CN110824281A - Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator - Google Patents
Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator Download PDFInfo
- Publication number
- CN110824281A CN110824281A CN201911138356.0A CN201911138356A CN110824281A CN 110824281 A CN110824281 A CN 110824281A CN 201911138356 A CN201911138356 A CN 201911138356A CN 110824281 A CN110824281 A CN 110824281A
- Authority
- CN
- China
- Prior art keywords
- fault
- phase modulator
- synchronous phase
- stator
- rotor
- 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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R15/00—Details of measuring arrangements of the types provided for in groups G01R17/00 - G01R29/00, G01R33/00 - G01R33/26 or G01R35/00
- G01R15/14—Adaptations providing voltage or current isolation, e.g. for high-voltage or high-current networks
- G01R15/20—Adaptations providing voltage or current isolation, e.g. for high-voltage or high-current networks using galvano-magnetic devices, e.g. Hall-effect devices, i.e. measuring a magnetic field via the interaction between a current and a magnetic field, e.g. magneto resistive or Hall effect devices
- G01R15/202—Adaptations providing voltage or current isolation, e.g. for high-voltage or high-current networks using galvano-magnetic devices, e.g. Hall-effect devices, i.e. measuring a magnetic field via the interaction between a current and a magnetic field, e.g. magneto resistive or Hall effect devices using Hall-effect devices
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/0092—Arrangements for measuring currents or voltages or for indicating presence or sign thereof measuring current only
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
A synchronous phase modulator on-line monitoring and fault diagnosis method and system, make up the fault characteristic vector of the magnetic flux between the stator and rotor turn and fault characteristic vector of the vibration signal under the fault operation of the synchronous phase modulator and fault characteristic vector set, the fault characteristic vector set is regarded as the data training input of the classifier of the multiple faults; the multi-fault classifier sets data ranges of vector sets corresponding to different faults according to the basic data set and the fault feature vector set; inputting the characteristic vector set of the synchronous phase modulator under the condition of unknown fault into a vector machine classifier to obtain a corresponding fault category; and diagnosing the fault type of the synchronous phase modulator through the stator and rotor current characteristic vectors, if the diagnosis result is the same as the diagnosis result, judging the fault type, and comprehensively judging the specific fault mode of the synchronous phase modulator. The invention can realize multi-parameter comprehensive diagnosis of the synchronous phase modulator, solves the problem of simplification of fault diagnosis parameters of the existing synchronous phase modulator, and improves the accuracy of fault diagnosis.
Description
Technical Field
The invention belongs to the technical field of on-line monitoring and fault diagnosis of power equipment, relates to the technical field of monitoring of synchronous phase modulators, and particularly relates to a method and a system for on-line monitoring and fault diagnosis of a synchronous phase modulator.
Background
At present, a plurality of convertor stations in China are provided with large synchronous phase modulators. The maintenance work of the synchronous phase modulator is mainly to periodically perform preventive tests, and the running state of the synchronous phase modulator is further judged according to the test result, so that whether the synchronous phase modulator can continue to run or not is determined. However, with the complexity and diversification of the structure of the power system, the requirement index for the safe and reliable operation of the power system is higher and higher. The traditional simple test and diagnosis method cannot meet the market requirements, and is particularly important for carrying out online monitoring on the running state of the synchronous phase modulator.
At present, no mature synchronous phase modulator state monitoring product exists in China or even internationally, various documents and data are combined, at present, an improved system is not formed for the on-line monitoring and diagnosis technology of the synchronous phase modulator, most of the on-line monitoring and diagnosis technologies only aim at the on-line monitoring of single or a plurality of parameters, the fault diagnosis is not facilitated, and the storage and processing of data at the later stage are not required. A paper, "study of phase modulation fault diagnosis algorithm based on RBF neural network" (master paper of university of science and technology in china-2018) proposes a radial basis function neural network algorithm based on an improved K-means clustering algorithm, and the paper only studies a single fault through theoretical analysis without simulating fault data and considering the condition of multi-fault aliasing. The problem also exists in the patent "method for diagnosing faults of synchronous phase modulators based on longitudinal analysis and transverse correction" (application No. 201811154826.8).
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides an online monitoring and fault diagnosis system and a fault diagnosis method for a synchronous phase modulator, and solves the problem that the existing synchronous phase modulator can only aim at single or a plurality of parameters for online monitoring and is not beneficial to fault diagnosis.
In order to achieve the above purpose, the invention adopts the following technical scheme:
a synchronous phase modulator on-line monitoring and fault diagnosis method is characterized in that:
training a multi-fault classifier through fault characteristic vectors of turn-to-turn magnetic fluxes of a stator and a rotor and fault characteristic vectors of vibration signals under fault operation of a synchronous phase modulator; inputting the characteristic vector of the synchronous phase modulator under the condition of unknown fault into the trained multi-fault classifier to obtain a corresponding fault category; meanwhile, the fault type of the synchronous phase modulator is diagnosed through the stator and rotor current characteristic vectors, and the fault of the synchronous phase modulator is comprehensively judged.
An online monitoring and fault diagnosis method for a synchronous phase modulator is characterized by comprising the following steps:
the method comprises the following steps: collecting a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of a synchronous phase modulator under a normal running state of the synchronous phase modulator, and collecting a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator under a laboratory simulation synchronous phase modulator fault state;
step two: preprocessing the collected stator and rotor current signals, synchronous phase modulator vibration signals and magnetic flux signals in a normal operation state and a fault state;
step three: carrying out wavelet packet decomposition and reconstruction on the preprocessed stator and rotor current signals of the synchronous phase modulator to obtain stator and rotor current characteristic vectors of the synchronous phase modulator in a normal operation state and in different corresponding fault states;
carrying out empirical mode decomposition on the preprocessed vibration signals of the synchronous phase modulator, and extracting the characteristic vectors of the vibration signals in a normal operation state and different fault states;
the numerical change of the magnetic flux of the end part of the stator and rotor winding after pretreatment, namely the difference value of the magnetic flux of the end part of the stator and rotor winding under the normal operation state and the fault state is used as the fault characteristic vector of the turn-to-turn magnetic flux of the stator and rotor of the synchronous phase modulator;
step four: fault characteristic vectors of turn-to-turn magnetic fluxes of a stator and a rotor of a synchronous phase modulator and fault characteristic vectors of vibration signals, namely the vibration signal characteristic vectors in different fault states, form a fault characteristic vector set, and the fault characteristic vector set is input as a training sample set of a multi-fault classifier (hereinafter referred to as a multi-fault classifier) based on a least square support vector machine to train the multi-fault classifier;
taking the stator and rotor turn-to-turn magnetic flux characteristic vector and the vibration signal characteristic vector of the synchronous phase modulator in a normal operation state as a basic data set of normal operation, and setting different fault categories by a multi-fault classifier according to the basic data set and a fault characteristic vector set;
step five: acquiring a synchronous phase modulator vibration signal, a stator-rotor inter-turn magnetic flux signal and a stator-rotor current signal under the condition of unknown fault in real time; extracting vibration signal characteristic vectors and stator and rotor inter-turn magnetic flux characteristic vectors under the condition of unknown faults, forming a characteristic vector set, inputting the characteristic vector set into the multi-fault classifier obtained after the training in the step four, and obtaining the specific fault category of the synchronous phase modulator;
step six: calculating a stator and rotor current characteristic vector according to the mode of the step three based on the stator and rotor current signals of the synchronous phase modulator under the unknown fault condition acquired in the step five, diagnosing the fault type of the synchronous phase modulator through the stator and rotor current characteristic vector, and if the fault type diagnosis result is the same as the diagnosis result of the step five, judging that the synchronous phase modulator has the fault of the type; and if the results are different, data collection is carried out again, the first step to the sixth step are repeated again, and if the judgment results are still different, classification check is carried out on different faults diagnosed in the two modes during maintenance.
The invention further comprises the following preferred embodiments:
in the first step, the current sensor is a Hall current sensor;
the vibration signals of the synchronous phase modulator are measured through a plurality of groups of acceleration sensors, and the acceleration sensors are arranged on the outer surface of the phase modulator and at positions, which are rigidly connected, in the phase modulator.
The acceleration sensors are provided with 8 measuring points, wherein 6 acceleration sensors are arranged at the parts of the outer shell body with rigid connection, and the other two acceleration sensors are arranged at the two ends of the rotating shaft.
In step two, the pre-processing of the stator and rotor current signals of the synchronous phase modulator comprises: collecting current signals of a stator and a rotor through a current sensor, and then amplifying the signals through an isolation amplification module and removing ripple components;
the preprocessing of the vibration signal of the synchronous phase modulator comprises: the vibration signal is collected by the acceleration sensor, and then the interference signal is filtered by the low-pass filter circuit.
In the third step, wavelet decomposition is carried out on the stator and rotor current signals of the synchronous phase modulator simulated by the laboratory in different fault states, and characteristic quantities are extracted, wherein different faults correspond to different characteristic quantities, and characteristic quantity ranges corresponding to different faults are obtained.
In the third step, wavelet packet decomposition and reconstruction are carried out on the stator and rotor current signals according to the following steps:
(1) firstly, carrying out wavelet packet 3-layer decomposition on an acquired signal;
(2) reconstructing the wavelet packet decomposition coefficient, and extracting signals in each frequency band range; the reconstructed total signal S can be expressed as:
S＝S0+S1+S2+S3+S4+S5+S6+S7；
wherein S is0、S1、S2、S3、S4、S5、S6、S7Respectively representing the reconstructed signals of the wavelet decomposition coefficients from low frequency to high frequency;
(3) calculating the total energy of each frequency band signal, and setting Sj(j 0, 1.. 7.) corresponds to an energy Ej(j ═ 0,1,. 7), then:
wherein xjk(j-0, 1.. 7; k-1, 2.. n) denotes a reconstructed signal SjN is the number of discrete points;
(4) constructing a feature vector, and constructing a feature vector T by taking the energy of each frequency band signal as an element:
T＝[E0,E1,E2,E3,E4,E5,E6,E7]
carrying out normalization processing on the characteristic vector T; order to
The normalized vector T' is
T′＝[E0/E,E1/E,E2/E,E3/E,E4/E,E5/E,E6/E,E7/E]。
In the third step, the preprocessed vibration signals of the synchronous phase modulator are subjected to empirical mode decomposition according to the following steps, and the characteristic vectors of the vibration signals are extracted;
(1) respectively carrying out cubic spline interpolation on the maximum value point and the minimum value point of the vibration signal x (t) to obtain an upper envelope line B and a lower envelope line B1And B2The upper and lower envelope lines should envelop all data points;
(2) calculating the average value m of the upper envelope and the lower envelope1；
(3) Computing the eigenmode function h1：
h1＝x(t)-m1；
(4) If h is1Satisfies the following two conditions, then h1Is the first eigenmode function of x (t), denoted as c1Otherwise, to h1Repeating the steps (1) to (3) until the two conditions are met;
two conditions are satisfied: 1) the extreme points of the function are equal to or different from the zero point number by one; 2) at any moment, the mean value of the upper envelope line and the lower envelope line of the function is 0;
(5) c is to1Separate from x (t), i.e.:
r1＝x(t)-c1
continue to r1And (5) decomposing, repeating the steps (1) to (5), and cycling for h times to obtain vibration signal feature vectors, namely h intrinsic mode function components meeting the conditions and a remainder.
In step (5), the number of cycles h is preferably 8.
In step four, the multi-fault classifier based on the least square support vector machine takes the error square sum loss function as the empirical loss of the data training set, and converts the solving quadratic programming problem into the problem of solving the linear equation set, wherein the training is completed by the following formula:
s.t.yi(WTg(xi)+b)＝1-ξi
i＝1,...,M
wherein { (x)1,y1),(x2,y2),...,(xM,yM) Is the training set with M samples, sample xi(i 1.. M) corresponding to category yiE { -1,1}, then there exists an optimal classification hyperplane that satisfies the following condition:
w is the normal vector of the hyperplane, C is the planning factor, ξiIs an error variable; m is the number of training sample sets;
b is an offset; g (x) is a function that maps x from the input space to the feature space;
constructing l (l-1)/2 two-class classifiers, and taking the training sample of the mth class as one class with the class label y when constructing the mth classifier in the l classifiers i1, taking training samples of all the other classes except the m classes as one class, wherein the class labels are
The classification output function of the mth classifier is:
αiis Lagrange multiplier; k (x)iAnd x) is a kernel function of the support vector machine; inputting the test data sample x into the classifier, if the formula f is determined1(x) If 1, it is determined as class 1, and so on.
Wherein the value of l (l-1)/2 is preferably 6.
The application also discloses an online monitoring and fault diagnosis system of the synchronous phase modulator applying the fault diagnosis method, wherein the online monitoring and fault diagnosis system comprises a signal acquisition module, a signal acquisition preprocessing module, an A/D conversion module, a wireless transmission module, a wavelet packet transformation module, an empirical mode decomposition module, a numerical variable quantity calculation module, a fault mode identification module, a multi-fault classifier and a comprehensive judgment module; the method is characterized in that:
the signal acquisition module acquires a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator in a fault operation state and a normal operation state of the synchronous phase modulator, and transmits the acquired signals to the acquired signal preprocessing module;
the acquisition signal preprocessing module comprises an isolation amplifying circuit unit and a low-pass filter circuit unit, and is used for amplifying the stator and rotor current signals through the isolation amplifying circuit unit and removing ripple components; filtering the vibration signal by a low-pass filter circuit unit to remove interference signals;
the synchronous phase modulator stator and rotor current signals, the synchronous phase modulator vibration signals and the magnetic flux signals at the end of the synchronous phase modulator stator and rotor winding are input into an A/D conversion circuit module to be converted into corresponding digital signals, and then the digital signals are respectively uploaded to a wavelet transformation module, an empirical mode decomposition module and a numerical value change calculation module through a wireless transmission module;
the wavelet packet change module carries out wavelet packet decomposition and reconstruction on the preprocessed stator and rotor current signals of the synchronous phase modulator to obtain the characteristic vector of the stator and rotor current;
the empirical mode decomposition module is used for carrying out an empirical mode decomposition module on the preprocessed vibration signals of the synchronous phase modulator, and extracting the characteristic vectors of the vibration signals;
the numerical variation calculation module calculates the numerical variation of the magnetic flux at the end part of the stator and rotor winding as a fault characteristic vector of the turn-to-turn magnetic flux of the stator and rotor of the synchronous phase modulator;
uploading the feature vector of the vibration signal and the fault feature vector of the stator-rotor turn-to-turn magnetic flux to a multi-fault classifier for training, and realizing fault classification by the trained fault classifier based on the feature vector of the vibration signal and the fault feature vector of the stator-rotor turn-to-turn magnetic flux;
extracting characteristic quantities of stator and rotor currents through a wavelet packet conversion module, and uploading the characteristic quantities to a fault mode identification module, wherein the fault mode identification module identifies fault types according to the characteristic vectors of different stator and rotor currents;
and the comprehensive judgment module synthesizes the fault diagnosis results of the fault mode identification module and the multi-fault classifier, and finally realizes the identification and comprehensive diagnosis of the fault mode.
The signal acquisition module comprises a Hall sensor for acquiring current signals of the stator and the rotor, an acceleration sensor for acquiring vibration signals of the synchronous phase modulator, and a fluxmeter for acquiring magnetic flux signals of the end part of a stator winding and a rotor winding of the synchronous phase modulator.
The invention achieves the following beneficial effects:
1. the measured signal of the invention realizes the field digitization, which is beneficial to improving the measurement precision and avoiding the problems of attenuation, interference and the like of the long-distance transmission of the traditional analog signal;
2. the invention can realize multi-parameter comprehensive diagnosis of the synchronous phase modulator, solves the problem of simplification of fault diagnosis parameters of the existing synchronous phase modulator, and improves the accuracy of fault diagnosis.
Drawings
FIG. 1 is a schematic diagram of the on-line monitoring and fault diagnosis system of the synchronous phase modulator of the present invention;
FIG. 2 is a schematic flow chart of the on-line monitoring and fault diagnosis method of the synchronous phase modulator of the present invention;
FIG. 3 is a schematic diagram of the arrangement positions of vibration signal measuring points of the phase modulator according to the embodiment of the present invention;
FIG. 4 is a 3-layer exploded view of a wavelet packet according to an embodiment of the present invention;
FIG. 5 is a flowchart of a multi-classifier algorithm according to an embodiment of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
Fig. 2 is a schematic flow chart of the method for on-line monitoring and fault diagnosis of a synchronous phase modulator, and the method for on-line monitoring and fault diagnosis of a synchronous phase modulator disclosed by the invention specifically comprises the following steps:
the method comprises the following steps: collecting a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator in a fault operation state and a normal operation state of the synchronous phase modulator;
the current sensor is a Hall current sensor and is arranged at the current outlet end;
in the preferred embodiment of the application, 8 measuring points are arranged at the place where the outer surface and the inner part of the synchronous phase modifier are rigidly connected, and the vibration signal of the synchronous phase modifier is measured, as shown in figure 3;
the phase modulator vibration signal is provided with 8 measuring points, wherein No. 1-6 acceleration sensors are respectively arranged at the parts of the outer shell body with rigid connection, and No. 7 and No. 8 acceleration sensors are arranged at two ends of the rotating shaft.
And 3-5 groups of stator and rotor current signals, synchronous phase modifier vibration signals and synchronous phase modifier stator and rotor winding end magnetic flux signals of the synchronous phase modifier in fault operation states and in normal operation states are measured respectively.
Step two: preprocessing the collected stator and rotor current signals, the collected synchronous phase modifier vibration signals and the collected magnetic flux signals;
the preprocessing of the vibration signal of the synchronous phase modulator comprises: the vibration signal is collected by the acceleration sensor, then the interference signal is filtered by the low-pass filter circuit, and the interference signal is converted into a digital signal by the A/D conversion module and then is uploaded to the empirical mode decomposition module;
the preprocessing of the vibration signal of the synchronous phase modulator comprises: the vibration signal is collected by the acceleration sensor, then the interference signal is filtered by the low-pass filter circuit, and the interference signal is converted into a digital signal by the A/D conversion module and then is uploaded to the empirical mode decomposition module;
the pretreatment of the magnetic flux of the stator and rotor winding end comprises the following steps: the magnetic flux signals are collected by the fluxmeter and then converted into digital signals through the A/D conversion module, and the digital signals are used as fault characteristic vectors of turn-to-turn magnetic fluxes of the stator and the rotor.
Step three: carrying out wavelet packet decomposition and reconstruction on the preprocessed stator and rotor current signals of the synchronous phase modulator to obtain stator and rotor current characteristic vectors of the synchronous phase modulator;
wavelet packet analysis divides frequency bands in multiple layers, further decomposes high-frequency parts which are not subdivided in multi-resolution analysis, and can adaptively select corresponding frequency bands according to the characteristics of analysis signals to enable the corresponding frequency bands to be matched with signal frequency spectrums, so that time-frequency resolution is improved. Adopting a fault diagnosis mode algorithm of 'energy-fault':
(1) first, wavelet packet decomposition (taking 3-layer decomposition as an example) is performed on the acquired signal, and the decomposition structure is shown in fig. 1. Wherein, the (0,0) node represents the original signal S, the (1,0) node represents the first-layer low-frequency coefficient of wavelet packet decomposition, the (1,1) node represents the first-layer high-frequency coefficient, the (3,0) node represents the coefficient of the 0-th node of the 3 rd layer, the (3,1) node represents the coefficient of the 1-th node of the 3 rd layer, and the rest are similar.
(2) And reconstructing the wavelet packet decomposition coefficient, and extracting signals in each frequency band range. With S0、S1、S2、S3、S4、S5、S6、S7Respectively representing the reconstructed signals of the wavelet decomposition coefficients from low frequency to high frequency. The total signal S can be expressed as
S＝S0+S1+S2+S3+S4+S5+S6+S7
(3) And calculating the total energy of the signals of each frequency band. Let Sj(j 0, 1.. 7.) corresponds to an energy Ej(j ═ 0,1,. 7), then:
wherein xjk(j-0, 1.. 7; k-1, 2.. n) denotes a reconstructed signal SjThe amplitude of the discrete points of (a).
(4) A feature vector is constructed. Constructing a feature vector T for the element by energy:
T＝[E0,E1,E2,E3,E4,E5,E6,E7]
and when the energy is larger, normalizing the feature vector T. Order to
The normalized vector T' is
T′＝[E0/E,E1/E,E2/E,E3/E,E4/E,E5/E,E6/E,E7/E]
Performing Empirical Mode Decomposition (EMD) on the preprocessed vibration signals of the synchronous phase modulator, and extracting the characteristic vectors of the vibration signals;
the Intrinsic Mode Function (IMF) is proposed to make the instantaneous frequency of the signal have physical significance, and the IMF needs to satisfy two conditions: 1) the extreme points of the function are equal to or different from the zero point number by one; 2) at any time, the mean value of the upper envelope and the lower envelope of the function is 0.
The EMD starts from the local scale characteristics of signals, adopts the symmetrical rule of signal envelope lines, and comprises the following specific steps:
1) respectively carrying out cubic spline interpolation on the maximum value point and the minimum value point of the signal x (t) to obtain an upper envelope line B and a lower envelope line B1And B2The upper and lower envelope should envelope all data points.
2) Calculating the average value m of the upper envelope and the lower envelope1。
3) Calculating h1：h1＝x(t)-m1。
4) If h is1If IMF condition is satisfied, then h1Is the first IMF of x (t), denoted as c1Otherwise, to h1And (4) repeating the steps (1) to (3) until the IMF condition is met.
5) C is to1Separate from x (t), i.e.:
r1＝x(t)-c1
continue to r1And (4) decomposing, repeating the steps (1) to (5), and cycling for h times to obtain a vibration signal feature vector, namely h IMF components meeting the condition and a remainder (at most, only one extreme point).
In a particular embodiment of the present application, the number of cycles h is preferably 8.
And the numerical value change of the magnetic flux of the preprocessed stator and rotor winding end parts is used as a fault characteristic vector of the turn-to-turn magnetic flux of the stator and the rotor of the synchronous phase modulator.
Step four: fault characteristic vectors of turn-to-turn magnetic fluxes of a stator and a rotor under fault operation of a synchronous phase modulator and fault characteristic vectors of vibration signals form a fault characteristic vector set, the fault characteristic vector set is used as data training input of a multi-fault classifier (hereinafter referred to as a multi-fault classifier) based on a least square support vector machine, and the classifier of the vector machine is trained; taking stator and rotor inter-turn magnetic flux characteristic vectors and vibration signal characteristic vectors under the normal operation condition of a synchronous phase modulator as a basic data set of normal operation, and setting data ranges of vector sets corresponding to different faults by a vector machine classifier according to the basic data set and a fault characteristic vector set;
the multi-fault classifier (LS-SVM) based on the least square support vector machine changes the traditional inequality constraint into the equality constraint, takes the error square sum loss function as the experience loss of a training set, converts the problem of solving the quadratic programming into the problem of solving a linear equation set, and improves the solving speed and the convergence precision. The training is accomplished by the following formula:
s.t.yi(WTg(xi+b))＝1-ξi
i＝1,...,M
wherein { (x)1,y1),(x2,y2),...,(xM,yM) Is the training set with M samples, sample xi(i 1.. M) corresponding to category yiE { -1,1}, then there exists an optimal classification hyperplane that satisfies the following condition:
w is the normal vector of the hyperplane, C is the planning factor, ξiIs an error variable; m is the number of training sample sets;
b is an offset; g (x) is a function that maps x from the input space to the feature space;
constructing l (l-1)/2 two-class classifiers, and taking the training sample of the mth class as one class with the class label y when constructing the mth classifier in the l classifiers i1, taking training samples of all the other classes except the m classes as one class, wherein the class labels areThe classification output function of the mth classifier is:
αiis Lagrange multiplier; k (x)iAnd x) is a kernel function of the support vector machine; inputting the test data sample x into the classifier, if the formula f is determined1(x) If it is 1, it is determined as category 1, and so on, and if f ism(x) If 1, it is determined as the category m.
The classification flow is shown in fig. 5:
the fault categories can be classified into health conditions, rotor eccentric faults, rotor turn-to-turn short circuit faults and stator turn-to-turn short circuit faults, and are assumed to correspond to no fault, a fault category 1, a fault category 2 and a fault category 3 respectively.
Step five: inputting the characteristic vector set of the synchronous phase modulator under the condition of unknown fault into the multi-fault classifier obtained after training in the fourth step, wherein different faults of the synchronous phase modulator correspond to different stator and rotor vibration signal fault characteristic vectors and stator and rotor turn-to-turn magnetic flux fault characteristic vectors, so that fault categories output by the vector machine classifier are obtained;
step six: diagnosing the fault type of the synchronous phase modulator through the stator and rotor current characteristic vectors, and if the diagnosis result is the same as that in the step five, judging that the fault type of the synchronous phase modulator occurs; and if the results are different, data collection is carried out again, the first step to the sixth step are repeated again until the same diagnosis result is obtained, and the specific fault mode of the synchronous phase modulator is comprehensively judged.
The application also discloses a synchronous phase modulator on-line monitoring and fault diagnosis system based on the on-line monitoring and fault diagnosis method, as shown in the attached drawing 1, the on-line monitoring and fault diagnosis system comprises a signal acquisition module, a signal acquisition preprocessing module, a wireless transmission module, a wavelet packet transformation module, an empirical mode decomposition module, a numerical variable quantity calculation module, a fault mode identification module, a multi-fault classifier and a comprehensive judgment module.
The current signals of the stator and the rotor collected by the Hall sensor, the vibration signals collected by the acceleration sensor and the magnetic flux signals collected by the magnetic flowmeter are processed by the signal processing module and then converted into digital signals, the digital signals are transmitted to an upper computer through the wireless transmission module, the current of the stator and the rotor is extracted by the wavelet packet transformation module to obtain characteristic quantities, and if the fault diagnosis module diagnoses the fault, the fault type is specifically judged through fault mode identification. Extracting the feature vector of the vibration signal through an empirical mode decomposition module, forming a comprehensive feature vector by the feature vector formed by the feature vector and the numerical variation, inputting the comprehensive feature vector to a multi-fault classifier, classifying faults, and integrating fault mode identification and fault diagnosis results of the multi-fault classifier to finally realize identification and comprehensive diagnosis of the fault mode.
The signal acquisition module comprises a Hall current sensor, an acceleration sensor and a magnetic flowmeter. The signal acquisition module acquires a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator in a fault operation state and a normal operation state of the synchronous phase modulator, and transmits the acquired signals to the acquired signal preprocessing module;
the Hall current sensor is used for collecting current signals of the stator and the rotor, the current signals are amplified through the isolation amplification module and ripple components are removed, the signals are converted into digital signals through the A/D conversion module, and the digital signals are uploaded to an upper computer through the wireless transmission module;
the vibration signal acquisition process of the preprocessed synchronous phase modulator comprises the following steps: the vibration signal is collected by the acceleration sensor, then the interference signal is filtered by a low-pass filter circuit with the frequency of 10kHz, the influence of the out-of-band signal on the measurement result is reduced, the out-of-band signal is converted into a digital signal by the A/D conversion module, and the digital signal is uploaded to an upper computer by the wireless transmission module; the acceleration sensor is arranged on the rigid outer shell;
the magnetic flux of the pre-processed stator and rotor winding end is collected by a magnetic flowmeter and then converted into a digital signal through an A/D module, and the magnetic flowmeter is installed at a wire inlet end.
The data signals are conditioned through a conditioning circuit, and are stored after analog-to-digital conversion in a single chip microcomputer; after the single chip microcomputer receives a transmission instruction of the upper computer, the processed data information is sent to the upper computer through the ZigBee wireless transmission module;
the method and the device can realize multi-parameter comprehensive diagnosis of the synchronous phase modulator, solve the problem of simplification of fault diagnosis parameters of the conventional synchronous phase modulator, and improve the accuracy of fault diagnosis.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (13)
1. A synchronous phase modulator on-line monitoring and fault diagnosis method is characterized in that:
training a multi-fault classifier through fault characteristic vectors of turn-to-turn magnetic fluxes of a stator and a rotor and fault characteristic vectors of vibration signals under fault operation of a synchronous phase modulator; inputting the characteristic vector of the synchronous phase modulator under the condition of unknown fault into the trained multi-fault classifier to obtain a corresponding fault category; meanwhile, the fault type of the synchronous phase modulator is diagnosed through the stator and rotor current characteristic vectors, and the fault of the synchronous phase modulator is comprehensively judged.
2. An online monitoring and fault diagnosis method for a synchronous phase modulator is characterized by comprising the following steps:
the method comprises the following steps: collecting a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of a synchronous phase modulator under a normal running state of the synchronous phase modulator, and collecting a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator under a laboratory simulation synchronous phase modulator fault state;
step two: preprocessing the collected stator and rotor current signals, synchronous phase modulator vibration signals and magnetic flux signals in a normal operation state and a fault state;
step three: carrying out wavelet packet decomposition and reconstruction on the preprocessed stator and rotor current signals of the synchronous phase modulator to obtain stator and rotor current characteristic vectors of the synchronous phase modulator in a normal operation state and in different corresponding fault states;
carrying out empirical mode decomposition on the preprocessed vibration signals of the synchronous phase modulator, and extracting the characteristic vectors of the vibration signals in a normal operation state and different fault states;
the numerical change of the magnetic flux of the end part of the stator and rotor winding after pretreatment, namely the difference value of the magnetic flux of the end part of the stator and rotor winding under the normal operation state and the fault state is used as the fault characteristic vector of the turn-to-turn magnetic flux of the stator and rotor of the synchronous phase modulator;
step four: fault characteristic vectors of turn-to-turn magnetic fluxes of a stator and a rotor of a synchronous phase modulator and fault characteristic vectors of vibration signals, namely the vibration signal characteristic vectors in different fault states, form a fault characteristic vector set, and the fault characteristic vector set is input as a training sample set of a multi-fault classifier (hereinafter referred to as a multi-fault classifier) based on a least square support vector machine to train the multi-fault classifier;
taking the stator and rotor turn-to-turn magnetic flux characteristic vector and the vibration signal characteristic vector of the synchronous phase modulator in a normal operation state as a basic data set of normal operation, and setting different fault categories by a multi-fault classifier according to the basic data set and a fault characteristic vector set;
step five: acquiring a synchronous phase modulator vibration signal, a stator-rotor inter-turn magnetic flux signal and a stator-rotor current signal under the condition of unknown fault in real time; extracting vibration signal characteristic vectors and stator and rotor inter-turn magnetic flux characteristic vectors under the condition of unknown faults, forming a characteristic vector set, inputting the characteristic vector set into the multi-fault classifier obtained after the training in the step four, and obtaining the specific fault category of the synchronous phase modulator;
step six: calculating a stator and rotor current characteristic vector according to the mode of the step three based on the stator and rotor current signals of the synchronous phase modulator under the unknown fault condition acquired in the step five, diagnosing the fault type of the synchronous phase modulator through the stator and rotor current characteristic vector, and if the fault type diagnosis result is the same as the diagnosis result of the step five, judging that the synchronous phase modulator has the fault of the type; and if the results are different, data collection is carried out again, the first step to the sixth step are repeated again, and if the judgment results are still different, classification check is carried out on different faults diagnosed in the two modes during maintenance.
3. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 2, characterized in that:
in the first step, the current sensor is a Hall current sensor;
the vibration signals of the synchronous phase modulator are measured through a plurality of groups of acceleration sensors, and the acceleration sensors are arranged on the outer surface of the phase modulator and at positions, which are rigidly connected, in the phase modulator.
4. The synchronous phase modulator on-line monitoring and fault diagnosis method of claim 3, characterized in that:
the acceleration sensors are provided with 8 measuring points, wherein 6 acceleration sensors are arranged at the parts of the outer shell body with rigid connection, and the other two acceleration sensors are arranged at the two ends of the rotating shaft.
5. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 2, characterized in that:
in step two, the pre-processing of the stator and rotor current signals of the synchronous phase modulator comprises: collecting current signals of a stator and a rotor through a current sensor, and then amplifying the signals through an isolation amplification module and removing ripple components;
the preprocessing of the vibration signal of the synchronous phase modulator comprises: the vibration signal is collected by the acceleration sensor, and then the interference signal is filtered by the low-pass filter circuit.
6. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 2 or 5, characterized in that:
in the third step, wavelet decomposition is carried out on the stator and rotor current signals of the synchronous phase modulator simulated by the laboratory in different fault states, and characteristic quantities are extracted, wherein different faults correspond to different characteristic quantities, and characteristic quantity ranges corresponding to different faults are obtained.
7. The synchronous phase modulator on-line monitoring and fault diagnosis method of claim 6, characterized in that:
in the third step, wavelet packet decomposition and reconstruction are carried out on the stator and rotor current signals according to the following steps:
(1) firstly, carrying out wavelet packet 3-layer decomposition on an acquired signal;
(2) reconstructing the wavelet packet decomposition coefficient, and extracting signals in each frequency band range; the reconstructed total signal S can be expressed as:
S＝S0+S1+S2+S3+S4+S5+S6+S7；
wherein S is0、S1、S2、S3、S4、S5、S6、S7Respectively representing the reconstructed signals of the wavelet decomposition coefficients from low frequency to high frequency;
(3) calculating the total energy of each frequency band signal, and setting Sj(j 0, 1.. 7.) corresponds to an energy Ej(j ═ 0,1,. 7), then:
wherein xjk(j-0, 1.. 7; k-1, 2.. n) denotes a reconstructed signal SjN is the number of discrete points;
(4) constructing a feature vector, and constructing a feature vector T by taking the energy of each frequency band signal as an element:
T＝[E0,E1,E2,E3,E4,E5,E6,E7]
carrying out normalization processing on the characteristic vector T; order to
The normalized vector T' is
T′＝[E0/E,E1/E,E2/E,E3/E,E4/E,E5/E,E6/E,E7/E]。
8. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 2 or 5, characterized in that:
in the third step, the preprocessed vibration signals of the synchronous phase modulator are subjected to empirical mode decomposition according to the following steps, and the characteristic vectors of the vibration signals are extracted;
(1) respectively carrying out cubic spline interpolation on the maximum value point and the minimum value point of the vibration signal x (t) to obtain an upper envelope line B and a lower envelope line B1And B2The upper and lower envelope lines should envelop all data points;
(2) calculating the average value m of the upper envelope and the lower envelope1；
(3) Computing the eigenmode function h1：
h1＝x(t)-m1；
(4) If h is1Satisfies the following two conditions, then h1Is the first eigenmode function of x (t), denoted as c1Otherwise, to h1Repeating the steps (1) to (3) until the two conditions are met;
two conditions are satisfied: 1) the extreme points of the function are equal to or different from the zero point number by one; 2) at any moment, the mean value of the upper envelope line and the lower envelope line of the function is 0;
(5) c is to1Separated from the x (t) and the (c),namely:
r1＝x(t)-c1
continue to r1And (5) decomposing, repeating the steps (1) to (5), and cycling for h times to obtain vibration signal feature vectors, namely h intrinsic mode function components meeting the conditions and a remainder.
9. The synchronous phase modulator on-line monitoring and fault diagnosis method of claim 8, characterized in that:
in step (5), the number of cycles h is preferably 8.
10. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 2, characterized in that:
in step four, the multi-fault classifier based on the least square support vector machine takes the error square sum loss function as the empirical loss of the data training set, and converts the solving quadratic programming problem into the problem of solving the linear equation set, wherein the training is completed by the following formula:
s.t.yi(WTg(xi)+b)＝1-ξi
i＝1,...,M
wherein { (x)1,y1),(x2,y2),...,(xM,yM) Is the training set with M samples, sample xi(i 1.. M) corresponding to category yiE { -1,1}, then there exists an optimal classification hyperplane that satisfies the following condition:
w is the normal vector of the hyperplane, C is the planning factor, ξiIs an error variable; m is the number of training sample sets;
b is an offset; g (x) is a function that maps x from the input space to the feature space;
constructing l (l-1)/2 two-class classifiers, and taking the training sample of the mth class as one class with the class label y when constructing the mth classifier in the l classifiersi1, taking training samples of all the other classes except the m classes as one class, wherein the class labels are
The classification output function of the mth classifier is:
αiis Lagrange multiplier; k (x)iAnd x) is a kernel function of the support vector machine; inputting the test data sample x into the classifier, if the formula f is determined1(x) If 1, it is determined as class 1, and so on.
11. The synchronous phase modulator on-line monitoring and fault diagnosis method according to claim 10, characterized in that:
the value of l (l-1)/2 is preferably 6.
12. An online monitoring and fault diagnosis system of a synchronous phase modulation machine applying the fault diagnosis method of any one of claims 1 to 11, wherein the online monitoring and fault diagnosis system comprises a signal acquisition module, an acquired signal preprocessing module, an A/D conversion module, a wireless transmission module, a wavelet packet transformation module, an empirical mode decomposition module, a numerical variation calculation module, a fault mode identification module, a multi-fault classifier and a comprehensive judgment module; the method is characterized in that:
the signal acquisition module acquires a stator and rotor current signal, a synchronous phase modulator vibration signal and a magnetic flux signal of a stator and rotor winding end part of the synchronous phase modulator in a fault operation state and a normal operation state of the synchronous phase modulator, and transmits the acquired signals to the acquired signal preprocessing module;
the acquisition signal preprocessing module comprises an isolation amplifying circuit unit and a low-pass filter circuit unit, and is used for amplifying the stator and rotor current signals through the isolation amplifying circuit unit and removing ripple components; filtering the vibration signal by a low-pass filter circuit unit to remove interference signals;
the synchronous phase modulator stator and rotor current signals, the synchronous phase modulator vibration signals and the magnetic flux signals at the end of the synchronous phase modulator stator and rotor winding are input into an A/D conversion circuit module to be converted into corresponding digital signals, and then the digital signals are respectively uploaded to a wavelet transformation module, an empirical mode decomposition module and a numerical value change calculation module through a wireless transmission module;
the wavelet packet change module carries out wavelet packet decomposition and reconstruction on the preprocessed stator and rotor current signals of the synchronous phase modulator to obtain the characteristic vector of the stator and rotor current;
the empirical mode decomposition module is used for carrying out an empirical mode decomposition module on the preprocessed vibration signals of the synchronous phase modulator, and extracting the characteristic vectors of the vibration signals;
the numerical variation calculation module calculates the numerical variation of the magnetic flux at the end part of the stator and rotor winding as a fault characteristic vector of the turn-to-turn magnetic flux of the stator and rotor of the synchronous phase modulator;
uploading the feature vector of the vibration signal and the fault feature vector of the stator-rotor turn-to-turn magnetic flux to a multi-fault classifier for training, and realizing fault classification by the trained fault classifier based on the feature vector of the vibration signal and the fault feature vector of the stator-rotor turn-to-turn magnetic flux;
extracting characteristic quantities of stator and rotor currents through a wavelet packet conversion module, and uploading the characteristic quantities to a fault mode identification module, wherein the fault mode identification module identifies fault types according to the characteristic vectors of different stator and rotor currents;
and the comprehensive judgment module synthesizes the fault diagnosis results of the fault mode identification module and the multi-fault classifier, and finally realizes the identification and comprehensive diagnosis of the fault mode.
13. The synchronous phase modulator on-line monitoring and fault diagnosis system of claim 12, wherein:
the signal acquisition module comprises a Hall sensor for acquiring current signals of the stator and the rotor, an acceleration sensor for acquiring vibration signals of the synchronous phase modulator, and a fluxmeter for acquiring magnetic flux signals of the end part of a stator winding and a rotor winding of the synchronous phase modulator.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911138356.0A CN110824281A (en) | 2019-11-19 | 2019-11-19 | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911138356.0A CN110824281A (en) | 2019-11-19 | 2019-11-19 | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110824281A true CN110824281A (en) | 2020-02-21 |
Family
ID=69557036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911138356.0A Pending CN110824281A (en) | 2019-11-19 | 2019-11-19 | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110824281A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111474475A (en) * | 2020-03-22 | 2020-07-31 | 华南理工大学 | Motor fault diagnosis system and method |
| CN111783696A (en) * | 2020-07-06 | 2020-10-16 | 广东电网有限责任公司 | Edge calculation method for low-voltage branch topology real-time analysis based on PV relation |
| CN112327208A (en) * | 2020-11-02 | 2021-02-05 | 国网江苏省电力有限公司电力科学研究院 | Fault diagnosis method and device for turn-to-turn short circuit of phase modulator rotor winding |
| CN112465751A (en) * | 2020-11-14 | 2021-03-09 | 国网湖北省电力有限公司电力科学研究院 | Automatic detection method for physical surface in air gap of rotor of large phase modulator without pumping |
| CN112881909A (en) * | 2021-01-19 | 2021-06-01 | 哈尔滨理工大学 | Wavelet transform-based stator winding short-circuit fault diagnosis method for synchronous phase modulator |
| CN113155422A (en) * | 2021-03-30 | 2021-07-23 | 广州大学 | Fault detection method, device and storage medium |
| CN113156309A (en) * | 2021-03-29 | 2021-07-23 | 华北电力大学（保定） | Weak turn-to-turn short circuit fault diagnosis method for rotor winding of synchronous phase modulator |
| CN114062995A (en) * | 2021-11-15 | 2022-02-18 | 通号（长沙）轨道交通控制技术有限公司 | Mutual inductor fault diagnosis method, equipment and medium based on electric quantity multi-feature fusion |
| CN116400209A (en) * | 2023-03-10 | 2023-07-07 | 广东电网有限责任公司佛山供电局 | Method and device for diagnosing voltage regulating switch fault of distribution transformer |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04359622A (en) * | 1991-06-05 | 1992-12-11 | Toshiba Corp | Protector of synchronous phase modifier |
| CN108318815A (en) * | 2018-01-17 | 2018-07-24 | 河海大学 | A kind of doubly-fed wind turbine on-line monitoring and synthetic fault diagnosis method |
| CN208076675U (en) * | 2018-04-02 | 2018-11-09 | 国网湖南省电力有限公司 | A kind of trouble-shooter for large-scale phase modifier |
-
2019
- 2019-11-19 CN CN201911138356.0A patent/CN110824281A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04359622A (en) * | 1991-06-05 | 1992-12-11 | Toshiba Corp | Protector of synchronous phase modifier |
| CN108318815A (en) * | 2018-01-17 | 2018-07-24 | 河海大学 | A kind of doubly-fed wind turbine on-line monitoring and synthetic fault diagnosis method |
| CN208076675U (en) * | 2018-04-02 | 2018-11-09 | 国网湖南省电力有限公司 | A kind of trouble-shooter for large-scale phase modifier |
Non-Patent Citations (3)
| Title |
|---|
| 任芳: "《煤岩截割状态识别方法研究》", 31 December 2017, 煤炭工业出版社 * |
| 张玲玲 等: "《柴油发动机故障诊断技术》", 31 December 2015, 国防工业出版社 * |
| 杜经纬: "基于支持向量机的P2P流量管理模型设计", 《吉首大学学报（自然科学版）》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111474475A (en) * | 2020-03-22 | 2020-07-31 | 华南理工大学 | Motor fault diagnosis system and method |
| CN111474475B (en) * | 2020-03-22 | 2021-06-08 | 华南理工大学 | Motor fault diagnosis system and method |
| CN111783696A (en) * | 2020-07-06 | 2020-10-16 | 广东电网有限责任公司 | Edge calculation method for low-voltage branch topology real-time analysis based on PV relation |
| CN112327208A (en) * | 2020-11-02 | 2021-02-05 | 国网江苏省电力有限公司电力科学研究院 | Fault diagnosis method and device for turn-to-turn short circuit of phase modulator rotor winding |
| CN112465751A (en) * | 2020-11-14 | 2021-03-09 | 国网湖北省电力有限公司电力科学研究院 | Automatic detection method for physical surface in air gap of rotor of large phase modulator without pumping |
| CN112465751B (en) * | 2020-11-14 | 2022-08-23 | 国网湖北省电力有限公司电力科学研究院 | Automatic detection method for physical surface in air gap of rotor of large phase modulator without pumping |
| CN112881909A (en) * | 2021-01-19 | 2021-06-01 | 哈尔滨理工大学 | Wavelet transform-based stator winding short-circuit fault diagnosis method for synchronous phase modulator |
| CN113156309A (en) * | 2021-03-29 | 2021-07-23 | 华北电力大学（保定） | Weak turn-to-turn short circuit fault diagnosis method for rotor winding of synchronous phase modulator |
| CN113155422A (en) * | 2021-03-30 | 2021-07-23 | 广州大学 | Fault detection method, device and storage medium |
| CN113155422B (en) * | 2021-03-30 | 2022-07-26 | 广州大学 | Fault detection method, device and storage medium |
| CN114062995A (en) * | 2021-11-15 | 2022-02-18 | 通号（长沙）轨道交通控制技术有限公司 | Mutual inductor fault diagnosis method, equipment and medium based on electric quantity multi-feature fusion |
| CN116400209A (en) * | 2023-03-10 | 2023-07-07 | 广东电网有限责任公司佛山供电局 | Method and device for diagnosing voltage regulating switch fault of distribution transformer |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110824281A (en) | Method and system for on-line monitoring and fault diagnosis of synchronous phase modulator | |
| CN113255078A (en) | Bearing fault detection method and device under unbalanced sample condition | |
| CN110161343B (en) | Non-invasive real-time dynamic monitoring method for external powered device of intelligent train | |
| CN105841961A (en) | Bearing fault diagnosis method based on Morlet wavelet transformation and convolutional neural network | |
| Zhang et al. | Multi-fault diagnosis for rolling element bearings based on ensemble empirical mode decomposition and optimized support vector machines | |
| CN108318815A (en) | A kind of doubly-fed wind turbine on-line monitoring and synthetic fault diagnosis method | |
| CN102183366A (en) | Device and method for vibration measurement and failure analysis of rolling bearing | |
| CN108827605A (en) | A kind of mechanical breakdown characteristic automatic extraction method based on improvement sparseness filtering | |
| CN110307983B (en) | CNN-Bagging-based unmanned aerial vehicle bearing fault diagnosis method | |
| CN110333077B (en) | Rolling bearing fault diagnosis method based on approximate support vector | |
| Islam et al. | Motor bearing fault diagnosis using deep convolutional neural networks with 2d analysis of vibration signal | |
| CN112182490B (en) | Reactor state diagnosis method and system | |
| CN111855202A (en) | Gear box fault diagnosis method and system | |
| Han et al. | An intelligent fault diagnosis method of variable condition gearbox based on improved DBN combined with WPEE and MPE | |
| CN112052712B (en) | Power equipment state monitoring and fault identification method and system | |
| CN114755017A (en) | Variable-speed bearing fault diagnosis method of cross-domain data driving unsupervised field shared network | |
| CN116520140A (en) | Permanent magnet direct drive motor fault detection method integrating feature extraction and small sample classification | |
| CN115096581A (en) | Complex transmission device fault diagnosis tracing method and system based on time-frequency domain characteristics | |
| CN111795819A (en) | Gear box fault diagnosis method integrating vibration and current signal collaborative learning | |
| CN113092044B (en) | Rotary machine fault diagnosis method based on weighted level visible graph | |
| CN112686279B (en) | Gear box fault diagnosis method based on K-means clustering and evidence fusion | |
| Xue et al. | A novel framework for motor bearing fault diagnosis based on multi-transformation domain and multi-source data | |
| CN113758709A (en) | Rolling bearing fault diagnosis method and system combining edge calculation and deep learning | |
| Zheng et al. | A fault diagnosis method for planetary gear under multi-operating conditions based on adaptive extended bag-of-words model | |
| CN112014678A (en) | Three-phase voltage inverter online fault diagnosis method and device and electronic equipment |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200221 |