CN116449256A - Transformer state fault diagnosis system and method based on voiceprint sensing - Google Patents
Transformer state fault diagnosis system and method based on voiceprint sensing Download PDFInfo
- Publication number
- CN116449256A CN116449256A CN202310366244.0A CN202310366244A CN116449256A CN 116449256 A CN116449256 A CN 116449256A CN 202310366244 A CN202310366244 A CN 202310366244A CN 116449256 A CN116449256 A CN 116449256A
- Authority
- CN
- China
- Prior art keywords
- fault
- voiceprint
- transformer
- data
- faults
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000006870 function Effects 0.000 claims abstract description 28
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 239000004215 Carbon black (E152) Substances 0.000 claims abstract description 7
- HSFWRNGVRCDJHI-UHFFFAOYSA-N alpha-acetylene Natural products C#C HSFWRNGVRCDJHI-UHFFFAOYSA-N 0.000 claims abstract description 7
- 125000002534 ethynyl group Chemical group [H]C#C* 0.000 claims abstract description 7
- 229930195733 hydrocarbon Natural products 0.000 claims abstract description 7
- 150000002430 hydrocarbons Chemical class 0.000 claims abstract description 7
- 238000001514 detection method Methods 0.000 claims description 26
- 238000000605 extraction Methods 0.000 claims description 15
- 238000001228 spectrum Methods 0.000 claims description 13
- 238000009413 insulation Methods 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000004587 chromatography analysis Methods 0.000 claims description 7
- 238000010586 diagram Methods 0.000 claims description 6
- 230000004807 localization Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000011176 pooling Methods 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 238000012544 monitoring process Methods 0.000 abstract description 13
- 230000002159 abnormal effect Effects 0.000 abstract description 6
- 230000008569 process Effects 0.000 abstract description 5
- 230000004044 response Effects 0.000 abstract description 4
- 238000007499 fusion processing Methods 0.000 abstract description 2
- 238000011160 research Methods 0.000 abstract description 2
- 239000003921 oil Substances 0.000 description 14
- 239000007789 gas Substances 0.000 description 10
- 230000036541 health Effects 0.000 description 7
- 239000000126 substance Substances 0.000 description 7
- 238000004804 winding Methods 0.000 description 7
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical group [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 5
- 238000001816 cooling Methods 0.000 description 4
- MHAJPDPJQMAIIY-UHFFFAOYSA-N Hydrogen peroxide Chemical compound OO MHAJPDPJQMAIIY-UHFFFAOYSA-N 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 3
- 239000010721 machine oil Substances 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000013021 overheating Methods 0.000 description 2
- 230000003647 oxidation Effects 0.000 description 2
- 238000007254 oxidation reaction Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 230000005236 sound signal Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 206010067715 Gastrointestinal sounds abnormal Diseases 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000002555 auscultation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000004817 gas chromatography Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- JRZJOMJEPLMPRA-UHFFFAOYSA-N olefin Natural products CCCCCCCC=C JRZJOMJEPLMPRA-UHFFFAOYSA-N 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000001845 vibrational spectrum Methods 0.000 description 1
- 238000009736 wetting 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
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/62—Testing of transformers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention discloses a transformer state fault diagnosis system and a transformer state fault diagnosis method based on voiceprint sensing, which are used for acquiring oil chromatographic diagnosis data, voiceprint diagnosis data, working condition information record data, total hydrocarbon content data after load reduction operation and acetylene total amount data after load reduction operation of a transformer; constructing mass functions under various fault conditions, substituting various data into the mass functions under various fault conditions, and diagnosing the transformer fault by the maximum function value; real-time monitoring of the running state of the transformer is realized in a non-contact mode; through AI learning, the identification and the research of three-phase unbalance, load characteristics and other electrical parameters and partial discharge and no-load abnormal state information in the operation process of the transformer are combined with vibration and other parameters of the operation environment, and information fusion processing and early warning are timely carried out at the front end, so that the real-time performance of the fault response of the transformer is improved, the risk of the occurrence of the fault of the transformer is reduced, and the efficiency and the management and control capability of operation are improved.
Description
Technical Field
The invention relates to the technical field of transformer monitoring, in particular to a transformer fault diagnosis method.
Background
Transformers are important devices for transmitting and distributing electric energy in an electric power system, and the running stability and reliability of the transformers directly determine the reliability of power supply of users. The operation monitoring work of the existing power transformer is basically realized by a manual inspection mode, and the problems of large workload, low efficiency, slow response and the like exist.
Previously, a few experienced operators can judge whether the equipment is abnormal or not by means of sound when the equipment is running. However, the development of the power grid is gradually changed, and with the continuous increase of power equipment, the traditional manual auscultation mode can not meet the actual requirements of intelligent control. And is susceptible to external environmental interference such as background noise, personnel speaking sound, etc.; acoustic and vibration signals generated by power equipment such as transformers, circuit breakers and the like during operation contain a large amount of state information and have identification features like human fingerprints. The voiceprint of the device changes as it becomes defective or fails. And the voiceprint information is accurately identified, so that operation and maintenance personnel can diagnose equipment defects and lock fault reasons.
For example, the CN110864801a transformer noise and vibration on-line monitoring and fault analysis system is fixed in advance at each angle and position of the main transformer equipment through an external patch type acoustic wave sensor, collects and samples main transformer sound in real time, amplifies and filters sound signals and converts the sound signals into digital electric signals, then sends the data to an adjacent sink control processing unit through a wireless or bluetooth communication module, and then sends the data to an on-line monitoring total server and a monitoring background system through a switch, an optical fiber network and the like through the sink control unit. The method not only can analyze the possibility and the reason of various fault types, but also can comprehensively analyze and judge the faults, and can primarily judge the positions of the fault points according to the installation positions of the sensors and respective sampling data.
However, the fault accuracy of the transformer is low by a single sensor, and a conclusion can be obtained by combining the existing comprehensive analysis of multiple parameters.
Disclosure of Invention
The present invention is directed to a transformer fault diagnosis method, so as to solve the problems set forth in the background art.
In order to achieve the above purpose, the present invention provides the following technical solutions:
a transformer state fault diagnosis method based on voiceprint sensing is characterized in that: comprising the following steps:
acquiring oil chromatographic diagnosis data, voiceprint diagnosis data and working condition information recording data (mainly load conditions of the transformer) of the transformer, total hydrocarbon content data after load reduction operation and acetylene total amount data after load reduction operation;
and constructing mass functions under various fault conditions, substituting various data into the mass functions under various fault conditions, and diagnosing the transformer fault by the maximum function value.
Preferably, the faults include open circuit faults, insulation faults, circuit faults, magnetic circuit faults, and no faults;
broken fault mass function:
insulation fault mass function:
circuit fault mass function:
magnetic circuit fault mass function:
fault-free mass function:
wherein k=1-K, K being a normalization constant;
in breaking faults and insulating faults
In circuit faults, magnetic circuit faults and faults-free
Wherein m is 1 (e 1 ) For oil chromatography diagnosis e 1 Probability of occurrence of each fault, m 2 (e 2 ) Diagnosis of voiceprint as e 2 Probability of occurrence of each fault, m 3 (e 3 ) E is recorded as working condition information 3 Probability of occurrence of each fault, m 4 (e 4 ) For a total hydrocarbon content e after load-shedding operation 4 Probability of occurrence of each fault, m 5 (e 5 ) For the total amount of acetylene after load reduction operation to be e 5 And the probability of each fault occurrence is A, and A is the fault type.
A transformer state fault diagnosis system based on voiceprint sensing comprises a shell, and a channel sound source positioning system element, an anti-interference detection element, a voiceprint feature extraction element and a fault voiceprint judgment element which are arranged on the shell;
the channel sound source positioning system element is used for collecting voiceprint data of the transformer and sending the collected voiceprint data to the anti-interference detection element;
the anti-interference detection element carries out filtering processing on the received voiceprint data, and then sends the processed voiceprint data to the voiceprint feature extraction element;
the voiceprint feature extraction element is used for extracting fundamental frequencies and harmonic waves corresponding to a plurality of targets from the processed voiceprint data;
the fault voiceprint judging element is used for comparing the fundamental frequency and harmonic wave of each target with the fundamental frequency and harmonic wave preset by the target to calculate the similarity, judging the fault type of the transformer and outputting a result.
Preferably, the shell is of a horn-shaped structure with an opening at one end and hollow inside.
Preferably, the channel sound source localization system element is disposed at the open end of the housing.
Preferably, the channel sound source localization system component is a 64 channel sound source localization system, which can cover audible sound and ultrasonic frequency bands.
Preferably, the fault voiceprint judging element constructs a voiceprint detection model based on the MobileNet V3, and the voiceprint detection model comprises a voiceprint feature extraction module, a MobileNet V3 module and a circulation module; the voiceprint feature extraction module extracts features through a time domain diagram, the MobileNet V3 module classifies the extracted features, the extracted features are connected into the circulation module, modeling is carried out through information of front and rear frames, and model accuracy is improved.
Preferably, the fault voiceprint judging element introduces an attention mechanism to construct a voiceprint detection model based on MobileNetV3, performs pooling treatment on each channel of the obtained feature matrix, and then obtains an output vector through two full-connection layers, wherein the node number of the first full-connection layer is 1/4 of that of the input feature matrix channel, and the channels of the second full-connection layer are consistent with those of the feature matrix.
Preferably, the loss function is used in the process of training the voiceprint detection modelVerifying;
wherein m and n respectively represent the number of samples and the total number of categories,for cross entropy loss of each sample, p i The probability of being a sample of class i is thus obtained as a voiceprint characteristic of the transformer fault.
Preferably, the device also comprises a vibration sensor, wherein the vibration sensor is used for collecting vibration signals of the transformer during working, extracting corresponding frequency spectrum characteristics through a processor, and taking the obtained frequency spectrum characteristics of the transformer during various faults as comparison characteristics;
the processor analyzes and compares the frequency spectrum characteristic and the comparison characteristic, and diagnoses the working state or the fault type of the transformer.
Compared with the prior art, the invention has the beneficial effects that:
on the premise of not influencing the normal operation of the transformer, the real-time monitoring of the operation state of the transformer is realized in a non-contact mode;
through AI learning, the identification and the research of three-phase unbalance, load characteristics and other electrical parameters and partial discharge and no-load abnormal state information in the operation process of the transformer are combined with vibration and other parameters of the operation environment, and information fusion processing and early warning are timely carried out at the front end, so that the real-time performance of the fault response of the transformer is improved, the risk of the occurrence of the fault of the transformer is reduced, and the efficiency and the management and control capability of operation and detection are improved;
covering audible sound and ultrasonic frequency bands by means of a 64-channel sound source positioning system element, rapidly positioning an abnormal sound source position, analyzing defect types and greatly improving equipment operation and detection efficiency;
the voiceprint feature extraction element obtains the dispersion feature parameter distribution of the main transformer in the non-power-on level and the cooling mode by analyzing the voiceprint feature quantity of the transformer, the dispersion feature parameter distribution is basically similar to the voltage level and the cooling mode of the transformer, the spectrum dispersion feature parameter of the transformer approximately shows Gaussian normal distribution feature and the like, the threshold value parameter of the attention and the abnormality of the health state of the transformer is calculated by weighting each feature parameter, the health index parameter of the transformer is calculated for each group of received detection data, and the parameter is compared with the threshold value to judge the current health state of the transformer;
on the basis of collecting a large number of transformer voiceprint samples, the fault diagnosis accuracy is improved by means of an algorithm.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a voiceprint diagnostic apparatus according to the present invention;
FIG. 2 is a schematic view of a first direction structure of the housing of the present invention;
FIG. 3 is a schematic view of a second direction structure of the housing of the present invention;
FIG. 4 is a schematic diagram of the components of the channel sound source localization system of the present invention;
FIG. 5 is a schematic diagram of the vibration source and propagation path of the transformer according to the present invention.
1. A channel sound source localization system element; 2. a housing; 3. an anti-interference detection element; 4. a voiceprint feature extraction element; 5. and a fault voiceprint judging element.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1-5:
the opening part of the shell 2 is provided with a matched type channel sound source positioning system element 1 which is used in a combined and connected mode, an anti-interference detection element 3 is abutted to one line at the inner end of the channel sound source positioning system element 1, a fault voiceprint judging element 5 is arranged at one side of the voiceprint feature extracting element 4 at intervals, and the fault voiceprint judging element 5 is correspondingly embedded in a local groove point position on the surface of the shell 2.
The shell 2 is in a horn-shaped/cylinder/ear-like structure, so that sound can be collected, the channel sound source positioning system element 1 is a 64-channel sound source positioning system, audible sound and ultrasonic frequency bands can be covered, the abnormal sound source position can be positioned rapidly, the defect type can be analyzed, and the equipment operation and detection efficiency can be improved greatly;
the anti-interference detection element 3 monitors the acoustic signal characteristics of the transformer to obtain the stable signal of the transformer noise, and the fault voiceprint judgment element 5 calculates the health index parameters of the transformer for each group of received detection data;
the working principle of the anti-interference detection element 3 is as follows: the transformer noise obtained by monitoring the characteristics of the transformer acoustic signal is a stable signal, and is mainly concentrated in 50Hz even harmonic frequency components within the 700Hz range, the cooling fan noise is a stable signal, the frequency spectrum is mainly within the 2kHz range, and the energy distribution of other frequency components except the rotating frequency is relatively uniform. And there are many external disturbances such as borborygmus, frog borygmus, bird borygmus, vehicles, etc., which are generally sudden, intermittent and concentrated in a certain range, and the external environment sound disturbance is reduced by means of a packet spectrum method, a short time window method and a filter method.
The principle of operation of the texture feature extraction element 4 is as follows: different transformers, loads, ageing of cooling devices, winding deformation, different harmonic contents, direct current magnetic bias, other working conditions and the like, multi-dimensional transformer voiceprint characteristic quantities are obtained, frequency ratio characteristics, wavelet packet energy and dispersity characteristics are obtained, and concentration trend, discrete degree, bias state and kurtosis of the characteristic quantities are obtained.
The working principle of the fault voiceprint judging element 5 is as follows: and judging the fault type of the transformer by utilizing the characteristic combination, weighting each characteristic parameter, calculating threshold parameters of the attention and abnormality of the health state of the transformer, calculating health index parameters of the transformer for each group of received detection data, comparing the parameters with the threshold values, and judging the current health state of the transformer.
A voiceprint detection model based on MobileNet V3 is adopted, and the model is mainly divided into a voiceprint feature extraction module, a MobileNet V3 module and a circulation module. The voiceprint feature extraction module extracts features through a time domain diagram, the MobileNet V3 module classifies the features and is connected to the circulation module, and modeling is carried out through information of front and rear frames, so that the model accuracy is improved. Using loss functions during trainingPerforming verification that m, n respectively represent the number of samples and the total number of categories,/->For cross entropy loss of each sample, p i The probability that the sample is class i. Therefore, voiceprint characteristics of the faults of the transformer can be obtained, and the real-time performance of fault response of the transformer is improved by matching with electrical parameters such as three-phase unbalance, load characteristics and the like and partial discharge and no-load abnormal state information of the transformer in the operation process acquired in real time.
Compared with the prior model, the MobileNetv3 adds an attention mechanism (SE), performs pooling treatment on each channel of the obtained feature matrix, and then obtains output vectors through two fully-connected layers, wherein the node number of the first fully-connected layer is equal to 1/4 of that of the input feature matrix channel, and the channels of the second fully-connected layer are consistent with the channels of the feature matrix. Through the averaging pooling + two fully connected layers, the output eigenvectors can be understood as a weight relationship for each channel analysis of the eigenvalue matrix before SE, which considers that a more important channel will be given a greater weight and a less important channel dimension will be given a smaller weight. Therefore, compared with the prior model, the classification accuracy rate is increased by 3.2%, and the calculation delay is reduced by 20%.
The device also comprises a vibration sensor, and the vibration sensor is temporarily adsorbed on the outer surface of the transformer oil tank body to be detected through a permanent magnet when in use; the vibration sensor is used for collecting vibration signals of the transformer when the transformer works and extracting corresponding frequency spectrum characteristics through the processor;
as shown in fig. 5, the vibration of the transformer oil tank has multiple sources, the vibration frequencies of the different sources are different, and the vibration spectrum characteristics of the transformer oil tank caused by various conditions are prestored in the processor;
the processor compares the frequency spectrum characteristics of the transformer obtained on site with various pre-stored frequency spectrum characteristics, wherein the frequency spectrum characteristic with the highest similarity is the working state or vibration source of the transformer on site. Determining the source of vibration of the field transformer may provide a reference for subsequent transformer failures.
The fault types diagnosed by the numerical values obtained by the sensors are different, and meanwhile, the fault diagnosis of different monitoring systems often has overlapping parts, so that the fault range of the diagnosis can be narrowed by utilizing the cooperation monitoring of different sensors, for example, when the oil dissolved gas chromatographic monitoring system finds out the overheat fault, the monitoring data of the grounding current monitoring system can be combined for analysis so as to eliminate or confirm whether insulation overheat is caused by the multipoint grounding fault of the iron core, thereby generating corresponding gas. In this way, the scope of faults can be further reduced, and the fault type can be directly determined in some cases;
faults of power transformers are generally classified into external and internal faults. The root causes of internal faults are mainly mechanical loosening of the power supply winding and the insulating iron core, winding resonance, overheating, degradation of insulating oil, oxidation, wetting, pollution of various chemical solid substances of the insulating oil, partial discharge, design and manufacturing defects. External faults are caused by system faults such as short circuits, lightning strikes, system overload and system switching errors. The root causes of internal faults are mainly power supply winding and insulating iron core mechanical loosening, winding resonance, overheat, insulating oil degradation, oxidation, damp, various chemical solid matter pollution of insulating oil, partial discharge, design and manufacturing defects; the traditional method can not accurately detect the loosening faults of structural components such as windings, iron cores, wire clamps and the like, but can effectively and accurately monitor the faults by utilizing vibration.
e 1 For the obtained oil chromatography diagnostic data (using David triangle method), e 2 For voiceprint diagnostic data, e, acquired by the present system 3 Recording, e, for the acquired working condition information 4 Total hydrocarbon content after load-reducing operation e 5 The total amount of acetylene after load reduction operation is obtained; through e 1 -e 5 The method can be used for preliminarily judging and obtaining insulation faults, short circuit faults, discharge faults, protection faults and misoperation faults.
K is a normalization constant, k=1-K; wherein K is a conflict factor, which indicates the conflict between different evidences, and the larger K is, the larger the conflict between different evidences on the surface is, and 1/(1-K) is a normalization factor. m is m 1 、m 2 、m 3 、m 4 、m 5 E respectively 1 、e 2 、e 3 、e 4 、e 5 The corrected fault probability value;
the calculation process of data fusion:
and solving the mass function to obtain the mass function value of each fault, and diagnosing the fault of the transformer by the maximum value of the function.
Winding and core faults are determined and predicted by vibration sensors measuring and analyzing vibration signals of the transformer surface. The determination was performed by gas chromatography. The chemical composition of various chemical substances in the dissolved gas in the machine oil and the content of various gases in the machine oil can be completely and well judged and analyzed after the chemical composition and the content of various gases in the machine oil are detected through chromatography. For example, when chromatographic analysis is performed on the average hydrogen oxide content and other component chemical substance content contained in the dissolved gas in transformer oil, the main causes of gas failures such as iron core and overheating are shown in the analysis of these factors to consider that CH contained in the dissolved gas in transformer oil 4 And the average hydrogen oxide concentration of other components such as tetra-cyclic olefin is relatively high, and the content of CO and CO 2 The average hydrogen oxide concentration of the equal component gas has a smaller variation range than that of the prior gas, and the intermittent fault of multipoint continuous grounding is mainly shown in the chromatographic analysis and comprises the componentsSuch as C 2 H 2 And the like.
Embodiment one:
taking the three steps of circuit fault identification, magnetic circuit fault and no fault as examples:
m 4 (e 4 ) For a total hydrocarbon content e after load-shedding operation 4 Probability of occurrence of each fault, m 5 (e 5 ) For the total amount of acetylene after load reduction operation to be e 5 Is the probability of each fault occurring;
circuit fault mass function:
magnetic circuit fault mass function:
fault-free mass function:
after all formulas are calculated, the circuit fault value is the largest, and the circuit fault is the occurred fault;
through oil chromatography large guard triangle diagnosis, voiceprint diagnosis and working condition information recording, insulation faults, short circuit faults, discharge faults, protection and misoperation faults can be judged.
Embodiment two:
taking the identification of short-circuit faults, insulation faults and discharge faults as examples:
short circuit fault:
insulation failure:
discharge failure:
after all formulas are calculated, the insulation fault value is the largest, and the insulation fault is the generated fault.
In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (10)
1. A transformer state fault diagnosis system based on voiceprint sensing is characterized in that: the system comprises a shell (2), a channel sound source positioning system element (1), an anti-interference detection element (3), a voiceprint feature extraction element (4) and a fault voiceprint judgment element (5), wherein the channel sound source positioning system element (1), the anti-interference detection element (3) and the fault voiceprint judgment element (5) are arranged on the shell (2);
the channel sound source positioning system element (1) is used for collecting voiceprint data of the transformer and sending the collected voiceprint data to the anti-interference detection element (3);
the anti-interference detection element (3) carries out filtering processing on the received voiceprint data, and then sends the processed voiceprint data to the voiceprint feature extraction element (4);
the voiceprint feature extraction element (4) is used for extracting fundamental frequencies and harmonics corresponding to a plurality of targets from the processed voiceprint data;
the fault voiceprint judging element (5) is used for comparing the fundamental frequency and harmonic wave of each target with the fundamental frequency and harmonic wave preset by the target to calculate the similarity, judging the fault type of the transformer and outputting a result.
2. The voiceprint sensing based transformer condition fault diagnosis system of claim 1, wherein: the shell (2) is of a horn-shaped structure with an opening at one end and hollow inside.
3. The voiceprint sensing based transformer condition fault diagnosis system of claim 2, wherein: the channel sound source localization system component (1) is arranged at the open end of the housing (2).
4. The voiceprint sensing based transformer condition fault diagnosis system of claim 1, wherein: the channel sound source positioning system element (1) is a 64-channel sound source positioning system and can cover audible sound and ultrasonic frequency bands.
5. The voiceprint sensing based transformer condition fault diagnosis system of claim 1, wherein: the fault voiceprint judging element (5) constructs a voiceprint detection model based on MobileNet V3, wherein the voiceprint detection model comprises a voiceprint feature extraction module, a MobileNet V3 module and a circulation module; the voiceprint feature extraction module extracts features through a time domain diagram, the MobileNet V3 module classifies the extracted features, the extracted features are connected into the circulation module, modeling is carried out through information of front and rear frames, and model accuracy is improved.
6. The voiceprint sensing based transformer condition fault diagnosis system of claim 5, wherein: the fault voiceprint judging element (5) introduces a attention mechanism to construct a voiceprint detection model based on MobileNetV3, carries out pooling treatment on each channel of the obtained feature matrix, and then obtains an output vector through two full-connection layers, wherein the node number of the first full-connection layer is 1/4 of that of the input feature matrix channel, and the channels of the second full-connection layer are consistent with those of the feature matrix.
7. The voiceprint sensing based transformer condition fault diagnosis system of claim 6, wherein: using loss functions in voiceprint detection model trainingVerifying;
wherein m and n respectively represent the number of samples and the total number of categories,for cross entropy loss of each sample, p i The probability of being a sample of class i is thus obtained as a voiceprint characteristic of the transformer fault.
8. The voiceprint sensing based transformer condition fault diagnosis system of claim 7, wherein: the device also comprises a vibration sensor, wherein the vibration sensor is used for collecting vibration signals of the transformer during working, extracting corresponding frequency spectrum characteristics through a processor and taking the obtained frequency spectrum characteristics of the transformer during various faults as comparison characteristics;
the processor analyzes and compares the frequency spectrum characteristic and the comparison characteristic, and diagnoses the working state or the fault type of the transformer.
9. A transformer state fault diagnosis method based on voiceprint sensing is characterized in that: comprising the following steps:
acquiring oil chromatographic diagnosis data, voiceprint diagnosis data, working condition information recording data, total hydrocarbon content data after load reduction operation and acetylene total amount data after load reduction operation of a transformer;
and constructing mass functions under various fault conditions, substituting various data into the mass functions under various fault conditions, and diagnosing the transformer fault by the maximum function value.
10. The method for diagnosing a state fault of a transformer based on voiceprint sensing according to claim 9, wherein the method comprises the steps of: the faults include open circuit faults, insulation faults, circuit faults, magnetic circuit faults and no faults;
broken fault mass function:
insulation fault mass function:
circuit fault mass function:
magnetic circuit fault mass function:
fault-free mass function:
wherein k=1-K, K being a normalization constant;
among the open circuit faults and insulation faults:
circuit failure, magnetic circuit failure, and no failure:
wherein m is 1 (e 1 ) For oil chromatography diagnosis e 1 Probability of occurrence of each fault, m 2 (e 2 ) Diagnosis of voiceprint as e 2 Probability of occurrence of each fault, m 3 (e 3 ) E is recorded as working condition information 3 Probability of occurrence of each fault, m 4 (e 4 ) For a total hydrocarbon content e after load-shedding operation 4 Probability of occurrence of each fault, m 5 (e 5 ) For the total amount of acetylene after load reduction operation to be e 5 And the probability of each fault occurrence is A, and A is the fault type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310366244.0A CN116449256A (en) | 2023-04-07 | 2023-04-07 | Transformer state fault diagnosis system and method based on voiceprint sensing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310366244.0A CN116449256A (en) | 2023-04-07 | 2023-04-07 | Transformer state fault diagnosis system and method based on voiceprint sensing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116449256A true CN116449256A (en) | 2023-07-18 |
Family
ID=87125039
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310366244.0A Pending CN116449256A (en) | 2023-04-07 | 2023-04-07 | Transformer state fault diagnosis system and method based on voiceprint sensing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116449256A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118335108A (en) * | 2024-06-12 | 2024-07-12 | 国网江西省电力有限公司超高压分公司 | Transformer abnormal sound fault identification method based on tail monkey search algorithm |
| CN118362943A (en) * | 2024-06-19 | 2024-07-19 | 国网山东省电力公司聊城供电公司 | Transformer monitoring device and method based on non-electric quantity comprehensive characteristic information |
| CN118519070A (en) * | 2024-07-23 | 2024-08-20 | 江西华莱电技术有限公司 | Transformer detection method and system |
-
2023
- 2023-04-07 CN CN202310366244.0A patent/CN116449256A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118335108A (en) * | 2024-06-12 | 2024-07-12 | 国网江西省电力有限公司超高压分公司 | Transformer abnormal sound fault identification method based on tail monkey search algorithm |
| CN118362943A (en) * | 2024-06-19 | 2024-07-19 | 国网山东省电力公司聊城供电公司 | Transformer monitoring device and method based on non-electric quantity comprehensive characteristic information |
| CN118519070A (en) * | 2024-07-23 | 2024-08-20 | 江西华莱电技术有限公司 | Transformer detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116449256A (en) | Transformer state fault diagnosis system and method based on voiceprint sensing | |
| CN104749468B (en) | A kind of GIS fault diagnosis systems and its method | |
| CN103576059B (en) | A kind of transformer turn-to-turn electric discharge resultant fault diagnostic method and system | |
| CN113124929A (en) | Transformer substation multi-parameter signal acquisition comprehensive analysis system and method | |
| CN109188197B (en) | Evaluation system and method for health index of high-voltage cable | |
| CN202404166U (en) | On-line monitoring system for vibration performance of transformer | |
| CN110320467B (en) | Low-voltage direct-current circuit breaker fault diagnosis method | |
| CN108828406A (en) | The fault recognition method and its system of non-intrusion type user power utilization | |
| CN111856224A (en) | Partial discharge multi-source fusion detection system and method for high-voltage switch cabinet | |
| CN109443190B (en) | Transient traveling wave-based transformer winding deformation online monitoring method and device | |
| CN112880750A (en) | Transformer multidimensional comprehensive online monitoring intelligent diagnosis system | |
| CN110231529A (en) | A kind of control cabinet intelligent Fault Diagnose Systems and method for diagnosing faults | |
| CN110794301A (en) | Motor life value judgment method | |
| CN110347717A (en) | A kind of big data storage method based on urban electric power equipment monitoring | |
| CN116184265A (en) | Lightning arrester leakage current detection method and system based on multi-classification SVM | |
| CN117763460A (en) | Transformer fault detection method and system based on voiceprint analysis | |
| CN117153193B (en) | Power equipment fault voiceprint recognition method integrating physical characteristics and data diagnosis | |
| CN112684293A (en) | Power distribution system insulation fault diagnosis device | |
| CN108919041A (en) | A kind of transformer winding state on-line monitoring method based on clustering | |
| CN113075512B (en) | Transformer discharge fault diagnosis method and system based on acoustic detection | |
| CN217060335U (en) | Voltage transformer fault monitoring system | |
| RU2814857C1 (en) | System for monitoring and diagnosing state of turbine generator | |
| CN215894820U (en) | 10kV high tension cable exchanges withstand voltage and partial discharge test system | |
| CN114113939B (en) | Transient vibration signal-based transformer fault detection method | |
| WO2024138994A1 (en) | Electrical device sound recognition method based on characteristic frequency band selection |
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 |