CN112329914B - Fault diagnosis method and device for buried transformer substation and electronic equipment - Google Patents
Fault diagnosis method and device for buried transformer substation and electronic equipment Download PDFInfo
- Publication number
- CN112329914B CN112329914B CN202011157113.4A CN202011157113A CN112329914B CN 112329914 B CN112329914 B CN 112329914B CN 202011157113 A CN202011157113 A CN 202011157113A CN 112329914 B CN112329914 B CN 112329914B
- Authority
- CN
- China
- Prior art keywords
- data
- quantity data
- sample
- electrical quantity
- voiceprint
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000012545 processing Methods 0.000 claims abstract description 44
- 238000007781 pre-processing Methods 0.000 claims abstract description 35
- 238000012549 training Methods 0.000 claims description 46
- 230000006870 function Effects 0.000 claims description 15
- 230000008030 elimination Effects 0.000 claims description 8
- 238000003379 elimination reaction Methods 0.000 claims description 8
- 230000010354 integration Effects 0.000 claims description 8
- 230000009467 reduction Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 7
- 230000008447 perception Effects 0.000 abstract description 3
- 238000013528 artificial neural network Methods 0.000 description 11
- 238000004458 analytical method Methods 0.000 description 10
- 238000012360 testing method Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 210000002569 neuron Anatomy 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 5
- 238000009432 framing Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012847 principal component analysis method Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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
Abstract
The embodiment of the invention provides a fault diagnosis method and device for a buried substation and electronic equipment, wherein the method comprises the following steps: acquiring electric quantity data and non-electric quantity data of the buried transformer substation; respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input; and inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried substation. The electric quantity data and the non-electric quantity data of the buried transformer substation are predicted through the pre-trained meta-integrated learning model, so that accurate perception of the running state of the buried transformer is realized, the fault diagnosis accuracy is improved, and the running reliability and safety of the buried transformer substation are further improved.
Description
Technical Field
The present invention relates to the field of data processing, and in particular, to a fault diagnosis method and apparatus for a buried substation, and an electronic device.
Background
The advent of buried substations has provided solutions to the problems of increased power demand and land resource shortage resulting from the rapid evolution of cities, but has also brought many new problems. Most of the underground substations are preassembled, the internal space is narrow, and the overhaul access opening only allows one person to pass through, so that the underground transformer in the pit is very inconvenient to overhaul, replace and the like. Therefore, it is important to sense and diagnose the comprehensive state of the buried transformer station and accurately sense the operation state of the buried transformer so as to improve the operation reliability and safety of the buried transformer station.
The existing technologies for transformer state sensing and fault diagnosis mainly comprise an analysis technology of dissolved gas in oil, a partial discharge monitoring technology and the like. The analysis technology of the dissolved gas in the oil has the problems that the coding boundary is too absolute and the diagnostic equipment is large in size. And partial discharge monitoring techniques are susceptible to electromagnetic interference. Meanwhile, most transformer state sensing and diagnosis technologies only diagnose through a single state quantity, and reliability and accuracy are to be improved. Because most of the buried transformer substations are preassembled and the fully-sealed box body is adopted on the box body, the buried transformer is positioned in a closed and narrow buried space, the area of an overhaul access opening is small, and the traditional transformer fault diagnosis mode is not suitable for the buried transformer substations.
Disclosure of Invention
The embodiment of the invention provides a fault diagnosis method for an underground transformer substation, which can diagnose faults of the underground transformer substation through electric quantity data and non-electric quantity data, can accurately sense the running state of the underground transformer, and improves the fault diagnosis accuracy, thereby improving the running reliability and safety of the underground transformer substation.
In a first aspect, an embodiment of the present invention provides a fault diagnosis method for a buried substation, including:
acquiring electric quantity data and non-electric quantity data of the buried transformer substation;
respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input;
and inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried substation.
Optionally, the non-electrical quantity data includes voiceprint data of the buried substation.
Optionally, the step of preprocessing the non-electrical quantity data specifically includes:
carrying out noise elimination treatment on the voiceprint data;
and extracting features of the noise-removed voiceprint data to obtain time-frequency features of the voiceprint data.
Optionally, the meta integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network and an integrated network, and the step of inputting the data to be input into the pre-trained meta integrated learning model for processing and outputting the fault diagnosis result of the buried substation specifically includes:
inputting the time-frequency characteristics of the voiceprint data into the sub-classification network to obtain a first output;
inputting the electrical quantity data and non-electrical quantity data into the integrated weight dynamic iteration network to obtain second output, wherein the non-electrical quantity data comprises time-frequency characteristics of the voiceprint data;
and integrating the first output and the second output through an integrated network, and outputting to obtain a fault diagnosis result of the buried substation.
Optionally, the training of the meta integrated learning model includes the following steps:
acquiring sample electric quantity data and sample non-electric quantity data of the buried substation under different running states;
respectively preprocessing the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data;
training the meta-integrated learning model based on the sample set data.
Optionally, the sample non-electrical quantity data includes sample voiceprint data of the buried substation in different operation states, and the step of preprocessing the sample electrical quantity data and the sample non-electrical quantity data respectively to obtain sample set data specifically includes:
carrying out noise elimination treatment on the sample voiceprint data;
performing feature extraction on the denoised sample voiceprint data to obtain time-frequency features of the sample voiceprint data;
and respectively carrying out standardization processing and dimension reduction processing on the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data, wherein the sample set data comprises sample electrical quantity data and sample non-electrical quantity data, and the sample non-electrical quantity data comprises time-frequency characteristics of the sample voiceprint data.
Optionally, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iteration network and an integrated network, and the step of training the meta-integrated learning model based on the sample set data specifically includes:
inputting the time-frequency characteristics of the sample voiceprint data into the sub-classification network to obtain a first training output;
inputting the sample electrical quantity data and the sample non-electrical quantity data into the integrated weight dynamic iteration network to obtain a second training output;
integrating the first training output and the second training output through an integrated network to obtain a forward propagation result;
calculating a loss function of the meta-integrated learning model based on the forward propagation result;
and minimizing a loss function of the meta-integrated learning model through iterative training to obtain a trained meta-integrated learning model.
In a second aspect, an embodiment of the present invention further provides a fault diagnosis apparatus for a buried substation, where the apparatus includes:
the first acquisition module is used for acquiring electric quantity data and non-electric quantity data of the buried substation;
the first preprocessing module is used for respectively preprocessing the electric quantity data and the non-electric quantity data to obtain data to be input;
and the output module is used for inputting the data to be input into a pre-trained meta-integrated learning model for processing and outputting a fault diagnosis result of the buried substation.
In a third aspect, an embodiment of the present invention provides an electronic device, including: the system comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor realizes the steps in the fault diagnosis method of the buried substation provided by the embodiment of the invention when executing the computer program.
In the embodiment of the invention, the electrical quantity data and the non-electrical quantity data of the buried transformer substation are acquired; respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input; and inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried substation. The electric quantity data and the non-electric quantity data of the buried transformer substation are predicted through the pre-trained meta-integrated learning model, so that accurate perception of the running state of the buried transformer is realized, the fault diagnosis accuracy is improved, and the running reliability and safety of the buried transformer substation are further improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the 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 flowchart of a fault diagnosis method of a buried substation provided by an embodiment of the present invention;
FIG. 2 is a flowchart of a method for preprocessing voiceprint data according to an embodiment of the present invention;
FIG. 3 is a flow chart of another method for diagnosing faults of a buried substation provided by an embodiment of the present invention;
FIG. 4 is a schematic diagram of a meta-integrated learning model according to an embodiment of the present invention;
FIG. 5 is a flowchart of a method for training a meta-integrated learning model according to an embodiment of the present invention;
FIG. 5a is a graph showing the comparison of diagnostic accuracy of different fault diagnosis methods according to an embodiment of the present invention;
fig. 6 is a block diagram of a fault diagnosis apparatus of a buried substation according to an embodiment of the present invention;
fig. 7 is a block diagram of a fault diagnosis apparatus of another buried substation according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
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.
Referring to fig. 1, fig. 1 is a flowchart of a fault diagnosis method of an underground substation according to an embodiment of the present invention, as shown in fig. 1, including the following steps:
s1, acquiring electric quantity data and non-electric quantity data of the buried transformer substation.
In the embodiment of the present invention, the electrical quantity data may include electrical quantity data such as voltage, current, power, etc., and the non-electrical quantity data may include non-electrical quantity data such as voiceprint, temperature, humidity, etc.
The electrical quantity data and the non-electrical quantity data can be acquired by sensors such as a sound sensor, a voltage sensor, a current sensor, a power sensor, a temperature sensor and a temperature sensor which are installed in the buried transformer substation in advance.
Optionally, the electrical quantity data and the non-electrical quantity data may be data collected in real time, and are further divided by a preset time interval to obtain a plurality of groups of electrical quantity data and non-electrical quantity data, and each group of electrical quantity data and non-electrical quantity data is numbered to distinguish different groups of electrical quantity data and non-electrical quantity data. For example, the acquired electric quantity data and the non-electric quantity data may be divided at intervals of 10 seconds.
S2, preprocessing the electrical quantity data and the non-electrical quantity data respectively to obtain data to be input.
In the embodiment of the invention, the preprocessing can be data standardization and data dimension reduction.
Specifically, the electrical quantity data and the non-electrical quantity data may be normalized respectively, and the electrical quantity data and the non-electrical quantity data may be converted into dimensionless data, so that the electrical quantity data and the non-electrical quantity data may be in the same number level.
The electrical quantity data and the non-electrical quantity data can be subjected to dimension reduction processing respectively through a principal component analysis method, and the high-dimension data is reduced to a preset dimension, for example, the data higher than 3 dimensions is reduced to the data of 3 dimensions. The calculation speed can be improved by reducing the dimension of the data.
Optionally, the non-electrical quantity data comprises voiceprint data of the buried substation. The preprocessing of the non-electrical quantity data further includes preprocessing voiceprint data. Specifically, referring to fig. 2, fig. 2 is a flowchart of a voiceprint data preprocessing method according to an embodiment of the present invention, as shown in fig. 2, including the following steps:
s21, noise elimination processing is carried out on the voiceprint data.
In the embodiment of the invention, the noise elimination processing is performed on the voiceprint data so as to eliminate the environmental noise of the buried transformer substation.
S22, carrying out feature extraction on the voice print signal after noise cancellation to obtain the time-frequency feature of voice print data.
In the embodiment of the invention, since the voiceprint data acquired by the sound sensor is a time domain signal, the available characteristics are less, so that the voiceprint data can be converted from the time domain signal to the frequency domain signal, the voiceprint data can be subjected to time-frequency analysis, and the time-frequency characteristics of the voiceprint data can be extracted. It can be appreciated that the time-frequency characteristics include the time-domain characteristics and the frequency-domain characteristics of the voiceprint data, so that the available characteristics of the voiceprint data are enriched.
Specifically, the time-frequency analysis method of the voiceprint data can be used as a Short-time Fourier transform (Short-time Fourier transform, STFT), so that the frequency characteristic of the voiceprint data can be considered, and the time sequence change of the voiceprint data can be considered.
Furthermore, when the short-time Fourier transform is performed on the voiceprint data, the window width and the window sliding step length can be set in advance according to experience or needs, and are not changed in the processing process of the voiceprint data. The above-described sliding step length can be understood as the overlapping length between two adjacent frames of voiceprint data.
It will be appreciated that the time domain resolution may be considered a fixed value since the window length and window sliding step size are no longer changed during processing of the voiceprint data. Meanwhile, according to the Senburg inaccuracy principle, the frequency domain resolution is also a fixed value, that is to say, the time domain resolution and the frequency domain resolution cannot be dynamically adjusted along with the frequency.
In the embodiment of the invention, the time-frequency analysis can be performed on the voiceprint signal by adopting the short-time Fourier transform of the self-adaptive time scale. Specifically, first, the voiceprint data is processed with an initial window type, window length, sliding step size, and fast fourier transform (Fast Fourier transform, FFT) point number. The window type can select a window with better sidelobe suppression, the sliding step length can be 100%, and the number of the fast Fourier transform points can be smaller. Therefore, the voiceprint data can be rapidly processed, and the processing speed of the voiceprint data is improved. And then, according to the frequency of the voiceprint data, adjusting the window type, the window length, the sliding step length and the number of fast Fourier transform points to meet the time-frequency analysis requirement. Specifically, a window with a narrower main lobe width can be changed, and meanwhile, a sliding step length of 30% -50% is adopted, and the number of fast Fourier transform points is increased for adjustment. And obtaining the window type, window length, sliding step length and fast Fourier transform point number which meet the time-frequency analysis requirement through the adjustment of the self-adaptive time scale. And then carrying out short-time Fourier transform processing on the voiceprint data according to the window type, the window length, the sliding step length and the fast Fourier transform points which meet the time-frequency analysis requirements.
More specifically, the method of overlapping segmentation can be adopted to carry out frame processing on the voiceprint data, so that the voiceprint frames are in smooth transition, and the continuity of the voiceprint data is maintained. Let the length of each piece of voiceprint data be:
N＝f s ×t
wherein f s And t is the sampling time length of the voiceprint data.
Further, if the frame length of the short-time fourier transform is nfft and the overlap length between two adjacent voiceprint frames is overlap, the voiceprint data after framing is:
{y 1 ,y 2 ,…,y m }
wherein, m is the total frame number of the total voiceprint frame, and can be calculated by the following formula:
after the voiceprint data subjected to framing processing is obtained, windowing processing can be performed on the voiceprint data subjected to framing processing, and windowed voiceprint frame data is obtained. Specifically, the window function y can be obtained by calculation according to the adaptive time scale rule window And window length, multiply the voiceprint data after framing with window function to get the voiceprint frame data after windowing:
wherein y is i For the ith voiceprint frame data after framing,and (5) windowing the corresponding ith voiceprint frame data.
After the windowed voiceprint frame data is obtained, fast Fourier transform point number processing can be carried out on the windowed voiceprint frame data, and the voiceprint data is converted from a time domain signal to a frequency domain signal through the fast Fourier transform point number processing, so that frequency and amplitude information corresponding to the voiceprint data at each moment are obtained, and the time-frequency characteristics of the voiceprint data in the embodiment of the invention are obtained.
And S3, inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting a fault diagnosis result of the buried substation.
In an embodiment of the present invention, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network, and an integrated network. Referring to fig. 3, fig. 3 is a flowchart of another fault diagnosis method of an underground substation provided by the embodiment of the present invention, the non-electrical data includes voiceprint data of the underground substation, where the voiceprint data is a time-frequency characteristic, and specifically includes the following steps:
s31, inputting the time-frequency characteristics of the voiceprint data into a sub-classification network to obtain a first output.
Wherein the number of the sub-classification networks is n, and n is greater than or equal to 1. The above-mentioned sub-classification network may also be referred to as a classifier, where n sub-classification network outputs result in n first outputs.
S32, inputting the electrical quantity data and the non-electrical quantity data into the integrated weight dynamic iteration network to obtain a second output.
In an embodiment of the present invention, the non-electrical data includes a time-frequency characteristic of voiceprint data. The integrated weight dynamic iteration network comprises an input layer, a hidden layer and an output layer, wherein the input layer is used for inputting electric quantity data and non-electric quantity data, the hidden layer is used for calculating intermediate hidden characteristics, and the output layer comprises n neurons and is used for outputting n second outputs.
And S33, integrating the first output and the second output through an integrated network, and outputting a fault diagnosis result of the buried substation.
Specifically, taking FIG. 4 as an example, FIG. 4 is an example of the present inventionIn the architecture diagram of the meta-integrated learning model provided in the embodiment, as shown in fig. 4, the time-frequency characteristics of voiceprint data are input into a sub-classification network 1, sub-classification networks 2, … and a sub-classification network n to respectively obtain a first output c 1 First output c 2 …, first output c n The method comprises the steps of carrying out a first treatment on the surface of the Inputting the electrical quantity data and the non-electrical quantity data into an integrated weight dynamic iteration network, and respectively obtaining a second output O through n neurons of an output layer 1 Second output O 2 …, second output O n The method comprises the steps of carrying out a first treatment on the surface of the Will output a first output c 1 And a second output O 1 Performs a multiplication operation to output a first output c 2 And a second output O 2 The multiplication is performed until the first output c n And a second output O n And performing multiplication operation, accumulating, completing integration processing, and outputting a fault diagnosis result of the buried substation.
In the embodiment of the invention, the electrical quantity data and the non-electrical quantity data of the buried transformer substation are acquired; respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input; and inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried substation. The electric quantity data and the non-electric quantity data of the buried transformer substation are predicted through the pre-trained meta-integrated learning model, so that accurate perception of the running state of the buried transformer is realized, the fault diagnosis accuracy is improved, and the running reliability and safety of the buried transformer substation are further improved.
It should be noted that, the fault diagnosis method of the buried substation provided by the embodiment of the invention can be applied to devices such as a mobile phone, a monitor, a computer, a server and the like which can perform fault diagnosis of the buried substation.
Optionally, referring to fig. 5, fig. 5 is a flowchart of a meta-integrated learning model training method provided by an embodiment of the present invention, as shown in fig. 5, specifically including the following steps:
s51, acquiring sample electrical quantity data and sample non-electrical quantity data of the buried substation under different running states.
In the embodiment of the invention, the different operation states comprise a normal operation state, a fault state and the like.
The sample electrical quantity data may include electrical quantity data such as voltage, current, and power, and the sample non-electrical quantity data may include non-electrical quantity data such as voiceprint, temperature, and humidity.
The sample electrical quantity data and the sample non-electrical quantity data can be acquired by sensors such as a sound sensor, a voltage sensor, a current sensor, a power sensor, a temperature sensor and the like which are pre-installed in the buried transformer substation.
Optionally, the sample electrical quantity data and the sample non-electrical quantity data may be data collected in real time, and further divided by a preset time interval to obtain multiple groups of sample electrical quantity data and sample non-electrical quantity data, and numbering each group of sample electrical quantity data and sample non-electrical quantity data to distinguish different groups of sample electrical quantity data and sample non-electrical quantity data. For example, the acquired sample electrical quantity data and the sample non-electrical quantity data may be divided at intervals of 10 seconds.
S52, respectively preprocessing the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data.
In the embodiment of the invention, the preprocessing can be data standardization and data dimension reduction.
Specifically, the sample electrical quantity data and the sample non-electrical quantity data may be normalized respectively, and the sample electrical quantity data and the sample non-electrical quantity data may be converted into dimensionless data, which may be lower in that the sample electrical quantity data and the sample non-electrical quantity data are both in the same number level.
The electrical quantity data and the non-electrical quantity data of the sample can be subjected to dimension reduction processing respectively through a principal component analysis method, and the high-dimension data is reduced to a preset dimension, for example, the data higher than 3 dimensions is reduced to the data of 3 dimensions. The calculation speed can be improved by reducing the dimension of the data.
Optionally, the non-electrical quantity data comprises voiceprint data of the buried substation. The preprocessing of the sample non-electrical quantity data further comprises preprocessing sample voiceprint data. Preprocessing the sample voiceprint data may implement steps S21 and S22 of the corresponding portion with reference to fig. 2.
In the embodiment of the invention, the processed sample electrical quantity data and sample non-electrical quantity data can be subjected to data set division, and the sample electrical quantity data and the sample non-electrical quantity data are divided into a training set and a testing set, wherein the training set and the testing set are called sample set data.
The training set is used for training the meta-integrated learning model, and the testing set is used for testing the performance of the meta-integrated learning model.
And S53, training the meta integrated learning model based on the sample set data.
In an embodiment of the present invention, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network, and an integrated network. Specifically, the number of the sub-classification networks is n, and n is greater than or equal to 1. The above-mentioned sub-classification network may also be referred to as a classifier, where n sub-classification network outputs result in n first outputs. The integrated weight dynamic iteration network comprises an input layer, a hidden layer and an output layer, wherein the input layer is used for inputting electric quantity data and non-electric quantity data, the hidden layer is used for calculating intermediate hidden characteristics, and the output layer comprises n neurons and is used for outputting n second outputs.
In the first stage of training, n sub-classification networks are trained. And inputting the time-frequency characteristics of the preprocessed voiceprint data into n sub-classification networks, wherein the output of each classification network is integer data representing the buried transformation state. The n sub-classification networks may be similar classifiers with different super parameters, such as n XGBOOST classification networks with different super parameters, and the n sub-classification networks may also be different classification networks, such as XGBOOST, support vector machine, random forest, etc.
In the second training stage, the integrated weight dynamic iterative network comprises an integrated weight dynamic decision network, and the preprocessed sample electrical quantity data and sample are subjected toThe non-electrical quantity data input integrated weight dynamic decision network is realized through a BP neural network, a linear activation function is used by a hidden layer, a Softmax is used as an activation function by an output layer, and n neurons are shared by the output layer. The n outputs of BP neural network are O respectively 1 ,O 2 ,…,O n The following conditions are satisfied:
O 1 +O 2 +…+O n ＝1
then, the outputs of the n sub-classification networks are integrated with the n outputs of the BP neural network to obtain the final state of the buried transformer. The weight of the integration layer is fixed to be 1, the bias is fixed to be 0, and the weight and the bias are not updated in the training process of the meta-integration learning model. The output of the classification network is integrated with the output of the BP neural network as shown in the following equation:
wherein y is the final state of the buried transformer, O p E, R is the p-th output of the BP neural network; c (C) q ∈R m For the output of the sub-classification network q, m is the number of samples.
Training the model by adopting a meta-integration machine learning method, mainly by updating the weights and the biases of all layers in the BP neural network so as to influence the output O of the BP neural network p Ultimately affecting the determination of the buried state. Training can be performed by the following loss function:
l (W, b) is a loss function; n is the number of samples of a batch of training sets; x is x (k) The method is the input of a k sample of a batch of training set in the BP neural network (namely buried variable electric quantity data and non-electric quantity data); h is a W,b (x) Is a forward propagation function; y is (k) Is the label of the kth sample in the training set. The updating of weights and biases in the BP neural network follows the following rules:
wherein,is the weight of the (j) th neuron of the (1-1) th layer of the BP neural network to the (i) th neuron of the (l) th layer of the BP neural network,>is the bias of the ith neuron of the first layer; alpha is the learning rate.
In the embodiment of the invention, the trained model is used for comprehensive state sensing and diagnosis of the buried transformer, and according to the fault diagnosis method of the buried transformer substation provided by the embodiment of the invention, the fault diagnosis method is realized by programming in Python, testing is carried out by using a testing set, and the comprehensive state sensing and diagnosis accuracy of the buried transformer is recorded. The test set information used is shown in table 1:
TABLE 1
Each set of samples in table 1 was tested and compared with IEC three ratio method, support vector machine (Support vectormachine, SVM) algorithm based on analysis of dissolved gas in oil, XGBOOST algorithm based on voiceprint, respectively, and the test results obtained are shown in table 2:
TABLE 2
As can be seen from table 2, the diagnosis accuracy of the meta-integrated machine learning method based on the voiceprint signal and comprehensively considering the electric quantity and the non-electric quantity data is higher than that of the IEC three-ratio method which does not consider the voiceprint and uses a single feature quantity and the SVM algorithm based on the analysis of the dissolved gas in the oil. Meanwhile, as known by a comparison meta-integration machine learning method and a voice print-based XGBOOST algorithm, the fault diagnosis based on voice print signals is the same, and the diagnosis accuracy of the meta-integration machine learning is higher than that of a single XGBOOST classifier. In conclusion, the method has higher accuracy in comprehensive state sensing and diagnosis of the buried transformer substation. As shown in fig. 5a, with the increase of the number of samples, the fault diagnosis method for the buried substation provided by the embodiment of the invention has more obvious advantages in terms of large-scale state sensing and diagnosis.
Referring to fig. 6, fig. 6 is a schematic structural diagram of a fault diagnosis device of an underground substation according to an embodiment of the present invention, as shown in fig. 6, the device includes:
a first obtaining module 601, configured to obtain electrical quantity data and non-electrical quantity data of a buried substation;
a first preprocessing module 602, configured to respectively preprocess the electrical quantity data and the non-electrical quantity data to obtain data to be input;
and the output module 603 is configured to input the data to be input into a pre-trained meta-integrated learning model for processing, and output a fault diagnosis result of the buried substation.
Optionally, the non-electrical quantity data includes voiceprint data of the buried substation.
Optionally, the pair of first preprocessing modules 602 specifically includes:
the first denoising unit is used for denoising the voiceprint data;
the first extraction unit is used for carrying out feature extraction on the noise-removed voiceprint data to obtain time-frequency features of the voiceprint data.
Optionally, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network and an integrated network, and the first output module 603 specifically includes:
the first processing unit is used for inputting the time-frequency characteristics of the voiceprint data into the sub-classification network to obtain a first output;
the second processing unit is used for inputting the electrical quantity data and the non-electrical quantity data into the integrated weight dynamic iteration network to obtain second output, wherein the non-electrical quantity data comprises the time-frequency characteristics of the voiceprint data;
and the third processing unit is used for carrying out integrated processing on the first output and the second output through an integrated network and outputting to obtain a fault diagnosis result of the buried substation.
Optionally, as shown in fig. 7, the apparatus further includes:
the second obtaining module 604 is configured to obtain sample electrical quantity data and sample non-electrical quantity data of the buried substation under different operation states;
a second preprocessing module 605 respectively preprocessing the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data;
a training module 606 is configured to train the meta-integrated learning model based on the sample set data.
Optionally, the sample non-electrical quantity data includes sample voiceprint data of the buried substation under different operation states, and the second preprocessing module 605 specifically includes:
the second denoising unit is used for denoising the sample voiceprint data;
the second extraction unit is used for extracting the characteristics of the denoised sample voiceprint data to obtain the time-frequency characteristics of the sample voiceprint data;
and the fourth processing unit is used for respectively carrying out standardization processing and dimension reduction processing on the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data, wherein the sample set data comprises sample electrical quantity data and sample non-electrical quantity data, and the sample non-electrical quantity data comprises time-frequency characteristics of the sample voiceprint data.
Optionally, the training module 606 specifically includes:
the fifth processing unit is used for inputting the time-frequency characteristics of the sample voiceprint data into the sub-classification network to obtain a first training output;
the sixth processing unit is used for inputting the sample electric quantity data and the sample non-electric quantity data into the integrated weight dynamic iteration network to obtain a second training output;
a seventh processing unit, configured to integrate the first training output and the second training output through an integrated network to obtain a forward propagation result;
a loss calculation unit for calculating a loss function of the meta-integrated learning model based on the forward propagation result;
and the iteration unit is used for obtaining the trained meta-integrated learning model by iteration training and minimizing the loss function of the meta-integrated learning model.
The fault diagnosis device for the buried substation provided by the embodiment of the invention can be applied to equipment such as a mobile phone, a monitor, a computer, a server and the like which can perform fault diagnosis of the buried substation.
The fault diagnosis device of the buried substation provided by the embodiment of the invention can realize each process realized by the fault diagnosis method of the buried substation in the method embodiment, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
Referring to fig. 8, fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 8, including: a memory 802, a processor 801, and a computer program stored on the memory 802 and executable on the processor 801, wherein:
the processor 801 is configured to call a computer program stored in the memory 802, and execute the following steps:
acquiring electric quantity data and non-electric quantity data of the buried transformer substation;
respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input;
and inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried substation.
Optionally, the non-electrical quantity data includes voiceprint data of the buried substation.
Optionally, the step of preprocessing the non-electrical quantity data performed by the processor 801 specifically includes:
carrying out noise elimination treatment on the voiceprint data;
and extracting features of the noise-removed voiceprint data to obtain time-frequency features of the voiceprint data.
Optionally, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network and an integrated network, and the step of inputting the data to be input into the pre-trained meta-integrated learning model for processing, and outputting the fault diagnosis result of the buried substation, which is executed by the processor 801, specifically includes:
inputting the time-frequency characteristics of the voiceprint data into the sub-classification network to obtain a first output;
inputting the electrical quantity data and non-electrical quantity data into the integrated weight dynamic iteration network to obtain second output, wherein the non-electrical quantity data comprises time-frequency characteristics of the voiceprint data;
and integrating the first output and the second output through an integrated network, and outputting to obtain a fault diagnosis result of the buried substation.
Optionally, the training of the meta-integrated learning model performed by the processor 801 includes the steps of:
acquiring sample electric quantity data and sample non-electric quantity data of the buried substation under different running states;
respectively preprocessing the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data;
training the meta-integrated learning model based on the sample set data.
Optionally, the sample non-electrical quantity data includes sample voiceprint data of the buried substation in different operation states, and the step of preprocessing the sample electrical quantity data and the sample non-electrical quantity data performed by the processor 801 to obtain sample set data specifically includes:
carrying out noise elimination treatment on the sample voiceprint data;
performing feature extraction on the denoised sample voiceprint data to obtain time-frequency features of the sample voiceprint data;
and respectively carrying out standardization processing and dimension reduction processing on the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data, wherein the sample set data comprises sample electrical quantity data and sample non-electrical quantity data, and the sample non-electrical quantity data comprises time-frequency characteristics of the sample voiceprint data.
Optionally, the meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network and an integrated network, and the step of training the meta-integrated learning model based on the sample set data performed by the processor 801 specifically includes:
inputting the time-frequency characteristics of the sample voiceprint data into the sub-classification network to obtain a first training output;
inputting the sample electrical quantity data and the sample non-electrical quantity data into the integrated weight dynamic iteration network to obtain a second training output;
integrating the first training output and the second training output through an integrated network to obtain a forward propagation result;
calculating a loss function of the meta-integrated learning model based on the forward propagation result;
and minimizing a loss function of the meta-integrated learning model through iterative training to obtain a trained meta-integrated learning model.
The electronic device may be a mobile phone, a monitor, a computer, a server, or the like, which can be used for fault diagnosis of a buried transformer substation.
The electronic equipment provided by the embodiment of the invention can realize each process realized by the fault diagnosis method of the buried substation in the embodiment of the method, can achieve the same beneficial effects, and is not repeated here for avoiding repetition.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.
Claims (8)
1. The fault diagnosis method of the buried substation is characterized by comprising the following steps of:
acquiring electric quantity data and non-electric quantity data of the buried transformer substation;
respectively preprocessing the electrical quantity data and the non-electrical quantity data to obtain data to be input;
inputting the data to be input into a pre-trained meta-integrated learning model for processing, and outputting to obtain a fault diagnosis result of the buried transformer substation, wherein the method comprises the following steps: inputting the time-frequency characteristics of the voiceprint data into a sub-classification network to obtain a first output; inputting the electrical quantity data and non-electrical quantity data into an integrated weight dynamic iteration network to obtain second output, wherein the non-electrical quantity data comprises time-frequency characteristics of the voiceprint data; and integrating the first output and the second output through an integration network, and outputting to obtain a fault diagnosis result of the buried substation, wherein the meta-integrated learning model comprises at least one sub-classification network, an integration weight dynamic iteration network and an integration network.
2. The fault diagnosis method of a buried substation according to claim 1, wherein the non-electrical quantity data includes voiceprint data of the buried substation.
3. The method for diagnosing a fault in a buried substation according to claim 2, wherein said step of preprocessing said non-electrical quantity data specifically includes:
carrying out noise elimination treatment on the voiceprint data;
and extracting features of the noise-removed voiceprint data to obtain time-frequency features of the voiceprint data.
4. The fault diagnosis method of a buried substation according to claim 1, wherein the training of the meta integrated learning model comprises the steps of:
acquiring sample electric quantity data and sample non-electric quantity data of the buried substation under different running states;
respectively preprocessing the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data;
training the meta-integrated learning model based on the sample set data.
5. The method for diagnosing faults in a buried substation as set forth in claim 4, wherein said sample non-electrical data includes sample voiceprint data of said buried substation in different operation states, and said step of preprocessing said sample electrical data and sample non-electrical data, respectively, to obtain sample set data specifically includes:
carrying out noise elimination treatment on the sample voiceprint data;
performing feature extraction on the denoised sample voiceprint data to obtain time-frequency features of the sample voiceprint data;
and respectively carrying out standardization processing and dimension reduction processing on the sample electrical quantity data and the sample non-electrical quantity data to obtain sample set data, wherein the sample set data comprises sample electrical quantity data and sample non-electrical quantity data, and the sample non-electrical quantity data comprises time-frequency characteristics of the sample voiceprint data.
6. The method for diagnosing a fault in a buried substation according to claim 5, wherein said meta-integrated learning model includes at least one sub-classification network, an integrated weight dynamic iterative network and an integrated network, and said training said meta-integrated learning model based on said sample set data comprises:
inputting the time-frequency characteristics of the sample voiceprint data into the sub-classification network to obtain a first training output;
inputting the sample electrical quantity data and the sample non-electrical quantity data into the integrated weight dynamic iteration network to obtain a second training output;
integrating the first training output and the second training output through an integrated network to obtain a forward propagation result;
calculating a loss function of the meta-integrated learning model based on the forward propagation result;
and minimizing a loss function of the meta-integrated learning model through iterative training to obtain a trained meta-integrated learning model.
7. A fault diagnosis device for a buried substation, the device comprising:
the first acquisition module is used for acquiring electric quantity data and non-electric quantity data of the buried substation;
the first preprocessing module is used for respectively preprocessing the electric quantity data and the non-electric quantity data to obtain data to be input;
the output module is used for inputting the data to be input into a pre-trained meta-integrated learning model for processing, outputting a fault diagnosis result of the buried substation, and comprises the following steps: inputting the time-frequency characteristics of the voiceprint data into a sub-classification network to obtain a first output; inputting the electrical quantity data and non-electrical quantity data into an integrated weight dynamic iteration network to obtain second output, wherein the non-electrical quantity data comprises time-frequency characteristics of the voiceprint data; and integrating the first output and the second output through an integration network, and outputting to obtain a fault diagnosis result of the buried substation, wherein the meta-integrated learning model comprises at least one sub-classification network, an integration weight dynamic iteration network and an integration network.
8. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the fault diagnosis method of the buried substation according to any one of claims 1 to 6 when the computer program is executed.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011157113.4A CN112329914B (en) | 2020-10-26 | 2020-10-26 | Fault diagnosis method and device for buried transformer substation and electronic equipment |
| PCT/CN2021/091830 WO2022088643A1 (en) | 2020-10-26 | 2021-05-06 | Fault diagnosis method and apparatus for buried transformer substation, and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011157113.4A CN112329914B (en) | 2020-10-26 | 2020-10-26 | Fault diagnosis method and device for buried transformer substation and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112329914A CN112329914A (en) | 2021-02-05 |
| CN112329914B true CN112329914B (en) | 2024-02-02 |
Family
ID=74311701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011157113.4A Active CN112329914B (en) | 2020-10-26 | 2020-10-26 | Fault diagnosis method and device for buried transformer substation and electronic equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112329914B (en) |
| WO (1) | WO2022088643A1 (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112329914B (en) * | 2020-10-26 | 2024-02-02 | 华翔翔能科技股份有限公司 | Fault diagnosis method and device for buried transformer substation and electronic equipment |
| US11677480B2 (en) * | 2021-04-13 | 2023-06-13 | Samsung Electronics Co., Ltd. | Systems, methods, and apparatus for symbol timing recovery based on machine learning |
| CN113866571A (en) * | 2021-08-06 | 2021-12-31 | 厦门欧易奇机器人有限公司 | Partial discharge source positioning method, device and equipment |
| CN115114448B (en) * | 2022-06-02 | 2023-07-14 | 中国电力科学研究院有限公司 | Intelligent multi-mode fusion power consumption inspection method, device, system, equipment and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107634988A (en) * | 2017-08-25 | 2018-01-26 | 深圳市盛路物联通讯技术有限公司 | A kind of method and server of the Fault Diagnosis for Substation based on Internet of Things |
| CN109586239A (en) * | 2018-12-10 | 2019-04-05 | 国网四川省电力公司电力科学研究院 | Intelligent substation real-time diagnosis and fault early warning method |
| CN109740523A (en) * | 2018-12-29 | 2019-05-10 | 国网陕西省电力公司电力科学研究院 | A kind of method for diagnosing fault of power transformer based on acoustic feature and neural network |
| KR101965159B1 (en) * | 2018-11-30 | 2019-08-13 | 문경훈 | System and method of estimating load with null data correction |
| CN110929847A (en) * | 2019-11-15 | 2020-03-27 | 国网浙江省电力有限公司电力科学研究院 | Converter transformer fault diagnosis method based on deep convolutional neural network |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103728551B (en) * | 2013-01-30 | 2016-03-09 | 中国人民解放军海军航空工程学院 | A kind of analog-circuit fault diagnosis method based on cascade integrated classifier |
| CN106597260B (en) * | 2016-12-29 | 2020-04-03 | 合肥工业大学 | Analog circuit fault diagnosis method based on continuous wavelet analysis and ELM network |
| CN107784325B (en) * | 2017-10-20 | 2020-06-23 | 河北工业大学 | Spiral fault diagnosis method based on data-driven incremental fusion |
| CN111653291B (en) * | 2020-06-01 | 2023-02-14 | 莫毓昌 | Intelligent health monitoring method for power equipment based on voiceprint |
| CN112329914B (en) * | 2020-10-26 | 2024-02-02 | 华翔翔能科技股份有限公司 | Fault diagnosis method and device for buried transformer substation and electronic equipment |
-
2020
- 2020-10-26 CN CN202011157113.4A patent/CN112329914B/en active Active
-
2021
- 2021-05-06 WO PCT/CN2021/091830 patent/WO2022088643A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107634988A (en) * | 2017-08-25 | 2018-01-26 | 深圳市盛路物联通讯技术有限公司 | A kind of method and server of the Fault Diagnosis for Substation based on Internet of Things |
| KR101965159B1 (en) * | 2018-11-30 | 2019-08-13 | 문경훈 | System and method of estimating load with null data correction |
| CN109586239A (en) * | 2018-12-10 | 2019-04-05 | 国网四川省电力公司电力科学研究院 | Intelligent substation real-time diagnosis and fault early warning method |
| CN109740523A (en) * | 2018-12-29 | 2019-05-10 | 国网陕西省电力公司电力科学研究院 | A kind of method for diagnosing fault of power transformer based on acoustic feature and neural network |
| CN110929847A (en) * | 2019-11-15 | 2020-03-27 | 国网浙江省电力有限公司电力科学研究院 | Converter transformer fault diagnosis method based on deep convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112329914A (en) | 2021-02-05 |
| WO2022088643A1 (en) | 2022-05-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112329914B (en) | Fault diagnosis method and device for buried transformer substation and electronic equipment | |
| CN109357749B (en) | DNN algorithm-based power equipment audio signal analysis method | |
| CN110261109B (en) | Rolling bearing fault diagnosis method based on bidirectional memory cyclic neural network | |
| Han et al. | A recursive sparse representation strategy for bearing fault diagnosis | |
| CN110534118A (en) | Transformer/reactor method for diagnosing faults based on Application on Voiceprint Recognition and neural network | |
| CN111044814B (en) | Method and system for identifying transformer direct-current magnetic bias abnormality | |
| Talmon et al. | Single-channel transient interference suppression with diffusion maps | |
| CN111312273A (en) | Reverberation elimination method, apparatus, computer device and storage medium | |
| CN112735456A (en) | Speech enhancement method based on DNN-CLSTM network | |
| CN115758082A (en) | Fault diagnosis method for rail transit transformer | |
| CN111291918B (en) | Rotating machine degradation trend prediction method based on stationary subspace exogenous vector autoregression | |
| Hwang et al. | Application of cepstrum and neural network to bearing fault detection | |
| CN113345399A (en) | Method for monitoring sound of machine equipment in strong noise environment | |
| CN115410589A (en) | Attention generation confrontation voice enhancement method based on joint perception loss | |
| CN116106013A (en) | Bearing fault feature recognition method and system, electronic equipment and storage medium | |
| Rahman et al. | Dynamic time warping assisted svm classifier for bangla speech recognition | |
| CN116773952A (en) | Transformer voiceprint signal fault diagnosis method and system | |
| CN114813129B (en) | Rolling bearing acoustic signal fault diagnosis method based on WPE and EMD | |
| CN111025100A (en) | Transformer ultrahigh frequency partial discharge signal mode identification method and device | |
| Tran et al. | Denoising induction motor sounds using an autoencoder | |
| CN112735477B (en) | Voice emotion analysis method and device | |
| Reshma et al. | A survey on speech emotion recognition | |
| CN115064182A (en) | Fan fault feature identification method of self-adaptive Mel filter in strong noise environment | |
| CN114121025A (en) | Voiceprint fault intelligent detection method and device for substation equipment | |
| Shi et al. | Extraction method of weak underwater acoustic signal based on the combination of wavelet transform and empirical mode decomposition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210303 Address after: 413002 No. 1 Xuefu Road West, Longling Industrial Park, Heshan District, Yiyang City, Hunan Province Applicant after: Huaxiang XiangNeng Technology Co.,Ltd. Address before: 413002 No. 1 Xuefu Road West, Longling Industrial Park, Heshan District, Yiyang City, Hunan Province Applicant before: Huaxiang XiangNeng Technology Co.,Ltd. Applicant before: HUNAN University |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |