CN110059437B - GIS vibration signal characteristic quantity extraction method based on variation modal decomposition - Google Patents
GIS vibration signal characteristic quantity extraction method based on variation modal decomposition Download PDFInfo
- Publication number
- CN110059437B CN110059437B CN201910350197.4A CN201910350197A CN110059437B CN 110059437 B CN110059437 B CN 110059437B CN 201910350197 A CN201910350197 A CN 201910350197A CN 110059437 B CN110059437 B CN 110059437B
- Authority
- CN
- China
- Prior art keywords
- gis
- vibration signal
- imf
- frequency
- decomposition
- 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
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 80
- 238000000605 extraction Methods 0.000 title claims abstract description 9
- 239000013598 vector Substances 0.000 claims abstract description 32
- 241000282461 Canis lupus Species 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims description 49
- 238000005457 optimization Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 15
- 230000008859 change Effects 0.000 claims description 10
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 6
- 241000282421 Canidae Species 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 241000894007 species Species 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 4
- 229910018503 SF6 Inorganic materials 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- SFZCNBIFKDRMGX-UHFFFAOYSA-N sulfur hexafluoride Chemical compound FS(F)(F)(F)(F)F SFZCNBIFKDRMGX-UHFFFAOYSA-N 0.000 description 3
- 229960000909 sulfur hexafluoride Drugs 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The application discloses a GIS vibration signal characteristic quantity extraction method based on variation modal decomposition, which solves the problems that the GIS vibration signal is different in source, and useful characteristic information cannot be accurately obtained under the condition of different characteristics, and information loss caused by excessive decomposition exists. The application uses the gray wolf predation algorithm to optimize the parameters of the variation modal decomposition algorithm; decomposing the GIS mechanical vibration signal by using the optimized variation modal decomposition algorithm, and effectively extracting an optimal IMF component bearing rich characteristic information; extracting the sum of the maximum amplitude, root mean square value, kurtosis, 100Hz duty ratio, 50Hz odd frequency multiplication duty ratio and 100Hz frequency multiplication duty ratio of the optimal IMF component as a characteristic quantity, and constructing a characteristic vector; and through the comparison of the feature vectors under the normal and fault states, the diagnosis of GIS mechanical faults is realized.
Description
Technical Field
The application relates to the technical field of GIS fault detection, in particular to a GIS vibration signal characteristic quantity extraction method based on variation modal decomposition.
Background
The sulfur hexafluoride closed type combined electrical appliance, internationally called as gas insulated switchgear (Gas Insulated Switchgear) is called GIS for short, and organically combines primary equipment except a transformer in a transformer substation, including a breaker, a disconnecting switch, a grounding switch, a voltage transformer, a current transformer, a lightning arrester, a bus, a cable terminal, a wire inlet and outlet sleeve and the like into a whole through optimized design. The sulfur hexafluoride totally-enclosed Gas Insulated Switchgear (GIS) has the advantages of small occupied area, high reliability, strong safety, short installation period, small operation and maintenance workload and the like, is successfully popularized and applied worldwide from the appearance to the present, and is widely applied to substations of different grades in China.
Mechanical vibration phenomena with different degrees can be generated in the GIS operation process. The mechanical vibration is generated by various reasons, for example, when the equipment has mechanical defects such as unbalanced butt joint of a shell, abnormal contact of a switch contact, slight bending of a guide rod and the like, the equipment can generate mechanical motion under the action of the factors such as mechanical force of switch operation, alternating electric power generated by load current and the like. Especially when certain defects occur in the GIS, the GIS can generate abnormal vibration, the damage of the abnormal vibration to the GIS equipment body is very large, faults such as sulfur hexafluoride gas leakage, damage of a basin-type insulator or an insulating support column, suspension of a shell grounding point and the like can be possibly caused, and accidents can be possibly caused by long-term development. The GIS vibration signals are generated from different sources and have different characteristics. Therefore, a corresponding signal processing method is required to be used for extracting characteristic quantities with representative characteristics, analyzing signal characteristics of the characteristic quantities and deeply researching the characteristics of the GIS shell vibration signals.
The current methods for signal extraction mainly comprise Fourier transformation, wavelet decomposition, wavelet packet decomposition, empirical mode decomposition, improved empirical mode decomposition and the like. The Empirical Mode Decomposition (EMD) and the improved empirical mode decomposition (EEMD) become a relatively wide decomposition method used in the field of mechanical fault diagnosis by virtue of the self-adaptive characteristics. However, the method has a certain disadvantage when processing the GIS vibration fault signal, because the center frequency and the bandwidth of the signal frequency band carrying the fault information are not determined, if the frequency band carrying the fault signal is in the frequency band of the first decomposition component, excessive interference may be introduced due to the frequency band of the first decomposition component, and if the fault related frequency band is in the frequency band of the higher order decomposition component, important characteristic information may be omitted due to the too narrow frequency band of the component.
Disclosure of Invention
The technical problems to be solved by the application are as follows: the application provides a GIS vibration signal characteristic quantity extraction method based on variation modal decomposition, which solves the problems that the central frequency and the bandwidth of a signal frequency band carrying fault information are uncertain when a GIS vibration fault signal is processed by a conventional Empirical Mode Decomposition (EMD) method and an improved empirical mode decomposition (EEMD) method, so that excessive interference can be introduced due to the fact that the frequency band carrying the fault signal is in the frequency band of a first decomposition component, if the fault related frequency band is in the frequency band of a higher-order decomposition component, important characteristic information can be omitted due to the fact that the frequency band of the fault related frequency band is too narrow, the useful characteristic information which cannot be accurately obtained can not be obtained, information loss caused by excessive decomposition exists, and the like. And through the comparison of the feature vectors under the normal and fault states, the diagnosis of GIS mechanical faults is realized.
The application is realized by the following technical scheme:
a GIS vibration signal characteristic quantity extraction method based on variation modal decomposition comprises the following steps:
step 1: the reasonable GIS vibration signal acquisition scheme is designed, vibration signals of the GIS under normal and fault states are acquired, a section of GIS vibration signals is intercepted, and key parameters in the variation modal decomposition method are optimized by using a gray wolf predation algorithm; setting variation modal decomposition parameters according to the optimization result, decomposing the GIS vibration signal to obtain K modal components IMF, and selecting the component with the largest amplitude as the optimal IMF characteristic component;
step 2: extracting the maximum amplitude, root mean square value, kurtosis, 100Hz duty ratio, 50Hz odd frequency multiplication duty ratio and 100Hz frequency multiplication duty ratio of the optimal IMF characteristic components as characteristic quantities, and constructing characteristic vectors of GIS vibration signals; comparing the feature vector of the vibration signal in the same GIS fault state with the feature vector of the normal vibration signal to obtain the change trend of the fault feature quantity; and comparing the feature vector of the vibration signal corresponding to the GIS in the unknown operation state with the feature vector of the normal vibration signal to obtain the change trend of the feature vector of the GIS in the unknown operation state, and comparing the change trend of the feature vector of the GIS in the unknown operation state with the change trend of the fault feature vector to judge whether the GIS in the unknown operation state is abnormal or not.
After the GIS mechanical vibration signal is processed, the maximum amplitude, the root mean square value, the kurtosis, the 100Hz duty ratio, the 50Hz and odd frequency multiplication duty ratio thereof and the sum of the 100Hz frequency multiplication duty ratio of the optimal IMF characteristic component are extracted as characteristic quantities, characteristic vectors are constructed, and GIS fault discrimination is realized through characteristic vector comparison; reasonable IMF number can be decomposed according to signal characteristics through the variation modal decomposition optimized by the gray wolf predation algorithm, and the optimal IMF characteristic components bearing rich characteristic information can be effectively extracted; the characteristic vector is formed by the sum of the amplitude, the average value, the kurtosis, the 100Hz duty ratio, the 50Hz odd frequency multiplication duty ratio and the 100Hz frequency multiplication duty ratio of the optimal IMF characteristic components, and the frequency domain and time domain peak characteristics can be reflected simultaneously; the method can accurately extract the useful characteristic information and avoid information loss caused by excessive decomposition; meanwhile, the states of the GIS vibration signals are recognized by considering the time domain and the frequency domain multi-feature quantity of the GIS vibration signals, and the judgment accuracy is improved.
Further, aiming at the GIS vibration signal intercepted in the step 1, a gray wolf predation optimization algorithm is used for optimizing key parameters in the variation modal decomposition method, wherein the key parameters comprise a penalty factor alpha and the number K of modal components IMF components.
The optimization procedure of the variational modal decomposition (K, alpha) by the gray wolf predation optimization algorithm comprises the following steps:
(1) Initializing the gray wolf population, randomly generating the positions of n intelligent individuals, initializing various parameters in a gray wolf predation optimization algorithm, and determining an adaptability function in the optimizing process;
(2) Carrying out variation modal decomposition calculation on the signals at different species positions, and calculating the corresponding fitness value of each gray wolf position;
(3) Performing comparison analysis on the magnitude of each fitness value, and updating the optimal intelligent individual position of the gray wolves;
(4) And (3) carrying out loop iteration, returning to the step (2), and outputting the optimal fitness value to correspond to the position (K, alpha) of the wolf after the iteration times reach a preset value.
Further, in the step 1, according to the optimization result, setting variation modal decomposition parameters, and decomposing GIS vibration signals to obtain IMF components with corresponding numbers; the method specifically comprises the following steps:
the variation modal decomposition algorithm transfers the GIS vibration signal decomposition process into a variation frame, achieves GIS vibration signal self-adaptive decomposition by searching the constraint variation model optimal solution, further obtains IMF components, the frequency center and the bandwidth of each IMF component are continuously updated in the process of iteratively solving the variation model, and finally, the self-adaptive subdivision of the signal frequency band is completed according to the frequency domain characteristics of the actual signal, and a plurality of narrow-band IMF components are obtained.
The specific implementation process comprises the following steps: in the variational modal decomposition algorithm, the eigenmode function (IMF) is redefined as an amplitude-frequency modulated signal expressed as:
u k (t)＝A k (t)cos(φ k (t)) (1)
(1) Wherein: a is that k (t) is u k Instantaneous amplitude, ω, of (t) k (t) is u k (t) is the instantaneous frequency of the signal,A k (t) and ω k (t) relative phase phi k (t) is slowly varying, so u k (t) is regarded as an amplitude A k (t) frequency is ω k A harmonic signal of (t);
the variation modal decomposition algorithm is to transfer the GIS vibration signal decomposition process into a variation frame, realize GIS vibration signal self-adaptive decomposition by searching the constraint variation model optimal solution, further obtain IMF components, the frequency center and bandwidth of each IMF component are continuously updated in the process of iteratively solving the variation model, and finally, the self-adaptive subdivision of the signal frequency band can be completed according to the frequency domain characteristics of the actual signal, and a plurality of narrow-band IMF components are obtained. Assuming that the original GIS vibration signal is decomposed into k IMF components, the expression of the constraint variation model is:
(2) In the formula, { u k (t)}＝{u 1 ,...,u k -k IMF components resulting from the decomposition; { omega k (t)}＝{ω 1 ,...,ω k -the center frequency of each component; sigma (t) is a given one of the impact functions; f is the original input signal;is given as an estimated center frequency; />t represents time.
And, introduce the following form and enlarge Lagrange function to solve the above-mentioned constraint variation model problem optimal solution, the expression is:
(3) Wherein alpha is penalty factor, and lambda is Lagrange multiplier; sigma (t) is a given one of the impact functions; f (t) is the original input signal;is given as an estimated center frequency; />t represents time; u (u) k (t) is a amplitude A k (t) frequency is ω k (t) a harmonic signal.
And obtaining an optimal solution of the constraint variation model through the above method, so that the original GIS vibration signal is decomposed into K narrow-band IMF components.
The specific implementation process of decomposing the original GIS vibration signal into K narrow-band IMF components is as follows:
(1) Initialization ofλ 1 And n is 0, wherein->Eigenmode function representing the initial moment, +.>Represents the frequency, lambda, of the eigenmode function at the moment of origin 1 As Lagrange multiplier at initial moment, n is the number of loops, and the numbers of the superscripts below represent the number of iterations;
(2) n=n+1, performing the whole cycle;
(3) Executing the first cycle of the inner layer according toUpdating u k ；
(4) Repeating the step (3) with k=k+1, wherein K represents the number of IMF decomposition until k=k, and ending the first cycle of the inner layer;
(5) Performing a second cycle of the inner layer according toUpdating omega k ；
(6) Repeating step (5) for k=k+1, ending the second cycle of the inner layer until k=k;
(7) According toUpdating lambda;
(8) Repeating steps (2) to (7) until the iteration stop condition is satisfiedWherein epsilon represents an error allowable value, the whole cycle is ended, and a result is output to obtain K narrow-band IMF components.
Further, IMF components with the largest amplitude are selected from the K decomposed narrow-band IMF components to serve as optimal components, and the sum of the largest amplitude, the root mean square value, the kurtosis, the 100Hz duty ratio, the 50Hz and odd frequency multiplication duty ratio thereof and the 100Hz frequency multiplication duty ratio of the optimal IMF characteristic components is extracted to serve as the characteristic quantity.
The application has the following advantages and beneficial effects:
1. the gray wolf predation optimization algorithm adopted by the application can carry out combination optimization on punishment factors alpha and component number K in the variation modal decomposition algorithm according to the characteristics of the GIS mechanical vibration signals; the optimized variation modal decomposition algorithm can decompose the variation modal decomposition algorithm into preset K components according to the characteristics of the GIS vibration signals;
2. the vibration signal can be reasonably decomposed by utilizing the variation modal decomposition optimized by the gray wolf predation algorithm, and the optimal IMF characteristic component bearing rich characteristic information can be effectively extracted; the variation modal decomposition method decomposes the GIS vibration signal into a plurality of IMF components with different bandwidths from low to high, wherein the extracted optimal IMF is a component capable of reflecting the main frequency band of the GIS, and then the optimal IMF component is taken as an analysis object, so that the rest components are filtered, and the purpose of filtering high-frequency and low-frequency component interference is achieved; not only can the useful characteristic information be extracted correctly and the low frequency and high frequency interference be removed, but also the information loss caused by excessive decomposition is avoided; the mechanical state of the GIS can be more truly represented by extracting the time domain and frequency domain characteristics of the optimal IMF component;
3. the application provides a new thought for processing GIS mechanical vibration signals, which is an important supplement in the aspect of GIS mechanical vibration signal research;
4. the amplitude-average-kurtosis-100 Hz duty ratio-100 Hz frequency multiplication duty ratio sum-50 Hz and the odd frequency multiplication duty ratio feature vector thereof extracted by the method comprise the time domain and frequency domain characteristics of GIS mechanical vibration, and the application is favorable for completely reflecting signal features.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:
FIG. 1 is a flow chart of the method of the present application.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present application, the present application will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present application and the descriptions thereof are for illustrating the present application only and are not to be construed as limiting the present application.
Examples
As shown in fig. 1, a method for extracting feature quantity of a GIS vibration signal based on variation modal decomposition includes the following steps:
step 1: GIS mechanical vibration signal processing
Step 1.1: initializing each parameter of a gray wolf predation algorithm, and searching an optimal parameter combination (K, alpha) of variation modal decomposition, wherein a penalty factor alpha and the number K of modal component IMF components;
the method is simple in operation, few in setting parameters, strong in robustness, faster in convergence speed, higher in solving precision and good in global optimizing capability.
Specifically, the optimization procedure of the variational modal decomposition (K, α) by the wolf predation optimization algorithm comprises the following steps:
(1) Initializing the gray wolf population, randomly generating the positions of n intelligent individuals, initializing various parameters in a gray wolf predation optimization algorithm, and determining an adaptability function in the optimizing process;
(2) Carrying out variation modal decomposition calculation on the signals at different species positions, and calculating the corresponding fitness value of each gray wolf position;
(3) Performing comparison analysis on the magnitude of each fitness value, and updating the optimal intelligent individual position of the gray wolves;
(4) And (3) carrying out loop iteration, returning to the step (2), and outputting the optimal fitness value to correspond to the position (K, alpha) of the wolf after the iteration times reach a preset value.
Obtaining that the maximum iteration number of the gray wolf predation algorithm is 10 and the population scale is 20;
step 1.2: according to the set of optimal parameters (K 0 ，α 0 ) Setting the penalty factor alpha of the variational modal decomposition as alpha 0 The number of IMF components K is K 0 Analyzing the GIS vibration signal by using an optimized algorithm; intercepting a section of GIS vibration signal, and analyzing and decomposing the GIS vibration signal by using an optimized variation modal decomposition algorithm;
the variation modal decomposition method is characterized in that a GIS vibration signal decomposition process is transferred into a variation frame, GIS vibration signal self-adaptive decomposition is realized by searching for an optimal solution of a constraint variation model, so that IMF components are obtained, the frequency center and the bandwidth of each IMF component are continuously updated in the process of iteratively solving the variation model, and finally, the self-adaptive subdivision of a signal frequency band can be completed according to the frequency domain characteristics of an actual signal, and a plurality of narrow-band IMF components are obtained;
the optimized variation modal decomposition algorithm is used for analyzing and decomposing the GIS vibration signal more reasonably than the existing method, and the optimal IMF characteristic component bearing rich characteristic information can be effectively extracted; the variation modal decomposition method decomposes the GIS vibration signal into a plurality of IMF components with different bandwidths from low to high, wherein the extracted optimal IMF is a component capable of reflecting the main frequency band of the GIS, and then the optimal IMF component is taken as an analysis object, so that the rest components are filtered, and the purpose of filtering high-frequency and low-frequency component interference is achieved. Thus, not only can the useful characteristic information be extracted correctly, but also information loss caused by excessive decomposition is avoided effectively.
Step 1.3: according to the step 1.2, K is obtained after the optimized variational modal decomposition 0 And (3) IMF components, wherein the component with the largest amplitude is taken as the optimal component.
Step 2: GIS mechanical vibration signal characteristic quantity extraction and fault diagnosis
Step 2.1: obtaining the maximum amplitude, average value, kurtosis, 100Hz duty ratio, the sum of 100Hz frequency multiplication duty ratios, 50Hz and odd frequency multiplication duty ratio of the best component in the step 1.3 as characteristic quantity to form characteristic vector;
wherein the feature quantity in step 2.1 is defined as follows:
(1) Amplitude value: the GIS vibration signal is a periodic signal, the amplitude of the GIS vibration signal is an important parameter (g, gravity acceleration) reflecting the running state, and researches show that the amplitude of the GIS vibration signal can be changed when the GIS is abnormally operated.
(2) Average value: for GIS vibration signals, the mean value can reflect the symmetry degree (g, gravity acceleration) of the waveform, the standard sinusoidal signal mean value is 0, and the closer the vibration signal mean value is to 0, the better the symmetry of the positive half cycle and the negative half cycle of the signal is considered, and the degree of distortion is smaller.
(3) Kurtosis: kurtosis is a mathematical statistic reflecting waveform kurtosis, is a dimensionless parameter, and can describe the distribution characteristics of signals. Kurtosis is denoted by the letter K, which is defined as:
where μ is the mean value of the signal x, σ is the standard deviation of the signal x, and E (t) represents the expected value of the variable t. When the GIS state changes, the waveform kurtosis of the vibration signal also changes correspondingly.
(4) 100Hz duty cycle: when the GIS operates normally, the fundamental frequency is 100Hz, and the amplitudes of other frequency components are almost negligible. When there is a fault, the amplitude of other frequencies increases correspondingly, and the 100Hz duty cycle changes. Therefore, this parameter is taken as one parameter in the feature vector.
(5) Sum of frequency multiplication duty cycle of 100 Hz: when the GIS operates normally, the frequency multiplication component of 100Hz in the vibration signal frequency spectrum is very small, and when a fault occurs, the amplitude of the frequency multiplication component of 100Hz can be obviously increased.
(6) 50Hz and odd multiple duty cycle: when the GIS operates normally, the frequency spectrum of the vibration signal does not contain 50Hz and odd frequency multiplication components, and when faults occur, corresponding components can appear, so that the change of the duty ratio can also reflect the occurrence of the faults.
Step 2.2: and comparing the feature vector of the vibration signal under the same GIS fault state with the feature vector of the normal vibration signal to obtain the change trend of the fault feature quantity.
Step 2.3: for the GIS with unknown running state, the relative ratio of the vibration signal characteristic vector to the normal state can be extracted through the steps, and whether the vibration signal characteristic vector is abnormal or not can be judged.
The working principle is as follows: after the GIS mechanical vibration signal is processed, the maximum amplitude, the root mean square value, the kurtosis, the 100Hz duty ratio, the 50Hz and odd frequency multiplication duty ratio thereof and the sum of the 100Hz frequency multiplication duty ratio of the optimal IMF characteristic component are extracted as characteristic quantities, characteristic vectors are constructed, and GIS fault discrimination is realized through characteristic vector comparison; reasonable IMF number can be decomposed according to signal characteristics through the variation modal decomposition optimized by the gray wolf predation algorithm, and the optimal IMF characteristic components bearing rich characteristic information can be effectively extracted; the characteristic vector is formed by the sum of the amplitude, the average value, the kurtosis, the 100Hz duty ratio, the 50Hz odd frequency multiplication duty ratio and the 100Hz frequency multiplication duty ratio of the optimal IMF characteristic components, and the frequency domain and time domain peak characteristics can be reflected simultaneously; the optimized variation modal decomposition method decomposes the GIS vibration signal into a plurality of IMF components with different bandwidths from low to high, wherein the extracted optimal IMF is a component capable of reflecting the main frequency band of the GIS, and then the optimal IMF component is taken as an analysis object, so that the rest components are filtered out, and the purpose of filtering high-frequency and low-frequency component interference is achieved. Thus, not only can the useful characteristic information be extracted correctly and the low-frequency and high-frequency interference be removed, but also the information loss caused by excessive decomposition is avoided.
The method can accurately extract useful characteristic information, remove low-frequency and high-frequency interference and avoid information loss caused by excessive decomposition; meanwhile, the states of the GIS vibration signals are recognized by considering the time domain and the frequency domain multi-feature quantity of the GIS vibration signals, and the judgment accuracy is improved.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the application, and is not meant to limit the scope of the application, but to limit the application to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the application are intended to be included within the scope of the application.
Claims (4)
1. A GIS vibration signal characteristic quantity extraction method based on variation modal decomposition is characterized by comprising the following steps: the method comprises the following steps:
step 1: collecting vibration signals of a GIS in normal and fault states, intercepting a section of GIS vibration signals, and optimizing key parameters in a variation modal decomposition method by using a gray wolf predation algorithm; setting variation modal decomposition parameters according to the optimization result, decomposing the GIS vibration signal to obtain K modal components IMF, and selecting the component with the largest amplitude as the optimal IMF characteristic component;
step 2: extracting the maximum amplitude, root mean square value, kurtosis, 100Hz duty ratio, the sum of 100Hz frequency multiplication duty ratios, 50Hz and odd frequency multiplication duty ratio thereof of the optimal IMF characteristic components as characteristic quantities, and constructing characteristic vectors of GIS vibration signals; comparing the feature vector of the vibration signal in the same GIS fault state with the feature vector of the normal vibration signal to obtain the change trend of the fault feature quantity; comparing the feature vector of the vibration signal corresponding to the GIS in the unknown operation state with the feature vector of the normal vibration signal to obtain the change trend of the feature vector of the GIS in the unknown operation state, comparing the change trend of the feature vector of the GIS in the unknown operation state with the change trend of the fault feature vector, and judging whether the GIS in the unknown operation state is abnormal or not;
setting variation modal decomposition parameters according to an optimization result, and decomposing GIS vibration signals to obtain IMF components with corresponding numbers; the method specifically comprises the following steps:
the variation modal decomposition algorithm transfers the GIS vibration signal decomposition process into a variation frame, achieves GIS vibration signal self-adaptive decomposition by searching the constraint variation model optimal solution, further obtains IMF components, the frequency center and the bandwidth of each IMF component are continuously updated in the process of iteratively solving the variation model, and finally, the self-adaptive subdivision of a signal frequency band is completed according to the frequency domain characteristics of an actual signal, and a plurality of narrow-band IMF components are obtained;
in the variational modal decomposition algorithm, the eigenmode function IMF is redefined as an amplitude-frequency modulated signal, and the expression is:
u k (t)＝A k (t)cos(φ k (t))
wherein: a is that k (t) is u k Instantaneous amplitude, ω, of (t) k (t) is u k (t) is the instantaneous frequency of the signal,A k (t) and ω k (t) relative phase phi k (t) is slowly varying, so u k (t) is regarded as an amplitude A k (t) frequency is ω k A harmonic signal of (t);
the expression of the constraint variation model is:
in the formula, { u k (t)}＝{u 1 ,...,u k -k IMF components resulting from the decomposition; { omega k (t)}＝{ω 1 ,...,ω k -the center frequency of each component; sigma (t) is a given one of the impact functions; f is the original input signal;is given as an estimated center frequency; />t represents time;
and, introduce the following form and enlarge Lagrange function to solve the above-mentioned constraint variation model problem optimal solution, the expression is:
wherein alpha is penalty factor, and lambda is Lagrange multiplier; sigma (t) is a given one of the impact functions; f (t) is the original input signal;is given as an estimated center frequency; />t represents time; u (u) k (t) is a amplitude A k (t) frequency is ω k A harmonic signal of (t);
obtaining an optimal solution of the constraint variation model through the above method, so that an original GIS vibration signal is decomposed into K narrow-band IMF components;
the specific process for decomposing the original GIS vibration signal into K narrow-band IMF components comprises the following steps:
(1) Initializingλ 1 And n is 0, wherein->Eigenmode function representing the initial moment, +.>Represents the frequency, lambda, of the eigenmode function at the moment of origin 1 As Lagrange multiplier at initial time, n is number of cycles, and the following superscript numerals are usedIteration times;
(2) N=n+1, performing the whole cycle;
(3) Executing the first cycle of the inner layer according toUpdating u k ；
(4) Repeating the step (3) with k=k+1, wherein K represents the number of IMF decomposition until k=k, and ending the first cycle of the inner layer;
(5) Performing a second cycle of the inner layer according toUpdating omega k ；
(6) Repeating step (5) for k=k+1, ending the second cycle of the inner layer until k=k;
(7) According toUpdating lambda;
(8) Repeating steps (2) to (7) until the iteration stop condition is satisfiedWherein epsilon represents an error allowable value, the whole cycle is ended, and a result is output to obtain K narrow-band IMF components.
2. The method for extracting the GIS vibration signal characteristic quantity based on variation modal decomposition according to claim 1, wherein the method comprises the following steps: aiming at the GIS vibration signal intercepted in the step 1, a gray wolf predation optimization algorithm is used for optimizing key parameters in the variation modal decomposition method, wherein the key parameters comprise a penalty factor alpha and the number K of modal component IMF components.
3. The method for extracting the GIS vibration signal characteristic quantity based on variation modal decomposition according to claim 2, wherein the method comprises the following steps: the optimization procedure of the variational modal decomposition (K, alpha) by the gray wolf predation optimization algorithm comprises the following steps:
(1) Initializing the gray wolf population, randomly generating the positions of n intelligent individuals, initializing various parameters in a gray wolf predation optimization algorithm, and determining an adaptability function in the optimizing process;
(2) Carrying out variation modal decomposition calculation on the signals at different species positions, and calculating the corresponding fitness value of each gray wolf position;
(3) Performing comparison analysis on the magnitude of each fitness value, and updating the optimal intelligent individual position of the gray wolves;
(4) And (3) carrying out loop iteration, returning to the step (2), and outputting the optimal fitness value to correspond to the position (K, alpha) of the wolf after the iteration times reach a preset value.
4. The method for extracting the GIS vibration signal characteristic quantity based on variation modal decomposition according to claim 1, wherein the method comprises the following steps: and selecting an IMF component with the largest amplitude from the K decomposed narrow-band IMF components as an optimal component, and extracting the sum of the largest amplitude, root mean square value, kurtosis, 100Hz duty ratio, 50Hz and odd frequency multiplication duty ratio thereof and 100Hz frequency multiplication duty ratio of the optimal IMF characteristic component as a characteristic quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910350197.4A CN110059437B (en) | 2019-04-28 | 2019-04-28 | GIS vibration signal characteristic quantity extraction method based on variation modal decomposition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910350197.4A CN110059437B (en) | 2019-04-28 | 2019-04-28 | GIS vibration signal characteristic quantity extraction method based on variation modal decomposition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110059437A CN110059437A (en) | 2019-07-26 |
| CN110059437B true CN110059437B (en) | 2023-12-01 |
Family
ID=67321304
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910350197.4A Active CN110059437B (en) | 2019-04-28 | 2019-04-28 | GIS vibration signal characteristic quantity extraction method based on variation modal decomposition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110059437B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110441018B (en) * | 2019-08-14 | 2021-07-13 | 清华大学 | Time-frequency analysis method for fire impact response data |
| CN110569478B (en) * | 2019-09-12 | 2021-02-02 | 北京航空航天大学 | Improved variational modal decomposition method for encoder signal analysis |
| CN110646714A (en) * | 2019-09-26 | 2020-01-03 | 杭州电力设备制造有限公司 | Feature extraction method of ultrasonic partial discharge signal and related equipment |
| CN111160315B (en) * | 2020-01-03 | 2023-05-05 | 云南电网有限责任公司电力科学研究院 | Method and device for acquiring intrinsic frequency band of post insulator |
| CN111965456A (en) * | 2020-08-17 | 2020-11-20 | 红相股份有限公司 | Method for diagnosing mechanical fault of electrical equipment |
| CN112307950B (en) * | 2020-10-29 | 2024-04-23 | 重庆大学 | Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification |
| CN112307963A (en) * | 2020-10-30 | 2021-02-02 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Converter transformer running state identification method based on vibration signals |
| CN112557894B (en) * | 2020-12-07 | 2023-03-31 | 四川轻化工大学 | Breaker fault diagnosis method |
| CN113191524B (en) * | 2021-01-29 | 2023-07-11 | 河北建投新能源有限公司 | Wind speed prediction method, device, storage medium and processor |
| CN113625164A (en) * | 2021-08-02 | 2021-11-09 | 南京航空航天大学 | Aviation generator fault feature extraction method, system, medium and computing device |
| CN117370847B (en) * | 2023-12-08 | 2024-02-13 | 深圳宇翊技术股份有限公司 | Deep learning-based disconnecting switch detection method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106840637A (en) * | 2017-03-24 | 2017-06-13 | 国网山东省电力公司莱芜供电公司 | Based on the GIS mechanical oscillation signal Time-Frequency Analysis Methods for improving HHT algorithms |
| CN107084854A (en) * | 2017-04-17 | 2017-08-22 | 四川大学 | Self-adapting random resonant Incipient Fault Diagnosis method based on grey wolf optimized algorithm |
| CN107702908A (en) * | 2017-10-12 | 2018-02-16 | 国网山东省电力公司莱芜供电公司 | GIS mechanical oscillation signal Time-Frequency Analysis Methods based on VMD self adapting morphologies |
| CN108426715A (en) * | 2018-06-13 | 2018-08-21 | 福州大学 | Rolling bearing Weak fault diagnostic method based on PSO-VMD-MCKD |
| CN108663202A (en) * | 2018-05-03 | 2018-10-16 | 国家电网公司 | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system |
| CN108692936A (en) * | 2018-03-27 | 2018-10-23 | 四川大学 | Mechanical failure diagnostic method based on parameter adaptive VMD |
-
2019
- 2019-04-28 CN CN201910350197.4A patent/CN110059437B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106840637A (en) * | 2017-03-24 | 2017-06-13 | 国网山东省电力公司莱芜供电公司 | Based on the GIS mechanical oscillation signal Time-Frequency Analysis Methods for improving HHT algorithms |
| CN107084854A (en) * | 2017-04-17 | 2017-08-22 | 四川大学 | Self-adapting random resonant Incipient Fault Diagnosis method based on grey wolf optimized algorithm |
| CN107702908A (en) * | 2017-10-12 | 2018-02-16 | 国网山东省电力公司莱芜供电公司 | GIS mechanical oscillation signal Time-Frequency Analysis Methods based on VMD self adapting morphologies |
| CN108692936A (en) * | 2018-03-27 | 2018-10-23 | 四川大学 | Mechanical failure diagnostic method based on parameter adaptive VMD |
| CN108663202A (en) * | 2018-05-03 | 2018-10-16 | 国家电网公司 | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system |
| CN108426715A (en) * | 2018-06-13 | 2018-08-21 | 福州大学 | Rolling bearing Weak fault diagnostic method based on PSO-VMD-MCKD |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110059437A (en) | 2019-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110059437B (en) | GIS vibration signal characteristic quantity extraction method based on variation modal decomposition | |
| CN110068759B (en) | Fault type obtaining method and device | |
| CN107546855B (en) | Non-invasive decomposition method for residential electricity load | |
| CN109029960B (en) | Method for detecting mechanical state of circuit breaker | |
| Cai et al. | Intelligent DC arc-fault detection of solar PV power generation system via optimized VMD-based signal processing and PSO–SVM classifier | |
| CN108399147B (en) | Transformer excitation inrush current identification method based on MEEMD algorithm | |
| CN111428755A (en) | Non-invasive load monitoring method | |
| CN113269082A (en) | Partial discharge threshold denoising method based on improved variational modal decomposition | |
| Wang et al. | Regional nonintrusive load monitoring for low voltage substations and distributed energy resources | |
| Zhang et al. | Series AC arc fault diagnosis based on data enhancement and adaptive asymmetric convolutional neural network | |
| Jin et al. | Mode identification of low‐frequency oscillations in power systems based on fourth‐order mixed mean cumulant and improved TLS‐ESPRIT algorithm | |
| Mishra et al. | Loss of main detection in distribution generation system based on hybrid signal processing and machine learning technique | |
| Mishra et al. | Intrinsic time decomposition based fault location scheme for unified power flow controller compensated transmission line | |
| CN113949069B (en) | Method and system for determining transient voltage stability of high-proportion new energy power system | |
| CN114397569A (en) | Circuit breaker fault arc detection method based on VMD parameter optimization and sample entropy | |
| CN110174553A (en) | A kind of dense frequencies harmonic wave/harmonic detection method decomposed based on resolution modalities | |
| CN110794210B (en) | Method and device for judging voltage harmonic isolation effect, power supply system, computer equipment and storage medium | |
| Dinesh et al. | Non-intrusive load monitoring based on low frequency active power measurements | |
| CN111666870A (en) | Power signal reconstruction method and system by utilizing quadratic constraint | |
| CN109194367B (en) | Power signal reconstruction method in energy decomposition | |
| CN113030570A (en) | Harmonic electric energy detection method and adaptive filter | |
| CN109344814A (en) | A kind of GIS machine performance signal characteristic extracting methods | |
| Gonzalez-Cordoba et al. | Automatic detection and classification of electrical disturbances by means of empirical mode decomposition method | |
| CN110531149B (en) | Power signal filtering method and system based on waveform regularization | |
| Han et al. | Improvement of VMD for anomalous collision disturbance based on nonlinear l 1/2 norm |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200617 Address after: 637000 Nanchong city in Sichuan province Fu Road No. 228 Applicant after: NANCHONG POWER SUPPLY COMPANY, STATE GRID SICHUAN ELECTRIC POWER Co.,Ltd. Address before: 637000 Nanchong city in Sichuan province Fu Road No. 228 Applicant before: NANCHONG POWER SUPPLY COMPANY, STATE GRID SICHUAN ELECTRIC POWER Co.,Ltd. Applicant before: SICHUAN University |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |