CN109765003A - Blade imbalance fault signal characteristics extracting method based on Hilbert transform - Google Patents
Blade imbalance fault signal characteristics extracting method based on Hilbert transform Download PDFInfo
- Publication number
- CN109765003A CN109765003A CN201910047324.3A CN201910047324A CN109765003A CN 109765003 A CN109765003 A CN 109765003A CN 201910047324 A CN201910047324 A CN 201910047324A CN 109765003 A CN109765003 A CN 109765003A
- Authority
- CN
- China
- Prior art keywords
- signal
- frequency
- mean
- electric signal
- hilbert transform
- 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.)
- Granted
Links
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The blade imbalance fault signal characteristics extracting method based on Hilbert transform that the invention discloses a kind of, include: that Hilbert transform is carried out to the electric signal of input, obtains the corresponding instantaneous frequency signal of electric signal and ideally corresponding bearing rotation instantaneous frequency；Calculate the corresponding time series of electrical signal peak；Calculate corresponding means frequency of each electric signal period and corresponding means frequency of each bearing rotary period；The electric signal means frequency and bearing rotation means frequency obtain to step 3 carries out cubic spline interpolation, obtains interpolated signal identical with original signal samples frequency；Five, the difference for the interpolated signal for asking step 4 to obtain, obtains fault characteristic signals.This method is related to the key technology in generation of electricity by new energy status monitoring field, solves the problems, such as that conventional failure detection method is difficult to accurately extract blade imbalance fault feature using power generation dynamoelectric signal.
Description
Technical field:
The present invention relates to the key technologies in generation of electricity by new energy status monitoring field, and in particular to one kind is become based on Hilbert
The blade imbalance fault signal characteristics extracting method changed.
Background technique:
In recent years as wind energy, energy by ocean current are more and more paid attention to, wind-driven generator and ocean current generator etc.
New energy equipment is widely used.Wherein wind power generation and energy by ocean current power generation in the most common failure problems just
It is because of situations such as blade attachment, corrosion of blade, the caused blade imbalance problem of blade wear, this kind of failure can be by drawing
Entering unbalance mass, influences the operation of system, therefore detects machine state and repair or replace failure blade to avoid system damage
Bad is highly important.The status monitoring carried out using electric signal is in maintenance cost, installation and system stability all than passing
The detection method based on vibration signal of system is advantageously.However, the low noise of electric signal and imbalance fault characteristic frequency
Than allowing imbalance fault detection to face huge challenge.The conventional failure detection technique being widely used at present is difficult to realize for not
The feature extraction of balancing fault, causes that diagnostic method is excessively single, also very limited to the Utilization ability of diagnostic message, therefore
Showing be out of order poor correct detectability, rate of failing to report and the higher problem of false alarm rate in use seriously restricts power generation
Equipment Condition Monitoring System efficiency gives full play to.
Feature extraction step is generally in the prime of fault detection, which will extract the feature with following property:
Characteristic attribute or numerical value from same category of different samples should be very close, the feature from different classes of sample
Attribute or numerical value should have biggish difference.In addition, the link also needs to extract the feature of most distinguishing ability, these
Feature transformation incoherent for classification information has invariance.It is all often non-flat as handled by condition monitoring system
Surely, nonlinear data, therefore it is contemplated that carry out design error failure feature extraction scheme using Hilbert transform, to assist shape
State detection system reaches the fault diagnosis effect of low rate of failing to report Yu high accuracy.
Summary of the invention:
The present invention is difficult to accurately extract to solve in condition monitoring system conventional failure detection method using electric signal
The problem of blade imbalance fault feature, proposes a kind of blade imbalance fault signal characteristics based on Hilbert transform
Extracting method.
Detailed process is as follows by the present invention:
Step 1: Hilbert transform is carried out to the electric signal of input, obtains the corresponding instantaneous frequency signal f of electric signale
(t) and ideally corresponding bearing rotates instantaneous frequency fr(t)；
Step 2: the corresponding time series S of electrical signal peak is calculatedpeak(k)；
Step 3: corresponding means frequency f of each electric signal period is calculatede,mean(i) and each bearing rotary period pair
The means frequency f answeredr,mean(j)；
Step 4: the f that step 3 is obtainede,mean(i) and fr,mean(j) carry out cubic spline interpolation, obtain with it is original
The identical interpolated signal of signal sampling frequenciesAnd
Step 5: the interpolated signal for asking step 4 to obtainWithDifference signal, obtain fault signature letter
Number fim(n)。
The generator rotational frequency information that the present invention is implied from electric signal is started with, and estimates that power generation is electromechanical by means such as HT
Signal transient frequency and rotational frequency, and blade imbalance fault feature is extracted on the basis of instantaneous frequency, it is one
The blade imbalance fault feature extracting method that kind is efficient, stability is high.
The invention has the following advantages over the prior art:
Signal characteristic extracting methods proposed by the invention utilize the methods of HT, cubic spline interpolation, to Generator Status
Monitoring system raw electrical signal collected is handled, and is started with from electric signal instantaneous frequency and bearing rotation instantaneous frequency,
Blade imbalance fault characteristic signal is extracted, adaptivity possessed by the program and stability are other feature extracting methods
It is unable to reach.
Signal characteristic extracting methods proposed by the invention are inherently to the instantaneous frequency estimated by initial data
Rate is handled, and does not carry out any change to original signal, so that completely remaining in entire characteristic extraction procedure original
All information of signal, while but also subsequent fault detection in condition monitoring system, that diagnosis ring is brought is convenient, it is higher
The algorithm of level or decision application are brought conveniently.
Present invention blade imbalance fault signal characteristics suitable for generation of electricity by new energy condition detecting system extract field.
Detailed description of the invention:
Fig. 1 is that the present invention is based on the blade imbalance fault signal characteristics extracting method flow charts of Hilbert transform
Fig. 2 is ocean current generator voltage under non-failure conditions
Fig. 3 is ocean current generator failure characteristic signal under non-failure conditions
Fig. 4 is ocean current generator voltage in the case of blade imbalance fault
Fig. 5 is ocean current generator fault characteristic signals in the case of blade imbalance fault
Fig. 6 is ocean current generator failure characteristic signal spectral contrast figure under different situations
Specific embodiment:
Specific embodiment 1: embodiment is described with reference to Fig. 1, a kind of blade based on Hilbert transform is uneven
Failure signal characteristics extracting method, detailed process is as follows:
Step 1: Hilbert transform is carried out to the electric signal of input, obtains the corresponding instantaneous frequency signal f of electric signale
(t) and ideally corresponding bearing rotates instantaneous frequency fr(t)；
Step 2: the corresponding time series S of electrical signal peak is calculatedpeak(k)；
Step 3: corresponding means frequency f of each electric signal period is calculatede,mean(i) and each bearing rotary period pair
The means frequency f answeredr,mean(j)；
Step 4: the f that step 3 is obtainede,mean(i) and fr,mean(j) carry out cubic spline interpolation, obtain with it is original
The identical interpolated signal of signal sampling frequenciesAnd
Step 5: the interpolated signal of step 4 acquisition is soughtWithDifference signal, obtain fault signature
Signal fim(n)。
Hilbert transform is by carrying out convolution algorithm for 1/ π t of signal and function, so that the local characteristics of x (t) are obtained,
Such as following formula:
Wherein CPV is Cauchy principal value integral (Cauchy Principle Value, CPV), from the viewpoint of frequency domain:
A complex signal z (t) can be formed using x (t) and y (t):
Z (t)=x (t)+jy (t)
The frequency domain form of z (t) is as follows:
Z (t) is an analytical function, be can be represented by the formula under polar coordinates:
Z (t)=x (t)+jy (t)=a (t) ejθ(t)
A (t) indicates the envelope amplitude size of x (t), and θ (t) indicates the phase angle of x (t):
The instantaneous frequency of x (t) is defined as:
When carrying out the feature extraction of blade imbalance fault for actual electric signal, it is contemplated that bearing rotational frequency is relatively steady
It is fixed, it can not consider to ask in instantaneous frequency using above formula, it may be as caused by influence of the DC component with multiple frequency ingredient
Interference directly calculates the instantaneous frequency of electric signal and bearing rotation using Hilbert transform.One, blade is rotated simultaneously
Period is regarded as a working condition, and acquire two kinds of instantaneous frequencys are successively asked according to electric signal period and bearing rotation period
Mean value, can weaken directly using Hilbert transform solve instantaneous frequency introduce interference, while estimate practical electric signal with
And bearing rotational frequency, then extract fault characteristic signals relevant to blade imbalance fault.
Entire fault characteristic signals extraction process has raw electrical signal itself to drive, and has complete adaptivity, and do not have
Have and any processing is carried out to original signal, completely remain all information of signal, is subsequent event in condition monitoring system
It is convenient that barrier detects, diagnosis is brought, and brings conveniently for the application of higher level algorithm or decision.
Specific embodiment 2: present embodiment is the further explanation to specific embodiment one, to defeated in step 1
The monitoring electric signal entered carries out Hilbert transform, obtains the corresponding instantaneous frequency signal f of electric signale(t) and perfect condition
Under corresponding bearing rotate instantaneous frequency fr(t) process are as follows:
The electric signal being originally inputted is set as x (t), time t=1,2 ..., N,
Step a: discrete convolution is carried out to raw electrical signal x (t), its Hilbert transform y (t) is obtained, is shown below:
Step b: the envelope amplitude a (t) of analytic signal c (the t)+jy (t) of raw electrical signal is calculated, is shown below:
Step c: phase angle theta (t) the instantaneous frequency f corresponding with electric signal of analytic signal c (t)+jy (t) is calculatede(t), such as
Shown in following formula:
Obtain the instantaneous frequency signal f of raw electrical signale(t)；
Step d: ideally generator bearing rotational frequency is p times of its signal frequency, thus finds out step c institute
The f found oute(t) bearing corresponding to rotates instantaneous frequency fr(t)=p*fe(t) the original telecommunications acquired in condition monitoring system
Number it is typically all non-stationary, nonlinear properties, is directly handled using the signal of Hilbert transform pairs this type, it is original
Entrained DC component and multiple frequency impact processing result at branch in signal, and the present invention does not consider these shadows
Ring, carry out processing solver instantaneous frequency using Hilbert transform pairs raw electrical signal, then based on the electric signal period and
Bearing rotation period averages on instantaneous frequency to weaken above-mentioned influence.Due to using Hilbert transform to calculate instantaneous frequency
It is easy to produce end effect in the process, needs exist for the head of shielding power-off signal instantaneous frequency, bearing turn signal instantaneous frequency
The data in portion 30%, tail portion 10%, i.e. these data are no longer used.
Specific embodiment 3: present embodiment is the further explanation to specific embodiment one, calculated in step 2
Time series S corresponding to raw electrical signal peak valuepeak(k) process are as follows:
Step a: peak point all among raw electrical signal is hunted out；
Step b: the corresponding time point information of all peak points among step a is found out, time series S is obtainedpeak(k),
Middle k=1,2 ..., K, k indicate that k-th of peak point, K indicate the peak point total number that the raw electrical signal is included.
Specific embodiment 4: present embodiment is the further explanation to specific embodiment one, calculated in step 3
Each electric signal period corresponding means frequency fe,mean(i) and corresponding means frequency f of each bearing rotary periodr,mean
(j) process are as follows:
The signal for including between adjacent two peak value is considered as to the period of an electric signal, then the bearing rotary period is electric signal
P times of period,
Step a: corresponding electric signal means frequency f of each electric signal period is calculatede,mean(i) as electric signal frequency
The estimation of rate；
Step b: corresponding bearing rotary means frequency f of each bearing rotary period is calculatedr,mean(j) it is used as motor bearings
The estimation of speed.
Specific embodiment 5: present embodiment is the further explanation to specific embodiment one, to step in step 4
Rapid three f obtainede,mean(i) and fr,mean(j) cubic spline interpolation is carried out, insert identical with original signal samples frequency is obtained
Value signalAndProcess are as follows:
Using cubic spline interpolation, respectively to fe,mean(i) and fr,mean(j) interpolation is carried out, both allows difference result
Sample frequency is equal with raw electrical signal sample frequency.
Specific embodiment 6: present embodiment is the further explanation to specific embodiment one, step is sought in step 5
Rapid four interpolated signals obtainedWithDifference as fault characteristic frequency signal fim(n) process are as follows:
Calculate the equal interpolated signal of sample frequency, the data volume obtained among step 4WithDifference
Score value, as fault characteristic signals fim(n)。
The influence of electric signal instantaneous frequency bring directly is calculated using Hilbert transform in order to weaken, it can be by this
Instantaneous frequency successively calculates the mean value of each cycle time section of electric signal, weakens DC component and multiple frequency ingredient to reach
The purpose of interference.One swing circle of blade can be regarded as the one of the equipment by rotating machinery this kind of for generator
A working condition, therefore mean value of the instantaneous frequency under each blade swing circle can be rotated by calculating bearing, come with this
Estimate generator bearing rotational frequency.The period of electric signal, can by the time difference between the peak point of raw electrical signal come
It is estimated, blade swing circle can then be simply viewed as being P times of the electric signal period.Therefore by find raw electrical signal it
In time to peak sequence, that is, can reach to the electric signal period and blade swing circle estimation purpose, so as to use
Two kinds of cycle informations complete the estimation to electric signal instantaneous frequency and bearing rotary instantaneous frequency.In order to extract spy of being out of order
Reference fim(n), two kinds of instantaneous frequency f estimation obtainede,mean(i) and fr,mean(j) cubic spline interpolation is carried out, is obtained
To sample frequency interpolated signal identical with raw electrical signalAnd, finally calculateWithDifference
Value can extract blade imbalance fault characteristic signal fim(n)。
The fault characteristic signals extracted are able to reflect out blade imbalance fault feature: the failure being under malfunction
The frequency spectrum of characteristic signal spike can occur near 1p frequency.Since what be can determine knows that the blade finally extracted is uneven
Spike can occur near 1p frequency in the frequency spectrum of fault characteristic signals, therefore can be by fault characteristic signals among practical application
It is transformed among frequency domain, then the information directly near interception 1p frequency is analyzed, and can be monitored the subsequent event of system
The operations such as barrier detection, diagnosis.
With certain ocean current generator experiment porch condition monitoring system institute, collected generator unit stator voltage signal is below
Embodiment illustrates a specific embodiment of the invention:
Ocean current generator is commonly called as " seabed windmill ", is a kind of turbine, blade and electric component etc. by being fixed on seabed
It constitutes, flows moving blade using seawater, kinetic force is switched to the power generator of electric energy, with land wind-driven generator principle base
This is consistent.From ocean current generator experiment porch, condition monitoring system collects the interception of the present embodiment original signal in the process of running
Underwater ocean current generator unit stator voltage signal, which is direct drive permanent magnetic synchronous generator, and number of pole-pairs is
8, and two groups of initial data, flow rate of water flow 1.2m/s are respectively acquired under unfaulty conditions and blade imbalance fault state respectively
Left and right, sample frequency 1KHz, when sampling a length of 10s.Ocean current generator under normal condition and blade imbalance fault state
Rotational frequency is approximately uniform, is all 1.98Hz or so.
The present embodiment has chosen each two groups of sampled datas under fault-free and faulty state as input data, holds respectively
Row following steps, and implementation procedure is completely the same, wherein the specific implementation procedure of 1 group of input data is as follows:
It executes step 1: Hilbert transform (Hilbert Transform) being carried out to input data, obtains voltage signal
Corresponding instantaneous frequency signal and ideally corresponding bearing rotation instantaneous frequency, are denoted as f respectivelye(t)、fr(t), divide
Portion executes as follows:
Input signal x (t), time t=0.001,0.002 ..., 10.000；
Discrete convolution is carried out to the voltage signal x (t) of acquired original, its Hilbert transform y (t) is obtained, such as following formula institute
Show:
The envelope amplitude a (t) for calculating analytic signal c (the t)+jy (t) of raw voltage signals, is shown below:
Calculate phase angle theta (t) the instantaneous frequency f corresponding with voltage signal of analytic signal c (t)+jy (t)e(t), such as following formula
It is shown:
The instantaneous frequency signal f of raw voltage signals can be obtainede(t)；
Ideally generator bearing rotational frequency is 8 times of its signal frequency, thus finds out raw voltage signals
Instantaneous frequency fe(t) bearing corresponding to rotates instantaneous frequency fr(t)=8*fe(t)；Shield power-off signal instantaneous frequency fe(t)、
Bearing turn signal instantaneous frequency fr(t) data of stem 30%, tail portion 10%.
It executes step 2: calculating time series S corresponding to raw voltage signals peak valuepeak(k), distribution executes as follows:
Search searches out peak point all among raw voltage signals；
The corresponding temporal information of each peak point is found out, time series S corresponding with peak point is obtainedpeak(k), wherein k
=1,2 ..., K, k indicate k-th of peak point, K indicates the peak point total number that the raw electrical signal is included；
Time series Speak(k) each time interval can indicate a voltage signal cycles in, and every 8 time intervals are
It can indicate a bearing rotary period.
It executes step 3: calculating voltage transient frequency signal fe(t) the mean value signal f under each voltage cyclee,mean
(i) and bearing instantaneous frequency signal fr(t) the mean value signal f under each bearing rotary periodr,mean(j), distribution executes
It is as follows:
Calculate the corresponding average voltage frequency f of each voltage signal cyclese,mean(i) as voltage signal frequency
Estimation；
Calculate corresponding bearing means frequency f of each bearing rotary periodr,mean(j) as motor bearings speed
Estimation.
Execute step 4: using cubic spline interpolation to fe,mean(i) and fr,mean(j) interpolation is carried out, is successively obtained
Interpolated signalWithMake the two sample frequency equal with raw voltage signals sample frequency.
It executes step 5: calculating interpolated signal by following formulaWithDifference as fault characteristic frequency
Signal fim(n):
Execute step 6: to the fault characteristic signals f extractedim(n) Fourier analysis is carried out, the frequency domain of the signal is used
Feature carries out fault detection.
Separately below with the feature extraction effect of fault-free and faulty two kinds of situation analysis verifying embodiment.Shown in Fig. 2
Non-failure conditions under ocean current generator voltage, execute Hilbert transform after obtain voltage transient frequency under the state, together
When estimate perfect condition lower bearing rotational frequency.It finds the peak information of voltage signal under the state and finds out corresponding time sequence
Then column are made using the time interval between every two peak point as the time interval between voltage cycle, every nine peak points
For the bearing rotary period, voltage transient frequency cycle mean value and bearing rotary frequency Periodic Mean are asked, cubic spline is reused
Interpolation method carries out interpolation to the two Periodic Means with this, and the interpolation result sample frequency made and raw voltage signals sample
Frequency is equal, the voltage transient frequency and bearing rotary instantaneous frequency estimated.Then to the voltage transient estimated
Frequency and bearing rotary instantaneous frequency seek difference, extract fault characteristic signals, as shown in Figure 3.Equally, to leaf shown in Fig. 4
Ocean current generator voltage in the case of piece imbalance fault obtains result shown in fig. 5 after executing all steps of the present invention.
From the fault characteristic signals finally extracted, we are difficult to intuitively evaluating characteristic extraction effect and fault detection
Effect, the fault characteristic signals extracted are carried out Fourier analysis, are transformed into frequency domain and are analyzed by us.It will be by upper
It states under the non-failure conditions that step extracts and ocean current generator fault characteristic signals carries out frequency domain point respectively in faulty situation
Analysis, and respective spectrogram finally is drawn on same figure, as shown in Figure 6.Failure is special under malfunction as can be seen from Figure 6
The spectrogram of reference number is controlled at 1P frequency (1.98Hz) there is apparent spike, and fault characteristic signals under unfaulty conditions
Spectrogram it is then very steady here, failure has obtained good detection.It can be seen that utilizing blade imbalance fault of the invention
Signal characteristics extracting method can be good at carrying out blade imbalance fault feature extraction and can be applied to failure well
Among detection, and there is biggish sense margin, reflects the validity of this feature extracting method.
Claims (6)
1. a kind of blade imbalance fault signal characteristics extracting method based on Hilbert transform, it is characterised in that specific mistake
Journey is as follows:
Step 1: Hilbert transform is carried out to the electric signal of input, obtains the corresponding instantaneous frequency signal f of electric signale(t) with
And ideally corresponding bearing rotates instantaneous frequency fr(t)；
Step 2: the corresponding time series S of electrical signal peak is calculatedpeak(k)；
Step 3: corresponding means frequency f of each electric signal period is calculatede,mean(i) and each bearing rotary period is corresponding
Means frequency fr,mean(j)；
Step 4: the f that step 3 is obtainede,mean(i) and fr,mean(j) cubic spline interpolation is carried out, is obtained and original signal
The identical interpolated signal of sample frequencyAnd
Step 5: the interpolated signal for asking step 4 to obtainWithDifference signal, obtain fault characteristic signals fim
(n)。
2. a kind of blade imbalance fault signal characteristics extraction side based on Hilbert transform according to claim 1
Method, it is characterised in that Hilbert transform is carried out to the electric signal of input in step 1, obtains the corresponding instantaneous frequency of electric signal
Signal fe(t) and ideally corresponding bearing rotates instantaneous frequency fr(t) process are as follows:
The electric signal being originally inputted is set as x (t), time t=1,2 ..., N,
Step a: discrete convolution is carried out to raw electrical signal x (t), its Hilbert transform y (t) is obtained, is shown below:
Step b: the envelope amplitude a (t) of analytic signal c (the t)+jy (t) of raw electrical signal is calculated, is shown below:
Step c: phase angle theta (t) the instantaneous frequency f corresponding with electric signal of analytic signal c (t)+jy (t) is calculatede(t), such as following formula institute
Show:
Obtain the instantaneous frequency signal f of raw electrical signale(t)；
Step d: ideally generator bearing rotational frequency is p times of number of pole-pairs of its signal frequency, thus finds out step c
Calculated fe(t) bearing corresponding to rotates instantaneous frequency fr(t)=p*fe(t)。
3. a kind of blade imbalance fault signal characteristics extraction side based on Hilbert transform according to claim 1
Method, it is characterised in that time series S corresponding to raw electrical signal peak value is calculated in step 2peak(k) process are as follows:
Step a: peak point all among raw electrical signal is hunted out；
Step b: the corresponding time point information of all peak points among step a is found out, time series S is obtainedpeak(k), wherein k=
1,2 ..., K, k indicate that k-th of peak point, K indicate the peak point total number that the raw electrical signal is included.
4. a kind of blade imbalance fault signal characteristics extraction side based on Hilbert transform according to claim 1
Method, it is characterised in that corresponding means frequency f of each electric signal period is calculated in step 3e,mean(i) and each bearing rotary
Period corresponding means frequency fr,mean(j) process are as follows:
The signal for including between adjacent two peak value is considered as to the period of an electric signal, then the bearing rotary period is the electric signal period
P times,
Step a: corresponding electric signal means frequency f of each electric signal period is calculatede,mean(i) as signal frequency
Estimation；
Step b: corresponding bearing rotary means frequency f of each bearing rotary period is calculatedr,mean(j) it is rotated as motor bearings
The estimation of frequency.
5. a kind of blade imbalance fault signal characteristics extraction side based on Hilbert transform according to claim 1
Method, it is characterised in that the f that step 3 is obtained in step 4e,mean(i) and fr,mean(j) cubic spline interpolation is carried out, is obtained
Interpolated signal identical with original signal samples frequencyAndProcess are as follows:
Using cubic spline interpolation, respectively to fe,mean(i) and fr,mean(j) interpolation is carried out, the two interpolation result is allowed to sample
Frequency is equal with raw electrical signal sample frequency.
6. a kind of blade imbalance fault signal characteristics extraction side based on Hilbert transform according to claim 1
Method, it is characterised in that the interpolated signal of step 4 acquisition is sought in step 5WithDifference signal, obtain therefore
Hinder characteristic signal fim(n) process are as follows:
Calculate the equal interpolated signal of sample frequency, the data volume obtained among step 4WithDifference value,
As fault characteristic signals fim(n)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910047324.3A CN109765003B (en) | 2019-01-18 | 2019-01-18 | Method for extracting characteristics of blade imbalance fault electrical signals based on Hilbert transform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910047324.3A CN109765003B (en) | 2019-01-18 | 2019-01-18 | Method for extracting characteristics of blade imbalance fault electrical signals based on Hilbert transform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109765003A true CN109765003A (en) | 2019-05-17 |
| CN109765003B CN109765003B (en) | 2021-02-23 |
Family
ID=66454163
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910047324.3A Active CN109765003B (en) | 2019-01-18 | 2019-01-18 | Method for extracting characteristics of blade imbalance fault electrical signals based on Hilbert transform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109765003B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104089778A (en) * | 2014-07-12 | 2014-10-08 | 东北电力大学 | Water turbine vibration fault diagnosis method |
| CN104184383A (en) * | 2014-09-17 | 2014-12-03 | 重庆大学 | Doubly-fed wind power generator stator current diagnosis method for impeller imbalance fault |
| CN107191339A (en) * | 2017-07-31 | 2017-09-22 | 上海电气风电集团有限公司 | Wind-driven generator group wind-wheel imbalance monitoring method |
| CN107256546A (en) * | 2017-05-23 | 2017-10-17 | 上海海事大学 | Ocean current machine blade attachment method for diagnosing faults based on PCA convolution pond SOFTMAX |
| CN107559153A (en) * | 2017-08-07 | 2018-01-09 | 浙江运达风电股份有限公司 | A kind of double-fed fan motor unit impeller imbalance detection system |
-
2019
- 2019-01-18 CN CN201910047324.3A patent/CN109765003B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104089778A (en) * | 2014-07-12 | 2014-10-08 | 东北电力大学 | Water turbine vibration fault diagnosis method |
| CN104184383A (en) * | 2014-09-17 | 2014-12-03 | 重庆大学 | Doubly-fed wind power generator stator current diagnosis method for impeller imbalance fault |
| CN107256546A (en) * | 2017-05-23 | 2017-10-17 | 上海海事大学 | Ocean current machine blade attachment method for diagnosing faults based on PCA convolution pond SOFTMAX |
| CN107191339A (en) * | 2017-07-31 | 2017-09-22 | 上海电气风电集团有限公司 | Wind-driven generator group wind-wheel imbalance monitoring method |
| CN107559153A (en) * | 2017-08-07 | 2018-01-09 | 浙江运达风电股份有限公司 | A kind of double-fed fan motor unit impeller imbalance detection system |
Non-Patent Citations (4)
| Title |
|---|
| MILU ZHANG ET AL: "An imbalance fault detection method basedon data normalization and EMD for marine current turbines", 《ISA TRANSACTIONS》 * |
| MILU ZHANG ET AL: "Imbalance Fault Detection of Marine Current Turbine under Condition of Wave and Turbulence", 《IEEE》 * |
| 张米露等: "海流发电系统叶片附着物故障检测与特性研究", 《中国海洋学会2015年学术论文集》 * |
| 张米露等: "直驱式永磁同步海流机叶轮不平衡故障建模与实验研究", 《电机与控制学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109765003B (en) | 2021-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107132450B (en) | A kind of sea double feedback electric engine stator winding inter-turn short circuit initial failure discrimination method | |
| CN104677623B (en) | A kind of blade of wind-driven generator fault acoustics in place diagnostic method and monitoring system | |
| CN109061474A (en) | A kind of motor bearings trouble-shooter | |
| Chen et al. | Bearing corrosion failure diagnosis of doubly fed induction generator in wind turbines based on stator current analysis | |
| CN103713237B (en) | A kind of power system transmission line short trouble diagnostic method | |
| CN102262215B (en) | Method for detecting stator and rotor air gap eccentric faults of large generator | |
| CN103471841B (en) | A kind of rotating machinery vibrating failure diagnosis method | |
| CN105569932A (en) | Dynamic unbalance online testing and fault identification method and system for wind turbine generators | |
| CN105699080A (en) | Wind turbine generator set bearing fault feature extraction method based on vibration data | |
| CN103713235A (en) | Turbine generator rotor interturn short-circuit fault diagnosis method based on end distortion effect | |
| CN101977008A (en) | Judgment method of key sensor fault of double-fed wind generating set | |
| CN103234702B (en) | Method for diagnosing imbalance faults of blades | |
| CN102820665B (en) | Method for rapidly identifying sub-synchronous oscillation in wind power integrated system | |
| CN102087139B (en) | Method for analyzing frequency components of low-frequency vibration of steam turbine generator unit in real time | |
| Tian et al. | A review of fault diagnosis for traction induction motor | |
| CN115955161A (en) | Method, apparatus, device and medium for estimating slip of adaptive asynchronous induction motor | |
| CN113391235A (en) | System and method for detecting dynamic turn-to-turn short circuit fault of synchronous generator rotor | |
| CN108278184A (en) | Impeller of wind turbine set imbalance monitoring method based on empirical mode decomposition | |
| CN109765003A (en) | Blade imbalance fault signal characteristics extracting method based on Hilbert transform | |
| CN108732522A (en) | The test device and method of generator rotor interturn short-circuit fault monitoring device | |
| Gong et al. | Current-based online bearing fault diagnosis for direct-drive wind turbines via spectrum analysis and impulse detection | |
| CN103149047A (en) | Cooling tower acoustic diagnosis method based on nonlinear mixed model | |
| CN102879084B (en) | On-line low-frequency vibration unstable state early warning method for steam turbine generator unit | |
| CN102879085B (en) | Real-time low-frequency vibration unstable state early warning method for steam turbine generator unit | |
| CN110441081B (en) | Intelligent diagnosis method and intelligent diagnosis system for faults of rotating machinery |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |