CN115979598A - Electric reactor mechanical defect early warning method and system based on deep learning - Google Patents
Electric reactor mechanical defect early warning method and system based on deep learning Download PDFInfo
- Publication number
- CN115979598A CN115979598A CN202210459165.XA CN202210459165A CN115979598A CN 115979598 A CN115979598 A CN 115979598A CN 202210459165 A CN202210459165 A CN 202210459165A CN 115979598 A CN115979598 A CN 115979598A
- Authority
- CN
- China
- Prior art keywords
- signal
- estimated
- vibration
- original
- sound
- 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
Images
Classifications
-
- 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
- Y02E30/00—Energy generation of nuclear origin
- Y02E30/30—Nuclear fission reactors
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a reactor mechanical defect early warning method and system based on deep learning. Wherein, the method comprises the following steps: extracting an original sound signal and an original vibration signal of the reactor; filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal; calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal; according to the SRU neural network, the cepstrum coefficients of the sound signals and the vibration signals are deeply learned, and the mechanical defects of the reactor are early warned. After interference signal filtering and time frequency spectrum dimensionality reduction, a data-driven reactor mechanical defect early warning model is established through a deep learning algorithm, the running state of the reactor is accurately mastered, and the safe running level of the reactor is improved.
Description
Technical Field
The invention relates to the technical field of reactor defect early warning, in particular to a reactor mechanical defect early warning method and system based on deep learning.
Background
The conventional reactor mechanical defect early warning mainly adopts a defect diagnosis method based on signal analysis, utilizes a signal analysis theory to obtain a plurality of characteristic vectors of a system in a deep level in a time domain and a frequency domain, utilizes the relation between the characteristic vectors and system defects to early warn the defects, but is very difficult to establish an analysis model of the characteristic vectors and the defects.
Disclosure of Invention
According to the invention, the reactor mechanical defect early warning method and system based on deep learning are provided, so as to solve the technical problem that in the prior art, a defect diagnosis method based on signal analysis needs to obtain a plurality of characteristic vectors, but an analysis model of the characteristic vectors and defects is very difficult to establish.
According to a first aspect of the invention, a reactor mechanical defect early warning method based on deep learning is provided, and comprises the following steps:
extracting an original sound signal and an original vibration signal of the reactor;
filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal;
calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
according to the SRU neural network, the cepstrum coefficients of the sound signals and the vibration signals are deeply learned, and the mechanical defects of the reactor are early warned.
Optionally, filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal, includes:
decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet respectively to obtain a sound sub-band signal and a vibration sub-band signal;
and calculating mutual information values of the sound subband signal and the vibration subband signal.
Optionally, the method further includes filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal, and further includes:
selecting subspace signals in the mutual information value, reconstructing the subspace signals, and determining reconstructed signals;
performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix, and determining an estimated sound source signal and an estimated vibration source signal.
Optionally, calculating cepstral coefficients of the estimated sound source signal and the estimated vibration source signal comprises:
framing the estimated sound source signal and the estimated vibration source signal, and determining framing data;
applying a Hamming window to each frame of data to carry out windowing, and determining each frame of data after windowing;
and carrying out fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and carrying out modulus squaring on the frequency spectrum of each frame of data to obtain a power spectrum.
Optionally, calculating cepstral coefficients of the estimated sound source signal and the estimated vibration source signal further comprises:
smoothing the power spectrum according to a triangular band-pass filter, and filtering harmonic waves;
and solving the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
According to another aspect of the invention, a reactor mechanical defect early warning system based on deep learning is also provided, and comprises:
the original signal extracting module is used for extracting an original sound signal and an original vibration signal of the electric reactor;
the interference signal filtering module is used for filtering interference signals in the original sound signal and the original vibration signal and determining an estimated sound source signal and an estimated vibration source signal;
a cepstrum coefficient calculating module for calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
and the early warning mechanical defect module is used for deeply learning the cepstrum coefficients of the sound signal and the vibration signal according to the SRU neural network and early warning the mechanical defect of the reactor.
Optionally, the interference signal filtering module includes:
the original signal decomposition submodule is used for decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet to obtain a sound subband signal and a vibration subband signal;
and the mutual information value calculating submodule is used for calculating the mutual information values of the sound subband signal and the vibration subband signal.
Optionally, the module for filtering an interference signal further includes:
a signal reconstruction determining submodule for selecting subspace signals in the mutual information value, reconstructing the subspace signals and determining reconstructed signals;
a separation matrix obtaining submodule for performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and the estimated sound source signal determining submodule is used for performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix and determining an estimated sound source signal and an estimated vibration source signal.
Optionally, the module for calculating cepstral coefficients includes:
the framing submodule is used for framing the estimated sound source signal and the estimated vibration source signal and determining framing data;
the windowing submodule is used for applying a Hamming window to each frame of data to carry out windowing processing and determining each frame of data after windowing;
and the power spectrum obtaining submodule is used for carrying out fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and obtaining the power spectrum by carrying out modulus calculation and squaring on the frequency spectrum of each frame of data.
Optionally, the module for calculating cepstral coefficients further includes:
the smoothing power spectrum submodule is used for smoothing the power spectrum according to the triangular band-pass filter and filtering harmonic waves;
and the cepstrum coefficient obtaining submodule is used for solving the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
According to another aspect of the embodiments of the present invention, there is provided a computer-readable storage medium storing a computer program for executing the method according to any one of the above-mentioned embodiments of the present invention.
Therefore, the physical mechanism of the mechanical defects of the reactor does not need to be mastered, an analytic model of the relation between the characteristics of the sound signals and the vibration signals and the mechanical defects of the reactor does not need to be established, after interference signal filtering and time-frequency spectrum dimension reduction, a data-driven reactor mechanical defect early warning model is established through a deep learning algorithm, the running state of the reactor is accurately mastered, and the safe running level of the reactor is improved.
Drawings
A more complete understanding of exemplary embodiments of the present invention may be had by reference to the following drawings in which:
fig. 1 is a schematic flow chart of a reactor mechanical defect early warning method based on deep learning according to the embodiment;
fig. 2 is a schematic flow chart of a reactor mechanical defect warning method according to the embodiment;
fig. 3 is a schematic diagram of a reactor mechanical defect early warning system based on deep learning according to the embodiment.
Detailed Description
The exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, however, the present invention may be embodied in many different forms and is not limited to the embodiments described herein, which are provided for complete and complete disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention. In the drawings, the same units/elements are denoted by the same reference numerals.
Unless otherwise defined, terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Further, it will be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense.
According to a first aspect of the present invention, there is provided a method 100, as illustrated with reference to fig. 1, the method 100 comprising:
s101, extracting an original sound signal and an original vibration signal of the reactor;
s102, filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal; (ii) a
S103, calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
and S104, deeply learning the cepstrum coefficients of the sound signal and the vibration signal according to the SRU neural network, and early warning the mechanical defect of the reactor.
Specifically, in step S101, a reactor sound signal and a vibration signal are extracted;
in step S102, an SDICA blind source separation algorithm is applied to filter out interference signals in original sound signals and vibration signals;
in step S103, a Mel-time frequency spectrum is applied, mel cepstrum coefficients of the sound signal and the vibration signal are calculated, and dimension reduction processing is performed;
in step S104, the SRU neural network is applied to deeply learn the cepstrum coefficients of the sound signal and the vibration signal, so that the early warning of the mechanical defect of the reactor is realized.
The SDICA blind source separation algorithm comprises the following steps:
decomposing the sound signal and the vibration signal of the reactor by using a wavelet packet to obtain a sub-band signal;
calculating mutual information value (MI) of each subspace of the wavelet decomposed signals;
selecting one or more subspaces with the minimum MI value, namely relatively independent subspaces, and reconstructing the subspaces;
performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and blind source separation is carried out on the original sound signal and the vibration signal by applying a separation matrix.
The Mel-time spectrum calculation process comprises the following steps:
framing the signals after blind source separation, and framing by taking the signals with the duration of 1s as a sample, wherein the frame length of each frame is 0.04s, and the frame shift is 0.01s;
applying a Hamming window to each frame of data to perform windowing;
performing fast Fourier transform on each frame of windowed data to obtain a frequency spectrum of each frame of data, and performing modulo squaring on the frequency spectrum to obtain a power spectrum;
smoothing the power spectrum by using a 50Hz frequency-doubled triangular band-pass filter, and filtering harmonic waves;
logarithm is obtained from the output of each filter bank, and then Discrete Cosine Transform (DCT) is performed to obtain the cepstrum coefficient (FMCC) of each frame data.
Referring to fig. 2, performing wavelet packet decomposition on the extracted sound signal and vibration signal of the reactor, extracting a self-contained signal, then extracting the self-contained signal to calculate a mutual information value, selecting 1 or more subspace reconstruction signals with the minimum mutual information value, calculating the reconstruction signals by applying a fastICA decomposition algorithm to obtain a separation matrix, and then calculating the sound signal and the vibration signal to obtain an estimation source signal; estimating a power spectrum of each frame of data obtained by calculating a source signal through framing, windowing and fast Fourier transform, filtering by a 50Hz frequency-doubled triangular band-pass filter, solving the logarithm and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data; and (3) taking cepstrum coefficients of the sound and vibration signals as input of an SRU neural network to perform early warning on mechanical defects of the reactor.
Optionally, filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal, includes:
decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet to obtain a sound subband signal and a vibration subband signal;
and calculating mutual information values of the sound subband signal and the vibration subband signal.
Optionally, the method further includes filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal, and further includes:
selecting subspace signals in the mutual information value, reconstructing the subspace signals, and determining reconstructed signals;
performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix, and determining an estimated sound source signal and an estimated vibration source signal.
Optionally, calculating cepstral coefficients of the estimated sound source signal and the estimated vibration source signal comprises:
framing the estimated sound source signal and the estimated vibration source signal, and determining frame data;
applying a Hamming window to each frame of data to perform windowing processing, and determining each frame of data after windowing;
and performing fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and performing modulo squaring on the frequency spectrum of each frame of data to obtain a power spectrum.
Optionally, calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal further comprises:
smoothing the power spectrum according to a triangular band-pass filter, and filtering harmonic waves;
and (4) calculating the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
Therefore, the physical mechanism of the mechanical defects of the reactor does not need to be mastered, an analytic model of the relation between the characteristics of the sound signals and the vibration signals and the mechanical defects of the reactor does not need to be established, after interference signal filtering and time-frequency spectrum dimension reduction, a data-driven reactor mechanical defect early warning model is established through a deep learning algorithm, the running state of the reactor is accurately mastered, and the safe running level of the reactor is improved.
According to another aspect of the present invention, there is also provided a reactor mechanical defect warning system 300 based on deep learning, and referring to fig. 3, the system 300 includes:
an original signal extracting module 310, configured to extract an original sound signal and an original vibration signal of the reactor;
a filtering interference signal module 320, configured to filter interference signals in the original sound signal and the original vibration signal, and determine an estimated sound source signal and an estimated vibration source signal;
a cepstrum coefficient calculating module 330 for calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
and the early warning mechanical defect module 340 is used for deeply learning the cepstrum coefficients of the sound signal and the vibration signal according to the SRU neural network and early warning the mechanical defect of the reactor.
Optionally, the interference signal filtering module 320 includes:
the original signal decomposition submodule is used for decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet to obtain a sound sub-band signal and a vibration sub-band signal;
and the mutual information value calculating submodule is used for calculating the mutual information values of the sound subband signal and the vibration subband signal.
Optionally, the interference signal filtering module 320 further includes:
a signal reconstruction determining submodule for selecting subspace signals in the mutual information value, reconstructing the subspace signals and determining reconstructed signals;
a separation matrix obtaining submodule for performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and the estimated sound source signal determining submodule is used for performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix and determining an estimated sound source signal and an estimated vibration source signal.
Optionally, the module 340 for calculating cepstral coefficients includes:
the framing submodule is used for framing the estimated sound source signal and the estimated vibration source signal and determining framing data;
the windowing submodule is used for applying a Hamming window to each frame of data to carry out windowing processing and determining each frame of data after windowing;
and the power spectrum obtaining submodule is used for carrying out fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and carrying out modulo calculation on the frequency spectrum of each frame of data to obtain a power spectrum.
Optionally, the module for calculating cepstral coefficients 340 further includes:
the smoothing power spectrum submodule is used for smoothing the power spectrum according to the triangular band-pass filter and filtering harmonic waves;
and the cepstrum coefficient obtaining submodule is used for solving the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
The reactor mechanical defect early warning system 300 based on deep learning according to the embodiment of the present invention corresponds to the reactor mechanical defect early warning method 100 based on deep learning according to another embodiment of the present invention, and is not described herein again.
According to another aspect of the embodiments of the present invention, there is provided a computer-readable storage medium storing a computer program for executing the method according to any one of the above-mentioned embodiments of the present invention.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The scheme in the embodiment of the application can be implemented by adopting various computer languages, such as object-oriented programming language Java and transliterated scripting language JavaScript.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.
Claims (11)
1. A reactor mechanical defect early warning method based on deep learning is characterized by comprising the following steps:
extracting an original sound signal and an original vibration signal of the reactor;
filtering interference signals in the original sound signal and the original vibration signal, and determining an estimated sound source signal and an estimated vibration source signal;
calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
according to the SRU neural network, the cepstrum coefficients of the sound signals and the vibration signals are deeply learned, and the mechanical defects of the reactor are early warned.
2. The method of claim 1, wherein filtering the original sound signal and the original vibration signal to remove interfering signals, and determining an estimated sound source signal and an estimated vibration source signal comprises:
decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet respectively to obtain a sound sub-band signal and a vibration sub-band signal;
and calculating mutual information values of the sound subband signal and the vibration subband signal.
3. The method of claim 2, wherein filtering the original sound signal and the original vibration signal to remove interfering signals, determining an estimated sound source signal and an estimated vibration source signal, further comprises:
selecting subspace signals in the mutual information value, reconstructing the subspace signals, and determining reconstructed signals;
performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix, and determining an estimated sound source signal and an estimated vibration source signal.
4. The method according to claim 1, wherein calculating cepstral coefficients of the estimated sound source signal and the estimated vibration source signal comprises:
framing the estimated sound source signal and the estimated vibration source signal, and determining framing data;
applying a Hamming window to each frame of data to perform windowing processing, and determining each frame of data after windowing;
and carrying out fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and carrying out modulus squaring on the frequency spectrum of each frame of data to obtain a power spectrum.
5. The method according to claim 4, wherein calculating cepstral coefficients of the estimated sound source signal and the estimated vibration source signal further comprises:
smoothing the power spectrum according to a triangular band-pass filter, and filtering harmonic waves;
and (4) calculating the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
6. The utility model provides a reactor mechanical defect early warning system based on degree of depth study which characterized in that includes:
the original signal extracting module is used for extracting an original sound signal and an original vibration signal of the electric reactor;
the interference signal filtering module is used for filtering interference signals in the original sound signal and the original vibration signal and determining an estimated sound source signal and an estimated vibration source signal;
a cepstrum coefficient calculating module for calculating cepstrum coefficients of the estimated sound source signal and the estimated vibration source signal;
and the early warning mechanical defect module is used for deeply learning the cepstrum coefficients of the sound signal and the vibration signal according to the SRU neural network and early warning the mechanical defect of the reactor.
7. The system of claim 6, wherein the interference signal filtering module comprises:
the original signal decomposition submodule is used for decomposing the original sound signal and the original vibration signal by utilizing a wavelet packet to obtain a sound sub-band signal and a vibration sub-band signal;
and the mutual information value calculating submodule is used for calculating the mutual information values of the sound subband signal and the vibration subband signal.
8. The system of claim 7, wherein the interference signal filtering module further comprises:
a signal reconstruction determining submodule for selecting subspace signals in the mutual information value, reconstructing the subspace signals and determining reconstructed signals;
a separation matrix obtaining submodule for performing fastICA decomposition on the reconstructed signal to obtain a separation matrix;
and the estimated sound source signal determining submodule is used for performing blind source separation on the original sound signal and the original vibration signal according to the separation matrix and determining an estimated sound source signal and an estimated vibration source signal.
9. The system of claim 6, the calculate cepstral coefficients module, comprising:
the framing submodule is used for framing the estimated sound source signal and the estimated vibration source signal and determining framing data;
the windowing submodule is used for applying a Hamming window to each frame of data to carry out windowing processing and determining each frame of data after windowing;
and the power spectrum obtaining submodule is used for carrying out fast Fourier transform on each frame of windowed data to obtain the frequency spectrum of each frame of data, and obtaining the power spectrum by carrying out modulus calculation and squaring on the frequency spectrum of each frame of data.
10. The system of claim 9, wherein the means for calculating cepstral coefficients further comprises:
the smoothing power spectrum submodule is used for smoothing the power spectrum according to the triangular band-pass filter and filtering harmonic waves;
and the cepstrum coefficient obtaining submodule is used for solving the logarithm of the output of each triangular band filter group, and performing discrete cosine transform to obtain the cepstrum coefficient of each frame of data.
11. A computer-readable storage medium, characterized in that the storage medium stores a computer program for performing the method of any of the preceding claims 1-5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210459165.XA CN115979598A (en) | 2022-04-27 | 2022-04-27 | Electric reactor mechanical defect early warning method and system based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210459165.XA CN115979598A (en) | 2022-04-27 | 2022-04-27 | Electric reactor mechanical defect early warning method and system based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115979598A true CN115979598A (en) | 2023-04-18 |
Family
ID=85968680
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210459165.XA Pending CN115979598A (en) | 2022-04-27 | 2022-04-27 | Electric reactor mechanical defect early warning method and system based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115979598A (en) |
-
2022
- 2022-04-27 CN CN202210459165.XA patent/CN115979598A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107705801B (en) | Training method of voice bandwidth extension model and voice bandwidth extension method | |
EP3262639B1 (en) | Apparatus and method for processing an audio signal to obtain a processed audio signal using a target time-domain envelope | |
KR101378696B1 (en) | Determining an upperband signal from a narrowband signal | |
CN106816158B (en) | Voice quality assessment method, device and equipment | |
CN110322891B (en) | Voice signal processing method and device, terminal and storage medium | |
CN110459241B (en) | Method and system for extracting voice features | |
US20100161319A1 (en) | Device and method for generating a complex spectral representation of a discrete-time signal | |
CN112820315B (en) | Audio signal processing method, device, computer equipment and storage medium | |
CN108564965B (en) | Anti-noise voice recognition system | |
EP4189677B1 (en) | Noise reduction using machine learning | |
CN104919524A (en) | Method and apparatus for determining encoding mode, method and apparatus for encoding audio signals, and method and apparatus for decoding audio signals | |
US20130253920A1 (en) | Method and apparatus for robust speaker and speech recognition | |
CN113096682A (en) | Real-time voice noise reduction method and device based on mask time domain decoder | |
Hasan et al. | Preprocessing of continuous bengali speech for feature extraction | |
CN112599148A (en) | Voice recognition method and device | |
Kumaresan et al. | On representing signals using only timing information | |
Kantamaneni et al. | Speech enhancement with noise estimation and filtration using deep learning models | |
CN115979598A (en) | Electric reactor mechanical defect early warning method and system based on deep learning | |
CN113571074B (en) | Voice enhancement method and device based on multi-band structure time domain audio frequency separation network | |
Joy et al. | Deep scattering power spectrum features for robust speech recognition | |
CN116913307A (en) | Voice processing method, device, communication equipment and readable storage medium | |
CN104715756A (en) | Audio data processing method and device | |
CN117288444A (en) | GIS defect determination method and system based on sound vibration signal deep learning | |
Vumanthala et al. | Modified nonlocal means filtering for speech enhancement and its FPGA prototype | |
Ganapathy | Robust speech processing using ARMA spectrogram models |
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 |