CN114358077A - Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system - Google Patents

Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system Download PDF

Info

Publication number
CN114358077A
CN114358077A CN202111674484.4A CN202111674484A CN114358077A CN 114358077 A CN114358077 A CN 114358077A CN 202111674484 A CN202111674484 A CN 202111674484A CN 114358077 A CN114358077 A CN 114358077A
Authority
CN
China
Prior art keywords
fault diagnosis
fault
classifier
working conditions
permanent magnet
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
Application number
CN202111674484.4A
Other languages
Chinese (zh)
Inventor
谢金平
张晓飞
黄凤琴
宋殿义
唐瑶
龙卓
唐镜博
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 CN202111674484.4A priority Critical patent/CN114358077A/en
Publication of CN114358077A publication Critical patent/CN114358077A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Control Of Electric Motors In General (AREA)

Abstract

The invention discloses a method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor, a fault diagnosis method and a system, which are suitable for multiple working conditions and variable working conditions, wherein, the method takes the motor surface magnetic flux leakage signal as the original signal of fault diagnosis and expands the signal into a symmetrical dot matrix image, further extracting the multi-class local high-level features of the symmetric lattice image, fusing, and finally constructing a classifier by using the fused features, particularly preferably constructing a fuzzy multi-model classifier for fault diagnosis, so that the precision of the fault diagnosis result can be effectively improved, the symmetric dot matrix image has strong invariant adaptability to the working conditions, the validity of the characteristics is improved through the fusion of the multi-type local high-level characteristics of the image, the accuracy of the model is further guaranteed, the demagnetization fault is identified with high precision, and the problem of demagnetization fault diagnosis of the permanent magnet synchronous motor under multiple working conditions and variable working conditions is effectively solved.

Description

Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system
Technical Field
The invention relates to fault diagnosis of a permanent magnet synchronous motor, in particular to a method for constructing a demagnetization fault diagnosis model of the permanent magnet synchronous motor, a fault diagnosis method and a system, and is particularly suitable for demagnetization fault diagnosis of the permanent magnet synchronous motor under multiple working conditions and variable working conditions.
Background
The permanent magnet synchronous motor adopts the permanent magnet to generate a motor magnetic field, has simple structure and high efficiency and control precision, and is widely applied to the fields of numerical control machines, electric automobiles, aerospace, wind power generation and the like. The magnetic steel sheet of the permanent magnet synchronous motor is mostly made of neodymium iron boron permanent magnet material, and the Curie temperature of the permanent magnet synchronous motor is low. Therefore, overload of the motor, damage of the heat dissipation system, and the like may cause magnetic loss of the permanent magnet. The armature reaction magnetic field and the permanent magnet magnetic field are opposite in direction, so that the permanent magnet has an essential demagnetization effect, and the permanent magnet is extremely easy to demagnetize under the conditions of large load working condition and stator winding short circuit fault. Finally, manufacturing and installation defects, variable working conditions and environments can also accelerate motor damage and aging, thereby causing motor failure. The demagnetization fault can aggravate torque ripple and motor loss, seriously reduce the equipment performance, and cause property loss and casualties in serious cases.
The permanent magnet synchronous motor can change the operation condition along with the external wind speed or road conditions and the like in the application fields of wind turbines, electric automobiles and the like. Most of the existing diagnosis methods realize fault diagnosis by comparing different state signals among motors. However, when the motor is in a variable condition, the state signal is also changed, which makes fault diagnosis more difficult. The state signals for demagnetization fault diagnosis at present are motor stator current, voltage, torque, magnetic flux and the like. The change of the state signals is caused by the indirect influence of the demagnetization fault of the permanent magnet, and is easily interfered by other faults. Compared with other state signals, the occurrence of demagnetization faults can directly cause the magnetic field of the motor to change. The cogging magnetic flux and the air gap magnetic flux signals realize demagnetization fault diagnosis of the permanent magnet motor. However, the acquisition of these magnetic flux signals requires disassembly of the motor, which is costly and prone to human interference.
The motor status signal not only contains fault characteristics, but also is mixed with redundant signals and other interference signals. Therefore, the extraction of the fault characteristics is important in the motor fault diagnosis. The existing theory and method focus on the detail feature extraction of time domain, frequency domain or wavelet domain. The main feature extraction methods include short-time Fourier transform, continuous wavelet transform and the like. These methods extract only the detail features in one dimension, and cannot obtain the two-dimensional features in the state signal. And the state signal is expanded to a two-dimensional image, so that fault diagnosis is facilitated. The local features of the image and their descriptors are local domain compact vector representations. Common local features include angular features and speckle features, among others. The characterization methods used for classification are mainly heat maps, shapes, etc. However, since the local feature numbers of different images are not fixed, the extracted feature set is unbalanced, and it is difficult to meet the requirements of most classification algorithms on input features.
In summary, the existing diagnostic methods have the following technical problems:
1) the motor state diagnosis signal is easy to be interfered, and the acquisition cost is high;
2) the one-dimensional detail features cannot meet the requirements of most classification algorithms, are not beneficial to fault diagnosis, and particularly are influenced by variable working conditions, signals of the same motor are variable, so that the accuracy requirement of fault diagnosis cannot be met by single feature extraction.
In addition, most of the existing diagnostic methods are effective only under fixed working conditions or limited conditions, and the application of the diagnostic methods in the variable working condition industrial field is limited. Therefore, a need exists for a method and a system for diagnosing demagnetization faults of a permanent magnet synchronous motor, which are suitable for multiple working conditions and variable working conditions.
Disclosure of Invention
The invention aims to solve at least part of technical problems in the prior art, and further provides a method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor, and a fault diagnosis method and system. According to the method, the magnetic flux leakage signals of the fault motor under different working conditions are selected and extracted to serve as original signals, the one-dimensional time domain signals are converted into two-dimensional images, the fault signal characteristics are enriched, and the problem that accurate fault diagnosis cannot be met by one-dimensional detail characteristics and single characteristic extraction is effectively solved.
On one hand, the invention provides a method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor, which comprises the following steps: acquiring signals, performing two-dimensional image conversion on the acquired signals, extracting and fusing the characteristics of the two-dimensional images, and constructing a fault diagnosis classifier by utilizing the fused characteristics, which comprises the following steps:
step 1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
step 2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
and step 3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
and 4, step 4: taking the fused features as the input of a fault diagnosis classifier, taking the fault type as the output of the fault diagnosis classifier, and performing network training by using the fused features of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier;
the classifier is applied to demagnetization fault diagnosis of the permanent magnet synchronous motor.
The invention selects the leakage magnetic signal as the original signal of the fault diagnosis, considers that the leakage magnetic signal can directly reflect the demagnetization fault, compared with other voltage and current signals which indirectly reflect the demagnetization fault, the effectiveness of the fault characteristic can be improved, the interference by other faults is less, and secondly, the leakage magnetic signal outside the motor shell is generally measured, so the motor does not need to be disassembled, and the artificial interference can not be introduced.
Moreover, the magnetic leakage signal is converted into the symmetrical dot matrix image, the time domain one-dimensional signal can be converted into a polar coordinate by considering the symmetrical dot matrix image, the fault characteristics are reflected by the size of an image arm, the curvature and the like, two-dimensional effective information hidden in the one-dimensional signal can be shown, the diagnosis effect is improved, the symmetrical dot matrix image has multiple working conditions, and the stability and the anti-interference performance under the variable working conditions are the key for realizing the diagnosis under the variable working conditions.
In a second aspect, the invention provides a method for diagnosing demagnetization fault of a permanent magnet synchronous motor based on the model building method, which includes: the method comprises the following steps of constructing a fault diagnosis classifier and diagnosing faults based on the fault diagnosis classifier, wherein the fault diagnosis classifier specifically comprises the following steps:
s1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
s2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
s3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
s4: using the fused features as the input of a fault diagnosis classifier, using the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of different working conditions/variable working conditions of various faults in S3 to obtain the fault diagnosis classifier;
s5: and acquiring a magnetic leakage signal of the permanent magnet synchronous motor to be diagnosed, performing feature extraction and feature fusion on the fault motor under the current working condition/variable working condition according to the steps S2-S3, and inputting the fused features into a fault diagnosis classifier to obtain a fault diagnosis result.
Further optionally, in step 2, an accelerated robust speckle feature method is respectively adopted to acquire accelerated robust speckle features of the symmetric dot matrix image, an accelerated segment test corner feature method is adopted to acquire accelerated segment test corner features, and the accelerated robust speckle features and the accelerated segment test corner features of the symmetric dot matrix image under the same type of faults and under the same working condition/different working conditions/variable working conditions are fused.
The research of the invention finds that when the accelerated steady spot characteristic and the accelerated sectional test corner characteristic are selected for fusion, the diagnosis effect is optimal, the two characteristics have complementary effects, when a single characteristic is selected, one or two fault diagnosis effects are poor and staggered, and the two characteristics are fused to obtain better diagnosis effect.
Further optionally, in step 2, the process of expanding the leakage magnetic signal into a symmetric lattice image is as follows: converting signal points of the magnetic leakage signal into a polar coordinate space, and constructing a symmetrical dot matrix image based on coordinates of the polar coordinate space of all the signal points;
the conversion formula for converting the signal point of the leakage magnetic signal into the polar coordinate space is as follows:
Figure BDA0003451138730000031
Figure BDA0003451138730000032
Figure BDA0003451138730000033
in the formula, sjRepresents a certain signal point in the leakage signal s, and r (j), α (j), β (j) represent the signal point sjCoordinates in polar coordinate space, where r (j) is the signal point sjThe radius of the polar coordinate is mapped, and alpha (j) and beta (j) are respectively an anticlockwise rotating angle and a clockwise rotating angle of a mirror symmetry plane; sminAnd smaxRespectively the minimum value and the maximum value of the amplitude in the leakage magnetic signal s, wherein the counterclockwise rotation angle alpha (j) and the clockwise rotation angle beta (j) of the mirror symmetry plane are respectively equal to sjSignal amplitude s after the phase difference kj+kMapping, k is any positive integer, θwAngle of w-th plane of symmetry, θw＝360m/n，m＝0,1,2…n-1, n is the number of mirror symmetry planes, λ is the gain angle, λ ≦ θw。
Further optionally, the fault diagnosis classifier is a fuzzy multi-model classifier, and a training process of the model multi-model classifier is as follows:
inputting the fusion characteristics of the data samples into a fuzzy network; updating the fuzzy network according to the following process until all data samples are trained; each data sample is classified and determined according to a maximum confidence coefficient winning principle, wherein the leakage magnetic signal of a fault motor with a known fault type is used as the data sample, and the fusion characteristics of the data samples are obtained according to the methods of the steps 2 to 3;
wherein, firstly, the ith fault ithOf the τ -th data sample τthFusion feature g ofτCarrying out normalization;
Figure BDA0003451138730000041
wherein the content of the first and second substances,
Figure BDA0003451138730000042
represents the normalized feature, | gτ| is gτNorm of the feature;
then, the ith fault i is calculated according to the following formulathOf the data samples τthCorresponding mean value
Figure BDA0003451138730000043
And average scalar
Figure BDA0003451138730000044
Based on the mean value
Figure BDA0003451138730000045
And average scalar
Figure BDA0003451138730000046
Calculating the Euclidean distance
Figure BDA0003451138730000047
Figure BDA0003451138730000048
Figure BDA0003451138730000049
Figure BDA00034511387300000410
Wherein the content of the first and second substances,
Figure BDA00034511387300000411
judging whether a new fuzzy rule and a new data cloud are generated or not according to the following condition 1; if condition 1 is satisfied, then the feature is
Figure BDA00034511387300000412
Forming a data cloud by a new fuzzy rule around; otherwise, judging whether the condition 2 is met; if condition 2 is satisfied, the feature is utilized
Figure BDA00034511387300000413
Updating distance features
Figure BDA00034511387300000414
Fuzzy rules of the most recent data cloud; otherwise, in the feature
Figure BDA00034511387300000415
Forming a data cloud by a new fuzzy rule around;
condition 1:
Figure BDA00034511387300000416
in the formula, HiFor fault i of type ithThe number of data samples in the data cloud of (1),
Figure BDA00034511387300000417
characterizing class i faults ithThe Euclidean distance corresponding to each data sample in the data cloud;
condition 2:
Figure BDA00034511387300000418
wherein the content of the first and second substances,
Figure BDA00034511387300000419
for the distance characteristic sought
Figure BDA00034511387300000420
The feature points of the closest cloud are,
Figure BDA00034511387300000421
is a characteristic point
Figure BDA00034511387300000422
The state parameter of the data cloud.
Common intelligent classification models include support vector machines, decision trees, neural networks, and the like. The classification model can obtain better diagnosis effect, and aiming at the fusion characteristics of the invention, the existing network can be adopted to construct the classifier, the construction process is not optimized, and the prior art can be referred.
However, the present invention further contemplates that the internal parameters and hyper-parameters of the above network models strongly influence their performance. And the hyper-parameters of the classification model are different for different classification objects. Therefore, for different fault diagnoses, a professional and complex hyper-parameter optimization adjustment process is often required, such as the maximum number of partitions of a decision tree or the frame constraints of a support vector machine. Although neural networks exist as high-level networks with good mobility, most of them have a deep network layer, a long training process, and poor transparency and interpretability. In view of the shortcomings of the network, in order to further improve the performance of the fault diagnosis classifier, the invention further considers that the fuzzy multi-model classifier constructed based on the fuzzy network is applicable. Wherein the fuzzy multi-classifier is a non-parametric classifier driven entirely by data. The diagnosis process is completely independent without advanced assumption, all the element parameters in the model come from data and are automatically updated recursively, a complex professional super-parameter adjustment optimization process is not needed, and the model is high in transparency and strong in interpretability. Therefore, the technical combination of the symmetrical dot matrix image and the fuzzy multi-classifier is optimized, the symmetrical dot matrix can realize the stability of the fault characteristics under the changeable working condition, and the fusion characteristics of the symmetrical dot matrix and the fuzzy multi-classifier provide effective fault characteristics for the multi-fuzzy classifier to further improve the diagnosis effect.
Further optionally, when the classification is based on the maximum confidence winning principle, the calculation formula of the confidence score is as follows:
Figure BDA0003451138730000051
Figure BDA0003451138730000052
wherein, γiFor confidence, i is 1,2 … q, q is the number of sub-classifiers corresponding to the number of fault types, Label is the fault type,
Figure BDA0003451138730000053
cloud center point features representing a type i fault.
From the above formula, gτ:
Figure BDA0003451138730000054
Indicates g if it is to be diagnosedτIf the sample is at the ith sub-classifier (i.e. the ith fault), the "confidence score" of the class is calculated, and finally the comparison g is carried outτTaking the maximum confidence score of all the sub-classifiers as the gτIt is determined to belong to the class of fault.
In a third aspect, the present invention provides a system based on the above fault diagnosis model building method or fault diagnosis method, including:
the signal acquisition module is used for acquiring/obtaining magnetic leakage signals of the fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
the image conversion module is used for expanding the magnetic leakage signal into a symmetrical dot matrix image;
the characteristic extraction module is used for extracting local high-level characteristics of the symmetrical dot matrix image by adopting different characteristic point detection methods;
the characteristic fusion module is used for fusing local high-level characteristics of the symmetrical dot matrix images under the same type of faults and under the same working condition/different working conditions/variable working conditions;
the fault diagnosis classifier building module is used for taking the fused features as the input of the fault diagnosis classifier, taking the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier;
and the diagnosis module is used for diagnosing the fault by utilizing the fault diagnosis classifier and the fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result.
In a fourth aspect, the present invention provides an electronic terminal, comprising a processor and a memory, which are connected to each other, wherein the processor is programmed or configured to execute the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor or the steps of the method for diagnosing the demagnetization fault of the permanent magnet synchronous motor.
In a fifth aspect, the present invention provides a system based on the above fault diagnosis model building method or fault diagnosis method, including: the system comprises a fault diagnosis upper computer and an experiment platform with a plurality of fault motors, wherein the experiment platform simulates the operation of permanent magnet synchronous motor equipment;
if the experimental platform is provided with a processor and a memory, the processor is programmed or configured to execute the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor, and the fault diagnosis classification model is uploaded to a fault diagnosis upper computer so that the fault diagnosis upper computer can carry out fault diagnosis by using the fault diagnosis classifier and fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result;
or, if be equipped with treater and memory in the fault diagnosis host computer, the experiment platform will simulate the magnetic leakage signal during PMSM class equipment operation and upload to the fault diagnosis host computer, the treater in the fault diagnosis host computer is programmed or the configuration in order to carry out PMSM demagnetization fault diagnosis method.
In a sixth aspect, the present invention provides a readable storage medium, wherein the readable storage medium stores a computer program programmed or configured to execute the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor or the method for diagnosing the demagnetization fault of the permanent magnet synchronous motor.
Advantageous effects
1. The invention directly converts the motor state signal into the two-dimensional symmetrical dot matrix image, presents more effective fault high-dimensional characteristics and avoids the need of signal processing with strong professional knowledge.
2. The invention provides a local high-rise fusion feature extraction method by adopting innovativeness, and effectively solves the problem of low diagnosis effect caused by variable signals under multiple working conditions and variable working conditions.
3. The invention provides a method for diagnosing demagnetization faults by adopting a fuzzy multi-model classifier, the diagnosis process is completely independent, model parameters are completely derived from data, and a higher diagnosis effect can be obtained on the basis of greatly simplifying the diagnosis process.
Drawings
Fig. 1 is a schematic diagram of a demagnetization fault experiment platform of a permanent magnet synchronous motor in the embodiment of the invention.
Fig. 2 is a schematic diagram of a demagnetization fault diagnosis method of a permanent magnet synchronous motor suitable for multiple working conditions and variable working conditions according to the method of the embodiment of the invention.
FIG. 3 is a diagram of a symmetry point of different motors under no-load and variable-speed conditions in the embodiment of the present invention.
FIG. 4 is a partial high-level fusion feature diagram of two features in an embodiment of the present invention.
FIG. 5 is a flow chart of training of the fuzzy multi-model classifier in an embodiment of the present invention.
Detailed Description
The purpose and effect of the present invention will be more apparent from the following further description of the present invention with reference to the accompanying drawings. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
In this embodiment, the invention is described in detail by taking a built demagnetization fault experiment platform of a permanent magnet synchronous motor as an example. Fig. 1 is a schematic diagram of an experimental platform provided in an embodiment of the present invention. As shown in fig. 1, the platform includes: the upper computer monitoring and diagnosing platform 1 issues a control instruction to the motor driver 2, the motor driver 2 respectively controls the fault motor 3 and the load motor 4, and the two motors are dragged to simulate the operation of permanent magnet synchronous motor equipment. And a flux leakage signal on the surface of the fault motor 3 is measured by adopting a flux sensor 5, the flux leakage signal is uploaded to an upper computer monitoring and diagnosing platform 1 through a signal acquisition device 6, and the upper computer monitoring and diagnosing platform 1 realizes the fault diagnosis of the motor based on a semi-supervised depth rule-based permanent magnet synchronous motor demagnetization fault diagnosis method.
Example 1:
as shown in fig. 2, the method for diagnosing demagnetization fault of a permanent magnet synchronous motor applicable to multiple working conditions and variable working conditions according to the embodiment includes three parts, namely signal acquisition, two-dimensional image expansion, local high-level feature fusion extraction and fuzzy multi-model classifier diagnosis. In this embodiment, the fuzzy multi-model classifier is selected and constructed, and it should be understood that the fuzzy multi-model classifier is the best example of the present invention, but the present invention is not limited thereto, and other classification network models may be selected without departing from the concept of the present invention.
In this embodiment, a method for diagnosing a demagnetization fault of a permanent magnet synchronous motor specifically includes the following steps:
1) and collecting a motor surface magnetic flux leakage signal as an original signal for fault diagnosis, and expanding the original signal into a symmetrical dot matrix image.
Adopt alternating current magnetic flux sensor to gather the magnetic leakage signal of trouble motor under different operating modes in this embodiment, the operating mode type of selecting for use includes multiple stable condition, variable rotational speed operating mode, variable load operating mode. For example, according to the experimental platform shown in fig. 1 in this embodiment, 1000r/min,1300r/min and the variation working conditions between the rotation speeds of 2 demagnetization faulty motors (i.e., demagnetization faulty motor 1 (demagnetization 10%) and demagnetization faulty motor 2 (demagnetization 30%)) and 1 normal motor in the experimental platform are performed, and meanwhile, under different load working conditions, the non-contact ac magnetic sensor is used to measure the magnetic flux leakage signal on the surface of the motor.
The acquired magnetic leakage signal is a one-dimensional time domain signal, and can be converted into a two-dimensional image by adopting a symmetric lattice pattern generation method. The leakage flux signals of the motor 3 under various and variable working conditions are converted into a two-dimensional image data set which can be used for fault diagnosis, the expansion from one-dimensional time domain leakage flux signals to two-dimensional images is realized, and fig. 3 is a symmetrical point pattern of the demagnetization fault motor 1 at no-load time-varying rotating speed.
Wherein the signal point s of the time domain is divided intojWhen converting to S (r (j), α (j), β (j)) in polar coordinate space, r (j) is mapped to the radius of the polar coordinate, and the formula can be expressed as:
Figure BDA0003451138730000081
wherein s isminAnd smaxRespectively the minimum and maximum of the amplitude in the time domain signal s. The counterclockwise rotation angle alpha (j) and the clockwise rotation angle beta (j) of the mirror symmetry plane are defined byjBy amplitude s after instant kj+kMapping, the formula is:
Figure BDA0003451138730000082
Figure BDA0003451138730000083
wherein, thetawAngle of w-th plane of symmetry, θw＝360m/n(m＝0,1,2…n-1, n is the number of mirror symmetry planes, typically 6); λ is the gain angle (λ ≦ θ)w) α (j) and β (j) together determine the pattern range of the polar coordinates. Based on the formula conversion, the method is suitable for the SDP image to reflect the hidden morphological characteristics in the one-dimensional magnetic flux leakage signal.
2) And respectively extracting and fusing the acceleration robust speckle characteristic of the image and the local high-level characteristic of the acceleration segmentation test corner characteristic.
It should be noted that a group of magnetic leakage signals collected by the invention may be under the same working condition, or under variable working conditions or under different working conditions, so that when the acceleration steady spot features of a group of symmetric lattice images of the magnetic leakage signals and the local high-level features of the acceleration segmented test corner features are fused, the obtained fusion features may be obtained by the fault motor under the same working condition, under variable working conditions or under different working conditions; in addition, the invention also does not restrict the fusion of the characteristics of a plurality of groups of magnetic leakage signals under different working conditions under the same type of faults or the fusion of the characteristics of a plurality of groups of magnetic leakage signals under the same working condition under the same type of faults.
The local high-level feature extraction is divided into feature point detection and feature description, different feature point detection methods determine which feature is extracted, and the feature description method determines the specific representation form of the feature. In this embodiment, an accelerated robust blob feature method and an accelerated segment testing corner feature method are selected to detect the image, specifically:
2.1) detecting the accelerated robust speckle characteristics of the image by adopting an accelerated robust speckle characteristic method, and describing the speckle characteristics by adopting an accelerated robust description method;
2.2) detecting the corner features of the image by adopting an accelerated segment testing corner feature method, and describing the corner features by adopting an accelerated robust description method.
In order to fully utilize different types of local high-level features, the invention further fuses the different types of local high-level features, and the clustering algorithm is selected for fusion in the embodiment, and more specifically, the Kmean + + clustering algorithm is selected as follows:
2.3) respectively carrying out N-type clustering on the two characteristics by adopting a Kmean + + clustering algorithm, wherein N is the number of clusters, and a clustering center is used as a characteristic vector;
and 2.4) fusing the two local feature vectors, and taking the fused feature vector as the feature vector of the image, thereby realizing the local high-level fusion feature of the image.
It should be understood that in other possible embodiments, other fusion techniques may be used for feature fusion, and the present invention is not limited thereto.
The accelerated robust blob feature of the image is detected with respect to the accelerated robust blob feature method selected in this embodiment. The determination of the accelerated robust speckle is that the concerned characteristic point position is determined by comparing three adjacent Hessian matrixes, and the calculation formula of the Hessian matrix can be expressed as:
Figure BDA0003451138730000091
where δ is the filter constant, fxxRepresenting the convolution of the second order differential of the Gaussian function at a pixel point (x, y), fxyAnd fyyThe definitions of (a) and (b) are consistent.
The method for testing corner features in accelerated segments selected in this embodiment detects corner features in accelerated segments of an image. Wherein the specific extraction process is as follows: a pixel P is selected from the image, and whether the pixel P is an accelerated segment test corner feature is determined. Setting a judgment threshold value t, taking a point P as a center, and taking 3 pixels with the radius as a discretization circle, wherein the periphery of the circle has 16 pixels; and comparing the brightness values of all 16 pixel points with the pixel value of a point P according to a threshold value t, wherein when the point P is a corner feature point of accelerated segment test, the point P is defined as a maximum brightness point of 16 points around or a minimum brightness point of 16 points around.
3) And taking the fused local high-level feature sample as input, and training and diagnosing the fuzzy multi-model classifier. In this embodiment, the fault diagnosis classifier is selected as a fuzzy multi-model classifier, and it should be understood that the fuzzy multi-model classifier is an optimal implementation manner, which is that the performance of most existing classification models depends on internal parameters and hyper-parameters, and the hyper-parameters of the classification models are different for different classification objects. The fuzzy multi-model classifier does not have the problems when facing different fault types, but the invention is not limited to the network model.
In this embodiment, training and diagnosing the fuzzy multi-model classifier by using the fused local high-level feature sample as an input includes:
3.1) dividing the local high-level fusion characteristic data set into a training set verification set;
3.2) using the training set data for training the fuzzy multi-model classifier;
3.3) using the validation set data for validation of the fuzzy multi-model classifier.
Wherein the i-th type fault ithOf the τ -th data sample τthFusion feature g ofτThe training process is illustrated by way of example. Through gτNormalized by the norm, the formula is as follows:
Figure BDA0003451138730000101
then, τ is updatedthMean of class data, i.e. about
Figure BDA0003451138730000102
Is updated to
Figure BDA0003451138730000103
Due to gτAverage scalar presence of normalized data
Figure BDA0003451138730000104
No update is required. Euclidean distance
Figure BDA0003451138730000105
Expressed as:
Figure BDA0003451138730000106
wherein the mean value
Figure BDA0003451138730000107
And average scalar
Figure BDA0003451138730000108
With recursive operations, one-way densities can be obtained without loops and iterations, and their calculations can be expressed as:
Figure BDA0003451138730000109
Figure BDA00034511387300001010
calculation by equation (6) includes
Figure BDA00034511387300001011
All the identified features therein
Figure BDA00034511387300001012
Is a Euclidean distance of, wherein HiIs the number of features that have been identified. Judging whether a new fuzzy rule and a data cloud thereof are generated by the following condition 1:
Figure BDA00034511387300001013
if the condition is satisfied, then
Figure BDA00034511387300001014
And forming a data cloud nearby by using a new fuzzy rule, wherein relevant parameters are as follows:
Figure BDA00034511387300001015
the specific explanation is as follows:
Hi←Hi+1 in HiThe number of data samples in the data cloud of the i-th type fault is already, so that the formula is used for adding 1 to the number of data samples representing the first type fault.
Figure BDA00034511387300001016
In (1),
Figure BDA00034511387300001017
all data samples of the i-th fault are characterized, and therefore the formula indicates that the features of the t-th data sample to be determined are classified into the i-th fault.
Figure BDA0003451138730000111
In (1),
Figure BDA0003451138730000112
the number of the newly added data clouds in the whole fuzzy system is shown, and the number is increased by 1 because the new data clouds are generated.
Figure BDA0003451138730000113
In, r0A parameter representing an initial state of the new data cloud,
Figure BDA0003451138730000114
and representing the state parameters of the data cloud corresponding to the ith type of fault.
If condition 1 is not satisfied, then find the off-feature
Figure BDA0003451138730000115
Recent features
Figure BDA0003451138730000116
Figure BDA0003451138730000117
Then, the characteristics are measured
Figure BDA0003451138730000118
Categorizing features
Figure BDA0003451138730000119
Before the corresponding data cloud, judging the characteristics through the condition 2
Figure BDA00034511387300001110
Whether or not to interact with a feature
Figure BDA00034511387300001111
The corresponding data cloud is closest;
condition 2:
Figure BDA00034511387300001112
the above conditions represent: if it is satisfied with
Figure BDA00034511387300001113
The data cloud classified to the nearest is classified, and the update formula of the fuzzy rule corresponding to the nearest data cloud can be expressed as:
Figure BDA00034511387300001114
Figure BDA00034511387300001115
in (1),
Figure BDA00034511387300001116
number representing that the i-th fault corresponds to in the whole fuzzy systemThe number of clouds, i.e. not currently new, remains the same.
Otherwise, it is also according to the formula (12)
Figure BDA00034511387300001117
A new data cloud is generated nearby. The multi-model structure has no updated rule, and the original rule is reserved.
Each failure verification sample is sent to all sub-fuzzy rules and a confidence score is calculated according to the following rules:
Figure BDA00034511387300001118
wherein, gτ:
Figure BDA00034511387300001119
Indicates g if it is to be diagnosedτThe sample is at the ith sub-classifier (i.e. the ith fault), and the corresponding confidence score is gammaiFormula (II)
Figure BDA00034511387300001120
And then according to a judgment formula, identifying the test sample as the fault type with the maximum confidence coefficient, thereby realizing fault diagnosis. The decision formula is:
Figure BDA0003451138730000121
wherein, γlAnd (4) determining the confidence, the number of the fuzzy models with the q value, and Label as the fault type.
Based on the theoretical statement, a diagnosis model (fault diagnosis classifier) which can be used for realizing demagnetization fault diagnosis of the permanent magnet synchronous motor is constructed, and then based on the fault diagnosis classifier, demagnetization fault diagnosis can be realized on the motor.
4) Acquiring a magnetic leakage signal of the permanent magnet synchronous motor to be diagnosed, extracting and fusing the characteristics according to the modes of 1) -2), and inputting the fused characteristics into a fault diagnosis classifier to obtain a fault diagnosis result.
Example 2:
the present embodiment provides a system based on the above fault diagnosis model building method or fault diagnosis method, which includes:
the signal acquisition module is used for acquiring/obtaining magnetic leakage signals of the fault motor under various faults and different working conditions/variable working conditions of the fault motor, and the magnetic leakage signals are time domain signals. The signal acquisition module can be realized by a software module, namely, used for acquiring a magnetic leakage signal acquired by hardware, and can also be realized by hardware, such as a magnetic flux sensor.
And the image conversion module is used for expanding the magnetic leakage signal into a symmetrical dot matrix image, and the implementation process of the image conversion module can refer to the content of the method.
The characteristic extraction module is used for extracting local high-level characteristics of the symmetrical dot matrix image by adopting different characteristic point detection methods;
the characteristic fusion module is used for fusing local high-level characteristics of the symmetrical dot matrix images under the same type of faults and under the same working condition/different working conditions/variable working conditions;
the fault diagnosis classifier building module is used for taking the fused features as the input of the fault diagnosis classifier, taking the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier;
and the diagnosis module is used for diagnosing the fault by utilizing the fault diagnosis classifier and the fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result.
It should be understood that, the specific implementation process of the above unit module refers to the method content, and the present invention is not described herein in detail, and the division of the above functional module unit is only a division of a logic function, and there may be another division manner in the actual implementation, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or may not be executed. Meanwhile, the integrated unit can be realized in a hardware form, and can also be realized in a software functional unit form.
Example 3:
the embodiment provides an electronic terminal, which comprises a processor and a memory which are connected with each other, wherein the processor is programmed or configured to execute the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor or the steps of the demagnetization fault diagnosis method of the permanent magnet synchronous motor.
When the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor is executed, the following steps are specifically executed:
step 1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
step 2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
and step 3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
and 4, step 4: and performing network training by using the fusion characteristics of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier.
When the demagnetization fault diagnosis method of the permanent magnet synchronous motor is executed, the following steps are specifically executed:
s1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
s2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
s3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
s4: using the fused features as the input of a fault diagnosis classifier, using the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of different working conditions/variable working conditions of various faults in S3 to obtain the fault diagnosis classifier;
s5: and acquiring a magnetic leakage signal of the permanent magnet synchronous motor to be diagnosed, performing feature extraction and feature fusion on the fault motor under the current working condition/variable working condition according to the steps S2-S3, and inputting the fused features into a fault diagnosis classifier to obtain a fault diagnosis result.
The specific implementation process of each step refers to the explanation of the foregoing method.
It should be understood that in the embodiments of the present invention, the Processor may be a Central Processing Unit (CPU), and the Processor may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The memory may include both read-only memory and random access memory, and provides instructions and data to the processor. The portion of memory may also include non-volatile random access memory. For example, the memory may also store device type information.
Example 4:
the present embodiment provides a fault diagnosis system, which includes: the fault diagnosis system comprises an upper fault diagnosis computer and an experiment platform with a plurality of fault motors, wherein the experiment platform simulates the operation of permanent magnet synchronous motor equipment, and the reference is made to figure 1.
One implementation manner is as follows:
if a processor and a memory are arranged in the experimental platform, the processor is programmed or configured to execute the permanent magnet synchronous motor demagnetization fault diagnosis model construction method according to claim 1, and the fault diagnosis classification model is uploaded to a fault diagnosis upper computer so that the fault diagnosis upper computer can carry out fault diagnosis by using the fault diagnosis classifier and fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result;
the other realization mode is as follows:
if a processor and a memory are arranged in the fault diagnosis upper computer, the experimental platform uploads a magnetic leakage signal simulating the running of the permanent magnet synchronous motor equipment to the fault diagnosis upper computer, and the processor in the fault diagnosis upper computer is programmed or configured to execute the demagnetization fault diagnosis method of the permanent magnet synchronous motor according to claim 2.
The specific implementation process of each step refers to the explanation of the foregoing method.
It should be understood that in the embodiments of the present invention, the Processor may be a Central Processing Unit (CPU), and the Processor may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The memory may include both read-only memory and random access memory, and provides instructions and data to the processor. The portion of memory may also include non-volatile random access memory. For example, the memory may also store device type information.
Example 5:
the present embodiment provides a readable storage medium, wherein the readable storage medium stores a computer program programmed or configured to execute the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor or the method for diagnosing the demagnetization fault of the permanent magnet synchronous motor.
When the method for constructing the demagnetization fault diagnosis model of the permanent magnet synchronous motor is executed, the following steps are specifically executed:
step 1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
step 2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
and step 3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
and 4, step 4: and performing network training by using the fusion characteristics of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier.
When the demagnetization fault diagnosis method of the permanent magnet synchronous motor is executed, the following steps are specifically executed:
s1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
s2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
s3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
s4: using the fused features as the input of a fault diagnosis classifier, using the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of different working conditions/variable working conditions of various faults in S3 to obtain the fault diagnosis classifier;
s5: and acquiring a magnetic leakage signal of the permanent magnet synchronous motor to be diagnosed, performing feature extraction and feature fusion on the fault motor under the current working condition/variable working condition according to the steps S2-S3, and inputting the fused features into a fault diagnosis classifier to obtain a fault diagnosis result.
The specific implementation process of each step refers to the explanation of the foregoing method.
The readable storage medium is a computer readable storage medium, which may be an internal storage unit of the controller according to any of the foregoing embodiments, for example, a hard disk or a memory of the controller. The readable storage medium may also be an external storage device of the controller, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the controller. Further, the readable storage medium may also include both an internal storage unit of the controller and an external storage device. The readable storage medium is used for storing the computer program and other programs and data required by the controller. The readable storage medium may also be used to temporarily store data that has been output or is to be output.
Based on such understanding, the technical solution of the present invention essentially or partially contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned readable storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The present application is directed to methods, apparatus (systems), and computer program products according to embodiments of the application, wherein the instructions that execute via the flowcharts and/or processor of the computer program product create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
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 method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor is characterized by comprising the following steps: the method comprises the following steps: acquiring signals, performing two-dimensional image conversion on the acquired signals, extracting and fusing the characteristics of the two-dimensional images, and constructing a fault diagnosis classifier by utilizing the fused characteristics, which comprises the following steps:
step 1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
step 2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
and step 3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
and 4, step 4: taking the fused features as the input of a fault diagnosis classifier, taking the fault type as the output of the fault diagnosis classifier, and performing network training by using the fused features of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier;
the classifier is applied to demagnetization fault diagnosis of the permanent magnet synchronous motor.
2. A demagnetization fault diagnosis method for a permanent magnet synchronous motor based on the method of claim 1, characterized in that: the method comprises the following steps: the method comprises the following steps of constructing a fault diagnosis classifier and diagnosing faults based on the fault diagnosis classifier, wherein the fault diagnosis classifier specifically comprises the following steps:
s1: collecting magnetic leakage signals of a fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
s2: expanding the magnetic leakage signal into a symmetrical dot matrix image;
s3: adopting different feature point inspection methods to perform local high-level feature extraction and feature fusion on the symmetric dot matrix image;
s4: using the fused features as the input of a fault diagnosis classifier, using the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of different working conditions/variable working conditions of various faults in S3 to obtain the fault diagnosis classifier;
s5: and acquiring a magnetic leakage signal of the permanent magnet synchronous motor to be diagnosed, performing feature extraction and feature fusion on the fault motor under the current working condition/variable working condition according to the steps S2-S3, and inputting the fused features into a fault diagnosis classifier to obtain a fault diagnosis result.
3. The method of claim 1, wherein: and 2, respectively extracting the acceleration steady spot characteristics of the symmetrical dot matrix image by adopting an acceleration steady spot characteristic method, extracting the acceleration subsection testing corner characteristics by adopting an acceleration subsection testing corner characteristic method, and fusing the acceleration steady spot characteristics and the acceleration subsection testing corner characteristics of the symmetrical dot matrix image under the same working condition/different working conditions/variable working conditions of the same type of faults.
4. The method of claim 1, wherein: the process of expanding the magnetic leakage signal into a symmetrical lattice image in the step 2 is as follows: converting signal points of the magnetic leakage signal into a polar coordinate space, and constructing a symmetrical dot matrix image based on coordinates of the polar coordinate space of all the signal points;
the conversion formula for converting the signal point of the leakage magnetic signal into the polar coordinate space is as follows:
Figure FDA0003451138720000021
Figure FDA0003451138720000022
Figure FDA0003451138720000023
in the formula, sjRepresents a certain signal point in the leakage signal s, and r (j), α (j), β (j) represent the signal point sjCoordinates in polar coordinate space, where r (j) is the signal point sjThe radius of the polar coordinate is mapped, and alpha (j) and beta (j) are respectively an anticlockwise rotating angle and a clockwise rotating angle of a mirror symmetry plane; sminAnd smaxRespectively the minimum value and the maximum value of the amplitude in the leakage magnetic signal s, and the counterclockwise rotation angle alpha (j) and the clockwise rotation angle beta (j) of the mirror symmetry plane are respectively equal to sjSignal amplitude s after the phase difference kj+kMapping, θwAngle of w-th plane of symmetry, θw＝360m/n，m＝0，1，2…n-1, n is the number of mirror symmetry planes, λ is the gain angle, λ ≦ θw。
5. The method of claim 1, wherein: the fault diagnosis classifier is a fuzzy multi-model classifier, and the training process of the model multi-model classifier is as follows: inputting the fusion characteristics of the data samples into a fuzzy network; updating the fuzzy network according to the following process until all data samples are trained; each data sample is classified and determined according to a maximum confidence coefficient winning principle, wherein the leakage magnetic signal of a fault motor with a known fault type is used as the data sample, and the fusion characteristics of the data samples are obtained according to the methods of the steps 2 to 3;
wherein, firstly, the ith fault ithOf the τ -th data sample τthFusion feature g ofτCarrying out normalization;
Figure FDA0003451138720000024
wherein the content of the first and second substances,
Figure FDA0003451138720000025
represents the normalized feature, | gτ| is gτNorm of the feature;
then, the ith fault i is calculated according to the following formulathOf the data samples τthCorresponding mean value
Figure FDA0003451138720000026
And average scalar
Figure FDA0003451138720000027
Based on the mean value
Figure FDA0003451138720000028
And average scalar
Figure FDA0003451138720000029
Calculating the Euclidean distance
Figure FDA00034511387200000210
Figure FDA00034511387200000211
Figure FDA00034511387200000212
Figure FDA00034511387200000213
Wherein the content of the first and second substances,
Figure FDA00034511387200000214
judging whether a new fuzzy rule and a new data cloud are generated or not according to the following condition 1; if condition 1 is satisfied, then the feature is
Figure FDA0003451138720000031
Forming a data cloud by a new fuzzy rule around; otherwise, judging whether the condition 2 is met; if condition 2 is satisfied, the feature is utilized
Figure FDA0003451138720000032
Updating distance features
Figure FDA0003451138720000033
Fuzzy rules of the most recent data cloud; otherwise, in the feature
Figure FDA0003451138720000034
Forming a data cloud by a new fuzzy rule around;
condition 1:
Figure FDA0003451138720000035
or
Figure FDA0003451138720000036
In the formula, HiFor fault i of type ithThe number of data samples in the data cloud of (1),
Figure FDA0003451138720000037
characterizing class i faults ithThe Euclidean distance corresponding to each data sample in the data cloud;
condition 2:
Figure FDA0003451138720000038
wherein the content of the first and second substances,
Figure FDA0003451138720000039
for the distance characteristic sought
Figure FDA00034511387200000310
The feature points of the closest cloud are,
Figure FDA00034511387200000311
is a characteristic point
Figure FDA00034511387200000312
The state parameter of the data cloud.
6. The method of claim 5, wherein: when the maximum confidence coefficient winning principle is classified, the calculation formula of the confidence coefficient score is as follows:
Figure FDA00034511387200000313
Figure FDA00034511387200000314
wherein, γiFor confidence, i is 1,2 … q, q is the number of sub-classifiers corresponding to the number of fault types, Label is the fault type,
Figure FDA00034511387200000315
cloud center point features representing a type i fault.
7. A system according to claim 1 or 2, characterized in that: the method comprises the following steps:
the signal acquisition module is used for acquiring/obtaining magnetic leakage signals of the fault motor under various faults and different working conditions/variable working conditions of the fault motor, wherein the magnetic leakage signals are time domain signals;
the image conversion module is used for expanding the magnetic leakage signal into a symmetrical dot matrix image;
the characteristic extraction module is used for extracting local high-level characteristics of the symmetrical dot matrix image by adopting different characteristic point detection methods;
the characteristic fusion module is used for fusing local high-level characteristics of the symmetrical dot matrix images under the same type of faults and under the same working condition/different working conditions/variable working conditions;
the fault diagnosis classifier building module is used for taking the fused features as the input of the fault diagnosis classifier, taking the fault types as the output of the fault diagnosis classifier, and performing network training by using the fused features of various faults of known fault types under different working conditions/variable working conditions to obtain the fault diagnosis classifier;
and the diagnosis module is used for diagnosing the fault by utilizing the fault diagnosis classifier and the fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result.
8. An electronic terminal comprising a processor and a memory connected to each other, wherein the processor is programmed or configured to execute the steps of the method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor according to claim 1 or the method for diagnosing a demagnetization fault of a permanent magnet synchronous motor according to claim 2.
9. A system based on the method of claim 1 or claim 2, comprising: the system comprises a fault diagnosis upper computer and an experiment platform with a plurality of fault motors, wherein the experiment platform simulates the operation of permanent magnet synchronous motor equipment;
if a processor and a memory are arranged in the experimental platform, the processor is programmed or configured to execute the permanent magnet synchronous motor demagnetization fault diagnosis model construction method according to claim 1, and the fault diagnosis classification model is uploaded to a fault diagnosis upper computer so that the fault diagnosis upper computer can perform fault diagnosis by using the fault diagnosis classifier and fusion characteristics corresponding to the permanent magnet synchronous motor to be diagnosed to obtain a fault diagnosis result;
or, if be equipped with treater and memory in the fault diagnosis host computer, the experiment platform will simulate the magnetic leakage signal of PMSM class equipment during operation and upload to the fault diagnosis host computer, the treater in the fault diagnosis host computer is programmed or is configured in order to carry out claim 2 PMSM demagnetization fault diagnosis method.
10. A readable storage medium, characterized in that the readable storage medium stores therein a computer program programmed or configured to execute the method for constructing a demagnetization fault diagnosis model of a permanent magnet synchronous motor according to claim 1 or the method for diagnosing a demagnetization fault of a permanent magnet synchronous motor according to claim 2.
CN202111674484.4A 2021-12-31 2021-12-31 Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system Pending CN114358077A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111674484.4A CN114358077A (en) 2021-12-31 2021-12-31 Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111674484.4A CN114358077A (en) 2021-12-31 2021-12-31 Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system

Publications (1)

Publication Number Publication Date
CN114358077A true CN114358077A (en) 2022-04-15

Family

ID=81105802

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111674484.4A Pending CN114358077A (en) 2021-12-31 2021-12-31 Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system

Country Status (1)

Country Link
CN (1) CN114358077A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116597167A (en) * 2023-06-06 2023-08-15 中国人民解放军92942部队 Permanent magnet synchronous motor small sample demagnetization fault diagnosis method, storage medium and system

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116597167A (en) * 2023-06-06 2023-08-15 中国人民解放军92942部队 Permanent magnet synchronous motor small sample demagnetization fault diagnosis method, storage medium and system
CN116597167B (en) * 2023-06-06 2024-02-27 中国人民解放军92942部队 Permanent magnet synchronous motor small sample demagnetization fault diagnosis method, storage medium and system

Similar Documents

Publication Publication Date Title
Hang et al. Application of multi-class fuzzy support vector machine classifier for fault diagnosis of wind turbine
Maraaba et al. Convolutional neural network-based inter-turn fault diagnosis in LSPMSMs
Haroun et al. Multiple features extraction and selection for detection and classification of stator winding faults
Seera et al. Application of the fuzzy min–max neural network to fault detection and diagnosis of induction motors
Li et al. A new intelligent fault diagnosis method of rotating machinery under varying-speed conditions using infrared thermography
CN115510962A (en) Water pump permanent magnet synchronous motor with fault self-monitoring function and method thereof
CN114358077A (en) Permanent magnet synchronous motor demagnetization fault diagnosis model construction method and fault diagnosis method and system
CN113420837A (en) Fault diagnosis method based on multi-source compressed sensing
Zaman et al. Greedy-gradient max cut-based fault diagnosis for direct online induction motors
CN117269754A (en) IPSM rotor demagnetizing and eccentric fault diagnosis method based on convolutional neural network operation
Hussain et al. Stator winding fault detection and classification in three-phase induction motor
Kim et al. Hybrid data-scaling method for fault classification of compressors
Al-Haddad et al. Improved UAV blade unbalance prediction based on machine learning and ReliefF supreme feature ranking method
CN114881071A (en) Synchronous motor rotor winding turn-to-turn short circuit fault diagnosis method based on multi-source information
Thakur et al. Prediction of unknown fault of induction motor using SVM following decision-directed acyclic graph
Chen et al. Multi-dimensional color image recognition and mining based on feature mining algorithm
Demirel et al. Autonomous fault detection and diagnosis for permanent magnet synchronous motors using combined variational mode decomposition, the Hilbert-Huang transform, and a convolutional neural network
Dai et al. Fault diagnosis of permanent magnet synchronous motor based on improved probabilistic neural network
CN116597167B (en) Permanent magnet synchronous motor small sample demagnetization fault diagnosis method, storage medium and system
Amiri Ahouee et al. Inter-turn fault detection in PM synchronous motor by neuro-fuzzy technique
CN117332340A (en) PMSM fault diagnosis method and system based on multi-sensor visual feature fusion
Karakose Reinforcement learning based artificial immune classifier
CN114254674A (en) Permanent magnet synchronous motor demagnetization fault diagnosis method and system based on semi-supervised classifier
CN116754230A (en) Bearing abnormality detection and fault diagnosis method based on deep convolution generation countermeasure network
Al-Greer et al. Autonomous fault detection and diagnosis for permanent magnet synchronous motors using combined variational mode decomposition, the Hilbert-Huang transform, and a convolutional neural network

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