CN114019434A - Transformer residual magnetism detection method, device, system and storage medium - Google Patents
Transformer residual magnetism detection method, device, system and storage medium Download PDFInfo
- Publication number
- CN114019434A CN114019434A CN202111320746.7A CN202111320746A CN114019434A CN 114019434 A CN114019434 A CN 114019434A CN 202111320746 A CN202111320746 A CN 202111320746A CN 114019434 A CN114019434 A CN 114019434A
- Authority
- CN
- China
- Prior art keywords
- residual magnetism
- voiceprint
- transformer
- preset
- sound signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000005389 magnetism Effects 0.000 title claims abstract description 197
- 238000001514 detection method Methods 0.000 title claims abstract description 130
- 230000005236 sound signal Effects 0.000 claims abstract description 106
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims description 33
- 238000012360 testing method Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 238000013136 deep learning model Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 7
- 230000001755 vocal effect Effects 0.000 claims description 6
- 238000012935 Averaging Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 8
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical group [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 6
- 238000004804 winding Methods 0.000 description 6
- 238000001228 spectrum Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- 238000002555 auscultation Methods 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 3
- 230000008030 elimination Effects 0.000 description 3
- 238000003379 elimination reaction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000013021 overheating Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000005347 demagnetization Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000001012 protector Effects 0.000 description 1
- 238000009666 routine test Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/12—Measuring magnetic properties of articles or specimens of solids or fluids
Abstract
The application discloses a method, a device and a system for detecting residual magnetism of a transformer and a storage medium, wherein the method comprises the following steps: acquiring a first sound signal when a transformer works; and detecting the first sound signal to obtain a detection result of whether the transformer has residual magnetism. Through the mode, residual magnetism of the transformer can be detected without shutdown detection.
Description
Technical Field
The application relates to the technical field of power equipment detection, in particular to a transformer residual magnetism detection method, device and system and a storage medium.
Background
When the transformer operates, residual magnetism in the iron core of the transformer enables the iron core to generate a large amount of harmonic waves in exciting current to form exciting inrush current, the exciting inrush current can increase reactive power consumption of the transformer, and meanwhile false triggering and even damage of a relay protector can be caused; in addition, the high saturation of the iron core increases magnetic leakage, which causes overheating of a metal structural member and an oil tank, and the local overheating can cause the insulation paper to age and promote the decomposition of transformer oil, thereby affecting the service life of the transformer; therefore, in order to ensure the operation safety of the winding and the stable operation of the power system, the influence caused by the residual magnetism of the transformer needs to be considered, and the problem to be solved urgently is to detect whether the residual magnetism exists in the transformer.
Disclosure of Invention
The application provides a method, a device and a system for detecting residual magnetism of a transformer and a storage medium, which can realize the residual magnetism detection of the transformer without shutdown detection.
In order to solve the technical problem, the technical scheme adopted by the application is as follows: the method for detecting the residual magnetism of the transformer based on the voiceprint comprises the following steps: acquiring a first sound signal when a transformer works; and detecting the first sound signal to obtain a detection result of whether the transformer has residual magnetism.
In order to solve the above technical problem, another technical solution adopted by the present application is: the residual magnetism detection device comprises a memory and a processor which are connected with each other, wherein the memory is used for storing a computer program, and the computer program is used for realizing the transformer residual magnetism detection method based on the voiceprint in the technical scheme when being executed by the processor.
In order to solve the above technical problem, another technical solution adopted by the present application is: the audio acquisition equipment is used for acquiring sound signals when the transformer works and sending the sound signals to the residual magnetism detection device so that the residual magnetism detection device can detect whether residual magnetism exists in the transformer or not based on the sound signals, wherein the residual magnetism detection device is the residual magnetism detection device in the technical scheme.
In order to solve the above technical problem, another technical solution adopted by the present application is: a computer-readable storage medium is provided, which is used for storing a computer program, and when the computer program is executed by a processor, the computer program is used for implementing the voiceprint-based transformer residual magnetism detection method in the above technical solution.
Through the scheme, the beneficial effects of the application are that: the method for detecting whether the transformer has residual magnetism based on the voiceprint is adopted, and a first sound signal is obtained when the transformer operates; then, the first sound signal is detected to obtain a detection result of whether the transformer has residual magnetism, so that early warning can be conveniently carried out in time when the residual magnetism of the transformer is detected in the follow-up process, related personnel can conveniently carry out demagnetization treatment on the transformer in time, and adverse effects caused by the residual magnetism are relieved; the residual magnetism detection scheme provided by the application can judge whether the transformer has residual magnetism through the sound signal when the transformer operates, and has low requirements on detection environment and good universality; and the transformer residual magnetism detection device does not need to be stopped for detection, can realize the real-time residual magnetism detection of the transformer, and has convenient use and lower cost.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts. Wherein:
FIG. 1 is a schematic flowchart of an embodiment of a voiceprint-based transformer residual magnetism detection method provided by the present application;
FIG. 2 is a schematic flow chart illustrating another embodiment of a voiceprint-based transformer residual magnetism detection method provided by the present application;
FIG. 3 is a schematic flow chart of step 23 in the embodiment shown in FIG. 2;
FIG. 4 is a schematic flow chart of step 31 in the embodiment shown in FIG. 3;
FIG. 5 is a schematic illustration of a first reconstruction error profile and a second reconstruction error profile provided herein;
FIG. 6 is another schematic flow chart of step 31 in the embodiment shown in FIG. 3;
FIG. 7 is a schematic structural diagram of an embodiment of a residual magnetism detecting device provided in the present application;
FIG. 8 is a schematic structural diagram of an embodiment of a residual magnetism detection system provided herein;
FIG. 9 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided in the present application.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be noted that the following examples are only illustrative of the present application, and do not limit the scope of the present application. Likewise, the following examples are only some examples and not all examples of the present application, and all other examples obtained by a person of ordinary skill in the art without any inventive work are within the scope of the present application.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the specification. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
It should be noted that the terms "first", "second" and "third" in the present application are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of indicated technical features. Thus, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of the feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless explicitly specifically limited otherwise. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.
The winding direct current resistance test essential in the routine test of transformer condition maintenance will make residual magnetism remain in the iron core of the transformer, therefore, before the transformer is put into operation or run, it is necessary to detect whether residual magnetism exists, there are many technical schemes about residual magnetism detection and elimination in the related art, and the commonly used residual magnetism detection scheme includes the following schemes:
the method comprises the steps of firstly, adopting residual magnetism detection equipment which does not need to intrude into the transformer, wherein the residual magnetism detection equipment can be equipment based on harmonic measurement, adopting Fast Fourier Transform (FFT) algorithm, carrying out Fourier decomposition and waveform recording on transformer no-load current by using an oscilloscope with FFT decomposition function, obtaining a distribution diagram of the content of each subharmonic including even subharmonic, and judging whether the transformer has residual magnetism by using the distribution diagram.
Secondly, frequency response of the low-voltage winding is checked by inputting a sweep frequency signal to the high-voltage winding of the transformer to obtain an amplitude-frequency response curve, and whether residual magnetism exists in the iron core of the transformer can be visually judged by comparing the amplitude-frequency response curve with a reference amplitude-frequency response curve in a state without residual magnetism.
For the technical scheme, the residual magnetism detection equipment has high requirements on detection environments, has the problem of poor universality, is difficult to be compatible with transformers of different brands and models, and has long time consumption and low efficiency in the whole detection process. For the second technical scheme, the transformer needs to be invaded, so that shutdown detection is needed, limitation is realized, and detection cost is high. In order to solve the problems, the present application provides a method for detecting whether residual magnetism exists in a transformer based on voiceprint, and the following describes the technical solution adopted in the present application in detail.
Referring to fig. 1, fig. 1 is a schematic flowchart illustrating an embodiment of a voiceprint-based transformer residual magnetism detection method provided in the present application, where the method includes:
step 11: and acquiring a first sound signal when the transformer works.
Carrying out audio acquisition processing on a working voltage device in the current environment by using audio acquisition equipment to obtain a sound signal (namely a first sound signal) of the transformer, and realizing voiceprint acquisition; specifically, the audio acquisition device may be a sensor disposed at a winding of the transformer, the sensor may acquire a sound signal of the transformer, and may transmit the sound signal to a detection platform (i.e., a residual magnetism detection device) through a Message queue Telemetry Transport protocol (MQTT), where the residual magnetism detection device may be a device with processing capability, such as a computer or a cloud server.
Step 12: and detecting the first sound signal to obtain a detection result of whether the transformer has residual magnetism.
After acquiring the first sound signal, the residual magnetism detection device may detect the first sound signal to determine whether residual magnetism exists in the current transformer, and generate a detection result, where the detection result may be that residual magnetism exists in the transformer or that residual magnetism does not exist in the transformer. Specifically, the currently acquired first sound signal may be compared with a sound signal when the transformer has no residual magnetism (recorded as a reference non-residual magnetism sound signal) to determine whether the transformer has residual magnetism according to the first sound signal.
In a specific embodiment, when residual magnetism exists in the transformer, early warning processing can be carried out based on a detection result; specifically, if the residual magnetism detection device determines that the current transformer has residual magnetism, the detection result of the residual magnetism of the transformer is directly displayed, or early warning is performed to remind related personnel to perform residual magnetism elimination in time, so that potential safety hazards are eliminated and energy consumption of the transformer is reduced; specifically, the reminding method may be to generate a warning message, send the warning message to an external electronic device (for example, a mobile terminal or a computer), display the warning message, play the warning message, or dial a phone of a preset contact (for example, an operator).
It is understood that in other embodiments, the sound signals generated when the transformer operates may be collected at intervals of a preset time to reduce the processing load of the residual magnetism detection device; for example: after the residual magnetism detection device and the audio acquisition equipment are deployed, the first sound signal can be acquired at an interval of 60 seconds, for example, the sound signal of the 0 th to 10 th seconds in the current minute is acquired, the sound signal of the 0 th to 10 th seconds in the next minute is acquired, and the like.
The embodiment provides a power transformer residual magnetism detection method based on voiceprint, which combines the characteristics of a power transformer, obtains a first sound signal when the transformer operates by pointing an audio acquisition device to a winding of the transformer, and can intuitively judge whether residual magnetism exists in an iron core of the transformer by comparing the first sound signal with a reference non-residual magnetism sound signal in a non-residual magnetism state; because the sound signals of the transformer in the residual magnetism state and the non-residual magnetism state are different, whether the transformer has residual magnetism can be smoothly judged according to the currently acquired first sound signal by learning the characteristic of the non-residual magnetism sound signal of the transformer in the non-residual magnetism state, and the universality is better because the requirement on the detection environment is not high; in addition, whether residual magnetism exists or not can be detected in real time in the operation process of the transformer, shutdown detection is not needed, and the transformer is convenient to use.
Referring to fig. 2, fig. 2 is a schematic flowchart illustrating another embodiment of a voiceprint-based transformer residual magnetism detection method provided in the present application, the method including:
step 21: and acquiring a first sound signal when the transformer works.
Step 22: and carrying out feature extraction processing on the first sound signal to obtain the current voiceprint feature.
Performing feature extraction processing on new data (namely a first sound signal) acquired currently by adopting a feature extraction method to generate current voiceprint features; specifically, the current voiceprint features include a short-time energy spectrum, a spectrum entropy, or a Mel-Frequency Cepstral Coefficients (MFCC), and the feature extraction method is a method capable of realizing a feature extraction function in the related art and is not described herein again.
Step 23: and detecting the current voiceprint characteristics by adopting a residual magnetism detection model to obtain a detection result.
The remanence detection model comprises a deep learning model or a signal processing model, wherein the deep learning model is a network model constructed by adopting a deep learning method, the signal processing model is a model for carrying out statistics or other processing on input data, and the signal processing model can be a non-machine learning model, namely the signal processing model does not need to be trained; specifically, the scheme shown in fig. 3 may be adopted to obtain the detection result, including the following steps:
step 31: and judging whether the current voiceprint characteristics meet the preset remanence condition or not based on the current voiceprint characteristics, the preset voiceprint library or the output result of the remanence detection model.
After obtaining the voiceprint feature (i.e., the current voiceprint feature) corresponding to the first sound signal, whether residual magnetism exists in the current transformer may be determined by combining the current voiceprint feature, a preset voiceprint library established in advance, or an output result of the residual magnetism detection model.
In a specific embodiment, the deep learning model is a self-coding model, the self-coding model includes an encoder and a decoder, and the scheme shown in fig. 4 can be adopted to determine whether the current voiceprint feature satisfies the predetermined remanence condition, including the following steps:
step 41: and coding the current voiceprint characteristics by adopting a coder to obtain coding characteristics.
The current voiceprint characteristics are input to the encoder such that the encoder encodes the current voiceprint characteristics, generating encoded characteristics, which are then input to the decoder.
Step 42: and decoding the coding characteristics by adopting a decoder to obtain reconstructed voiceprint characteristics.
After receiving the coding features, the decoder performs decoding processing on the coding features to generate features (namely reconstructed voiceprint features) obtained by reconstructing the input current voiceprint features.
Step 43: and judging whether the current voiceprint characteristics meet the preset remanence condition or not based on the reconstructed voiceprint characteristics and a preset voiceprint library.
The preset voiceprint library comprises first test sample data, the first test sample data comprises a plurality of remanent magnetism-free voiceprint characteristics, and the remanent magnetism-free voiceprint characteristics are voiceprint characteristics corresponding to sounds generated by operation when the transformer does not have remanent magnetism; determining the reference non-residual magnetism vocal print characteristics by utilizing all non-residual magnetism vocal print characteristics; then calculating the error between the reference non-residual magnetism vocal print characteristic and the reconstructed vocal print characteristic to obtain the current reconstruction error (namely the singularity); then judging whether the current reconstruction error is larger than a preset error threshold value, wherein the preset error threshold value is used for judging a newly input first sound signal by a residual magnetism detection device; if the current reconstruction error is larger than the preset error threshold value, the current reconstruction error is larger, and at the moment, the current voiceprint characteristic can be judged to meet the preset residual magnetism condition; and if the current reconstruction error is smaller than or equal to the preset error threshold, the current reconstruction error is smaller, and at the moment, the current voiceprint characteristic is judged not to meet the preset residual magnetism condition. Specifically, the preset error threshold may be a threshold set according to experience or actual application requirements, or determined by a training process of the self-coding model.
Further, one non-residual magnetism voiceprint characteristic can be randomly selected from all non-residual magnetism voiceprint characteristics directly as a reference non-residual magnetism voiceprint characteristic; or processing all the non-residual-magnetism voiceprint characteristics by adopting a self-coding model to obtain the reference non-residual-magnetism voiceprint characteristics, such as: and learning the historical input non-remanent magnetic voiceprint characteristics by adopting a hidden space encoding mode with memory slot storage, and storing the main components of the non-remanent magnetic voiceprint characteristics as the reference non-remanent magnetic voiceprint characteristics. Further, in the process of inputting the non-residual-magnetism voiceprint characteristics into the coding model each time, when a decoder is used for decoding, the coding characteristics corresponding to the non-residual-magnetism voiceprint characteristics are decoded by combining the reference non-residual-magnetism voiceprint characteristics to generate corresponding decoded characteristics.
In a specific embodiment, the training process of the self-coding model is as follows:
1) and constructing first test sample data.
Acquiring sound signals of the transformer in a non-residual magnetism state, and forming all the sound signals into a data set; then screening the data set to obtain screened sound signals without residual magnetism; and extracting the screened non-remanent-magnetism sound signals to obtain the voiceprint characteristics (including short-time energy spectrum, spectral entropy or MFCC and the like) of the non-remanent-magnetism sound signals so as to construct first test sample data, wherein the first test sample data comprises the non-remanent-magnetism sound signals and the non-remanent-magnetism voiceprint characteristics corresponding to the non-remanent-magnetism sound signals.
Furthermore, because equipment such as a transformer is deployed outdoors, various sounds may exist in the environment where the equipment is located, in order to improve the effect of the self-coding model, sound signals with large noise in a data set can be planed, and relatively pure sound signals can be screened out, so that the effectiveness of data required by training the self-coding model is ensured.
2) Inputting the first test sample data into the self-coding model to train the self-coding model.
Carrying out self-supervision training on the self-coding model by adopting first test sample data so as to learn the characteristics of the sound signal in a non-remanence state; after the self-coding model is trained by adopting the first test sample data, the self-coding model can output the reference non-remanence voiceprint characteristic.
In a specific embodiment, the preset voiceprint library further comprises second test sample data, wherein the second test sample data comprises a plurality of residual magnetic voiceprint characteristics, and the residual magnetic voiceprint characteristics are characteristics corresponding to sound signals generated when the transformer operates in a residual magnetic state; processing all the non-remanent-magnetism voiceprint characteristics by adopting a self-coding model to obtain first reconstruction error distribution; processing all residual magnetism voiceprint characteristics by adopting a self-coding model to obtain second reconstruction error distribution; and determining a preset error threshold value based on the first reconstruction error distribution and the second reconstruction error distribution.
Further, inputting each non-remanent magnetism voiceprint characteristic into a self-coding model for training to obtain a corresponding first reconstruction characteristic; calculating the error between each non-residual magnetism voiceprint characteristic and the corresponding first reconstruction characteristic to obtain a first reconstruction error; counting all the first reconstruction errors to obtain first reconstruction error distribution; inputting each residual magnetism voiceprint characteristic into a self-coding model for training to obtain a corresponding second reconstruction characteristic; calculating the error between each residual magnetic voiceprint feature and the corresponding second reconstruction feature to obtain a second reconstruction error; and counting all the second reconstruction errors to obtain a second reconstruction error distribution. Specifically, the first reconstruction error and the second reconstruction error respectively conform to a certain distribution and can be used for assisting in identifying and alarming, the first reconstruction error distribution is a relation between the first reconstruction error and a quantity/probability corresponding to the first reconstruction error, the second reconstruction error distribution is a relation between the second reconstruction error and a quantity/probability corresponding to the second reconstruction error, and an error which can distinguish the first reconstruction error distribution from the second reconstruction error distribution is used as a preset error threshold. For example: as shown in fig. 5, the first reconstruction error distribution is denoted as a, the second reconstruction error distribution is denoted as B, and since there is no overlapping area between the first reconstruction error distribution a and the second reconstruction error distribution B, the error C falling between the first reconstruction error distribution a and the second reconstruction error distribution B may be used as the preset error threshold. It can be understood that, when there is an overlapping region between the first reconstruction error distribution and the second reconstruction error distribution, a certain reconstruction error in the overlapping region may be used as a preset error threshold, and then the self-encoding model is verified by using verification sample data in a preset voiceprint library, and when a verified false alarm rate (i.e. a probability that a sound signal without residual magnetism is determined to have residual magnetism) is less than a preset false alarm rate, the verification is stopped; and when the verified false alarm rate is greater than the preset false alarm rate, adjusting the numerical value of the preset error threshold, and continuing to verify until the verified false alarm rate is less than the preset false alarm rate.
In another specific embodiment, a voiceprint short-time smoothing technique may be combined, and a signal processing model is used to process the first sound signal, that is, the scheme shown in fig. 6 is used, which specifically includes the following steps:
step 61: an energy average of the first sound signal is obtained.
The energy mean value of the first sound signal can be directly calculated by using the short-time energy spectrum in the current voiceprint characteristic; or dividing the first sound signal according to the frequency to obtain second sound signals of different frequency bands; acquiring energy values of second sound signals of different frequency bands to obtain sub-energy values; averaging all the sub-energy values to obtain an energy average value of the first sound signal; or calculating an energy value of the first sound signal in a preset frequency band to obtain an energy average value, where the preset frequency band is a frequency band where residual magnetism is likely to appear in a voiceprint, for example: 4KHz to 8 KHz.
Step 62: and judging whether the difference value between the energy mean value and the reference energy value exceeds a preset difference value or not.
The reference energy value is an energy average value corresponding to the reference non-residual magnetism voiceprint characteristic, and whether residual magnetism exists in the current transformer is determined by comparing the energy average value of the first sound signal with the energy average value corresponding to the reference non-residual magnetism voiceprint characteristic.
And step 63: and if the difference value between the energy mean value and the reference energy value exceeds a preset difference value, determining that the current voiceprint characteristics meet a preset residual magnetism condition.
If the difference value between the energy mean value of the first sound signal and the energy mean value corresponding to the reference non-residual magnetism voiceprint characteristic exceeds a preset difference value, the difference value between the energy mean value of the first sound signal and the energy mean value is larger, and at the moment, the current voiceprint characteristic can be judged to meet a preset residual magnetism condition, namely, residual magnetism exists in the transformer; if the difference value between the energy mean value of the first sound signal and the energy mean value corresponding to the reference non-residual magnetism voiceprint characteristic does not exceed the preset difference value, the difference value is not large, and at the moment, the current voiceprint characteristic can be judged not to meet the preset residual magnetism condition, namely, the transformer does not have residual magnetism; the preset difference value is a value set according to experience or application requirements, such as: and 0, namely when the energy mean value of the first sound signal is different from the energy mean value corresponding to the reference non-residual magnetism voiceprint characteristic, judging that the residual magnetism exists in the transformer and giving an alarm.
The energy mean value of the first sound signal is calculated through short-time energy spectrums of different frequency bands, and the energy mean value is compared with a reference energy value, so that the capturing capability of the instantaneous and abnormal sound signals can be enhanced, and the instantaneous voiceprint can be captured more accurately.
Step 32: and if the current voiceprint characteristics meet the preset remanence condition, determining that remanence exists in the transformer.
Step 33: and if the current voiceprint characteristics do not meet the preset remanence condition, determining that the transformer does not have remanence.
Step 24: when the transformer has residual magnetism, warning information is generated, and residual magnetism detection related information is displayed and/or played and/or transmitted to external electronic equipment.
The detection related information comprises warning information or a detection result, namely when the residual magnetism of the transformer is detected, the detection result can be displayed, played or sent to preset external electronic equipment; or the warning information can be displayed, played or sent to preset external electronic equipment to remind relevant personnel to perform whispering phrase elimination processing on the transformer, so that the running safety of the transformer is improved, and the service life of the transformer is prolonged.
It can be understood that all newly acquired data (i.e., sound signals) are input into the deep learning model or the signal processing model, and if the detection result of any one of the deep learning model and the signal processing model is that the residual magnetism exists in the transformer, the residual magnetism of the transformer is determined, so as to avoid missing reports and improve the accuracy of residual magnetism detection.
Step 25: and when the transformer has no residual magnetism, displaying and storing the information of the transformer.
If it is determined that the transformer does not have the residual magnetism, detailed information of the transformer, including a model of the transformer, a sound signal, a voiceprint characteristic, and a detection result of absence of the residual magnetism, may be displayed and stored.
In other specific embodiments, for convenience of processing, the first sound signal may be further divided to obtain a plurality of third sound signals; then, the third sound signal is input into the residual magnetism detection model, and the specific implementation process is similar to the process of processing the first sound signal in the above embodiment, and is not described herein again. For example, the length of the first sound signal is 10 seconds, after the remanence detection device receives the first sound signal, the segment extraction processing is performed on the first sound signal, so that the 10-second first sound signal is divided into 9 2-second third sound signals, that is, the first third sound signal is obtained in 0-2 seconds, the second third sound signal is obtained in 1-3 seconds, and so on, and finally 9 third sound signals are obtained, and the 9 third sound signals are input into a remanence detection model for remanence detection processing; it is understood that, if the detection result of any one of the 9 third acoustic signals is that the transformer has residual magnetism, the transformer is determined to have residual magnetism.
The embodiment provides a voiceprint-based power transformer residual magnetism detection method, wherein a first sound signal obtained at present is input into a deep learning model and/or a signal processing model, so that the deep learning model and/or the signal processing model compare the first sound signal with a reference non-residual magnetism sound signal, whether the difference between the first sound signal and the reference non-residual magnetism sound signal is larger is determined, and if the difference is larger, the transformer residual magnetism is determined; the scheme adopted by the embodiment can carry out residual magnetism detection when the voltage transformer operates, detection can be carried out without stopping the operation of the transformer, and real-time online detection can be carried out; and the detection results of the deep learning model and the signal processing model are combined, so that the accuracy of detection is improved, and missing detection is prevented.
Referring to fig. 7, fig. 7 is a schematic structural diagram of an embodiment of a residual magnetism detecting device provided in the present application, in which the residual magnetism detecting device 70 includes a memory 71 and a processor 72 connected to each other, the memory 71 is used for storing a computer program, and the computer program is used for implementing the transformer residual magnetism detecting method based on voiceprint in the above embodiment when being executed by the processor 72.
Referring to fig. 8, fig. 8 is a schematic structural diagram of an embodiment of a residual magnetism detection system provided in the present application, where the residual magnetism detection system 80 includes an audio acquisition device 81 and a residual magnetism detection device 82, the audio acquisition device 81 is configured to acquire an audio signal when the transformer operates, and send the audio signal to the residual magnetism detection device 82, so that the residual magnetism detection device 82 detects whether residual magnetism exists in the transformer based on the audio signal, where the residual magnetism detection device 82 is the residual magnetism detection device in the above embodiment.
As shown in fig. 8, the residual magnetism detecting system 80 further includes an electronic device 83, the electronic device 83 is connected to the residual magnetism detecting device 82, the electronic device 83 is configured to receive and display a detection result of the residual magnetism detecting device 82, and the electronic device 83 may be a device 81 with a display function, such as a mobile terminal or a computer.
In other embodiments, an electric auscultation device (not shown in the figures) may be further disposed in the residual magnetism detection system 80, and the electric auscultation device is connected to the audio acquisition device 81 and the residual magnetism detection device 82, and is used for storing the sound signals acquired by the audio acquisition device 81 and transmitting the sound signals to the residual magnetism detection device 82; it is to be understood that the electric auscultation apparatus may transmit the sound signal collected by the audio collecting device 81 to the residual magnetism detecting device 82 at intervals of a preset time.
Referring to fig. 9, fig. 9 is a schematic structural diagram of an embodiment of a computer-readable storage medium 90 provided by the present application, where the computer-readable storage medium 90 is used to store a computer program 91, and when the computer program 91 is executed by a processor, the computer program is used to implement the transformer residual magnetism detection method based on voiceprint in the foregoing embodiment.
The computer-readable storage medium 90 may be a server, a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and various media capable of storing program codes.
In the several embodiments provided in the present application, it should be understood that the disclosed method and apparatus may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of modules or units is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed.
Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings, or which are directly or indirectly applied to other related technical fields, are intended to be included within the scope of the present application.
Claims (16)
1. A transformer residual magnetism detection method based on voiceprint is characterized by comprising the following steps:
acquiring a first sound signal when a transformer works;
and detecting the first sound signal to obtain a detection result of whether the transformer has residual magnetism.
2. The voiceprint-based transformer residual magnetism detection method according to claim 1, wherein the step of detecting the first sound signal to obtain a detection result of whether the transformer has residual magnetism comprises:
carrying out feature extraction processing on the first sound signal to obtain current voiceprint features;
and detecting the current voiceprint characteristics by adopting a residual magnetism detection model to obtain the detection result, wherein the residual magnetism detection model comprises a deep learning model or a signal processing model.
3. The voiceprint-based transformer residual magnetism detection method according to claim 2, wherein the step of detecting the current voiceprint characteristics by using a residual magnetism detection model to obtain the detection result comprises:
judging whether the current voiceprint characteristics meet preset remanence conditions or not based on the current voiceprint characteristics, a preset voiceprint library or output results of the remanence detection model;
if so, determining that residual magnetism exists in the transformer;
if not, determining that the transformer has no residual magnetism.
4. The voiceprint-based transformer residual magnetism detection method according to claim 3, wherein the deep learning model is a self-coding model, the self-coding model comprises an encoder and a decoder, and the step of determining whether the current voiceprint feature satisfies a preset residual magnetism condition based on the current voiceprint feature, a preset voiceprint library or an output result of the residual magnetism detection model comprises:
coding the current voiceprint characteristics by adopting the coder to obtain coding characteristics;
decoding the coding features by adopting the decoder to obtain reconstructed voiceprint features;
and judging whether the current voiceprint characteristics meet a preset remanence condition or not based on the reconstructed voiceprint characteristics and the preset voiceprint library.
5. The voiceprint-based transformer residual magnetism detection method according to claim 4, wherein the preset voiceprint library comprises first test sample data, the first test sample data comprises a plurality of non-remanent magnetism voiceprint features, the non-remanent magnetism voiceprint features are voiceprint features when a transformer has no remanent magnetism, and the step of judging whether the current voiceprint features meet a preset remanent magnetism condition based on the reconstructed voiceprint features and the preset voiceprint library comprises:
determining a reference non-residual magnetism voiceprint characteristic by using all the non-residual magnetism voiceprint characteristics;
calculating the error between the reference non-residual magnetism vocal print characteristic and the reconstruction vocal print characteristic to obtain the current reconstruction error;
judging whether the current reconstruction error is larger than a preset error threshold value or not;
and if so, determining that the current voiceprint characteristics meet the preset remanence condition.
6. The voiceprint based transformer remanence detection method according to claim 5, wherein the preset voiceprint library further comprises second test sample data, the second test sample data comprises a plurality of remanence voiceprint features, the method further comprises:
processing all the non-residual magnetism voiceprint characteristics by adopting the self-coding model to obtain first reconstruction error distribution;
processing all the residual magnetism voiceprint characteristics by adopting the self-coding model to obtain second reconstruction error distribution;
determining the preset error threshold based on the first reconstruction error distribution and the second reconstruction error distribution.
7. The voiceprint based transformer residual magnetism detection method of claim 6 further comprising:
inputting each non-remanent magnetism voiceprint feature into the self-encoding model for training to obtain a corresponding first reconstruction feature; calculating the error between each non-residual magnetism voiceprint feature and the corresponding first reconstruction feature to obtain a first reconstruction error; counting all the first reconstruction errors to obtain first reconstruction error distribution;
inputting each residual magnetism voiceprint characteristic into the self-coding model for training to obtain a corresponding second reconstruction characteristic; calculating the error between each residual magnetic voiceprint feature and the corresponding second reconstruction feature to obtain a second reconstruction error; and counting all the second reconstruction errors to obtain the second reconstruction error distribution.
8. The voiceprint based transformer residual magnetism detection method of claim 3 further comprising:
acquiring an energy mean value of the first sound signal;
judging whether the difference value between the energy mean value and a reference energy value exceeds a preset difference value or not, wherein the reference energy value is the energy mean value corresponding to the reference non-residual magnetism voiceprint characteristic;
and if so, determining that the current voiceprint characteristics meet the preset remanence condition.
9. The voiceprint based transformer residual magnetism detection method of claim 8 wherein the step of obtaining an energy average of the first sound signal comprises:
dividing the first sound signal according to frequency to obtain a second sound signal with different frequency bands;
acquiring energy values of the second sound signals of different frequency bands to obtain sub energy values;
and averaging all the sub-energy values to obtain an energy average value of the first sound signal.
10. The voiceprint based transformer residual magnetism detection method of claim 2 further comprising:
dividing the first sound signal to obtain a plurality of third sound signals;
inputting the third sound signal into the residual magnetism detection model.
11. The voiceprint based transformer residual magnetism detection method of claim 2,
and when the transformer has residual magnetism, performing early warning processing based on the detection result.
12. The voiceprint-based transformer residual magnetism detection method according to claim 11, wherein the step of performing early warning processing based on the detection result comprises:
when the transformer has residual magnetism, generating residual magnetism detection related information, displaying and/or playing the residual magnetism detection related information, and/or sending the residual magnetism detection related information to external electronic equipment, wherein the detection related information comprises warning information or a detection result;
and when the transformer has no remanent magnetism, displaying and storing the information of the transformer.
13. A residual magnetism detecting apparatus, characterized by comprising a memory and a processor connected to each other, wherein the memory is configured to store a computer program, which when executed by the processor is configured to implement the voiceprint based transformer residual magnetism detecting method of any one of claims 1-12.
14. A residual magnetism detection system, comprising a residual magnetism detection device and an audio collection device, wherein the audio collection device is used for collecting sound signals when a transformer works and sending the sound signals to the residual magnetism detection device, so that the residual magnetism detection device detects whether residual magnetism exists in the transformer based on the sound signals, and the residual magnetism detection device is the residual magnetism detection device of claim 13.
15. The residual magnetism detection system of claim 14 wherein,
the residual magnetism detection system also comprises electronic equipment, and the electronic equipment is connected with the residual magnetism detection device and used for receiving and displaying the detection result of the residual magnetism detection device.
16. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, is configured to implement the voiceprint based transformer residual magnetism detection method of any one of claims 1 to 12.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111320746.7A CN114019434A (en) | 2021-11-09 | 2021-11-09 | Transformer residual magnetism detection method, device, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111320746.7A CN114019434A (en) | 2021-11-09 | 2021-11-09 | Transformer residual magnetism detection method, device, system and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114019434A true CN114019434A (en) | 2022-02-08 |
Family
ID=80062491
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111320746.7A Pending CN114019434A (en) | 2021-11-09 | 2021-11-09 | Transformer residual magnetism detection method, device, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114019434A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115032577A (en) * | 2022-08-10 | 2022-09-09 | 之江实验室 | Low-noise residual magnetism detection device and method for annular ferrite material |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CH381762A (en) * | 1959-03-10 | 1964-09-15 | Agfa Ag | Procedure for the detection of the presence of remanent magnetization on device parts of magnetic sound devices |
| DE4312764A1 (en) * | 1993-04-20 | 1994-10-27 | Fraunhofer Ges Forschung | Detection device for identifying the remanence level in the core of a transformer |
| CN111259921A (en) * | 2019-12-19 | 2020-06-09 | 杭州安脉盛智能技术有限公司 | Transformer sound anomaly detection method based on improved wavelet packet and deep learning |
| CN113296029A (en) * | 2021-05-11 | 2021-08-24 | 江苏翼昇科技有限公司 | Distribution transformer voiceprint monitoring method, device, equipment and storage medium |
| KR20210109416A (en) * | 2020-02-27 | 2021-09-06 | 한국전자통신연구원 | Machine-learning based abnormal signal detection using normal signal removing filter |
-
2021
- 2021-11-09 CN CN202111320746.7A patent/CN114019434A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CH381762A (en) * | 1959-03-10 | 1964-09-15 | Agfa Ag | Procedure for the detection of the presence of remanent magnetization on device parts of magnetic sound devices |
| DE4312764A1 (en) * | 1993-04-20 | 1994-10-27 | Fraunhofer Ges Forschung | Detection device for identifying the remanence level in the core of a transformer |
| CN111259921A (en) * | 2019-12-19 | 2020-06-09 | 杭州安脉盛智能技术有限公司 | Transformer sound anomaly detection method based on improved wavelet packet and deep learning |
| KR20210109416A (en) * | 2020-02-27 | 2021-09-06 | 한국전자통신연구원 | Machine-learning based abnormal signal detection using normal signal removing filter |
| CN113296029A (en) * | 2021-05-11 | 2021-08-24 | 江苏翼昇科技有限公司 | Distribution transformer voiceprint monitoring method, device, equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| 唐云辉;胡曦琳;: "基于声信号分析的开关柜局部放电检测算法研究", 电子器件, vol. 41, no. 03, pages 644 - 648 * |
| 朱玲: "一种分析变压器噪声异常原因的简易方法", 机电工程技术, vol. 50, no. 5, pages 262 - 264 * |
| 谭风雷: "基于声音在线监测的变压器异常判别方法", 湖北电力, vol. 45, no. 5, pages 26 - 31 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115032577A (en) * | 2022-08-10 | 2022-09-09 | 之江实验室 | Low-noise residual magnetism detection device and method for annular ferrite material |
| CN115032577B (en) * | 2022-08-10 | 2022-11-15 | 之江实验室 | Low-noise residual magnetism detection device and method for annular ferrite material |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105181925B (en) | Running water detection means | |
| CN102426835B (en) | Method for identifying local discharge signals of switchboard based on support vector machine model | |
| US8441264B2 (en) | Station for detecting winding products and method for detecting inter-turn short circuit | |
| CN110275094A (en) | A kind of high-tension cable shelf depreciation acousto-optic associated detecting method and system | |
| US10690713B2 (en) | Method of detecting an electric arc by analysis of its acoustic signature | |
| WO2012171300A1 (en) | Sound abnormality detection system for power device and method thereof | |
| CN104333752B (en) | Carry out the method and device of audio quality diagnosis | |
| CN102692887A (en) | Device for monitoring online state of power transformer based on audio identification technology | |
| CN110031703B (en) | Abnormal working condition judging method and system for ultra-high voltage transformer | |
| CN105100758A (en) | Method and equipment used for security and protection monitoring as well as camera | |
| CN113595251B (en) | Automatic monitoring and early warning system and method for substation power equipment | |
| CN110031089A (en) | A kind of filtering method and device of running state of transformer vibration sound detection signal | |
| CN104614648A (en) | Electroacoustic combined DC local discharging detecting device | |
| CN114019434A (en) | Transformer residual magnetism detection method, device, system and storage medium | |
| CN116165338A (en) | Method and system for monitoring carbon emission of park | |
| CN103558532A (en) | Partial discharge on-line detection system of high-voltage crosslinked polyethylene power cable | |
| CN108091352A (en) | A kind of audio file processing method, device and storage medium | |
| CN115618205A (en) | Portable voiceprint fault detection system and method | |
| CN202281934U (en) | Power transformer on-line state monitoring device based on audio frequency identification technique | |
| JP4926588B2 (en) | Insulation discharge sound discrimination method and apparatus | |
| CN110161351B (en) | Transformer winding fault test system and diagnosis method under oscillatory wave | |
| CN105425124A (en) | On-line monitoring system for partial discharging of transformer and monitoring method thereof | |
| CN110261741A (en) | Discharge position localization method, device and the terminal device of high-tension switch gear | |
| CN115731943A (en) | Plosive detection method, plosive detection system, storage medium and electronic equipment | |
| CN115219955A (en) | Transformer winding deformation detection system and detection method thereof |
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 |