CN116007912A - Generator slot wedge tightness detection method - Google Patents
Generator slot wedge tightness detection method Download PDFInfo
- Publication number
- CN116007912A CN116007912A CN202211423742.6A CN202211423742A CN116007912A CN 116007912 A CN116007912 A CN 116007912A CN 202211423742 A CN202211423742 A CN 202211423742A CN 116007912 A CN116007912 A CN 116007912A
- Authority
- CN
- China
- Prior art keywords
- signal
- follows
- frequency
- slot wedge
- generator
- 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
- 238000001514 detection method Methods 0.000 title description 9
- 238000000034 method Methods 0.000 claims abstract description 39
- 230000005236 sound signal Effects 0.000 claims abstract description 29
- 238000001228 spectrum Methods 0.000 claims description 33
- 238000004364 calculation method Methods 0.000 claims description 21
- 238000000354 decomposition reaction Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 5
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 3
- OVBPIULPVIDEAO-LBPRGKRZSA-N folic acid Chemical compound C=1N=C2NC(N)=NC(=O)C2=NC=1CNC1=CC=C(C(=O)N[C@@H](CCC(O)=O)C(O)=O)C=C1 OVBPIULPVIDEAO-LBPRGKRZSA-N 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 6
- 238000012360 testing method Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000005259 measurement Methods 0.000 abstract description 2
- 230000005284 excitation Effects 0.000 description 10
- 230000009466 transformation Effects 0.000 description 4
- 238000002790 cross-validation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000013016 damping Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
Images
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention belongs to the technical field of generators, and particularly relates to a method for detecting tightness of a slot wedge of a generator. The method comprises the following steps: step 1: pretreatment; step 2: extracting fault characteristics; step 3: the frequency domain characteristics of the knock signal are calculated. The invention has the beneficial effects that: the invention can complete diagnosis by only collecting the sound signal at the knocking hammer, and can rapidly and accurately diagnose whether the slot wedge has faults or not by carrying out time-frequency analysis on the collected sound signal. The method has the advantages of simple signal acquisition device, convenient operation, accurate diagnosis result and the like. Through comparison test and measurement, the judgment accuracy rate reaches more than 95%.
Description
Technical Field
The invention belongs to the technical field of generators, and particularly relates to a method for detecting tightness of a slot wedge of a generator.
Background
The generator is an important component of the electric power system and mainly comprises a stator and a rotor and other parts, wherein the slot wedge is a structure for fixing a stator bar in a slot. In the running process of the generator, the electrified stator bar can receive the radial electromagnetic force under the environment of the transverse magnetic field in the stator slot, so that vibration is generated. The long-term vibration of the stator bar can cause the slot wedge plate to loose. The slot wedge plate is loosened, the stator bar vibrates under the action of alternating electromagnetic force, and the insulation layer is damaged along with long-time running of the generator, so that the electric corrosion phenomenon is more severe, the main insulation layer of the stator bar is broken down, even the stator bar is stopped, and therefore serious potential safety hazards are caused.
There are cases showing that the problem of slot wedge loosening occurs when a generator is operated for many years. When the generator is overhauled, the tightness of the slot wedge needs to be checked, and the detection and the re-tightening of the tightness of the stator slot wedge become an important link for the maintenance of the generator. The existing detection method comprises a manual knocking mode and a measuring mode of measuring holes, but in the generator set, the number of the stator slot wedges can be as high as tens of thousands, the existing detection mode has great dependence on experience of operators, the judgment accuracy and the detection efficiency depend on the manual experience, and meanwhile, secondary damage is caused to the slot wedges in the detection process, so that the detection requirement cannot be effectively met.
Disclosure of Invention
The invention aims to provide a method for detecting the tightness of a slot wedge of a generator, which can accurately diagnose the fault of the tightness of the slot wedge of the generator.
The technical scheme of the invention is as follows: a method for detecting tightness of a slot wedge of a generator comprises the following steps:
step 1: pretreatment;
step 2: extracting fault characteristics;
step 3: the frequency domain characteristics of the knock signal are calculated.
The step 1 comprises the following steps:
step 11: a decomposition process;
step 12: and (5) reconstructing.
The step 11 includes the following steps:
let the data sequence s= { S (k), k e Z }, its specific decomposition process is as follows:
step 111: splitting
Splitting a data sequence into an odd sample sequence S 0 ={s 0 (k) K epsilon Z with even sample sequence S e ={s e (k) K e Z, where the odd samples are:
s 0 (k)=s(2k+1),k∈Z
the even samples are:
s e (k)=s(2k),k∈Z
step 112: prediction
Let P (·) be the prediction period, use even samples s e (k) Predicting odd samples s 0 (k) The deviation detail signal d that will be predicted to occur k The calculation is as follows:
d(k)=s 0 (k)-P[s e (k)],k∈Z
the sequence of the detail signal is d= { D (k), k∈z };
step 113: updating
Let U (·) be the updater, s based on the detail signal d (k) e (k) Updating, and using the updated sequence as approximation informationNumber C (k), the approximated signal sequence is c= { C (k), k e Z }, where C (k) is as follows:
c(k)=s e (k)+U[d(k)],k∈Z。
the step 12 includes the following steps:
the decomposition step of step 11 is reversed:
s e (k)=c(k)-U[d(k)],k∈Z
s 0 (k)=d(k)+P[s e (k)],k∈Z
and then combining the odd samples obtained in the steps with the even samples to reconstruct a transformed signal s.
The step 2 comprises the following steps:
extracting and screening the characteristics of the knock signal, extracting the time domain characteristics, the frequency spectrum characteristics and the power spectrum characteristics of the pre-processed knock signal,
time domain characteristics of the knock signal:
root Mean Square (RMS) may be used to represent the effective value of a signal, calculated as follows:
where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal;
variance VAR reflects the degree of a higher average value of the data and can measure the fluctuation degree of the knock signal;
where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is the average value of (2);
the waveform factor K is the ratio of the effective value to the rectification average value, and the calculation formula is as follows:
the peak factor C is the ratio of the peak value X to the root mean square value RMS and is calculated as follows:
kurtosis K q The numerical statistic for reflecting the kurtosis of the waveform is shown as follows:
where N is the number of sample points, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is the average value of (2);
the zero crossing rate ZCR is the number of times that the time domain signal passes through the zero point in a certain time, and represents the vibration number of the time domain signal, and the calculation formula is as follows:
wherein N represents the total number of points of the knock signal, T represents time, and N X represents: when X is true, the output is 1, otherwise, 0 is output.
The step 3 comprises the following steps:
step 31: frequency domain peak and frequency thereof
Respectively finding out the peak value and the corresponding frequency of each frequency interval according to the defined frequency interval;
step 32: area of band interval
And calculating the frequency band amplitude area according to the defined frequency interval, wherein the frequency band amplitude area is the sum of the frequency spectrum amplitudes, and the calculation formula is as follows:
wherein i and j respectively represent a start value and an end value of each interval;
step 33: power spectrum area.
The step 33 includes the following steps:
the power spectrum estimation method of the knocking sound signal by adopting the Welch method comprises the following steps of:
step 331: the pre-processed sequence x (n) of the knock signal is divided into L segments, each segment has a sequence length of M=512, and the overlap length between every two segments of data is Md=256, so that the following can be obtained
Step 332: the window function to be employed is the Hamming window, denoted herein as W (n);
step 333: the power spectrum of each segment of data is calculated respectively, summed and then averaged to obtain the required self-power spectrum estimation, and the calculation formula is as follows:
the normalization factor U in the formula is calculated as follows:
dividing frequency intervals according to the obtained image of the power spectrum, and respectively obtaining the power spectrum area of each frequency interval, wherein the calculation formula of the power spectrum area is as follows:
and the previous candidate feature parameters are all candidate feature parameters for initial selection.
The invention has the beneficial effects that: the invention can complete diagnosis by only collecting the sound signal at the knocking hammer, and can rapidly and accurately diagnose whether the slot wedge has faults or not by carrying out time-frequency analysis on the collected sound signal. The method has the advantages of simple signal acquisition device, convenient operation, accurate diagnosis result and the like. Through comparison test and measurement, the judgment accuracy rate reaches more than 95%.
Drawings
FIG. 1 is a schematic diagram of an embodiment of the present invention;
fig. 2 is a flowchart of a method for detecting tightness of a slot wedge of a generator.
Detailed Description
The invention will be described in further detail with reference to the accompanying drawings and specific examples.
The invention relates to a method for detecting the tightness of a slot wedge of a generator, which is completed by adopting equipment comprising a vibration exciter, a sensor, a collecting card and an upper computer, wherein a vibration excitation source is that a rod piece moves after an electromagnet is electrified, and a knocking head is driven to knock the slot wedge through a connecting plate so as to realize vibration excitation; the acoustic sensor converts acoustic signals generated by excitation into electrical signals; the acquisition card acquires the converted electric signals and transmits the electric signals to the upper computer through the Ethernet to display, store, process and analyze the signals; the upper computer performs feature extraction on sound signals generated by knocking the slot wedge by the vibration exciter through an algorithm, and judges the tightness condition of the slot wedge of the generator according to the result of the feature value. The excitation source utilizes an electromagnet to generate a power source, the electromagnet drives the connecting plate to drive the knocking head to knock the slot wedges, the position of an excitation point of the excitation source selects the end part and the middle part of each slot wedge as excitation points, the electromagnet is controlled to knock the slot wedges according to specific knocking frequency, effective excitation of the slot wedges is realized, and a sound signal is generated. The sound sensor collects slot wedge excitation sound, is connected with the data acquisition card, and the data acquisition card carries out A/D conversion processing on sound signals and then uploads the sound signals to the controller, and data are transmitted to the upper computer through the Ethernet. After the upper computer finishes storing the slot wedge sound signals, firstly, the sound signals are preprocessed by utilizing second generation wavelet transformation, noise interference in the sound signals is effectively removed, and filtering of the signals is finished. Then, extracting fault characteristics from the denoising signal, wherein the signal characteristics comprise frequency band amplitude area, frequency spectrum centroid, waveform factor, third peak frequency and fifth peak frequency, and can effectively represent the fault degree of slot wedge tightness. And finally, classifying fault characteristics through a SVM with optimized optimizing parameters by a cross-validation method, and completing fault diagnosis of slot wedge tightness. And extracting the characteristic value of the signal uploaded to the upper computer, and classifying the characteristic value through a classifier, so as to judge the degree of the slot wedge tightness.
The utility model provides a generator slot wedge elasticity detection method, adopts the electro-magnet to drive to strike the hammer and accomplishes, and the excitation source strikes the slot wedge of generator and produces sound, utilizes sound sensor to gather the sound signal that the slot wedge sent, carries out the preliminary treatment to the signal, then carries out time-frequency analysis to the signal after the filtering, draws slot wedge fault feature to judge whether the generator slot wedge takes place not loose trouble, specifically includes the following step:
step 1: pretreatment of
For the collected original signal, because the existence of environmental noise can influence the analysis and judgment of the signal in the later stage, the original signal needs to be preprocessed, namely noise reduction.
Compared with the traditional wavelet transformation, the second-generation wavelet transformation has the characteristics of symmetry and tight support, the wavelet shape is impact damping oscillation, the transformation phase is linear, and the method is suitable for extracting the characteristics of faults with impact.
The second generation wavelet transform can be divided into two steps of decomposition and reconstruction. The process of decomposition and reconstruction of the second generation wavelet transform based on interpolation subdivision is as follows:
step 11: the decomposition process comprises the following steps: splitting, predicting and updating
Let the data sequence s= { S (k), k e Z }, its specific decomposition process is as follows:
step 111: splitting
Splitting a data sequence into an odd sample sequence S 0 ={s 0 (k) K epsilon Z with even sample sequence S e ={s e (k) K e Z, where the odd samples are:
s 0 (k)=s(2k+1),k∈Z
the even samples are:
s e (k)=s(2k),k∈Z
step 112: prediction
Let P (·) be the prediction period, use even samples s e (k) Predicting odd samples s 0 (k) The deviation detail signal d that will be predicted to occur k The calculation is as follows:
d(k)=s 0 (k)-P[s e (k)],k∈Z
the sequence of the detail signal is d= { D (k), k∈z }.
Step 113: updating
Let U (·) be the updater, s based on the detail signal d (k) e (k) Updating, wherein the updated sequence is used as an approximation signal C (k), the approximation signal sequence is C= { C (k), and k epsilon Z }, and C (k) is as follows:
c(k)=s e (k)+U[d(k)],k∈Z
step 12: reconstruction
The reverse of the above decomposition step is performed:
s e (k)=c(k)-U[d(k)],k∈Z
s 0 (k)=d(k)+P[s e (k)],k∈Z
and then combining the odd samples obtained in the steps with the even samples to reconstruct a transformed signal s.
The method comprises the following steps: fault feature extraction
And extracting and screening characteristics of the knocking sound signals. The time domain feature, the frequency spectrum feature and the power spectrum feature of the pre-processed knock sound signal are extracted, and 9 feature parameters are extracted in total.
Time domain characteristics of the knock signal:
root Mean Square (RMS) may be used to represent the effective value of a signal, calculated as follows:
where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal.
The Variance (VAR) reflects the degree to which the data is biased toward a higher average, and may measure the degree of fluctuation of the percussive signal.
Where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is a mean value of (c).
The waveform factor K is the ratio of an effective value to a rectification average value, is insensitive to the impact load, and has the following calculation formula:
the peak factor C is the ratio of the peak value X (maximum absolute value) to the root mean square value RMS, and can reflect the impact characteristics of the signal, and the calculation formula is as follows:
kurtosis K q The numerical statistic used for reflecting the kurtosis of the waveform is sensitive to impact components, the kurtosis value under normal working conditions accords with normal distribution and is approximately 3, when the impact components appear, the kurtosis value is larger than 3, and the kurtosis calculation formula is as follows:
where N is the number of sample points, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is a mean value of (c).
The zero crossing rate ZCR is the number of times that a time domain signal passes through a zero point in a certain time, represents the vibration number of the time domain signal, and can be used for reflecting the speed degree of signal change, and the calculation formula is as follows:
wherein N represents the total number of points of the knock signal, T represents time, and N X represents: when X is true, the output is 1, otherwise, 0 is output.
Step 3: computing frequency domain features of a knock signal
Firstly, performing spectrum analysis on the knocked sound signal by adopting fast Fourier transform, and dividing a frequency interval according to peak distribution by observing a spectrogram.
Step 31: frequency domain peak and frequency thereof
And respectively finding out the peak value and the corresponding frequency of each frequency interval according to the defined frequency interval.
Step 32: area of band interval
And calculating the frequency band amplitude area according to the defined frequency interval, wherein the frequency band amplitude area is the sum of the frequency spectrum amplitudes, and the calculation formula is as follows:
where i, j represent the start and end values of each interval, respectively.
Step 33: area of power spectrum
The power spectrum estimation method of the knocking sound signal by adopting the Welch method comprises the following steps of:
step 331: the pre-processed sequence x (n) of the knock signal is divided into L segments, each segment has a sequence length of M=512, and the overlap length between every two segments of data is Md=256, so that the following can be obtained
Step 332: the window function to be employed is a hamming window (herein denoted as W (n));
step 333: the power spectrum of each segment of data is calculated respectively, summed and then averaged to obtain the required self-power spectrum estimation, and the calculation formula is as follows:
the normalization factor U in the formula is calculated as follows:
frequency intervals are divided according to the obtained image of the power spectrum, and the power spectrum area of each frequency interval is obtained. The power spectrum area can reflect the condition that the power changes along with the frequency, namely, the distribution condition of the power in the frequency domain, and the calculation is the sum of the amplitude values on the power spectrum. The calculation formula is as follows:
and the previous candidate feature parameters are all candidate feature parameters for initial selection. And then distinguishing conditions of different tightness states of the slot wedge according to each characteristic parameter, then respectively screening the redundancy among each characteristic pair by using an F-ratio method and a Pearson correlation coefficient method on the extracted 10 characteristic parameters, and carrying out characteristic screening based on the F-ratio and the Pearson correlation coefficient method to obtain vibration sound signal characteristics, and finally reserving 3 characteristic parameters. The vibration sound signal characteristics comprise a frequency band interval area, a power spectrum area and a waveform factor.
According to the invention, the 3 extracted characteristic parameters form an input characteristic matrix, and the SVM optimized by the optimizing parameters of the cross-validation method is used for classifying the test set, so that the fault diagnosis of the slot wedge tightness is completed. The test result shows that compared with the KNN algorithm and the BP neural network identification accuracy, the SVM model has better effect.
Claims (7)
1. The method for detecting the tightness of the slot wedge of the generator is characterized by comprising the following steps of:
step 1: pretreatment;
step 2: extracting fault characteristics;
step 3: the frequency domain characteristics of the knock signal are calculated.
2. The method for detecting tightness of slot wedge of generator as claimed in claim 1, wherein said step 1 comprises the steps of:
step 11: a decomposition process;
step 12: and (5) reconstructing.
3. The method for detecting tightness of slot wedge of generator as claimed in claim 2, wherein said step 11 comprises the steps of:
let the data sequence s= { S (k), k e Z }, its specific decomposition process is as follows:
step 111: splitting
Splitting a data sequence into an odd sample sequence S 0 ={s 0 (k) K epsilon Z with even sample sequence S e ={s e (k) K e Z, where the odd samples are:
s 0 (k)=s(2k+1),k∈Z
the even samples are:
s e (k)=s(2k),k∈Z
step 112: prediction
Let P (·) be the prediction period, use even samples s e (k) Predicting odd samples s 0 (k) The deviation detail signal d that will be predicted to occur k The calculation is as follows:
d(k)=s 0 (k)-P[s e (k)],k∈Z
the sequence of the detail signal is d= { D (k), k∈z };
step 113: updating
Let U (·) be the updater, s based on the detail signal d (k) e (k) Updating, wherein the updated sequence is used as an approximation signal C (k), the approximation signal sequence is C= { C (k), and k epsilon Z }, and C (k) is as follows:
c(k)=s e (k)+U[d(k)],k∈Z。
4. the method for detecting tightness of slot wedge of generator as claimed in claim 2, wherein said step 12 comprises the steps of:
the decomposition step of step 11 is reversed:
s e (k)=c(k)-U[d(k)],k∈Z
s 0 (k)=d(k)+P[s e (k)],k∈Z
and then combining the odd samples obtained in the steps with the even samples to reconstruct a transformed signal s.
5. The method for detecting tightness of slot wedge of generator as claimed in claim 1, wherein said step 2 comprises the following steps:
extracting and screening the characteristics of the knock signal, extracting the time domain characteristics, the frequency spectrum characteristics and the power spectrum characteristics of the pre-processed knock signal,
time domain characteristics of the knock signal:
root Mean Square (RMS) may be used to represent the effective value of a signal, calculated as follows:
where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal;
variance VAR reflects the degree of a higher average value of the data and can measure the fluctuation degree of the knock signal;
where N is the length of the knock data, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is the average value of (2);
the waveform factor K is the ratio of the effective value to the rectification average value, and the calculation formula is as follows:
the peak factor C is the ratio of the peak value X to the root mean square value RMS and is calculated as follows:
kurtosis K q The numerical statistic for reflecting the kurtosis of the waveform is shown as follows:
where N is the number of sample points, x i I=1, 2,3,4,5 … N is the time-domain amplitude of the striking sound signal and μ is x i Is the average value of (2);
the zero crossing rate ZCR is the number of times that the time domain signal passes through the zero point in a certain time, and represents the vibration number of the time domain signal, and the calculation formula is as follows:
wherein N represents the total number of points of the knock signal, T represents time, and N X represents: when X is true, the output is 1, otherwise, 0 is output.
6. The method for detecting tightness of slot wedge of generator as claimed in claim 1, wherein said step 3 comprises the following steps:
step 31: frequency domain peak and frequency thereof
Respectively finding out the peak value and the corresponding frequency of each frequency interval according to the defined frequency interval;
step 32: area of band interval
And calculating the frequency band amplitude area according to the defined frequency interval, wherein the frequency band amplitude area is the sum of the frequency spectrum amplitudes, and the calculation formula is as follows:
wherein i and j respectively represent a start value and an end value of each interval;
step 33: power spectrum area.
7. The method for detecting tightness of slot wedge of generator as claimed in claim 6, wherein said step 33 comprises the steps of:
the power spectrum estimation method of the knocking sound signal by adopting the Welch method comprises the following steps of:
step 331: the pre-processed sequence x (n) of the knock signal is divided into L segments, each segment has a sequence length of M=512, and the overlap length between every two segments of data is Md=256, so that the following can be obtained
Step 332: the window function to be employed is the Hamming window, denoted herein as W (n);
step 333: the power spectrum of each segment of data is calculated respectively, summed and then averaged to obtain the required self-power spectrum estimation, and the calculation formula is as follows:
the normalization factor U in the formula is calculated as follows:
dividing frequency intervals according to the obtained image of the power spectrum, and respectively obtaining the power spectrum area of each frequency interval, wherein the calculation formula of the power spectrum area is as follows:
and the previous candidate feature parameters are all candidate feature parameters for initial selection.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211423742.6A CN116007912A (en) | 2022-11-14 | 2022-11-14 | Generator slot wedge tightness detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211423742.6A CN116007912A (en) | 2022-11-14 | 2022-11-14 | Generator slot wedge tightness detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116007912A true CN116007912A (en) | 2023-04-25 |
Family
ID=86023712
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211423742.6A Pending CN116007912A (en) | 2022-11-14 | 2022-11-14 | Generator slot wedge tightness detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116007912A (en) |
-
2022
- 2022-11-14 CN CN202211423742.6A patent/CN116007912A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100523765C (en) | Vibrating detection method for detecting looseness of large-scale generator rotor slot wedge and apparatus therefor | |
Gong et al. | Application of tentative variational mode decomposition in fault feature detection of rolling element bearing | |
TWI407026B (en) | Diagnosis method of ball screw preload loss via hilbert-huang transform and apparatus therefor | |
CN101799354A (en) | Method for testing assembly performance of rotor of aircraft engine | |
CN103953490A (en) | Implementation method for monitoring status of hydraulic turbine set based on HLSNE | |
CN106053074B (en) | Rolling bearing sound signal fault feature extraction method based on STFT and rotational inertia entropy | |
CN111256993A (en) | Method and system for diagnosing fault type of main bearing of wind turbine generator | |
CN107657110A (en) | The fatigue damage evaluation method of large scale wind power machine blade | |
CN104564542A (en) | Fault diagnosis system and fault diagnosis method based on massive data technology | |
CN111024527A (en) | Crack propagation monitoring method based on multi-sensor data fusion | |
Li et al. | Mobile robot motor bearing fault detection and classification on discrete wavelet transform and LSTM network | |
CN110332952B (en) | Transformer core looseness detection method and platform based on acoustic images | |
Yang et al. | Crack detection in carbide anvil using acoustic signal and deep learning with particle swarm optimisation | |
Wang et al. | Identification of engine faults based on acoustic emission signals using a 1DCNN-ViT ensemble model | |
CN110332080A (en) | A kind of fan blade health method of real-time based on resonance response | |
CN116007912A (en) | Generator slot wedge tightness detection method | |
CN113869339A (en) | Deep learning classification model for fault diagnosis and fault diagnosis method | |
CN112364296A (en) | P wave arrival time automatic picking method based on deep learning | |
CN1403793A (en) | Detection method of loosening slot wedge of rotating motor | |
CN115273895A (en) | Transformer running state monitoring method and system based on wavelet time-frequency entropy and RUSboost | |
CN115828144A (en) | Signal sparse representation and fusion detection method, storage medium and electronic device | |
Zhao et al. | Vibration feature extraction and fault detection method for transmission towers | |
Al-hababi et al. | Time-frequency domain methods for the identification of breathing cracks in beam-like structures | |
CN117150216B (en) | Regression analysis method and system for power data | |
CN114924157B (en) | Parallel reactor state monitoring method and system based on 5G transmission |
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 |