CN111680665A - Motor mechanical fault diagnosis method based on data driving and adopting current signals - Google Patents

Motor mechanical fault diagnosis method based on data driving and adopting current signals Download PDF

Info

Publication number
CN111680665A
CN111680665A CN202010597800.1A CN202010597800A CN111680665A CN 111680665 A CN111680665 A CN 111680665A CN 202010597800 A CN202010597800 A CN 202010597800A CN 111680665 A CN111680665 A CN 111680665A
Authority
CN
China
Prior art keywords
signal
current
motor
fault
frequency domain
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
Application number
CN202010597800.1A
Other languages
Chinese (zh)
Other versions
CN111680665B (en
Inventor
王辉
孙梅迪
黄守道
刘平
龙卓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hunan University
Original Assignee
Hunan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hunan University filed Critical Hunan University
Priority to CN202010597800.1A priority Critical patent/CN111680665B/en
Publication of CN111680665A publication Critical patent/CN111680665A/en
Application granted granted Critical
Publication of CN111680665B publication Critical patent/CN111680665B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Mathematical Physics (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Control Of Electric Motors In General (AREA)

Abstract

The invention discloses a data-driven motor mechanical fault diagnosis method adopting current signals, which comprises the steps of obtaining stator current signals, wherein the length of each stator current signal is D and the stator current signals comprise N samples; carrying out noise signal reconstruction on the stator current signal to obtain a reconstructed signal; subtracting the reconstructed signal from the original stator current signal to obtain a residual current; converting the signal of the residual current into a frequency domain to obtain a frequency domain signal; and inputting the frequency domain signal of the residual current into a pre-trained machine learning feature extraction and fault classification model to obtain the state of the motor. The invention adopts the current signal to diagnose the motor fault, can be used for diagnosing the mechanical fault, does not need to additionally install a vibration sensor and a current sensor, can acquire data through a current transformer which is already installed in a protection system or a control system, is a non-invasive low-cost mode, and has the advantages of high diagnosis efficiency, high diagnosis accuracy and low false alarm rate.

Description

Motor mechanical fault diagnosis method based on data driving and adopting current signals
Technical Field
The invention relates to a motor mechanical fault diagnosis technology, in particular to a motor mechanical fault diagnosis method based on data driving and adopting current signals.
Background
The motor faults can be divided into mechanical faults and electrical faults, and vibration signals are mostly adopted for diagnosis in the prior art aiming at the mechanical faults. However, the vibration sensor is expensive and has limited application in small and medium-sized motors. And vibration sensor needs direct contact motor, has the installation to place the problem, and some equipment that has already put into production can't install vibration sensor additional. Therefore, the existing vibration-based sensor has a problem of limited application range.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: the invention adopts the current signal to diagnose the motor fault, can be used for diagnosing the mechanical fault, does not need to additionally install a vibration sensor and a current sensor, can acquire data through a current transformer which is already installed in a protection system or a control system, is a non-invasive low-cost mode, and has the advantages of high diagnosis efficiency, high diagnosis accuracy and low false alarm rate.
In order to solve the technical problems, the invention adopts the technical scheme that:
a motor mechanical fault diagnosis method based on data driving and adopting current signals comprises the following steps:
1) obtaining a stator current signal, wherein the stator current signal has a length of D and comprises N samples;
2) carrying out noise signal reconstruction on the stator current signal to obtain a reconstructed signal;
3) subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
4) converting the time domain signal of the residue current to a frequency domain to obtain a frequency domain signal;
5) and inputting the frequency domain signal of the residual current into a pre-trained machine learning feature extraction and fault classification model to obtain the state of the motor.
Optionally, the step of converting the time domain signal of the residue current to the frequency domain in step 4) includes: the Hilbert envelope spectrum of the residual current is first calculated, and then the FFT spectrum of the Hilbert envelope spectrum is calculated.
Optionally, the noise signal reconstruction of the stator current signal in the step 2) to obtain the reconstructed signal adopts a signal noise reconstruction model built by an automatic encoder, the training signal noise reconstruction model includes an input layer, a hidden layer and an output layer which are connected in sequence, the input layer includes D +1 nodes, the last node is an offset node, and the hidden layer D includes a node with a maximum value of D +1h+1 nodes and the last node being a bias node, the output layer comprising D nodes; the input layer and the hidden layer form an encoder, the hidden layer and the output layer form a decoder, and the weight matrixes of the encoder and the decoder are respectively W1(Dh× D) and W2(D×Dh)，DhIs the dimension of the hidden layer; the inputs to the input layer are: x ═ X(1),x(2),…,x(N)Where X is the input stator current signal, X comprises X(1)～x(N)N samples in total, any ith sample x(i)Is expressed as x(i)＝[x1,x2,…,xD]Wherein x is1～xDRespectively sampling values of the stator current signals; the functional expression of the hidden layer is: h is(i)＝f(W1x(i)+b1) Wherein h is(i)For the ith hidden layer, f is the activation function, W1Is the weight vector, x, of the ith hidden layer(i)Is the ith sample, b1Is a bias vector; the functional expression of the output layer is:
Figure BDA0002557987770000021
wherein
Figure BDA0002557987770000022
For the (i) th output layer,
Figure BDA0002557987770000023
to activate a function, W2Is the weight vector of the ith output layer, h(i)Is the ith hidden layer, b2Is a bias vector.
Optionally, step 2) is preceded by the step of training a signal noise reconstruction model: collecting motor in a healthy state and stator current signals under N motor mechanical faults and forming a sample library, wherein the length of each sample is D, and each motor state comprises N samples, so that the total data amount is D × N (N + 1); training a noise reconstruction model of a training signal through motor samples of a health state in a sample library, wherein the training targets are an input signal X and a reconstruction signal thereof
Figure BDA0002557987770000024
The mean square error of (a) is minimal.
Optionally, the machine learning feature extraction and fault classification model in step 5) includes a feature extraction model and a fault classification model, where the feature extraction model is formed by stacking M automatic encoders to realize automatic feature extraction, an input of the automatic encoder at the bottommost layer is a frequency domain signal of a residual current, an output of the automatic encoder at the topmost layer is connected with a Softmax classifier to form the fault classification model, and the Softmax classifier is configured to output the classified motor states.
Optionally, step 5) is preceded by the step of training a machine learning feature extraction and fault classification model formed by stacking M automatic encoders: inputting a frequency domain signal of residual current obtained by extracting samples in a sample library into a machine learning feature extraction model and a feature extraction model in a fault classification model to extract features, then finely adjusting parameters of automatic encoders at each layer in the feature extraction model and a Softmax classifier in the fault classification model through a back propagation algorithm according to the motor state obtained by classifying the features by the Softmax classifier, and repeating the process until the iteration times are equal to a preset threshold value or the error is lower than the preset threshold value.
In addition, the invention also provides a data-driven motor mechanical fault diagnosis system adopting current signals, which comprises:
a signal acquisition program unit for acquiring a stator current signal;
the noise reconstruction program unit is used for reconstructing a noise signal of the stator current signal to obtain a reconstructed signal;
the noise elimination program unit is used for subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
converting the time domain signal of the residue current to a frequency domain to obtain a frequency domain signal; and the characteristic extraction and classification diagnosis program unit is used for inputting the frequency domain signal of the residual current into a pre-trained machine learning characteristic extraction and fault classification model to obtain the state of the motor.
In addition, the invention also provides a data-drive-based electromechanical fault diagnosis system adopting the current signal, which comprises a computer device, wherein the computer device is programmed or configured to execute the steps of the data-drive-based electromechanical fault diagnosis method adopting the current signal.
In addition, the invention also provides a data-drive-based electromechanical fault diagnosis system adopting the current signal, which comprises a computer device, wherein a memory of the computer device is stored with a computer program programmed or configured to execute the data-drive-based electromechanical fault diagnosis method adopting the current signal.
In addition, the present invention also provides a computer readable storage medium having stored therein a computer program programmed or configured to execute the data-drive-based motor mechanical failure diagnosis method using a current signal.
Compared with the prior art, the invention has the following advantages:
1. for mechanical faults, vibration signals are mostly adopted for diagnosis in the prior art. However, the vibration sensor is expensive and has limited application in small and medium-sized motors. And vibration sensor needs direct contact motor, has the installation to place the problem, and some equipment that has already put into production can't install vibration sensor additional. The method for diagnosing the motor fault by adopting the current signal can be used for diagnosing the mechanical fault. The method does not need to additionally install a vibration sensor and a current sensor, can acquire data through a current transformer which is already installed in a protection system or a control system, and is a non-invasive low-cost mode.
2. In the prior art, a model-based method is mostly adopted, and features are manually extracted through a fault mechanism analysis and signal processing means. Because the motor is a complex electromechanical-electro-magnetic coupling system, the modeling difficulty is high, and the requirement on professional knowledge is high. Meanwhile, the fault may generate a linkage effect, and the difficulty of building an accurate model is increased. And the high-order model has the defects of parameter sensitivity, low robustness and the like. The motor fault diagnosis method based on data driving provided by the invention can utilize motor operation data, mine information contained in the data, automatically extract fault characteristics which are difficult to model, has high robustness, reduces the requirement on professional knowledge and improves the diagnosis efficiency.
3. When the diagnosis is performed based on the data driving mode, if the original time domain data or the data converted into the frequency domain are directly used, the diagnosis accuracy rate is low, and the false alarm rate is high. Because the fault information of the mechanical fault reflected in the current can be modulated by the current fundamental wave, the signal-to-noise ratio is low, and the noise interference is large. The signal contains not only fundamental wave and its odd harmonics, but also some unknown components such as slot harmonics, saturated harmonics, environmental noise, etc. The fault characteristics are not obvious, and the fault diagnosis difficulty is high. The invention provides a noise elimination method based on data driving. The "noise signal" is defined as a component unrelated to the fault signature. The running data of the healthy motor is utilized to build a reconstruction model of the current noise signal, the noise component is automatically extracted, and the signal to noise ratio is greatly improved. The accuracy of subsequent fault feature extraction and classification algorithms based on data driving is improved, and the false alarm rate is reduced.
Drawings
FIG. 1 is a schematic diagram of a basic flow of a method according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of a method according to an embodiment of the present invention.
FIG. 3 is a schematic structural diagram of an automatic encoder according to an embodiment of the present invention.
Fig. 4 is a schematic structural diagram of a machine learning feature extraction and fault classification model in the embodiment of the present invention.
Detailed Description
The principle of the motor mechanical fault diagnosis method based on data driving and adopting current signals is as follows:
when the motor has typical mechanical faults (rotor broken bars, eccentric faults and bearing faults), the motor generates periodic torque pulsation during operation, so that the current is influenced. Each fault will occur at a specific frequency, i.e. fault signature frequency, on the current spectrum as follows:
rotor broken bar fault characteristic frequency fbrb1And fbrb2Comprises the following steps:
fbrb1＝(1±2k1s)·fs,k1＝1,2,3,… (1)
fbrb2＝[(k2/p)(1-s)±s]·fs,k2/p＝1,2,3,… (2)
characteristic frequency f of eccentric faultecc1Comprises the following steps:
fecc1＝[(k·R±nd)·(1-s)/Pp±v]·fs,k＝1,2,3,…v＝1,3,5,… (3)
ndwhen 0 is a static eccentricity fault, n d1,2,3, … are dynamic eccentric faults.
Bearing fault characteristic frequency:
characteristic frequency f at inner ring faultIComprises the following steps: f. ofI＝(Nb/2)·fs·[1+bd·cos(β)/dp]/Pp(4)
Outer ring fault characteristic frequency fOComprises the following steps: f. ofO＝(Nb/2)·fs·[1-bd·cos(β)/dp]/Pp(5)
Characteristic frequency f of rolling element failureBComprises the following steps: f. ofB＝dp·fs{1-[bd·cos(β)/dp]2}/2/bd/Pp(6)
In the above formulae (1) to (6), fsIs the power frequency, s is the slip, R is the number of rotor slots, PpIs the number of pole pairs, NbIs the number of rolling elements, bdIs the diameter of the rolling elements, dpIs the rolling element pitch diameter and β is the contact angle.
From the aforementioned expressions (1) to (6), the characteristic frequency of each fault can be simplified to the expression (7), where fcIs a general representation of the characteristic frequencies of each fault described above.
f＝fs±k·fc,k＝1,2,3,… (7)
As can be seen from equation (7), the diagnosis of the mechanical failure of the motor can be performed by detecting a specific frequency. However, the above formula is derived based on the theory of an ideal motor, and in actual conditions, the current has odd harmonics and other unknown components caused by slot harmonics, saturated harmonics, environmental noise and the like besides the fundamental wave. These components are not related to the fault characteristics, and the existence of the components can interfere the extraction and judgment of the fault characteristics, so that the current fundamental wave and other unknown components are defined as a noise signal in the motor mechanical fault diagnosis method based on data driving and adopting the current signal.
As shown in fig. 1, the method for diagnosing a mechanical fault of a motor using a current signal based on data driving of the present embodiment includes:
1) obtaining a stator current signal, wherein the stator current signal has a length of D and comprises N samples;
2) carrying out noise signal reconstruction on the stator current signal to obtain a reconstructed signal;
3) subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
4) converting the time domain signal of the residue current to a frequency domain to obtain a frequency domain signal;
5) and inputting the frequency domain signal of the residual current into a pre-trained machine learning feature extraction and fault classification model to obtain the state of the motor.
Referring to fig. 1, the method of the present embodiment uses a data-driven manner, and uses motor stator current data to complete automatic noise elimination and fault feature extraction and identification. During the period, the mechanical fault diagnosis of the motor can be realized only by simply processing the residual current after noise elimination. The method mainly comprises five steps of motor stator current signal acquisition, noise reconstruction, noise elimination, frequency domain conversion, feature extraction and classification diagnosis.
In the present embodiment, a signal that is not related to the fault characteristic is defined as a noise signal. Referring to fig. 2 and 3, in this embodiment, a signal noise reconstruction model built by an automatic encoder is used to reconstruct a noise signal of a stator current signal in step 2) to obtain a reconstructed signal, the training signal noise reconstruction model includes an input layer, a hidden layer and an output layer, which are sequentially connected, the input layer includes D +1 nodes, a last node is an offset node, and the hidden layer D is a nodeh+1 nodes and the last node being a bias node, the output layer comprising D nodes; the input layer and the hidden layer form an encoder, the hidden layer and the output layer form a decoder, and the weight matrixes of the encoder and the decoder are respectively W1(Dh× D) and W2(D×Dh)，DhIs the dimension of the hidden layer; the inputs to the input layer are: x ═ X(1),x(2),…,x(N)Where X is the input stator current signal, X comprises X(1)～x(N)N samples in total, any ith sample x(i)Is expressed as x(i)＝[x1,x2,…,xD](sample 1 is taken as an example and x is represented(1)＝[x1,x2,…,xD]) Wherein x is1～xDRespectively sampling values of the stator current signals; the functional expression of the hidden layer is: h is(i)＝f(W1x(i)+b1) Wherein h is(i)For the ith hidden layer, f is the activation function, W1Is the weight vector, x, of the ith hidden layer(i)Is the ith sample, b1Is a bias vector; the functional expression of the output layer is:
Figure BDA0002557987770000051
wherein
Figure BDA0002557987770000052
For the (i) th output layer,
Figure BDA0002557987770000053
to activate a function, W2Is the weight vector of the ith output layer, h(i)Is the ith hidden layer, b2Is a bias vector. Taking sample 1 as an example, the output is a reconstructed signal for the sample
Figure BDA0002557987770000054
Referring to fig. 2 and fig. 3, in this embodiment, step 2) further includes a step of training a signal noise reconstruction model: collecting motor in a healthy state and stator current signals under N motor mechanical faults and forming a sample library, wherein the length of each sample is D, and each motor state comprises N samples, so that the total data amount is D × N (N + 1); training a noise reconstruction model of a training signal through a sample of a motor in a healthy state in a sample library, wherein the training targets an input signal X and a reconstruction signal thereof
Figure BDA0002557987770000055
The mean square error (reconstruction error) of (d) is minimal, i.e.:
Figure BDA0002557987770000056
in the above formula, J is the mean square error, X is the input signal,
Figure BDA0002557987770000057
the reconstructed signal is X. When healthy motor current data are reconstructed, the data only contain components of fundamental waves and other unknown sources because fault characteristics are not contained in the data. That is, the reconstructed signal is a noise signal. The method can automatically and effectively extract the noise signal through the operation data without establishing a complex mathematical model. Inputting the motor current signal into the trained reconstruction model to obtain the correspondingThe reconstructed signal of (2). Current signal X of fault motorFInputting the signal into a trained model to obtain a reconstructed signal of the fault motor current under the model
Figure BDA0002557987770000058
Since the model is a reconstruction of the noise signal, the current signal of the faulty motor passes through the model, and its characteristic fault signal is not reconstructed.
Step 3) subtracting the reconstructed signal from the original stator current signal to obtain the residual current, and subtracting the original signal and the reconstructed signal of the motor (health and fault) to eliminate noise to obtain corresponding residual current signals R and RFWhere R represents the residual current drawn by a healthy motor, RFRepresenting the residual current drawn by the failed motor.
Referring to fig. 2, the step of converting the time domain signal of the residual current to the frequency domain in step 4) to obtain a frequency domain signal includes: the Hilbert envelope spectrum of the residual current is first calculated, and then the FFT (fourier) spectrum of the Hilbert envelope spectrum is calculated.
Referring to fig. 4, the machine learning feature extraction and fault classification model in step 5) includes a feature extraction model and a fault classification model, where the feature extraction model is formed by stacking M automatic encoders to realize automatic feature extraction, where the input of the automatic encoder at the bottom layer is a frequency domain signal of residual current, and the output of the automatic encoder at the top layer is connected with a Softmax classifier to form the fault classification model, and the Softmax classifier is used to output the classified motor states.
Referring to fig. 4, step 5) is preceded by the step of training a machine learning feature extraction and fault classification model formed by stacking M automatic encoders: inputting a frequency domain signal of residual current obtained by extracting samples in a sample library into a machine learning feature extraction model and a feature extraction model in a fault classification model to extract features, then finely adjusting parameters of automatic encoders at each layer in the feature extraction model and a Softmax classifier in the fault classification model through a back propagation algorithm according to the motor state obtained by classifying the features by the Softmax classifier, and repeating the process until the iteration times are equal to a preset threshold value or the error is lower than the preset threshold value.
As shown in fig. 4, in this embodiment, a feature extraction and classification model of the residue current is built by stacking an automatic encoder algorithm. The stacked automatic encoder is a deep learning algorithm and is formed by stacking a plurality of automatic encoders. The first layer of automatic encoder inputs are residue currents R and RFEnvelope spectrum X ofPIs output as its reconstructed signal
Figure BDA0002557987770000061
Obtaining a hidden layer H(1). The hidden layer H(1)And the signal is used as the input of the automatic encoder of the next layer, and the training is carried out layer by layer. The layer-by-layer pre-training mode can improve the depth of the algorithm and avoid the situation that the gradient disappears due to the multi-layer structure. The resulting hidden layers are respectively H(1)，H(2)，…，H(M)And M is the number of layers of the stacked autoencoder. And adding a Softmax classifier at the top layer of the trained model, and finely adjusting parameters of each layer and the classifier through a back propagation algorithm to finish fault classification training. The diagnosis of the motor state can be realized.
In summary, the data-driven motor mechanical fault diagnosis method using the current signal according to the embodiment can detect a mechanical fault by using the current signal, does not need an additional sensor, and is a non-invasive low-cost method. Based on a data-driven mode, the 'noise' characteristic which is irrelevant to the fault is automatically extracted to eliminate the power fundamental wave and odd harmonics thereof as well as other harmonics which are not easy to model in the original signal. The automatic noise elimination processing based on data driving can greatly improve the signal to noise ratio, avoid a complex fault modeling process and reduce the requirement on professional knowledge. And after the denoised signal is subjected to simple signal processing, automatically extracting fault characteristics and finishing the classification and identification of the faults by combining a deep learning algorithm. The invention fully utilizes the advantages of the data driving method, improves the diagnosis accuracy, reduces the false alarm rate and the requirement on professional knowledge, and improves the diagnosis efficiency.
In addition, the present embodiment further provides a data-driven electromechanical fault diagnosis system using a current signal, including:
a signal acquisition program unit for acquiring a stator current signal;
the noise reconstruction program unit is used for reconstructing a noise signal of the stator current signal to obtain a reconstructed signal;
the noise elimination program unit is used for subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
converting the time domain signal of the residue current to a frequency domain to obtain a frequency domain signal;
and the characteristic extraction and classification diagnosis program unit is used for inputting the frequency domain signal of the residual current into a pre-trained machine learning characteristic extraction and fault classification model to obtain the state of the motor.
In addition, the present embodiment also provides a data-driven electromechanical fault diagnosis system using a current signal, which includes a computer device programmed or configured to execute the steps of the aforementioned data-driven electromechanical fault diagnosis method using a current signal.
In addition, the present embodiment also provides a data-driven electromechanical fault diagnosis system using a current signal, which includes a computer device, where a memory of the computer device stores a computer program programmed or configured to execute the foregoing data-driven electromechanical fault diagnosis method using a current signal.
The present embodiment also provides a computer-readable storage medium having stored therein a computer program programmed or configured to execute the aforementioned data-drive-based motor mechanical failure diagnosis method using a current signal.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may occur to those skilled in the art without departing from the principle of the invention, and are considered to be within the scope of the invention.

Claims (10)

1. A motor mechanical fault diagnosis method based on data driving and adopting current signals is characterized by comprising the following steps:
1) obtaining a stator current signal, wherein the stator current signal has a length of D and comprises N samples;
2) carrying out noise signal reconstruction on the stator current signal to obtain a reconstructed signal;
3) subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
4) converting the time domain signal of the residue current to a frequency domain to obtain a frequency domain signal;
5) and inputting the frequency domain signal of the residual current into a pre-trained machine learning feature extraction and fault classification model to obtain the state of the motor.
2. The method for diagnosing mechanical failure of motor using current signal based on data driving of claim 1, wherein the step of converting the time domain signal of the residual current to the frequency domain in step 4) comprises: the Hilbert envelope spectrum of the residual current is first calculated, and then the FFT spectrum of the Hilbert envelope spectrum is calculated.
3. The data-driven motor mechanical fault diagnosis method based on current signals, according to claim 1, is characterized in that noise signal reconstruction is performed on the stator current signals in the step 2) to obtain reconstructed signals, a signal noise reconstruction model built by an automatic encoder is adopted, the training signal noise reconstruction model comprises an input layer, a hidden layer and an output layer which are sequentially connected, the input layer comprises D +1 nodes, the last node is an offset node, and the hidden layer D is a nodeh+1 nodes and the last node being a bias node, the output layer comprising D nodes; the input layer and the hidden layer form an encoder, the hidden layer and the output layer form a decoder, and the weight matrixes of the encoder and the decoder are respectively W1(Dh× D) and W2(D×Dh)，DhIs the dimension of the hidden layer; the inputs to the input layer are: x ═ X(1),x(2),…,x(N)Where X is the input stator current signal, X comprises X(1)～x(N)N samples in total, any ith sample x(i)Is expressed as x(i)＝[x1,x2,…,xD]Wherein x is1～xDRespectively sampling values of the stator current signals; the functional expression of the hidden layer is: h is(i)＝f(W1x(i)+b1) Wherein h is(i)For the ith hidden layer, f is the activation function, W1Is the weight vector, x, of the ith hidden layer(i)Is the ith sample, b1Is a bias vector; the functional expression of the output layer is:
Figure FDA0002557987760000011
wherein
Figure FDA0002557987760000012
For the (i) th output layer,
Figure FDA0002557987760000013
to activate a function, W2Is the weight vector of the ith output layer, h(i)Is the ith hidden layer, b2Is a bias vector.
4. The method for diagnosing mechanical faults of a motor based on data driving and adopting a current signal according to claim 3, wherein the step 2) is preceded by a step of training a signal noise reconstruction model: collecting motor in a healthy state and stator current signals under N motor mechanical faults and forming a sample library, wherein the length of each sample is D, and each motor state comprises N samples, so that the total data amount is D × N (N + 1); training a noise reconstruction model of a training signal through a sample of a motor in a healthy state in a sample library, wherein the training targets an input signal X and a reconstruction signal thereof
Figure FDA0002557987760000014
The mean square error of (a) is minimal.
5. The data-driven electromechanical fault diagnosis method based on current signals of claim 1, wherein the machine learning feature extraction and fault classification model in step 5) comprises a feature extraction model and a fault classification model, wherein the feature extraction model is formed by stacking M automatic encoders to realize automatic feature extraction, the input of the automatic encoder at the bottom layer is a frequency domain signal of residual current, and the output end of the automatic encoder at the top layer is connected with a Softmax classifier to form the fault classification model, and the Softmax classifier is used for outputting the classified motor state.
6. The method for diagnosing mechanical faults of motor based on data driving and adopting current signals as claimed in claim 5, wherein the step 5) is preceded by a step of training a machine learning feature extraction and fault classification model formed by stacking M automatic encoders: inputting a frequency domain signal of residual current obtained by extracting samples in a sample library into a machine learning feature extraction model and a feature extraction model in a fault classification model to extract features, then finely adjusting parameters of automatic encoders at each layer in the feature extraction model and a Softmax classifier in the fault classification model through a back propagation algorithm according to the motor state obtained by classifying the features by the Softmax classifier, and repeating the process until the iteration times are equal to a preset threshold value or the error is lower than the preset threshold value.
7. A data-driven electromechanical fault diagnosis system using a current signal, comprising:
a signal acquisition program unit for acquiring a stator current signal;
the noise reconstruction program unit is used for reconstructing a noise signal of the stator current signal to obtain a reconstructed signal;
the noise elimination program unit is used for subtracting the reconstructed signal from the original stator current signal to obtain a residual current;
the frequency domain conversion program unit is used for converting the residual current signal of the time domain to a frequency domain to obtain a frequency domain signal;
and the characteristic extraction and classification diagnosis program unit is used for inputting the frequency domain signal of the residual current into a pre-trained machine learning characteristic extraction and fault classification model to obtain the state of the motor.
8. A data-driven electromechanical fault diagnosis system using current signals, comprising a computer device, wherein the computer device is programmed or configured to execute the steps of the data-driven electromechanical fault diagnosis method using current signals according to any one of claims 1 to 6.
9. A data-driven motor mechanical fault diagnosis system using current signals, comprising a computer device, wherein a computer program programmed or configured to execute the data-driven motor mechanical fault diagnosis method using current signals according to any one of claims 1 to 6 is stored in a memory of the computer device.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored therein a computer program programmed or configured to execute the data-drive-based electromechanical fault diagnosis method using a current signal according to any one of claims 1 to 6.
CN202010597800.1A 2020-06-28 2020-06-28 Motor mechanical fault diagnosis method adopting current signals based on data driving Active CN111680665B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010597800.1A CN111680665B (en) 2020-06-28 2020-06-28 Motor mechanical fault diagnosis method adopting current signals based on data driving

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010597800.1A CN111680665B (en) 2020-06-28 2020-06-28 Motor mechanical fault diagnosis method adopting current signals based on data driving

Publications (2)

Publication Number Publication Date
CN111680665A true CN111680665A (en) 2020-09-18
CN111680665B CN111680665B (en) 2023-06-20

Family

ID=72456808

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010597800.1A Active CN111680665B (en) 2020-06-28 2020-06-28 Motor mechanical fault diagnosis method adopting current signals based on data driving

Country Status (1)

Country Link
CN (1) CN111680665B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112347944A (en) * 2020-11-10 2021-02-09 南京智谷人工智能研究院有限公司 Mechanical transmission fault detection method based on zero-value proportional spectrum characteristics
CN112504678A (en) * 2020-11-12 2021-03-16 重庆科技学院 Motor bearing fault diagnosis method based on knowledge distillation
CN112541524A (en) * 2020-11-18 2021-03-23 湖南大学 BP-Adaboost multi-source information motor fault diagnosis method based on attention mechanism improvement

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0894499A (en) * 1994-09-20 1996-04-12 Ishikawajima Harima Heavy Ind Co Ltd Failure diagnostic system for rotating machine
CN106295023A (en) * 2016-08-15 2017-01-04 南京航空航天大学 A kind of diagnostic method of asynchronous machine rotor combined failure
CN107957551A (en) * 2017-12-12 2018-04-24 南京信息工程大学 Stacking noise reduction own coding Method of Motor Fault Diagnosis based on vibration and current signal
CN108414923A (en) * 2018-02-05 2018-08-17 武汉大学 A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization
CN108731945A (en) * 2018-08-02 2018-11-02 南昌航空大学 A kind of extracting method of aircraft engine rotor system fault-signal characteristic information
CN108761332A (en) * 2018-05-08 2018-11-06 郑州轻工业学院 A kind of set empirical mode decomposition current diagnostic method of motor broken bar fault
CN108919116A (en) * 2018-05-16 2018-11-30 上海海事大学 Ocean current generator imbalance stator current method for diagnosing faults based on MCCKAF-FFT-Softmax
CN109061474A (en) * 2018-10-15 2018-12-21 株洲中车时代电气股份有限公司 A kind of motor bearings trouble-shooter
CN109145886A (en) * 2018-10-12 2019-01-04 西安交通大学 A kind of asynchronous machine method for diagnosing faults of Multi-source Information Fusion
CN109211546A (en) * 2018-08-28 2019-01-15 电子科技大学 Rotary machinery fault diagnosis method based on noise reduction autocoder and incremental learning
CN109774740A (en) * 2019-02-03 2019-05-21 湖南工业大学 A kind of wheel tread damage fault diagnostic method based on deep learning
CN110110768A (en) * 2019-04-24 2019-08-09 西安电子科技大学 Fault Diagnosis of Roller Bearings based on Concurrent Feature study and multi-categorizer
CN111094927A (en) * 2017-09-26 2020-05-01 舍弗勒技术股份两合公司 Bearing fault diagnosis method and device, readable storage medium and electronic equipment

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0894499A (en) * 1994-09-20 1996-04-12 Ishikawajima Harima Heavy Ind Co Ltd Failure diagnostic system for rotating machine
CN106295023A (en) * 2016-08-15 2017-01-04 南京航空航天大学 A kind of diagnostic method of asynchronous machine rotor combined failure
CN111094927A (en) * 2017-09-26 2020-05-01 舍弗勒技术股份两合公司 Bearing fault diagnosis method and device, readable storage medium and electronic equipment
CN107957551A (en) * 2017-12-12 2018-04-24 南京信息工程大学 Stacking noise reduction own coding Method of Motor Fault Diagnosis based on vibration and current signal
CN108414923A (en) * 2018-02-05 2018-08-17 武汉大学 A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization
CN108761332A (en) * 2018-05-08 2018-11-06 郑州轻工业学院 A kind of set empirical mode decomposition current diagnostic method of motor broken bar fault
CN108919116A (en) * 2018-05-16 2018-11-30 上海海事大学 Ocean current generator imbalance stator current method for diagnosing faults based on MCCKAF-FFT-Softmax
CN108731945A (en) * 2018-08-02 2018-11-02 南昌航空大学 A kind of extracting method of aircraft engine rotor system fault-signal characteristic information
CN109211546A (en) * 2018-08-28 2019-01-15 电子科技大学 Rotary machinery fault diagnosis method based on noise reduction autocoder and incremental learning
CN109145886A (en) * 2018-10-12 2019-01-04 西安交通大学 A kind of asynchronous machine method for diagnosing faults of Multi-source Information Fusion
CN109061474A (en) * 2018-10-15 2018-12-21 株洲中车时代电气股份有限公司 A kind of motor bearings trouble-shooter
CN109774740A (en) * 2019-02-03 2019-05-21 湖南工业大学 A kind of wheel tread damage fault diagnostic method based on deep learning
CN110110768A (en) * 2019-04-24 2019-08-09 西安电子科技大学 Fault Diagnosis of Roller Bearings based on Concurrent Feature study and multi-categorizer

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
卢艳辉,尹泽勇: "基于小波包分析方法的航空发动机滚动轴承故障诊断" *
王辉斌;吴长利;邹桂丽;卢文秀;褚福磊;陈贞石;朱俊;任继顺;陈伟;: "水轮机故障声学诊断技术研究及其应用探讨" *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112347944A (en) * 2020-11-10 2021-02-09 南京智谷人工智能研究院有限公司 Mechanical transmission fault detection method based on zero-value proportional spectrum characteristics
WO2022100046A1 (en) * 2020-11-10 2022-05-19 南京智谷人工智能研究院有限公司 Mechanical transmission fault detection method based on zero-value ratio spectrum feature
CN112347944B (en) * 2020-11-10 2023-10-13 南京大学 Mechanical transmission fault detection method based on zero value proportion frequency spectrum characteristics
CN112504678A (en) * 2020-11-12 2021-03-16 重庆科技学院 Motor bearing fault diagnosis method based on knowledge distillation
CN112541524A (en) * 2020-11-18 2021-03-23 湖南大学 BP-Adaboost multi-source information motor fault diagnosis method based on attention mechanism improvement
CN112541524B (en) * 2020-11-18 2024-04-02 湖南大学 BP-Adaboost multisource information motor fault diagnosis method based on attention mechanism improvement

Also Published As

Publication number Publication date
CN111680665B (en) 2023-06-20

Similar Documents

Publication Publication Date Title
CN111680665B (en) Motor mechanical fault diagnosis method adopting current signals based on data driving
Jiang et al. A GAN-based anomaly detection approach for imbalanced industrial time series
CN106096562B (en) Gearbox of wind turbine method for diagnosing faults based on vibration signal blind sources separation and sparse component analysis
Heydarzadeh et al. Gear fault diagnosis using discrete wavelet transform and deep neural networks
CN105758644A (en) Rolling bearing fault diagnosis method based on variation mode decomposition and permutation entropy
CN112052796A (en) Permanent magnet synchronous motor fault diagnosis method based on deep learning
Zeng et al. SOSO boosting of the K-SVD denoising algorithm for enhancing fault-induced impulse responses of rolling element bearings
Zhang et al. A GOA-MSVM based strategy to achieve high fault identification accuracy for rotating machinery under different load conditions
CN113486868B (en) Motor fault diagnosis method and system
CN113391207A (en) Motor fault detection method, medium and system
CN109655266A (en) A kind of Wind turbines Method for Bearing Fault Diagnosis based on AVMD and spectral coherence analysis
CN114263621B (en) Test method and system for centrifugal pump cavitation fault diagnosis simulation
CN111678691A (en) Gear fault detection method based on improved sparse decomposition algorithm
CN114705432A (en) Method and system for evaluating health state of explosion-proof motor bearing
CN113758709A (en) Rolling bearing fault diagnosis method and system combining edge calculation and deep learning
CN113985276B (en) Fault diagnosis method and device for wind generating set
Zhao et al. TQWT-based multi-scale dictionary learning for rotating machinery fault diagnosis
CN114912480A (en) Fault diagnosis method and system based on depth residual variation self-encoder
CN115326396A (en) Bearing fault diagnosis method and device
CN106772032A (en) A kind of fault signature extracting method of turbine-generator units
CN112595515A (en) Power shafting bearing fault detection method and system
Wang et al. An integrative computational method for gearbox diagnosis
Zheng et al. Wavelet packet decomposition and neural network based fault diagnosis for elevator excessive vibration
CN117970105B (en) Early fault diagnosis method and system for motor bearing based on signal fusion
Hong et al. Adaptive fusion based on physics-constrained dictionary learning for fault diagnosis 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