CN113933658B - Dry-type transformer discharge detection method and system based on audible sound analysis - Google Patents
Dry-type transformer discharge detection method and system based on audible sound analysis Download PDFInfo
- Publication number
- CN113933658B CN113933658B CN202110997908.4A CN202110997908A CN113933658B CN 113933658 B CN113933658 B CN 113933658B CN 202110997908 A CN202110997908 A CN 202110997908A CN 113933658 B CN113933658 B CN 113933658B
- Authority
- CN
- China
- Prior art keywords
- frequency
- signal
- mel
- dry
- frame
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 47
- 238000004458 analytical method Methods 0.000 title claims abstract description 40
- 230000005236 sound signal Effects 0.000 claims abstract description 43
- 230000003068 static effect Effects 0.000 claims abstract description 26
- 238000001228 spectrum Methods 0.000 claims description 63
- 230000006870 function Effects 0.000 claims description 28
- 238000000354 decomposition reaction Methods 0.000 claims description 24
- 238000000034 method Methods 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000003595 spectral effect Effects 0.000 claims description 9
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 7
- 230000009191 jumping Effects 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000009432 framing Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 abstract description 7
- 230000002159 abnormal effect Effects 0.000 abstract description 3
- 230000007613 environmental effect Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 17
- 230000003287 optical effect Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000000605 extraction Methods 0.000 description 5
- 238000009413 insulation Methods 0.000 description 5
- 238000000691 measurement method Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 238000009434 installation Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000005672 electromagnetic field Effects 0.000 description 1
- 239000003822 epoxy resin Substances 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000002923 metal particle Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000009659 non-destructive testing Methods 0.000 description 1
- 238000005375 photometry Methods 0.000 description 1
- 229920000647 polyepoxide Polymers 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000002834 transmittance Methods 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/62—Testing of transformers
-
- 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
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The application discloses a dry-type transformer discharge detection method and a system based on audible sound analysis, wherein the dry-type transformer discharge detection method based on audible sound analysis comprises the steps of obtaining an audio signal when a detected dry-type transformer operates; extracting high-frequency signal components from the audio signal; extracting features of the high-frequency signal components to obtain high-frequency static features and high-frequency dynamic features; and judging whether the detected dry-type transformer has a discharge condition or not based on the high-frequency dynamic characteristics and the high-frequency static characteristics. The application aims to realize discharge characteristic detection of the dry-type transformer, has no damage to a test object, has strong environmental adaptability, fully considers the non-stationarity of the abnormal state signal of the transformer, combines the dynamic and static characteristics of the signal, has stronger identification and better fault tolerance, can more effectively and accurately reflect the discharge signal characteristics, and improves the precision of discharge fault detection.
Description
Technical Field
The application relates to a fault diagnosis technology of equipment in the power industry, in particular to a dry-type transformer discharge detection method and system based on audible sound analysis.
Background
The power transformer is the link most closely related to power consumers in the power system. And the operation fault of the power transformer is a key cause of large-area power failure of the power system. The operational life of a transformer depends on the state of its insulating material, which must cause the transformer to be shut down once it fails. Therefore, the insulation condition of the transformer is in a close and inseparable relationship with the partial discharge.
Epoxy resin of a dry-type transformer can cause defects such as burrs, air gaps, metal particles, poor contact and the like in insulation due to production technology limitation or improper operation in transportation, installation, operation and maintenance processes, and charge aggregation is easy to occur at the defect part, so that the electric field distribution in the transformer is uneven, and further partial discharge phenomenon is caused. The partial discharge can further damage the insulation performance, quicken the insulation aging, and then aggravate the partial discharge to form vicious circle until the insulation performance of the transformer is completely lost. The development of the discharge fault diagnosis of the power transformer has very important practical significance for guaranteeing the safe operation of the power grid.
At present, in order to detect the partial discharge condition of the dry-type transformer, the partial discharge detection method can be divided into an electrical measurement method and a non-electrical measurement method according to the types of state parameters. In the electrical measurement method, for example, a high-frequency detection method, an ultrahigh-frequency detection method and the like, the position where the sensor needs to be installed is inconvenient or corona interference is not suitable for discharge fault detection of the dry-type transformer. Therefore, the non-electrical measurement method is more suitable for the dry-type transformer.
The electroless method is mainly based on optical or acoustic detection. The photometry is to detect the light wave radiated when partial discharge occurs to realize the detection of the discharge fault of the transformer. The optical measurement method relies on the fact that various optical sensors are placed in the transformer, and because optical signals cannot be interfered by corona, noise and the like, the method has high detection sensitivity and can achieve partial discharge positioning. However, the transformer has a complex structure and poor light transmittance, and therefore effective detection of partial discharge inside the winding or outside the viewing distance of the optical sensor cannot be realized. The optical sensor is required to be installed inside the transformer, is not suitable for the put-into-operation transformer, and the installation position and the installation mode of the optical sensor can influence the detection performance of the optical sensor and the safety level of the transformer.
In acoustic-based transformer discharge fault detection, the most practical use is ultrasonic detection. The ultrasonic sensor fixed on the wall of the transformer tank is used for receiving ultrasonic signals generated by partial discharge, and the detection frequency range is 20kHz-200kHz, so that the positioning effect can be realized. However, ultrasonic waves are sensitive to vibration signals, and vibration of the transformer in a strong electromagnetic field causes greater attenuation of acoustic signals. In addition, the transformer is complex in structure, so that a great amount of refraction and reflection exist in the acoustic wave signal in the propagation process, and the detection sensitivity of the method is low.
Disclosure of Invention
The application aims to solve the technical problems: aiming at the problems in the prior art, the application provides a dry-type transformer discharge detection method and a system based on audible sound analysis, which aim to realize discharge characteristic detection of a dry-type transformer, have no damage to a test object and strong environmental adaptability, fully consider the non-stationarity of abnormal state signals of the transformer, combine the dynamic and static characteristics of the signals, have stronger identification and better fault tolerance, can more effectively and accurately reflect the characteristics of the discharge signals, and improve the precision of discharge fault detection.
In order to solve the technical problems, the application adopts the following technical scheme:
a dry-type transformer discharge detection method based on audible sound analysis, comprising:
1) Acquiring an audio signal of the detected dry-type transformer during operation;
2) Extracting high-frequency signal components from the audio signal;
3) Extracting features of the high-frequency signal components to obtain high-frequency static features and high-frequency dynamic features;
4) And judging whether the detected dry-type transformer has a discharge condition or not based on the high-frequency dynamic characteristics and the high-frequency static characteristics.
Optionally, extracting the high frequency signal component in step 2) for the audio signal means: and performing empirical mode decomposition on the audio signal, and taking a first eigenmode function obtained by the empirical mode decomposition as a high-frequency signal component in the obtained high-frequency signal.
Optionally, the performing empirical mode decomposition on the audio signal includes:
2.1 The audio signal at the original current time t is taken as a current audio signal x (t);
2.2 Finding all the upper and lower extreme points of the current audio signal x (t);
2.3 Finding the envelope e between all the upper and lower extreme points of the current audio signal x (t) max(t) and emin (t) obtaining the envelope e between the upper and lower extreme points max(t) and emin The average value between (t) is used for obtaining a mean envelope curve m (t), and the mean envelope curve m (t) is subtracted from the current audio signal x (t) to obtain an intermediate signal h (t);
2.4 Judging whether the intermediate signal h (t) is an eigenmode function, if so, jumping to the next step; otherwise, taking the intermediate signal h (t) as a new current audio signal x (t), and executing the step 2.2 in a jumping manner;
2.5 A residual signal r is obtained by subtracting the sum of all intermediate signals h (t) obtained by subtracting the audio signal at the original current time t n (t) judging the residual Signal r n (t) whether a monotonic sequence or a constant sequence is established, and if not, determining the residual signal r n (t) as a new current audio signal x (t), performing step 2.2) in a jump; otherwise, judging that the empirical mode decomposition is finished.
Optionally, when feature extraction is performed on the high-frequency signal component in step 3), obtaining the high-frequency static feature therein refers to calculating mel-frequency cepstrum coefficients of the high-frequency signal component as the high-frequency static feature.
Optionally, the step of calculating mel-frequency cepstral coefficients of the high-frequency signal component includes:
3.1 Pre-emphasis processing the high frequency signal component to raise the high frequency part;
3.2 Frame-dividing the pre-emphasis processed high-frequency signal component, and dividing the longer sound signal into a plurality of small segments;
3.3 Windowing is carried out on the result after the framing processing, so that the condition that signals are discontinuous at two ends of each frame of signal is avoided;
3.4 Performing short-time Fourier transform on each frame of data X (i, n) subjected to windowing to obtain frequency domain data X (i, k), wherein i represents a sequence number of a frame, the frame data X (i, n) is a signal sequence, n represents nth data in the signal sequence, and k represents a sequence number of a kth spectral line;
3.5 Calculating the energy E (i, m) of the frequency domain data X (i, k) passing through the Mel filter;
in the above formula, N represents the number of frames, H m (k) The M-th filter is a Mel filter, M is the number of groups of Mel filters;
3.6 Taking the logarithm of the frequency domain data X (i, k) through the energy E (i, m) of the Mel filter, performing discrete cosine transform, and calculating Mel frequency cepstrum coefficient based on the following formula;
in the above expression, MFCC (i, n) represents a mel-frequency cepstrum coefficient, i represents a frame number, n represents a spectral line after discrete cosine transform, M is the number of sets of mel filters, E (i, M) is the energy of the frequency domain data X (i, k) passing through the mel filters, M is the mel-frequency cepstrum coefficient number, and L is the dimension of the mel-frequency cepstrum coefficient.
Optionally, when feature extraction is performed on the high-frequency signal component in step 3), obtaining the high-frequency dynamic feature refers to calculating a dynamic differential spectrum of the mel frequency cepstrum coefficient as the high-frequency dynamic feature.
Optionally, the dynamic differential spectrum for calculating the mel-frequency cepstrum coefficient includes a first-order differential spectrum and a second-order differential spectrum, and the function expression for calculating the first-order differential spectrum of the mel-frequency cepstrum coefficient is:
the expression of the calculation function for calculating the second-order differential spectrum of the mel frequency cepstrum coefficient is as follows:
wherein delta (i) represents the ith row of the first-order differential spectrum matrix, corresponding to the spectrum information of the ith frame, i represents the frame number, coeffs (i) represents the ith row of the Mel frequency cepstrum coefficient, corresponding to the ith frame, coeffs (i-1) represents the ith-1 row of the Mel frequency cepstrum coefficient, corresponding to the ith-1 frame; delta (i) represents the ith row of the second order differential spectrum matrix, corresponding to the ith frame, and Delta (i-1) represents the ith-1 row of the first order differential spectrum matrix, corresponding to the ith-1 frame.
Optionally, in step 4), determining whether the detected dry transformer has a discharge condition based on the high-frequency dynamic characteristic and the high-frequency static characteristic means that: and generating a characteristic spectrum image by the obtained high-frequency dynamic characteristics and the high-frequency static characteristics, comparing the characteristic spectrum image with the characteristic spectrum image under the normal state of the detected dry-type transformer, and judging whether the detected dry-type transformer has a discharge condition or not if the fluctuation amplitude and the fluctuation quantity change quantity respectively exceed corresponding preset threshold degrees.
In addition, the application also provides a dry-type transformer discharge detection system based on audible sound analysis, which comprises a microprocessor and a memory which are connected with each other, wherein the microprocessor is programmed or configured to execute the steps of the dry-type transformer discharge detection method based on audible sound analysis.
Furthermore, the present application provides a computer readable storage medium having stored therein a computer program programmed or configured to perform the dry transformer discharge detection method based on audible sound analysis.
Compared with the prior art, the application has the following advantages:
1. the application detects the discharge characteristics of the dry-type transformer based on audible sound analysis, and can achieve the effect of nondestructive testing of a test object by measuring the discharge characteristics of the dry-type transformer through non-contact measurement without a load effect, thereby solving the problem that the transformer is damaged in the current transformer test.
2. The method has the advantages that the detection object (the detected dry-type transformer) is detected based on the audio signal during operation, the environmental adaptability is high, and compared with the traditional dry-type transformer discharge detection method, the audio acquisition mode enables detection to be realized in a severe environment.
3. The application extracts the high-frequency signal component from the audio signal as an analysis object, fully considers the non-stationarity of the abnormal state signal of the transformer, has more definite physical meaning and has stronger pertinence to the discharge condition.
4. The method comprises the steps of extracting the characteristics of the high-frequency signal component, obtaining the high-frequency dynamic characteristics and the high-frequency static characteristics, combining the dynamic and static characteristics of the signals, having stronger identification and better fault tolerance, being capable of reflecting the characteristics of the discharge signals more effectively and accurately and improving the accuracy of discharge fault detection.
Drawings
FIG. 1 is a schematic diagram of a core flow chart of a method according to an embodiment of the present application.
FIG. 2 is a schematic diagram of a complete process of the method according to the embodiment of the application.
Fig. 3 is a time domain and frequency domain diagram of an audio signal according to an embodiment of the present application.
Fig. 4 is an IMF diagram obtained by empirical mode decomposition in an embodiment of the present application.
Fig. 5 is a two-dimensional plot of mel-frequency cepstrum coefficients in an embodiment of the application.
Fig. 6 is a three-dimensional plot of mel-frequency cepstrum coefficients in an embodiment of the application.
Fig. 7 is a first order difference spectrum of mel-frequency cepstrum coefficients in an embodiment of the application.
Fig. 8 is a first-order differential spectral dimension diagram of mel-frequency cepstrum coefficients in an embodiment of the application.
Fig. 9 is a second order differential spectrum of mel-frequency cepstrum coefficients in an embodiment of the application.
Fig. 10 is a second order differential spectral dimension diagram of mel-frequency cepstrum coefficients in an embodiment of the application.
FIG. 11 is a graph showing the first-order difference spectrum of the mel-frequency cepstrum coefficient according to the embodiment of the present application.
Fig. 12 is a graph showing the second order difference spectrum of mel-frequency cepstrum coefficients in the embodiment of the present application.
Fig. 13 is a schematic structural diagram of an apparatus according to an embodiment of the present application, wherein reference numerals include: 1. an acquisition module; 2. a processor; 21. a signal processing module; 22. a signal analysis module; 23. a feature operation module; 24. a feature analysis module; 3. and a display screen.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the application; all other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
As shown in fig. 1, the dry-type transformer discharge detection method based on audible sound analysis of the present embodiment includes:
1) Acquiring an audio signal of the detected dry-type transformer during operation;
2) Extracting high-frequency signal components from the audio signal;
3) Extracting features of the high-frequency signal components to obtain high-frequency static features and high-frequency dynamic features;
4) And judging whether the detected dry-type transformer has a discharge condition or not based on the high-frequency dynamic characteristics and the high-frequency static characteristics.
In this embodiment, the method further includes a step of performing noise reduction processing on the audio signal after the step 1) and before the step 2), and through the noise reduction processing, the anti-interference capability of the method of this embodiment can be effectively improved, and the accuracy of discharge detection of the dry-type transformer can be improved.
In this embodiment, extracting the high-frequency signal component in the audio signal in step 2) means: and performing empirical mode decomposition on the audio signal, and taking a first eigenmode function obtained by the empirical mode decomposition as a high-frequency signal component in the obtained high-frequency signal. Eigenmode functions (Intrinsic Mode Function, IMF), i.e. the signal components obtained after decomposition. One eigenmode function must fulfil the following two conditions: the method includes the steps that the number of local extreme points and zero crossing points of a function in the whole time range must be equal or at most differ by one; the envelope of the local maxima (upper envelope) and the envelope of the local minima (lower envelope) must be zero on average at any point in time. Since each eigenmode function obtained by empirical mode decomposition is a component arranged from high frequency to low frequency, and the difference between the discharge fault and the normal state is mainly represented by a high frequency part, the first eigenmode function signal obtained by empirical mode decomposition is analyzed to have pertinence to the discharge fault, the physical meaning of the first eigenmode function corresponds to the high frequency signal component of the original signal and is used for highlighting the high frequency characteristic generated by discharge, and therefore the first eigenmode function obtained by empirical mode decomposition is selected to be subjected to characteristic extraction. In this embodiment, the time domain and frequency domain diagrams of the audio signal are shown in fig. 3, and fig. 4 is a schematic diagram of each eigenmode function (IMF) obtained by empirical mode decomposition.
The empirical mode decomposition (Empirical Mode Decomposition, EMD) performs signal decomposition according to the time scale characteristics of the data itself, and adaptively performs signal principal component analysis without presetting any basis functions. The time-frequency analysis method based on EMD is suitable for the analysis of nonlinear and non-stationary signals and the analysis of linear and stationary signals, and the analysis of the linear and stationary signals reflects the physical significance of the signals better than other time-frequency analysis methods. In this embodiment, performing empirical mode decomposition on an audio signal includes:
2.1 The audio signal at the original current time t is taken as a current audio signal x (t);
2.2 Finding all the upper and lower extreme points of the current audio signal x (t);
2.3 Finding the envelope e between all the upper and lower extreme points of the current audio signal x (t) max(t) and emin (t) obtaining the envelope e between the upper and lower extreme points max(t) and emin The average value between (t) is used for obtaining a mean envelope curve m (t), and the mean envelope curve m (t) is subtracted from the current audio signal x (t) to obtain an intermediate signal h (t);
2.4 Judging whether the intermediate signal h (t) is an eigenmode function, if so, jumping to the next step; otherwise, taking the intermediate signal h (t) as a new current audio signal x (t), and executing the step 2.2 in a jumping manner;
2.5 A residual signal r is obtained by subtracting the sum of all intermediate signals h (t) obtained by subtracting the audio signal at the original current time t n (t) judging the residual Signal r n (t) whether a monotonic sequence or a constant sequence is established, and if not, determining the residual signal r n (t) as a new current audio signal x (t), performing step 2.2) in a jump; otherwise, judging that the empirical mode decomposition is finished.
Based on the above empirical mode decomposition process, the audio signal x (t) at the original current time t can be decomposed into a series of eigenmode functions and linear superposition of the rest, and the usable functions are expressed as:
in the above formula, hi (t) represents the intermediate signal h (t), r obtained in the ith iteration n And (t) is the final residual signal.
In this embodiment, when feature extraction is performed on the high-frequency signal component in step 3), obtaining the high-frequency static feature therein refers to calculating mel-frequency cepstrum coefficient of the high-frequency signal component as the high-frequency static feature. Analysis of mel-frequency cepstral coefficient (MFCC) is based on human auditory mechanism, i.e., analysis of the spectrum of sound according to human auditory test results, and is currently used in research on audible sound recognition. In this embodiment, the step of calculating the mel-frequency cepstrum coefficient of the high-frequency signal component includes:
3.1 Pre-emphasis processing is performed on the high-frequency signal component to raise the high-frequency part, for example, a high-pass filter is used to raise the high-frequency part in the embodiment, and a filter or a processing method for raising the high-frequency part in other families can be also used as required;
3.2 Frame-dividing the pre-emphasis processed high-frequency signal component, and dividing the longer sound signal into a plurality of small segments;
3.3 Windowing is carried out on the result after the framing processing, so that the condition that signals are discontinuous at two ends of each frame of signal is avoided;
3.4 Performing short-time Fourier transform on each frame of data X (i, n) subjected to windowing to obtain frequency domain data X (i, k), wherein i represents a sequence number of a frame, the frame data X (i, n) is a signal sequence, n represents nth data in the signal sequence, and k represents a sequence number of a kth spectral line;
3.5 Calculating the energy E (i, m) of the frequency domain data X (i, k) passing through the Mel filter;
in the above formula, N represents the number of frames, H m (k) The M-th filter is a Mel filter, M is the number of groups of Mel filters;
3.6 Taking the logarithm of the frequency domain data X (i, k) through the energy E (i, m) of the Mel filter, performing discrete cosine transform, and calculating Mel frequency cepstrum coefficient based on the following formula;
in the above expression, MFCC (i, n) represents a mel-frequency cepstrum coefficient, i represents a frame number, n represents a spectral line after discrete cosine transform, M is the number of sets of mel filters, E (i, M) is the energy of the frequency domain data X (i, k) passing through the mel filters, M is the mel-frequency cepstrum coefficient number, and L is the dimension of the mel-frequency cepstrum coefficient.
The standard mel frequency cepstrum coefficient only reflects the static characteristics of the audio parameters, the dynamic characteristics of the audio can be described by using the differential spectrum of the static characteristics, and when the characteristic extraction is performed on the high-frequency signal component in the step 3) of the embodiment, the high-frequency dynamic characteristics are obtained by calculating the dynamic differential spectrum of the mel frequency cepstrum coefficient as the high-frequency dynamic characteristics, and the identification performance and accuracy of the system can be effectively improved by combining the dynamic characteristics and the static characteristics.
In this embodiment, the dynamic differential spectrum for calculating the mel-frequency cepstrum coefficient includes a first-order differential spectrum and a second-order differential spectrum, and the functional expression for calculating the first-order differential spectrum of the mel-frequency cepstrum coefficient is:
the expression of the calculation function for calculating the second-order differential spectrum of the mel frequency cepstrum coefficient is as follows:
wherein delta (i) represents the ith row of the first-order differential spectrum matrix, corresponding to the spectrum information of the ith frame, i represents the frame number, coeffs (i) represents the ith row of the Mel frequency cepstrum coefficient, corresponding to the ith frame, coeffs (i-1) represents the ith-1 row of the Mel frequency cepstrum coefficient, corresponding to the ith-1 frame; delta (i) represents the ith row of the second order differential spectrum matrix, corresponding to the ith frame, and Delta (i-1) represents the ith-1 row of the first order differential spectrum matrix, corresponding to the ith-1 frame.
In this embodiment, in step 4), determining whether the detected dry transformer has a discharge condition based on the high-frequency dynamic characteristic and the high-frequency static characteristic means that: and generating a characteristic spectrum image by the obtained high-frequency dynamic characteristics and the high-frequency static characteristics, comparing the characteristic spectrum image with the characteristic spectrum image under the normal state of the detected dry-type transformer, and judging whether the detected dry-type transformer has a discharge condition or not if the fluctuation amplitude and the fluctuation quantity change quantity respectively exceed corresponding preset threshold degrees. Based on the above-mentioned judging conditions, in this embodiment, by comparing the spectrum with the spectrum in the normal state, the dynamic differential spectrum of the dry-type transformer during discharge is significantly larger than the spectrum fluctuation in the normal state, and the spectrum fluctuation condition is significant, so that the occurrence of the discharge condition can be judged according to the state of the dynamic differential spectrum.
Fig. 5 is a two-dimensional diagram of the mel-frequency cepstrum coefficient calculated in the present embodiment, and fig. 6 is a three-dimensional diagram of the mel-frequency cepstrum coefficient calculated in the present embodiment. Fig. 7 shows a first-order differential spectrum of mel-frequency cepstrum coefficients calculated in this embodiment, fig. 8 shows a first-order differential spectrum of mel-frequency cepstrum coefficients calculated in this embodiment, fig. 9 shows a second-order differential spectrum of mel-frequency cepstrum coefficients calculated in this embodiment, and fig. 10 shows a second-order differential spectrum of mel-frequency cepstrum coefficients calculated in this embodiment. Referring to fig. 11-12, a first-order differential spectrum contrast chart and a second-order differential spectrum contrast chart are output in the present embodiment, wherein the left side in fig. 11 is a first-order dynamic differential spectrum chart when the dry-type transformer discharges, and the right side is a first-order dynamic differential spectrum chart when the dry-type transformer is in a normal state. Referring to fig. 12, the left side in fig. 12 is a second order dynamic differential spectrum of the dry-type transformer when discharging, and the right side is a second order dynamic differential spectrum of the dry-type transformer when in a normal state. In summary, the dynamic differential spectrogram of the dry-type transformer during discharge is obviously larger than the fluctuation of the dry-type transformer during normal discharge, and the spectrum fluctuation condition is obvious, so that the occurrence of the discharge condition can be judged according to the state of the dynamic differential spectrogram.
As shown in fig. 13, the present embodiment further provides a dry-type transformer discharge detection system based on audible sound analysis, including:
the acquisition module 1 comprises a microphone array and is used for acquiring acoustic signals at the working site of the dry-type transformer;
the processor 2 is used for extracting and calculating the signals acquired by the acquisition module 1;
and the display screen 3 is used for displaying all analysis results and operation results.
The processor 2 comprises a signal processing module 21, a signal analysis module 22, a characteristic operation module 23 and a characteristic analysis module 24; the signal processing module 21 is used for denoising the acquired acoustic signals; the signal analysis module 22 is configured to perform time-frequency domain analysis on the signal processed by the signal processing module 21; the feature operation module 23 is configured to perform Empirical Mode Decomposition (EMD) on the signal, and calculate a mel frequency cepstrum coefficient and a dynamic difference spectrum of a first eigenmode function (IMF) after the decomposition; the feature analysis module 24 is configured to analyze the features and draw a feature map based on the mel-frequency cepstral coefficient and the dynamic differential spectrum. The dry-type transformer discharge detection system based on audible sound analysis in the embodiment mainly comprises a microphone array, a processor 2 and a display screen 3, wherein the components are modularized and clear, the operation is simple, the use is convenient, the practicability is strong, the non-contact measurement is adopted, the measurement is free of load effect, the effect of testing the test object without damage can be achieved, the problem that the transformer is damaged when the transformer is tested at present is solved, and compared with the traditional dry-type transformer discharge detection method, the microphone acquisition mode enables detection to be realized in a severe environment.
In addition, the present embodiment also provides a dry-type transformer discharge detection system based on audible sound analysis, comprising a microprocessor and a memory, which are connected with each other, wherein the microprocessor is programmed or configured to execute the steps of the dry-type transformer discharge detection method based on audible sound analysis.
Further, the present embodiment also provides a computer-readable storage medium having stored therein a computer program programmed or configured to perform the aforementioned dry-type transformer discharge detection method based on audible sound analysis.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, the scope of which is defined in the appended claims and their equivalents.
Claims (6)
1. A dry-type transformer discharge detection method based on audible sound analysis, comprising:
1) Acquiring an audio signal of the detected dry-type transformer during operation;
2) Extracting high-frequency signal components from the audio signal;
3) Extracting features of the high-frequency signal components to obtain high-frequency static features and high-frequency dynamic features; wherein, the obtaining of the high-frequency static characteristic refers to calculating the mel frequency cepstrum coefficient of the high-frequency signal component as the high-frequency static characteristic, and the obtaining of the high-frequency dynamic characteristic refers to calculating the dynamic difference spectrum of the mel frequency cepstrum coefficient as the high-frequency dynamic characteristic; the dynamic differential spectrum for calculating the mel frequency cepstrum coefficient comprises a first-order differential spectrum and a second-order differential spectrum, and the function expression for calculating the first-order differential spectrum of the mel frequency cepstrum coefficient is as follows:
，
the expression of the calculation function for calculating the second-order differential spectrum of the mel frequency cepstrum coefficient is as follows:
，
wherein ,delta(i) Representing the first order differential spectrum matrixiLine corresponding toiThe spectral information of the frames is used to determine,ithe sequence number of the frame is indicated,Coeffs(i) First representing mel-frequency cepstral coefficientiLine corresponding toiThe frame of the frame is a frame of a frame,Coeffs(i-1) Representation ofMel frequency cepstrum coefficient (Mb)i-1 row, corresponding toi-1 frame;Ddelta(i) Representing the second order differential spectrum matrixiLine corresponding toiThe frame of the frame is a frame of a frame,delta(i- 1) Representing the first order differential spectrum matrixi-1 row, corresponding toi-1 frame;
4) Judging whether the detected dry-type transformer has a discharge condition or not based on the high-frequency dynamic characteristic and the high-frequency static characteristic: and generating a characteristic spectrum image by the obtained high-frequency dynamic characteristics and the high-frequency static characteristics, comparing the characteristic spectrum image with the characteristic spectrum image under the normal state of the detected dry-type transformer, and judging that the detected dry-type transformer has a discharge condition if the fluctuation amplitude and the fluctuation quantity change amount respectively exceed the corresponding preset threshold degrees.
2. The dry-type transformer discharge detection method based on audible sound analysis according to claim 1, wherein extracting high frequency signal components therein for an audio signal in step 2) means: and performing empirical mode decomposition on the audio signal, and taking a first eigenmode function obtained by the empirical mode decomposition as a high-frequency signal component in the obtained high-frequency signal.
3. The dry transformer discharge detection method based on audible sound analysis of claim 2, wherein the empirical mode decomposition of the audio signal comprises:
2.1 To the original current timetIs the current audio signalx(t)；
2.2 Finding the current audio signalx(t) All the upper and lower extreme points of (a);
2.3 Finding the current audio signalx(t) Envelope between all upper and lower extreme points of (a)e max (t) and e min (t) Obtaining the envelope between the upper and lower extreme pointse max (t) and e min (t) Average value between them to obtain average envelope curvem(t) At the current audio signalx(t) Subtracting mean value packageWinding wirem(t) Obtaining an intermediate signalh(t)；
2.4 Judging intermediate signalh(t) If the model is the eigenmode function, jumping to the next step; otherwise, the intermediate signalh(t) As a new current audio signalx(t) Step 2.2) is executed in a jumping mode;
2.5 To the original current timetAll intermediate signals obtained by subtracting the audio signal of (a)h(t) The sum gives the residual signalr n (t) Judging the residual signalr n (t) Whether a monotonic sequence or a constant sequence is established, and if not, the residual signal is obtainedr n (t) As a new current audio signalx(t) Step 2.2) is executed in a jumping mode; otherwise, judging that the empirical mode decomposition is finished.
4. The dry transformer discharge detection method based on audible sound analysis according to claim 1, wherein the step of calculating mel-frequency cepstral coefficients of high frequency signal components comprises:
3.1 Pre-emphasis processing the high frequency signal component to raise the high frequency part;
3.2 Frame-dividing the pre-emphasis processed high-frequency signal component, and dividing the longer sound signal into a plurality of small segments;
3.3 Windowing is carried out on the result after the framing processing, so that the condition that signals are discontinuous at two ends of each frame of signal is avoided;
3.4 For each frame data after windowingX(i,n) Performing short-time Fourier transform to obtain frequency domain dataX(i,k), wherein iRepresenting the sequence number of a frame, frame dataX(i,n) In the form of a signal sequence, the signal sequence,nrepresenting the nth data in the signal sequence,krepresent the firstkSerial numbers of spectral lines;
3.5 Calculating frequency domain dataX(i,k) Energy passing through mel filterE(i,m)；
，
In the above-mentioned method, the step of,Nthe number of frames is represented and,H m (k) Is the first of the Mel filtermThe number of filters to be used in the filter,Mthe number of groups of Mel filters;
3.6 To frequency domain dataX(i,k) Energy passing through mel filterE(i,m) Taking logarithms, performing discrete cosine transform, and calculating a Mel frequency cepstrum coefficient based on the following formula;
，
in the above formula, MFCC (i, n) represents mel-frequency cepstral coefficients,ithe sequence number of the frame is indicated,nrepresenting the spectral lines after the discrete cosine transform,Mfor the number of sets of mel-filters,E(i,m) Is frequency domain dataX(i,k) The energy that passes through the mel-filter,mfor the mel-filter sequence number,Lis the dimension of the mel-frequency cepstral coefficient.
5. A dry-type transformer discharge detection system based on audible sound analysis comprising a microprocessor and a memory interconnected, characterized in that the microprocessor is programmed or configured to perform the steps of the dry-type transformer discharge detection method based on audible sound analysis as claimed in any one of claims 1 to 4.
6. A computer readable storage medium having stored therein a computer program programmed or configured to perform the audible sound analysis based dry transformer discharge detection method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110997908.4A CN113933658B (en) | 2021-08-27 | 2021-08-27 | Dry-type transformer discharge detection method and system based on audible sound analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110997908.4A CN113933658B (en) | 2021-08-27 | 2021-08-27 | Dry-type transformer discharge detection method and system based on audible sound analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113933658A CN113933658A (en) | 2022-01-14 |
| CN113933658B true CN113933658B (en) | 2023-08-29 |
Family
ID=79274630
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110997908.4A Active CN113933658B (en) | 2021-08-27 | 2021-08-27 | Dry-type transformer discharge detection method and system based on audible sound analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113933658B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2003203519B2 (en) * | 1997-08-14 | 2005-10-27 | Hendry Mechnical Works | Electric arc monitoring systems |
| CN101499862A (en) * | 2009-03-06 | 2009-08-05 | 厦门红相电力设备股份有限公司 | Mobile phone noise signal distinguishing method in electric appliance extra-high-frequency local discharging detection |
| CN105788592A (en) * | 2016-04-28 | 2016-07-20 | 乐视控股（北京）有限公司 | Audio classification method and apparatus thereof |
| CN107179486A (en) * | 2017-05-24 | 2017-09-19 | 长沙理工大学 | A kind of GIS device monitors ultrahigh-frequency signal noise-reduction method on-line |
| CN107431254A (en) * | 2015-03-24 | 2017-12-01 | 李升揆 | Fuse cutout and battery control device and control method |
| CN109406965A (en) * | 2018-11-16 | 2019-03-01 | 国网江苏省电力有限公司盐城供电分公司 | A kind of partial discharge detecting system and its detection method based on sound signal collecting |
| CN110931022A (en) * | 2019-11-19 | 2020-03-27 | 天津大学 | Voiceprint identification method based on high-frequency and low-frequency dynamic and static characteristics |
| CN111650470A (en) * | 2020-05-21 | 2020-09-11 | 中国矿业大学(北京) | Method for rapidly and adaptively detecting and identifying faults of microgrid circuit sections |
| CN111830408A (en) * | 2020-06-23 | 2020-10-27 | 朗斯顿科技(北京)有限公司 | Motor fault diagnosis system and method based on edge calculation and deep learning |
| CN111983489A (en) * | 2020-08-18 | 2020-11-24 | 华中科技大学鄂州工业技术研究院 | Method for detecting discharge fault of SOFC (solid oxide Fuel cell) system with transition mode |
| CN112014692A (en) * | 2020-07-20 | 2020-12-01 | 国网安徽省电力有限公司电力科学研究院 | Partial discharge ultrahigh frequency signal blind source separation denoising method based on principal component analysis |
| CN113008361A (en) * | 2021-02-04 | 2021-06-22 | 国网湖南省电力有限公司 | Substation boundary noise anti-environmental interference detection method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080157783A1 (en) * | 2007-01-01 | 2008-07-03 | Maxwell Technologies, Inc. | Apparatus and method for monitoring high voltage capacitors |
| US8823383B2 (en) * | 2011-05-02 | 2014-09-02 | Kyung Jin Min | System and method for electrostatic discharge testing of devices under test |
-
2021
- 2021-08-27 CN CN202110997908.4A patent/CN113933658B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2003203519B2 (en) * | 1997-08-14 | 2005-10-27 | Hendry Mechnical Works | Electric arc monitoring systems |
| CN101499862A (en) * | 2009-03-06 | 2009-08-05 | 厦门红相电力设备股份有限公司 | Mobile phone noise signal distinguishing method in electric appliance extra-high-frequency local discharging detection |
| CN107431254A (en) * | 2015-03-24 | 2017-12-01 | 李升揆 | Fuse cutout and battery control device and control method |
| CN105788592A (en) * | 2016-04-28 | 2016-07-20 | 乐视控股（北京）有限公司 | Audio classification method and apparatus thereof |
| CN107179486A (en) * | 2017-05-24 | 2017-09-19 | 长沙理工大学 | A kind of GIS device monitors ultrahigh-frequency signal noise-reduction method on-line |
| CN109406965A (en) * | 2018-11-16 | 2019-03-01 | 国网江苏省电力有限公司盐城供电分公司 | A kind of partial discharge detecting system and its detection method based on sound signal collecting |
| CN110931022A (en) * | 2019-11-19 | 2020-03-27 | 天津大学 | Voiceprint identification method based on high-frequency and low-frequency dynamic and static characteristics |
| CN111650470A (en) * | 2020-05-21 | 2020-09-11 | 中国矿业大学(北京) | Method for rapidly and adaptively detecting and identifying faults of microgrid circuit sections |
| CN111830408A (en) * | 2020-06-23 | 2020-10-27 | 朗斯顿科技(北京)有限公司 | Motor fault diagnosis system and method based on edge calculation and deep learning |
| CN112014692A (en) * | 2020-07-20 | 2020-12-01 | 国网安徽省电力有限公司电力科学研究院 | Partial discharge ultrahigh frequency signal blind source separation denoising method based on principal component analysis |
| CN111983489A (en) * | 2020-08-18 | 2020-11-24 | 华中科技大学鄂州工业技术研究院 | Method for detecting discharge fault of SOFC (solid oxide Fuel cell) system with transition mode |
| CN113008361A (en) * | 2021-02-04 | 2021-06-22 | 国网湖南省电力有限公司 | Substation boundary noise anti-environmental interference detection method and device |
Non-Patent Citations (1)
| Title |
|---|
| 许渊."GIS绝缘子局部放电高灵敏测量方法及应用".《中国电机工程学报》.2020,全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113933658A (en) | 2022-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Pipeline leak detection based on variational mode decomposition and support vector machine using an interior spherical detector | |
| CN107610715B (en) | Similarity calculation method based on multiple sound characteristics | |
| JP6272433B2 (en) | Method and apparatus for detecting pitch cycle accuracy | |
| CN111239263B (en) | Method and system for detecting foreign matter defects in GIS equipment | |
| Hou et al. | Research on audio-visual detection method for conveyor belt longitudinal tear | |
| CN111625763A (en) | Operation risk prediction method and prediction system based on mathematical model | |
| CN112599134A (en) | Transformer sound event detection method based on voiceprint recognition | |
| CN116778956A (en) | Transformer acoustic feature extraction and fault identification method | |
| Dang et al. | Fault diagnosis of power transformer by acoustic signals with deep learning | |
| CN112052712A (en) | Power equipment state monitoring and fault identification method and system | |
| CN113933658B (en) | Dry-type transformer discharge detection method and system based on audible sound analysis | |
| Patil et al. | Effectiveness of Teager energy operator for epoch detection from speech signals | |
| Wan et al. | Adaptive asymmetric real Laplace wavelet filtering and its application on rolling bearing early fault diagnosis | |
| CN116796130A (en) | Bridge vibration low-frequency reconstruction denoising method, system, computer and storage medium | |
| Yan et al. | Noise recognition of power transformers based on improved MFCC and VQ | |
| CN115911623A (en) | Battery thermal runaway diagnosis method and system of energy storage system based on acoustic signals | |
| CN110487911B (en) | Method for detecting acoustic emission signals of pressure container based on blind source separation | |
| CN112114215A (en) | Transformer aging evaluation method and system based on error back propagation algorithm | |
| CN117849560B (en) | Valve side sleeve insulation monitoring method and system combining end screen voltage and partial discharge | |
| CN117232644B (en) | Transformer sound monitoring fault diagnosis method and system based on acoustic principle | |
| EP2986963A1 (en) | Pipe inspection system and related methods | |
| CN110455916B (en) | Solid material identification method based on acoustic dispersion phenomenon | |
| CN117968971B (en) | Gas leakage amount detection method and device and electronic equipment | |
| CN117496995A (en) | Power transformer fault detection method based on voiceprint feature screening | |
| CN117419915A (en) | Motor fault diagnosis method for multi-source information fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |