVM.
Description
Technical Field
The invention relates to the technical field of transformer winding diagnosis, in particular to a transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM.
Background
With the continuous development of electric power utilities in China, higher requirements are put forward on the stability of electric power equipment and the accuracy of state diagnosis. The power transformer is used as a core device in power transmission and transformation equipment, and the safe and stable operation of the power transformer is very important. Therefore, the state diagnosis technology of the power transformer is improved, latent faults inside the transformer are found in time, accidents are prevented, and the method has important significance for guaranteeing safe and stable operation of a power grid. The winding is the core component of the transformer, and the operation and maintenance of the transformer are directly affected by the state diagnosis technology level. In the prior art of the transformer winding mechanical state detection technology, the diagnosis technology research based on the transformer vibration signal is mainly divided into two major directions of simulation modeling and signal processing. The existing time-frequency analysis method applied to the field of transformer state diagnosis mainly comprises Hilbert-Huang transform based on empirical mode decomposition and time-frequency analysis based on wavelet decomposition. The time frequency spectrum has the main function of monitoring sudden change of a signal on a time axis, although a vibration signal can be changed when the state of the transformer is changed, the change of the state of the transformer is mostly accumulation of weak faults, the state change is a gradual change process, in a sampling period, transformer state information is contained in a periodic component of the vibration signal, and the frequency distribution rule of the periodic component of the vibration signal in the sampling period is not changed along with time. Meanwhile, although the existing research tests the vibration signals of the transformer at different positions, the independent analysis of the vibration signals at different test positions during the feature extraction ignores the characteristic difference of the vibration signals at different test positions of the transformer, so that the feature quantity may not have universality when the test positions are changed.
The method aims at the defects of the transformer winding vibration signal analysis method. The prior art uses the traditional WOA (Whaleoptimization Algorithm), but the method has the defects of low solution precision, slow convergence speed and easy falling into local optimization; there is also the use of EMD decomposition, but this method has a spurious component; there are other methods that attempt to solve the technical problem, but are computationally expensive, occupy a significant amount of resources of the processor of the diagnostic and prognostic system, and increase power consumption. Therefore, there is an urgent need in the art for a transformer winding fault diagnosis method that can accurately diagnose and predict transformer winding fault diagnosis, and has low computation amount, versatility of characteristic quantity, less occupation of processor resources of the diagnosis and prediction system, low power consumption of the processor of the diagnosis and prediction system, strong operability, no problem of false components, and is clear, reliable and simple.
Disclosure of Invention
In order to solve the defects in the prior art, the transformer winding fault diagnosis method based on the IVMD array entropy and the CWAA-SVM can accurately diagnose and predict transformer winding fault diagnosis, has obvious fault characteristics, simple result, strong operability, no false component problem, low computation amount, universality of characteristic quantity, less occupation of processor resources of a diagnosis and prediction system and low power consumption of a processor of the diagnosis and prediction system.
Specifically, the invention provides a transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) arrangement entropy and CWAA-SVM (continuous wave automatic-support vector machine), and introduces chaotic mapping into data processing aiming at the defect that the traditional WOA is easy to fall into local optimum. Firstly, the parameters of the VMD are optimized by utilizing the PSO, the arrangement entropy is used as a fitness function for optimization, the IVMD is applied to the extraction of the fault characteristics of the transformer winding and is used as the input of the CWAA-SVM to be compared with the WOA-SVM and the SVM. The method can effectively extract the fault characteristics of the transformer winding.
In order to achieve the aim, the invention provides a transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM, which adopts the following technical scheme:
loading an original signal of a transformer winding;
initializing signal sample data of a transformer;
optimizing the particle swarm;
VMD processing signals
Screening the optimal component;
extracting permutation entropy characteristic quantity;
and (4) performing training correct rate comparison on the CWAA-SVM input.
According to a further embodiment of the invention, a transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM is provided, which is characterized by comprising the following steps:
1) acquiring vibration signals of the winding in a normal state and vibration signals of the winding in a fault state through an acceleration sensor, and preprocessing the vibration signals;
2) based on the preprocessed vibration signals, the parameter L and the punishment coefficient α of the VMD are optimized by utilizing the PSO, so that the intervention of artificial subjective factors can be avoided, the parameter combination of the optimal component is automatically screened out, the acquired vibration signals are decomposed, and the fault characteristic quantity of the winding is extracted;
the IVMD comprises the following steps of optimizing parameters L and penalty coefficients α of a VMD (virtual machine model) by a PSO (particle swarm optimization) function according to a preprocessed vibration signal, using a permutation entropy as a fitness function and using the permutation entropy as a characteristic to enable VMD decomposition to be adaptive, wherein the improved VMD is called IVMD (improved variable mode decomposition), and the optimization comprises the following specific steps:
① assume that in a D-dimensional space, the population consisting of M particles is X ═ X (X)1,X2,X3,...,XD) The ith particle position is Xi=(Xi1,Xi2,Xi3,...,XiD) The velocity of the ith particle is Vi=(Vi1,Vi2,Vi3,...,ViD) A local extremum of a certain body being Pi=(pi1,pi2,pi3,…piD) Population global extreme G ═ G1,g2,...,gD) Each particle iteratively updates its own velocity and position through individual local extrema and population extrema, which are expressed as:
in the formula: w is the inertial weight; d ═ 1,2, …, D; 1,2, …, M; k is the current iteration number; c. C1And c2Is an acceleration factor, η is between [0,1 ]]Random number in between.
②, using the permutation entropy as a fitness function, representing the vibration time series of the transformer as [ x (H), H ═ 1, 2.., H ], and performing phase space reconstruction on the time series:
j=1,2,3,…K
in the formula, m and t are respectively an embedding dimension and a delay time, K ═ n- (m-1) t, and each row in the matrix can be used as a component after the time sequence of the transformer winding is reconstructed, and K reconstructed components are in total; rearranging the jth reconstruction component x (j), x (j + t), x (j + (m-1) t) in the winding time sequence X (h) in ascending order, j1,j2,...,jmThe index that represents the column in which each vector is located, namely:
x(i+(j1-1)t)≤x(i+(j2-1)t)≤…≤x(i+(jm-1)t)
therefore, for any vector consisting of time series data, a set of symbol sequences can be obtained:
S(l)=(j1,j2,…jm)
wherein l is 1,2, …, r, and r is less than or equal to m! M-dimensional phase space maps different symbol sequences (j)1,j2,…,jm) Total m! An arrangement of which the symbol sequence S (l) is one; calculating the probability P of each symbol sequence1,P2,...,Pr(ii) a At this time, the Permutation Entropy (PE) of the r different symbol sequences of the time series x (h) may be defined according to the shannon entropy, that is, the fitness function is:
usually will be Hp(m) performing a normalization process, namely:
Hpof valueSize means time series [ x (H), H ═ 1,2]Degree of randomness.
③ through step ①, optimize VMD with PSO function, find out parameter L and penalty coefficient α with the arrangement entropy as the fitness function of optimized VMD, introduce the optimized parameter into VMD for signal decomposition, optimize to obtain parameter VMD, find out the mode function u with the minimum sum of L estimated bandwidthsk(t), the sum of modes being the input signal f (t);
the variation constraint problem is as follows:
in the formula uk={u1,u2,···,uKIs the set of modal functions; omegak={ω1,ω2,···,ωKThe center frequency sets are used as the center frequencies;the partial derivative of the time t is calculated for the function; δ (t) is a unit pulse function; j is an imaginary unit; denotes convolution;
in order to solve the constraint optimization problem, a lagrange function ζ is introduced;
wherein α represents a penalty factor, λ (t) represents a Lagrange multiplier, and L represents the number of narrow-band IMF components.
3) And calculating the decomposition mode of the IVMD by using the permutation entropy, extracting the characteristic quantity, taking the characteristic quantity as the input quantity of the trained CWAA-SVM, and outputting the fault diagnosis result of the transformer winding.
In one embodiment, in step 1), the acceleration sensor is a piezoelectric acceleration sensor.
In one embodiment, in step 1), the sampling frequency is 25.6kHz when the transformer is subjected to vibration testing and signal acquisition.
In one embodiment, the position of the sensor in step 1) is placed on the top of the transformer.
In one embodiment, in step 2), optimizing the relevant parameters of the VMD using the PSO function includes optimizing two influencing parameters of the VMD in parallel.
In one embodiment, in step 3), multiple groups of data are respectively taken under four working conditions of normal winding, falling-off of a winding cushion block, loose winding and deformation of the winding of the transformer, part data in each group are used as samples, then the rest part data are used as test data, and the arrangement entropy obtained by the first four IMF component components is used as characteristic quantity.
In one embodiment, in step 3), for the working conditions of normal winding, falling-off of a winding cushion block, loose winding and deformation of the winding of the transformer, even-numbered groups of data are taken respectively under four working conditions, any data in each group with the number of half of the even-numbered groups is used as a sample, the rest data with the number of the other half of the even-numbered groups is used as test data, and the arrangement entropy obtained by the previous plurality of IMF components is used as a characteristic quantity to form a training group and a test group.
In one embodiment, in step 3), GWO-SVM and SVM are trained using the same data and the data are predicted.
In one embodiment, the method comprises the steps of respectively taking multiple groups of data under four working condition states of normal winding, falling-off of a winding cushion block, looseness of the winding and deformation of the winding of the transformer, using partial data in each group as samples, using the rest partial data as test data, and using arrangement entropy obtained by the first four IMF component components as characteristic quantities.
The invention achieves the following beneficial effects:
1. the transformer winding fault diagnosis method based on the IVMD permutation entropy and the CWAA-SVM is adopted, so that the operation amount can be reduced, the occupation of processor resources of a diagnosis and prediction system is reduced, and the power consumption of a processor of the diagnosis and prediction system is correspondingly reduced;
2. compared with an EMD decomposition method, the method has the advantages that the problem of false components does not exist, and the result is more accurate;
3. the method adopts the PSO to determine the modal number of the VMD, is clear, reliable and simple;
4. the feature quantity of the present invention has versatility.
Drawings
FIG. 1 is a flow chart of the transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM of the invention;
FIG. 2 is a graphical representation of K resulting from PSO optimized VMD of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following embodiments are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
FIG. 1 is a flow chart of the transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM of the present invention, which comprises the following steps;
loading an original signal of a transformer winding;
initializing signal sample data of a transformer;
optimizing the particle swarm;
VMD processing signals
Screening the optimal component;
extracting permutation entropy characteristic quantity;
and (4) performing training correct rate comparison on the CWAA-SVM input.
According to a further embodiment of the invention, the transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM comprises the following specific steps:
1) acquiring vibration signals of the winding in a normal state and vibration signals of the winding in a fault state through an acceleration sensor, and preprocessing the vibration signals;
preferably, when the position of the sensor in the step (1) is placed, the effect of placing the sensor on the top is better than the effect of placing the sensor on the side surface, so that the position of the sensor is arranged on the top.
2) Based on the preprocessed vibration signals, the parameter L and the penalty coefficient α of the VMD are optimized by utilizing PSO, specifically, two influence parameters of the VMD are optimized in parallel, so that the interference of artificial subjective factors can be avoided, the parameter combination of the optimal component is automatically screened out, the acquired vibration signals are decomposed, and the fault characteristic quantity of the winding is extracted, wherein the IVMD comprises the following steps:
for the preprocessed vibration signal, the PSO function optimizes the parameter L and penalty coefficient α of the VMD, uses the permutation entropy as the fitness function, and uses the permutation entropy as the feature to make the VMD decomposition adaptive, the improved VMD is called ivmd (improved variable mode decomposition), and the optimization specifically comprises the following steps:
① assume that in a D-dimensional space, the population consisting of M particles is X ═ X (X)1,X2,X3,...,XD) The ith particle position is Xi=(Xi1,Xi2,Xi3,...,XiD) The velocity of the ith particle is Vi=(Vi1,Vi2,Vi3,...,ViD) A local extremum of a certain body being Pi=(pi1,pi2,pi3,…piD) Population global extreme G ═ G1,g2,...,gD) Each particle iteratively updates its own velocity and position through individual local extrema and population extrema, which are expressed as:
in the formula: w is the inertial weight; d ═ 1,2, …, D; 1,2, …, M; k is the current iteration number; c. C1And c2Is an acceleration factor, η is between [0,1 ]]Random number in between.
②, using the permutation entropy as a fitness function, representing the vibration time series of the transformer as [ x (H), H ═ 1, 2.., H ], and performing phase space reconstruction on the time series:
j=1,2,3,…K
in the formula, m and t are respectively an embedding dimension and a delay time, K ═ n- (m-1) t, and each row in the matrix can be used as a component after the time sequence of the transformer winding is reconstructed, and K reconstructed components are in total; rearranging the jth reconstruction component x (j), x (j + t), x (j + (m-1) t) in the winding time sequence X (h) in ascending order, j1,j2,...,jmThe index that represents the column in which each vector is located, namely:
x(i+(j1-1)t)≤x(i+(j2-1)t)≤…≤x(i+(jm-1)t)
therefore, for any vector consisting of time series data, a set of symbol sequences can be obtained:
S(l)=(j1,j2,…jm)
wherein l is 1,2, r, and r is less than or equal to m! M-dimensional phase space maps different symbol sequences (j)1,j2,…,jm) Total m! An arrangement of which the symbol sequence S (l) is one; calculating the probability P of each symbol sequence1,P2,...,Pr(ii) a At this time, the Permutation Entropy (PE) of the r different symbol sequences of the time series x (h) may be defined according to the shannon entropy, that is, the fitness function is:
usually will be Hp(m) performing a normalization process, namely:
Hpthe size of the values indicates the time series [ x (H), H ═ 1,2]Degree of randomness.
③ through step ①, optimize VMD with PSO function, find out parameter L and punishment coefficient α with the arrangement entropy as the fitness function of optimized VMD, introduce the optimized parameter into VMD again for signal decomposition, optimize to obtain parameter VMD, find out the L estimated bandwidth with the maximum sumSmall mode function uk(t), the sum of modes being the input signal f (t);
the variation constraint problem is as follows:
in the formula uk={u1,u2,···,uKIs the set of modal functions; omegak={ω1,ω2,···,ωKThe center frequency sets are used as the center frequencies;the partial derivative of the time t is calculated for the function; δ (t) is a unit pulse function; j is an imaginary unit; denotes convolution;
in order to solve the constraint optimization problem, a lagrange function ζ is introduced;
wherein α represents a penalty factor, λ (t) represents a Lagrange multiplier, and L represents the number of narrow-band IMF components.
3) And calculating the decomposition mode of the IVMD by using the permutation entropy, extracting the characteristic quantity, taking the characteristic quantity as the input quantity of the trained CWAA-SVM, and outputting the fault diagnosis result of the transformer winding.
Preferably, the process of training the CWOA-SVM is as follows:
① chaotic initialization is carried out on the WOA population through Logistic chaotic mapping, namely CGWOA, thereby avoiding the WOA from falling into global optimum;
②, because the kernel function of the radial basis function only needs to determine a parameter delta, the RBF is selected as the function, and the optimal delta and C are optimized by CWA;
where C is a penalty factor, i.e., tolerance to errors. The higher C indicates that the error is less tolerable and is easily overfitted. The smaller C, the easier it is to under-fit. If C is too large or too small, the generalization ability is poor.
In the formula, gamma is a parameter carried by the RBF function after the RBF function is selected as the kernel. The distribution of the data after being mapped to a new feature space is determined implicitly, the larger the gamma is, the fewer the support vectors are, and the smaller the gamma value is, the more the support vectors are. The number of support vectors affects the speed of training and prediction.
③ samples are input into the CWAA-SVM, part trained, part tested, accuracy calculated, and compared to other SVM's.
Preferably, the process of training the CWOA-SVM further comprises: and feeding the result back to the CWAA-SVM for adaptation according to the comparison with other SVM.
In one embodiment of the invention, a PCB356A16 piezoelectric acceleration sensor and an NI9234 data acquisition instrument are adopted to carry out vibration test on 1 model transformer model S11-M-10/10, and the sampling frequency is 25.6 kHz. The method comprises the steps of respectively taking 30 groups of data under four working conditions of normal transformer winding, falling-off of a winding cushion block, looseness of the winding and deformation of the winding, using any 15 groups of data in each group as a sample, then using the remaining 15 groups of data as test data, using the arrangement entropy obtained by the first 4 IMF components as a characteristic quantity, and being equivalent to 60 groups of training groups and 60 groups of testing groups. Table 1 shows the training set data prediction results. The table 1 shows that the prediction results of the CWAA-SVM are compared with the prediction results of the SVM and the WOA-SVM, and the transformer winding fault diagnosis method based on the IVMD and the CWAA-SVM can effectively identify fault types, wherein the prediction accuracy is respectively improved by 7% -35.7%, 6.7% -29.8%, 10.02% -30.22% and 8.66% -25.88% aiming at four working conditions of normal transformer winding, falling-off winding cushion blocks, loose winding and winding deformation.
TABLE 1
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (10)
1. A transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM is characterized by comprising the following steps:
loading an original signal of a transformer winding;
initializing signal sample data of a transformer;
optimizing the particle swarm;
VMD processing signals
Screening the optimal component;
extracting permutation entropy characteristic quantity;
and (4) performing training correct rate comparison on the CWAA-SVM input.
2. A transformer winding fault diagnosis method based on IVMD arrangement entropy and CWAA-SVM is characterized by comprising the following steps:
1) acquiring and loading vibration signals of a winding in a normal state and vibration signals of a winding in a fault state through an acceleration sensor, and preprocessing the acquired and loaded vibration signals;
2) based on the preprocessed vibration signals, two parameters of the VMD are optimized by utilizing a PSO function, the collected vibration signals are decomposed, and the fault characteristic quantity of the winding is extracted:
for the preprocessed vibration signals, a PSO function optimizes parameters L and punishment coefficients α of the VMD, the arrangement entropy is used as a fitness function, the arrangement entropy is used as a characteristic to enable the VMD decomposition to have adaptivity, and the improved VMD is called IVMD;
3) and calculating the decomposition mode of the IVMD by using the permutation entropy, extracting the characteristic quantity, taking the characteristic quantity as the input quantity of the trained CWAA-SVM, and outputting the fault diagnosis result of the transformer winding.
3. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 2, wherein the step of optimizing in the step 2) comprises the following steps:
① assume that in a D-dimensional space, the population consisting of M particles is X ═ X (X)1,X2,X3,...,XD) The ith particle position is Xi=(Xi1,Xi2,Xi3,...,XiD) The velocity of the ith particle is Vi=(Vi1,Vi2,Vi3,...,ViD) A local extremum of a certain body being Pi=(pi1,pi2,pi3,…piD) Population global extreme G ═ G1,g2,...,gD) Each particle iteratively updates its own velocity and position through individual local extrema and population extrema, which are expressed as:
in the formula: w is the inertial weight; d ═ 1,2, …, D; 1,2, …, M; k is the current iteration number; c. C1And c2Is an acceleration factor, η is between [0,1 ]]Random number in between.
4. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 3, wherein the step of optimizing in the step 2) further comprises the following steps:
②, using the permutation entropy as a fitness function, representing the vibration time series of the transformer as [ x (H), H ═ 1, 2.., H ], and performing phase space reconstruction on the time series:
in the formula, m and t are respectively an embedding dimension and a delay time, K ═ n- (m-1) t, and each row in the matrix can be used as a component after the time sequence of the transformer winding is reconstructed, and K reconstructed components are in total; rearranging the jth reconstruction component x (j), x (j + t), x (j + (m-1) t) in the winding time sequence X (h) in ascending order, j1,j2,...,jmThe index that represents the column in which each vector is located, namely:
x(i+(j1-1)t)≤x(i+(j2-1)t)≤…≤x(i+(jm-1)t)
therefore, for any vector consisting of time series data, a set of symbol sequences can be obtained:
S(l)=(j1,j2,…jm)
wherein l is 1,2, r, and r is less than or equal to m! M-dimensional phase space maps different symbol sequences (j)1,j2,…,jm) Total m! An arrangement of which the symbol sequence S (l) is one; calculating the probability P of each symbol sequence1,P2,...,Pr(ii) a At this time, the Permutation Entropy (PE) of the r different symbol sequences of the time series x (h) may be defined according to the shannon entropy, that is, the fitness function is:
usually will be Hp(m) performing a normalization process, namely:
Hpthe size of the values indicates the time series [ x (H), H ═ 1,2]Degree of randomness.
5. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 4, wherein the step of optimizing in the step 2) further comprises the following steps:
③ through step ①, optimize VMD with PSO function, find out parameter L and penalty coefficient α with the arrangement entropy as the fitness function of optimized VMD, introduce the optimized parameter into VMD for signal decomposition, optimize to obtain parameter VMD, find out the mode function u with the minimum sum of L estimated bandwidthsk(t), the sum of modes being the input signal f (t);
the variation constraint problem is as follows:
in the formula uk={u1,u2,···,uKIs the set of modal functions; omegak={ω1,ω2,···,ωKThe center frequency sets are used as the center frequencies;the partial derivative of the time t is calculated for the function; δ (t) is a unit pulse function; j is an imaginary unit; denotes convolution;
in order to solve the constraint optimization problem, a lagrange function ζ is introduced;
wherein α represents a penalty factor, λ (t) represents a Lagrange multiplier, and L represents the number of narrow-band IMF components.
6. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 5, wherein: in step 2), optimizing the relevant parameters of the VMD by using the PSO function comprises performing parallel optimization on two influence parameters of the VMD.
7. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 5, wherein: in the step 3), multiple groups of data are respectively taken under four working conditions of normal winding, falling-off of a winding cushion block, looseness of the winding and deformation of the winding of the transformer, the partial data in each group is used as a sample, then the rest partial data is used as test data, and the arrangement entropy obtained by the first four IMF component components is used as a characteristic quantity.
8. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 6 or 7, wherein: in the step 3), even-numbered groups of data are taken under four working conditions of normal winding, falling-off of a winding cushion block, looseness of the winding and deformation of the winding of the transformer respectively, arbitrary data in each group, the number of which is half of that of the even-numbered groups, are used as samples, the remaining data in the other half of the groups, the number of which is the even-numbered groups, are used as test data, and the arrangement entropy obtained by the previous IMF components is used as characteristic quantity, so that a training group and a test group are formed.
9. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 8, wherein: in the step 1), a piezoelectric acceleration sensor is adopted as the acceleration sensor.
10. The method for diagnosing the fault of the transformer winding based on the IVMD arrangement entropy and the CWAA-SVM as claimed in claim 9, wherein: in the step 1), when the transformer is subjected to vibration test and signal acquisition, the sampling frequency is 25.6 kHz.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911212783.9A CN110927490A (en) | 2019-12-02 | 2019-12-02 | Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911212783.9A CN110927490A (en) | 2019-12-02 | 2019-12-02 | Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110927490A true CN110927490A (en) | 2020-03-27 |
Family
ID=69848083
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911212783.9A Pending CN110927490A (en) | 2019-12-02 | 2019-12-02 | Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110927490A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111751763A (en) * | 2020-06-08 | 2020-10-09 | 武汉大学 | Power transformer winding fault diagnosis method based on GSMallat-NIN-CNN network |
CN112101144A (en) * | 2020-08-27 | 2020-12-18 | 云南电网有限责任公司昭通供电局 | Self-adaptive method for improving transformer vibration signal processing precision |
CN112215279A (en) * | 2020-10-12 | 2021-01-12 | 国网新疆电力有限公司 | Power grid fault diagnosis method based on immune RBF neural network |
CN112557966A (en) * | 2020-12-02 | 2021-03-26 | 国网江苏省电力有限公司南京供电分公司 | Transformer winding looseness identification method based on local mean decomposition and support vector machine |
CN113203471A (en) * | 2021-05-07 | 2021-08-03 | 国网山西省电力公司电力科学研究院 | Transformer mechanical fault detection method based on wavelet neural network prediction |
CN113514783A (en) * | 2021-03-16 | 2021-10-19 | 国网江苏省电力有限公司南京供电分公司 | Method for detecting and judging mechanical state of transformer winding under no-load closing |
CN113987949A (en) * | 2021-11-03 | 2022-01-28 | 燕山大学 | Plate strip steel deformation resistance prediction method based on data driving |
CN115859201A (en) * | 2022-11-22 | 2023-03-28 | 淮阴工学院 | Chemical process fault diagnosis method and system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105717422A (en) * | 2015-12-04 | 2016-06-29 | 国家电网公司 | High-voltage power equipment partial discharge feature extraction method and apparatus |
CN105758644A (en) * | 2016-05-16 | 2016-07-13 | 上海电力学院 | Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy |
CN108205698A (en) * | 2017-06-15 | 2018-06-26 | 广东工业大学 | A kind of cloud resource load predicting method based on the double string whale optimization algorithms of just remaining chaos |
CN108760316A (en) * | 2018-08-16 | 2018-11-06 | 苏州大学 | Information fusion method is joined in the change of variation mode decomposition |
CN109580222A (en) * | 2018-12-04 | 2019-04-05 | 河北科技大学 | Based on variation mode decomposition-transfer entropy bearing degradation state recognition prediction technique |
CN110441018A (en) * | 2019-08-14 | 2019-11-12 | 清华大学 | Firer's shock response data Time-Frequency Analysis Method |
CN110516831A (en) * | 2019-06-18 | 2019-11-29 | 国网(北京)节能设计研究院有限公司 | A kind of short-term load forecasting method based on MWOA algorithm optimization SVM |
-
2019
- 2019-12-02 CN CN201911212783.9A patent/CN110927490A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105717422A (en) * | 2015-12-04 | 2016-06-29 | 国家电网公司 | High-voltage power equipment partial discharge feature extraction method and apparatus |
CN105758644A (en) * | 2016-05-16 | 2016-07-13 | 上海电力学院 | Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy |
CN108205698A (en) * | 2017-06-15 | 2018-06-26 | 广东工业大学 | A kind of cloud resource load predicting method based on the double string whale optimization algorithms of just remaining chaos |
CN108760316A (en) * | 2018-08-16 | 2018-11-06 | 苏州大学 | Information fusion method is joined in the change of variation mode decomposition |
CN109580222A (en) * | 2018-12-04 | 2019-04-05 | 河北科技大学 | Based on variation mode decomposition-transfer entropy bearing degradation state recognition prediction technique |
CN110516831A (en) * | 2019-06-18 | 2019-11-29 | 国网(北京)节能设计研究院有限公司 | A kind of short-term load forecasting method based on MWOA algorithm optimization SVM |
CN110441018A (en) * | 2019-08-14 | 2019-11-12 | 清华大学 | Firer's shock response data Time-Frequency Analysis Method |
Non-Patent Citations (5)
Title |
---|
任刚: "改进变分模态分解在柴油机故障诊断中的应用", 《军事交通学院学报》 * |
唐贵基 等: "参数优化变分模态分解方法在滚动轴承早期故障诊断中的应用", 《西安交通大学学报》 * |
张栋良: "VMD参数优化及其在轴承故障特征提取中的应用", 《北京理工大学学报》 * |
杨宇: "《机械故障诊断的变量预测模式识别方法》", 31 August 2017, 湖南大学出版社 * |
楚剑雄: "基于变分模态分解能量熵和支持向量机的电力变压器绕组故障诊断", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111751763A (en) * | 2020-06-08 | 2020-10-09 | 武汉大学 | Power transformer winding fault diagnosis method based on GSMallat-NIN-CNN network |
CN111751763B (en) * | 2020-06-08 | 2021-08-10 | 武汉大学 | Power transformer winding fault diagnosis method based on GSMallat-NIN-CNN network |
CN112101144A (en) * | 2020-08-27 | 2020-12-18 | 云南电网有限责任公司昭通供电局 | Self-adaptive method for improving transformer vibration signal processing precision |
CN112215279A (en) * | 2020-10-12 | 2021-01-12 | 国网新疆电力有限公司 | Power grid fault diagnosis method based on immune RBF neural network |
CN112215279B (en) * | 2020-10-12 | 2024-02-02 | 国网新疆电力有限公司 | Power grid fault diagnosis method based on immune RBF neural network |
CN112557966A (en) * | 2020-12-02 | 2021-03-26 | 国网江苏省电力有限公司南京供电分公司 | Transformer winding looseness identification method based on local mean decomposition and support vector machine |
CN113514783A (en) * | 2021-03-16 | 2021-10-19 | 国网江苏省电力有限公司南京供电分公司 | Method for detecting and judging mechanical state of transformer winding under no-load closing |
CN113514783B (en) * | 2021-03-16 | 2023-12-15 | 国网江苏省电力有限公司南京供电分公司 | Method for detecting and judging mechanical state of transformer winding under no-load closing |
CN113203471A (en) * | 2021-05-07 | 2021-08-03 | 国网山西省电力公司电力科学研究院 | Transformer mechanical fault detection method based on wavelet neural network prediction |
CN113987949A (en) * | 2021-11-03 | 2022-01-28 | 燕山大学 | Plate strip steel deformation resistance prediction method based on data driving |
CN113987949B (en) * | 2021-11-03 | 2024-05-07 | 燕山大学 | Data-driven-based plate strip steel deformation resistance prediction method |
CN115859201A (en) * | 2022-11-22 | 2023-03-28 | 淮阴工学院 | Chemical process fault diagnosis method and system |
CN115859201B (en) * | 2022-11-22 | 2023-06-30 | 淮阴工学院 | Chemical process fault diagnosis method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110927490A (en) | Transformer winding fault diagnosis method based on IVMD (integrated virtual machine direction) permutation entropy and CWAA-SVM (continuous wave operation-support vector machine) | |
CN111487046B (en) | Fault diagnosis method for circuit breaker voiceprint and vibration entropy feature fusion | |
CN110132554B (en) | Rotary machine fault diagnosis method based on deep Laplace self-coding | |
CN104807534B (en) | Equipment eigentone self study recognition methods based on on-line vibration data | |
CN109948194B (en) | High-voltage circuit breaker mechanical defect integrated learning diagnosis method | |
CN111077386A (en) | Early fault signal noise reduction method for electrical equipment | |
CN112083328A (en) | Fault diagnosis method, system and device for high-voltage circuit breaker | |
CN111695452A (en) | Parallel reactor internal aging degree evaluation method based on RBF neural network | |
CN112486137A (en) | Method and system for constructing fault feature library of active power distribution network and fault diagnosis method | |
CN114237046B (en) | Partial discharge pattern recognition method based on SIFT data feature extraction algorithm and BP neural network model | |
CN114662386A (en) | Bearing fault diagnosis method and system | |
CN114046903A (en) | Real-time temperature prediction method and system for high-voltage cable core | |
CN109682892A (en) | A kind of signal based on time frequency analysis removes drying method | |
CN116933152B (en) | Wave information prediction method and system based on multidimensional EMD-PSO-LSTM neural network | |
CN117290679A (en) | Running state detection method and device of current transformer and electronic equipment | |
CN114897032B (en) | Current transformer fault diagnosis method and device based on width learning | |
CN114778141B (en) | Method, device and equipment for formulating durability test of automobile electric drive assembly rack | |
CN108334822B (en) | Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics | |
CN113391244B (en) | VMD-based transformer closing vibration signal characteristic frequency calculation method | |
CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
CN111352813A (en) | Distributed storage use capacity estimation method and device | |
CN114329905B (en) | Method and device for evaluating reliability of full-range analog machine and computer equipment | |
Jingdong et al. | Fault Diagnosis of CRDM Based on EEMD-Wavelet Threshold Denoising and WOA-KELM | |
CN117708563A (en) | Vibration prediction method suitable for valve rod of valve under deep peak shaving working condition and related system | |
CN117093854B (en) | Transformer mechanical fault diagnosis method, equipment and storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200327 |
|
RJ01 | Rejection of invention patent application after publication |