CN116380235A - Method and device for detecting abnormal sound of generator rotor based on convolutional neural network - Google Patents
Method and device for detecting abnormal sound of generator rotor based on convolutional neural network Download PDFInfo
- Publication number
- CN116380235A CN116380235A CN202310308572.5A CN202310308572A CN116380235A CN 116380235 A CN116380235 A CN 116380235A CN 202310308572 A CN202310308572 A CN 202310308572A CN 116380235 A CN116380235 A CN 116380235A
- Authority
- CN
- China
- Prior art keywords
- sound
- data
- generator
- abnormal
- spectrogram
- 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
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 64
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 title claims description 37
- 238000012545 processing Methods 0.000 claims abstract description 37
- 238000006243 chemical reaction Methods 0.000 claims abstract description 20
- 238000001514 detection method Methods 0.000 claims abstract description 8
- 238000010606 normalization Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 39
- 239000013598 vector Substances 0.000 claims description 15
- 230000005236 sound signal Effects 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 10
- 230000005856 abnormality Effects 0.000 claims description 7
- 230000003750 conditioning effect Effects 0.000 claims description 6
- 238000004806 packaging method and process Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 4
- 238000005303 weighing Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 6
- 239000004973 liquid crystal related substance Substances 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/18—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being spectral information of each sub-band
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/24—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being the cepstrum
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/45—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of analysis window
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/20—Hydro energy
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Mathematical Physics (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- Human Computer Interaction (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Computational Mathematics (AREA)
- Computing Systems (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Algebra (AREA)
- Life Sciences & Earth Sciences (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The application provides a generator rotor abnormal sound detection method based on a convolutional neural network, which comprises the following steps: acquiring sound data of a hydraulic generator; carrying out normalization processing on sound data, and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram; inputting the spectrogram into an ONE-CLASS-SVM model to obtain model scores; and judging whether the hydro-generator is abnormal according to the model score. The invention adopting the scheme can accurately judge various fault types of the generator.
Description
Technical Field
The application relates to the technical field of hydroelectric generating set detection, in particular to a method and a device for detecting abnormal sound of a generator rotor based on a convolutional neural network.
Background
In the prior art, the hydroelectric generating set has strong background noise during operation and complex components, and almost covers the whole frequency domain. Traditional signal processing methods such as wavelet, EMD and the like cannot effectively extract fault signals with high signal-to-noise ratio. Under the condition of more fault forms, a situation-division research and discrimination mode is needed, so that the research difficulty is greatly improved.
The existing method for recognizing the state of the unit based on the sound signals adopts a VMD and probability density method to generate feature vectors, and the operation state of the unit is judged by comparing the feature vectors under different operation states. The existing power equipment discharge fault identification method based on wavelet analysis is used for analyzing acoustic signals of power equipment based on wavelet decomposition and a support vector machine. The methods are not realized in the environment of strong background noise, and the signal to noise ratio is high. In order to realize real-time analysis, the sampling rate of the method is only below 4 kilohertz, and the characteristics of the high-frequency signals cannot be fully utilized. And none of the above methods has the ability to determine faults or can only determine a single fault type.
In the existing another judging method, the acoustic sample is subjected to Fourier transformation, corresponding parameters are combined with neural network training to obtain a model, and the test sample is judged through the model. The above method does not fundamentally solve two problems: it is impossible to determine whether the sound source of the abnormal sound comes from inside or outside the generator range; the method only carries out model training on known faults, but actually the rotor has more scratch categories, and the abnormal detection is needed to be realized.
Disclosure of Invention
The present application aims to solve, at least to some extent, one of the technical problems in the related art.
Therefore, a first object of the present application is to provide a method for detecting abnormal sounds of a generator rotor based on a convolutional neural network, which solves the technical problems that the existing method cannot distinguish multiple faults and the fault judgment accuracy is low, and realizes accurate judgment of multiple fault types of a generator.
A second object of the present application is to provide a sound collection device for a hydro-generator.
A third object of the present application is to provide a device for detecting abnormal sounds of a generator rotor based on a convolutional neural network.
To achieve the above objective, an embodiment of a first aspect of the present application provides a method for detecting abnormal sound of a generator rotor based on a convolutional neural network, including: acquiring sound data of a hydraulic generator; carrying out normalization processing on sound data, and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram; inputting the spectrogram into an ONE-CLASS-SVM model to obtain model scores; and judging whether the hydro-generator is abnormal according to the model score.
According to the generator rotor abnormal sound detection method based on the convolutional neural network, sound data are collected through the sequential generator collection device, abnormal recognition is conducted on the generator set according to the sound data through the ONE-CLASS-SVM model, and the abnormal occurrence range is judged. According to the method and the device, the abnormal condition of the generator set is accurately identified through the identification model, and meanwhile, the position of fault occurrence is effectively judged by comprehensively considering the sound data collected by the plurality of sound collecting units.
Optionally, in one embodiment of the present application, the sound data of the hydro-generator is obtained by a hydro-generator collecting device, wherein the hydro-generator collecting device comprises a plurality of sound collecting units, a network relay, a data server, the sound collecting units comprise a sound pickup and a signal processing circuit,
the sound collection units are distributed in a preset range of the stator base and are used for collecting sound signals of the hydraulic generator through the sound pick-up, carrying out signal conditioning, analog-to-digital conversion and digital signal processing on the sound signals through the signal processing circuit, and carrying out TCP/IP protocol packaging on the processed digital signals;
the network relay is used for forwarding the packed data;
and the data server is used for receiving the packed data and storing the packed data.
Optionally, in an embodiment of the present application, MFCC transformation is performed on the processed data to obtain a spectrogram, including:
dividing sound data of the hydraulic generator by taking the normal working time length of the generator as the dividing total length according to the preset window length and the preset moving length to obtain a plurality of data windows;
and carrying out MFCC conversion on each data window to obtain MFCC coefficients of each data window, and generating a spectrogram according to all the MFCC coefficients.
Optionally, in ONE embodiment of the present application, inputting the spectrogram into the ONE-CLASS-SVM model, obtaining the model score includes:
converting the spectrogram into an SPD matrix;
calculating a Riemann distance matrix of the SPD matrix and the Riemann average, and generating a feature vector as a model score according to the Riemann distance matrix, wherein the Riemann average is a matrix which minimizes the Riemann distance of all the SPD matrices in the training data.
Alternatively, in one embodiment of the present application, the SPD matrix is expressed as:
where N is the number of data windows, z= [ Z ] 1 ,z 2 ,...,z N ],z i MFCC coefficients associated for the i-th segment;
the Riemann average is expressed as:
wherein C represents Riemann average, C i SPD matrix representing training data;
the Riemann distance matrix is expressed as:
wherein C represents an SPD matrix, and C represents a Riemann average;
the feature vector is expressed as:
s i =vec(S i )
wherein s is i Representing feature vectors, S i Represents a Riemann distance matrix, vec (S i ) Is a symmetric matrix S i Is a column-wise vector representation of the upper triangle of (c).
Optionally, in an embodiment of the present application, after determining that the hydro-generator has an abnormality, determining whether the occurrence range of the acoustic abnormal event is located in the generator range further includes:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range. .
To achieve the above object, a second aspect of the present invention provides a sound collection device of a hydraulic generator, including a plurality of sound collection units, a network relay, and a data server, where the sound collection units include a sound pickup and a signal processing circuit,
the sound collection units are distributed in a preset range of the stator base and are used for collecting sound signals of the hydraulic generator through the sound pick-up, carrying out signal conditioning, analog-to-digital conversion and digital signal processing on the sound signals through the signal processing circuit, and carrying out TCP/IP protocol packaging on the processed digital signals;
the network relay is used for forwarding the packed data;
and the data server is used for receiving the packed data and storing the packed data.
In order to achieve the above object, an embodiment of a third aspect of the present invention provides a device for detecting abnormal sounds of a generator rotor based on a convolutional neural network, which comprises an acquisition module, a processing module, a model score generation module, and an abnormality judgment module, wherein,
the acquisition module is used for acquiring sound data of the hydraulic generator by using the acquisition device;
the processing module is used for carrying out normalization processing on the sound data and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram;
the model score generation module is used for inputting the spectrogram into the ONE-CLASS-SVM model to obtain model scores;
and the abnormality judging module is used for judging whether the hydraulic generator is abnormal according to the model fraction.
Optionally, in an embodiment of the present application, the processing module is further configured to:
dividing sound data of the hydraulic generator by taking the normal working time length of the generator as the dividing total length according to the preset window length and the preset moving length to obtain a plurality of data windows;
and carrying out MFCC conversion on each data window to obtain MFCC coefficients of each data window, and generating a spectrogram according to all the MFCC coefficients.
Optionally, in one embodiment of the present application, further includes:
the range judging module is used for judging whether the occurrence range of the sound abnormal event is positioned in the generator range after judging that the hydro-generator is abnormal;
the range judging module is specifically configured to:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range.
Additional aspects and advantages of the application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
fig. 1 is a schematic flow chart of a method for detecting abnormal sounds of a generator rotor based on a convolutional neural network according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a generator rotor abnormal sound detection device based on a convolutional neural network according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and intended for the purpose of explaining the present application and are not to be construed as limiting the present application.
The following describes a method and a device for detecting abnormal sound of a generator rotor based on a convolutional neural network according to an embodiment of the present application with reference to the accompanying drawings.
Fig. 1 is a flow chart of a method for detecting abnormal sounds of a generator rotor based on a convolutional neural network according to an embodiment of the present application.
As shown in fig. 1, the method for detecting abnormal sound of a generator rotor based on a convolutional neural network comprises the following steps:
102, carrying out normalization processing on sound data, and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram;
and 104, judging whether the hydro-generator is abnormal according to the model score.
According to the generator rotor abnormal sound detection method based on the convolutional neural network, sound data are collected through the sequential generator collection device, abnormal recognition is conducted on the generator set according to the sound data through the ONE-CLASS-SVM model, and the abnormal occurrence range is judged. According to the method and the device, the abnormal condition of the generator set is accurately identified through the identification model, and meanwhile, the position of fault occurrence is effectively judged by comprehensively considering the sound data collected by the plurality of sound collecting units.
Optionally, in one embodiment of the present application, the sound data of the hydro-generator is obtained by a hydro-generator collecting device, wherein the hydro-generator collecting device is composed of twelve sets of sound collecting units, a network relay and a data server.
Twelve sets of sound acquisition units are distributed around the stator base.
Each set of acquisition unit consists of a pickup and a signal processing circuit.
The functions performed by the signal processing circuit include: signal conditioning: impedance transformation, amplification and filtering; analog-to-digital conversion and digital signal processing: a/D conversion, ARM signal processing.
The signals are packed by ARM chip to complete the final TCP/IP protocol
The network relay completes the forwarding of the packed data.
Optionally, in one embodiment of the present application, one of the sound collection units processes as follows: the data server normalizes the sound data;
obtaining a data window according to a preset window length, and performing MFCC (frequency-division multiplexing) conversion on the data window to obtain coefficients;
obtaining a next data window and MFCC coefficients of the window according to a preset moving length;
obtaining all MFCC coefficients according to the normal working time length of the generator, and obtaining a spectrogram;
finally, the same processing steps are used for obtaining the spectrograms of other acquisition units.
Optionally, in ONE embodiment of the present application, the spectrogram is input into an ONE-CLASS-SVM model with RBF kernel, resulting in model scores, which are feature vectors, and the feature vectors generated by the SVM model separate the abnormal data from the normal data in a high-dimensional kernel space. After the SVM training is completed, the anomaly score (SPD matrix) of the new data is defined as its respective feature s i Distance from the trained separation hyperplane, specifically:
converting the spectrogram into a symmetric positive SPD matrix;
calculating a Riemann distance matrix of the SPD matrix and the Riemann average, and generating a feature vector as a model score according to the Riemann distance matrix, wherein the Riemann average is a matrix which minimizes the Riemann distance of all the SPD matrices in the training data.
Alternatively, in one embodiment of the present application, the SPD matrix is expressed as:
wherein, the liquid crystal display device comprises a liquid crystal display device,n is the number of data windows, z= [ Z ] 1 ,z 2 ,...,z N ]Let z be i Is an hidden variable of AE, or an MFCC coefficient related to the ith segment;
the Riemann average is a matrix that makes all training samples matrix space C i ∈C samples Is the least Riemann distance. The method is obtained by solving the following optimization problems:
wherein, the liquid crystal display device comprises a liquid crystal display device,represents Riemann average, C i SPD matrix representing training samples;
measurement using the Riemann distance formula:
where logm is a matrix logarithm operator, C 1 ,C 2 Is a symmetric positive definite matrix (SPD).
In the embodiment of the application, the distance matrix of the SPD and the Riemann of the Riemann average is expressed as:
the eigenvector of sample i will be:
s i =vec(S i )
wherein s is i Representing feature vectors, S i Represents a Riemann distance matrix, vec (S i ) Is a symmetric matrix S i Upper triangle column stack of (2)And (5) vector representation.
In addition, to adjust the model to accommodate domain migration problems, several Riemann averages are used. These averages are calculated on feature matrices constructed by spectral clustering. The affinity metric is defined by the Riemann distance. Then, each sample matrix S is calculated in the same way i And calculating the nearest group average value by using the Riemann average calculation mode.
Optionally, in an embodiment of the present application, after determining that the hydro-generator has an abnormality, determining whether the occurrence range of the acoustic abnormal event is located in the generator range further includes: obtaining accurate time of occurrence of abnormal event, calculating difference value between abnormal sound event time of other sound collecting units and the reference by taking time of sensor for recording the abnormal event as reference, and maximum value<=v Sound velocity And judging that the abnormal event occurs in the generator range according to the rotor diameter.
The specific steps of judgment include:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range.
In order to achieve the above embodiment, the present application further provides a sound collection device of a hydraulic generator, including a plurality of sound collection units, a network relay, and a data server, where the sound collection units include a sound pickup and a signal processing circuit,
the sound collection units are distributed in a preset range of the stator base and are used for collecting sound signals of the hydraulic generator through the sound pick-up, carrying out signal conditioning, analog-to-digital conversion and digital signal processing on the sound signals through the signal processing circuit, and carrying out TCP/IP protocol packaging on the processed digital signals;
the network relay is used for forwarding the packed data;
and the data server is used for receiving the packed data and storing the packed data.
In order to achieve the above embodiment, the present application further provides a device for detecting abnormal sounds of a generator rotor based on a convolutional neural network.
Fig. 2 is a schematic structural diagram of a generator rotor abnormal sound detection device based on a convolutional neural network according to an embodiment of the present application.
As shown in fig. 2, the device for detecting abnormal sound of the generator rotor based on the convolutional neural network comprises an acquisition module, a processing module, a model score generation module and an abnormality judgment module, wherein,
the acquisition module is used for acquiring sound data of the hydraulic generator by using the acquisition device;
the processing module is used for carrying out normalization processing on the sound data and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram;
the model score generation module is used for inputting the spectrogram into the ONE-CLASS-SVM model to obtain model scores;
and the abnormality judging module is used for judging whether the hydraulic generator is abnormal according to the model fraction.
Optionally, in an embodiment of the present application, the processing module is further configured to:
dividing sound data of the hydraulic generator by taking the normal working time length of the generator as the dividing total length according to the preset window length and the preset moving length to obtain a plurality of data windows;
and carrying out MFCC conversion on each data window to obtain MFCC coefficients of each data window, and generating a spectrogram according to all the MFCC coefficients.
Optionally, in one embodiment of the present application, further includes:
the range judging module is used for judging whether the occurrence range of the sound abnormal event is positioned in the generator range after judging that the hydro-generator is abnormal;
the range judging module is specifically configured to:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range.
It should be noted that the foregoing explanation of the embodiment of the method for detecting abnormal sound of the generator rotor based on the convolutional neural network is also applicable to the device for detecting abnormal sound of the generator rotor based on the convolutional neural network of this embodiment, and will not be repeated here.
In the description of the present specification, a description referring to the terms "one embodiment," "some embodiments," "examples," "particular examples," or "some examples," etc., means 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 application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "plurality" is at least two, such as two, three, etc., unless explicitly defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and additional implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. As with the other embodiments, if implemented in hardware, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product.
The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like. Although embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives, and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (10)
1. The method for detecting the abnormal sound of the generator rotor based on the convolutional neural network is characterized by comprising the following steps of:
acquiring sound data of a hydraulic generator;
carrying out normalization processing on the sound data, and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram;
inputting the spectrogram into an ONE-CLASS-SVM model to obtain model scores;
and judging whether the hydro-generator is abnormal according to the model score.
2. The method of claim 1, wherein the sound data of the hydro-generator is obtained by a hydro-generator collection device, wherein the hydro-generator collection device comprises a plurality of sound collection units, a network relay, a data server, wherein the sound collection units comprise sound pickups and signal processing circuits,
the sound collection units are distributed in a preset range of the stator base and are used for collecting sound signals of the hydraulic generator through the sound pick-up, carrying out signal conditioning, analog-to-digital conversion and digital signal processing on the sound signals through the signal processing circuit, and carrying out TCP/IP protocol packaging on the processed digital signals;
the network relay is used for forwarding the packed data;
the data server is used for receiving the packed data and storing the packed data.
3. The method of claim 1, wherein performing MFCC transformation on the processed data to obtain a spectrogram comprises:
dividing sound data of the hydraulic generator by taking the normal working time length of the generator as the dividing total length according to the preset window length and the preset moving length to obtain a plurality of data windows;
and carrying out MFCC conversion on each data window to obtain MFCC coefficients of each data window, and generating the spectrogram according to all the MFCC coefficients.
4. The method of claim 3, wherein said inputting the spectrogram into an ONE-CLASS-SVM model to obtain a model score comprises:
converting the spectrogram into an SPD matrix;
calculating a Riemann distance matrix of the SPD matrix and the Riemann average, and generating a feature vector according to the Riemann distance matrix as the model score, wherein the Riemann average is a matrix which minimizes the Riemann distance of all the SPD matrices in training data.
5. The method of claim 4, wherein the SPD matrix is represented as:
where N is the number of data windows, z= [ Z ] 1 ,z 2 ,...,z N ],z i MFCC coefficients associated for the i-th segment;
the Riemann average is expressed as:
wherein C represents the Riemann average, C i An SPD matrix representing the training data;
the Riemann distance matrix is expressed as:
the feature vector is expressed as:
s i =vec(S i )
wherein s is i Representing the feature vector, S i Represents the Riemann distance matrix, vec (S i ) To pair(s)Weighing matrix S i Is a column-wise vector representation of the upper triangle of (c).
6. The method of claim 2, further comprising, after determining that the hydro-generator is abnormal, determining whether an occurrence range of an acoustic anomaly event is located within a generator range, the determining comprising:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range.
7. The sound collecting device of the hydraulic generator is characterized by comprising a plurality of sound collecting units, a network relay and a data server, wherein the sound collecting units comprise sound pickups and signal processing circuits,
the sound collection units are distributed in a preset range of the stator base and are used for collecting sound signals of the hydraulic generator through the sound pick-up, carrying out signal conditioning, analog-to-digital conversion and digital signal processing on the sound signals through the signal processing circuit, and carrying out TCP/IP protocol packaging on the processed digital signals;
the network relay is used for forwarding the packed data;
the data server is used for receiving the packed data and storing the packed data.
8. The generator rotor abnormal sound detection device based on the convolutional neural network is characterized by comprising an acquisition module, a processing module, a model score generation module and an abnormal judgment module, wherein,
the acquisition module is used for acquiring sound data of the hydraulic generator by using the acquisition device according to claim 7;
the processing module is used for carrying out normalization processing on the sound data and carrying out MFCC (frequency division multiplexing) conversion on the processed data to obtain a spectrogram;
the model score generation module is used for inputting the spectrogram into an ONE-CLASS-SVM model to obtain model scores;
and the abnormality judgment module is used for judging whether the hydraulic generator is abnormal according to the model score.
9. The apparatus of claim 8, wherein the processing module is further to:
dividing sound data of the hydraulic generator by taking the normal working time length of the generator as the dividing total length according to the preset window length and the preset moving length to obtain a plurality of data windows;
and carrying out MFCC conversion on each data window to obtain MFCC coefficients of each data window, and generating the spectrogram according to all the MFCC coefficients.
10. The apparatus as recited in claim 8, further comprising:
the range judging module is used for judging whether the occurrence range of the sound abnormal event is positioned in the generator range after judging that the hydro-generator is abnormal;
the range judging module is specifically configured to:
acquiring abnormal occurrence time according to the sound data; obtaining the shortest travel time T collected in a plurality of sound collection units min And a maximum travel time T max ;
Calculate T ref Rotor diameter/V Sound velocity If (T) min ) 2 +(T max ) 2 <(T ref ) 2 And judging that the abnormal event occurs in the generator range.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310308572.5A CN116380235A (en) | 2023-03-27 | 2023-03-27 | Method and device for detecting abnormal sound of generator rotor based on convolutional neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310308572.5A CN116380235A (en) | 2023-03-27 | 2023-03-27 | Method and device for detecting abnormal sound of generator rotor based on convolutional neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116380235A true CN116380235A (en) | 2023-07-04 |
Family
ID=86976227
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310308572.5A Pending CN116380235A (en) | 2023-03-27 | 2023-03-27 | Method and device for detecting abnormal sound of generator rotor based on convolutional neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116380235A (en) |
-
2023
- 2023-03-27 CN CN202310308572.5A patent/CN116380235A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110307994B (en) | Abnormal sound detection device and abnormal sound detection method | |
US8533131B2 (en) | Method and device for classification of sound-generating processes | |
CN111354366B (en) | Abnormal sound detection method and abnormal sound detection device | |
CN109034046B (en) | Method for automatically identifying foreign matters in electric energy meter based on acoustic detection | |
Skowronski et al. | Acoustic detection and classification of microchiroptera using machine learning: lessons learned from automatic speech recognition | |
CN110044623B (en) | Intelligent rolling bearing fault identification method based on empirical mode decomposition residual signal characteristics | |
CN109855874B (en) | Random resonance filter for enhancing detection of weak signals in vibration assisted by sound | |
Gundewar et al. | Bearing fault diagnosis using time segmented Fourier synchrosqueezed transform images and convolution neural network | |
CN110018322B (en) | Rotating speed detection method and system based on deep learning | |
CN111508517A (en) | Intelligent micro-motor product control method based on noise characteristics | |
CN112052712B (en) | Power equipment state monitoring and fault identification method and system | |
CN115618205A (en) | Portable voiceprint fault detection system and method | |
Madain et al. | Fault diagnosis in vehicle engines using sound recognition techniques | |
CN115467787A (en) | Motor state detection system and method based on audio analysis | |
CN114739671A (en) | Bearing fault diagnosis method based on improved generalized S transformation | |
CN109934136B (en) | Rolling bearing fault diagnosis method based on Duffing vibrator and eigen mode component | |
CN114689298A (en) | Fault detection method and device for walking part of coal mining machine | |
CN112420074A (en) | Method for diagnosing abnormal sound of motor of automobile rearview mirror | |
CN116380235A (en) | Method and device for detecting abnormal sound of generator rotor based on convolutional neural network | |
CN113392874A (en) | Rail vehicle abnormal state diagnosis method and device and terminal equipment | |
CN114242085A (en) | Fault diagnosis method and device for rotating equipment | |
CN113139430A (en) | Acoustic signal analysis method for fault detection, and method and device for fault detection | |
CN113963719A (en) | Deep learning-based sound classification method and apparatus, storage medium, and computer | |
CN116482526A (en) | Analysis system for non-fault phase impedance relay | |
JP2002323371A (en) | Sound diagnostic device and sound diagnostic method |
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 |