CN110596506A - Converter fault diagnosis method based on time convolution network - Google Patents
Converter fault diagnosis method based on time convolution network Download PDFInfo
- Publication number
- CN110596506A CN110596506A CN201911042913.9A CN201911042913A CN110596506A CN 110596506 A CN110596506 A CN 110596506A CN 201911042913 A CN201911042913 A CN 201911042913A CN 110596506 A CN110596506 A CN 110596506A
- Authority
- CN
- China
- Prior art keywords
- fault
- classifier
- sample
- convolution network
- time convolution
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
The invention relates to a power electronic converter fault diagnosis method based on a time convolution network technology, which comprises the following steps: step S1: collecting electric signals of a measuring point and carrying out noise reduction treatment to obtain sample data with fault information; step S2: carrying out dimension reduction treatment on sample data with fault information by adopting normalization, and establishing a data sample library by corresponding the obtained fault characteristics and fault types one by one; step S3: constructing a fault classifier based on a time convolution network, and training and testing according to a data sample base to obtain an optimal network structure parameter; step S4, reconstructing the fault classifier based on the time convolution network according to the optimal network structure parameters to obtain the fault classifier with the optimal parameters; step S5: and writing the fault classifier network with the optimal parameters into simulink, and performing real-time fault diagnosis and positioning on the power electronic converter in actual operation. The invention can judge the health condition of the converter more accurately and more reliably.
Description
Technical Field
The invention relates to the technical field of power electronics, in particular to a converter fault diagnosis method based on a time convolution network.
Background
With the advent of the industrial 4.0 era, power electronic technology has been widely applied to various fields of production and life, and accordingly, power electronic fault diagnosis technology is indispensable.
Firstly, the power electronic converter is mostly used as a control device or a core power supply, if the fault type can not be scientifically diagnosed, only the carbuncle is cured, and the self-isolation and self-recovery of the device fault are influenced. Moreover, with the expansion of the fault range, the functional failures are increased, and great potential safety hazards exist.
Secondly, as the complexity of power electronics increases, maintenance costs increase. The converter, the main body of energy conversion in the power electronic device, has high failure rate and serious error consequences. The fault diagnosis is carried out slightly, the fault in a larger range is avoided, the maintenance cost is reduced, and unnecessary economic loss is avoided.
Finally, the number of components of the power electronic circuit is large, time and labor are wasted in one-by-one diagnosis, the rapidity, fault tolerance and reliability of automatic fault diagnosis are improved, the downtime can be reduced, the predictive maintenance can be realized, and manpower and material resources are saved.
The traditional power electronic converter fault diagnosis method comprises a support vector machine, a fault dictionary method and the like. The support vector machine is relatively simple computationally, but is susceptible to noise from the sampled signal, causing erroneous decisions on the output result. The fault dictionary has strong anti-interference capability, but the required fault samples are large, so that a good effect can be achieved. The method based on the time convolution network can distinguish and locate the known fault and the normal condition under the condition of less samples, and can distinguish the unknown fault, the normal condition and the known fault; the method can adapt to frequency conversion detection, and can still utilize the network to identify and position faults under the condition of output frequency change; the cloud server can be used for carrying out real-time online cloud processing on the multi-machine fault data, and massive data are provided for a multi-scale and multi-level complex system.
Disclosure of Invention
In view of this, an object of the present invention is to provide a converter fault diagnosis method based on a time convolution network, which can more accurately and reliably judge the health condition of a converter, can identify an unknown fault, is adaptive to a frequency conversion detection fault, and also improves the accuracy of converter fault analysis.
In order to achieve the purpose, the invention adopts the following technical scheme:
a converter fault diagnosis method based on a time convolution network comprises the following steps:
step S1: collecting electric signals of a measuring point and carrying out noise reduction treatment to obtain sample data with fault information;
step S2: carrying out dimension reduction treatment on sample data with fault information by adopting normalization, and establishing a data sample library by corresponding the obtained fault characteristics and fault types one by one;
step S3: constructing a fault classifier based on a time convolution network, and training and testing according to a data sample base to obtain an optimal network structure parameter;
step S4, reconstructing the fault classifier based on the time convolution network according to the optimal network structure parameters to obtain the fault classifier with the optimal parameters;
step S5: and writing the fault classifier network with the optimal parameters into simulink, and performing real-time fault diagnosis and positioning on the power electronic converter in actual operation.
Further, the step S1 is specifically:
step S11: applying faults to circuit components according to the actual fault occurrence condition, and simulating the circuit faults under the actual condition to generate output waveforms;
step S12: collecting the electric signals of the measuring points by using a data acquisition card;
step S13: and removing noise of the sampling signal through a simulink module, acquiring original sample data, and obtaining a sample with fault information.
Furthermore, the acquisition card adopts a PCI-6229 acquisition card.
Further, step S3 specifically includes the following steps:
step S31: dividing a data sample library into a training sample and a test sample;
step S32: training the fault classifier by using a training sample as the input of the time convolution network fault classifier;
step S33: training the training samples by adopting different classifier functions, and using the classifier function with good effect through the training result.
Step S34: judging whether the training error meets the preset condition, if so, entering the step S35, otherwise, changing the classifier function for further adjustment;
step S35: acquiring a better classifier function and a hyper-parameter, assigning the better classifier function and the hyper-parameter to a time convolution network-based classifier, and testing the time convolution network fault classifier assigned with the better parameter by adopting a test sample;
step S36: and judging whether the testing accuracy meets the preset requirement, if so, taking the last selected better parameter as the optimal parameter and ending the process, otherwise, returning to the step S33.
Further, the hyper-parameters include a convolution kernel size k, a dilation coefficient d, and a network depth n.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention can judge the health condition of the converter more accurately and more reliably;
2. the method can identify unknown faults, is suitable for frequency conversion detection faults, and improves the accuracy of fault analysis of the converter.
Drawings
FIG. 1 is a schematic diagram of a data collection process according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a TCN model structure according to an embodiment of the present invention
Fig. 4 is a diagram illustrating a TCN residual structure according to an embodiment of the invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
Referring to fig. 2, the present invention provides a converter fault diagnosis method based on a time convolution network, including the following steps:
step S1: collecting electric signals of a measuring point and carrying out noise reduction treatment to obtain sample data with fault information;
step S2: carrying out dimension reduction treatment on sample data with fault information by adopting normalization, and establishing a data sample library by corresponding the obtained fault characteristics and fault types one by one;
step S3: constructing a fault classifier based on a time convolution network, accurately dividing a normal state and various types of faults contained in a training sample after off-line training, extracting better parameters of the fault classifier, directly giving the better parameters to the classifier network, carrying out classifier test work, and selecting optimal network structure parameters through testing;
step S4, reconstructing the fault classifier based on the time convolution network according to the optimal network structure parameters to obtain the fault classifier with the optimal parameters;
step S5: and writing the fault classifier network with the optimal parameters into simulink, and performing real-time fault diagnosis and positioning on the power electronic converter in actual operation.
In the embodiment, a probability threshold value is set for a time convolution network fault classifier, when a novel unknown fault is input, the identification network can detect and judge the novel unknown fault into an N + 1-th fault different from a known N-type fault, and the system can be ensured to run safely when the unknown fault occurs;
as shown in fig. 1, in this embodiment, the step S1 specifically includes:
step S11: applying faults to circuit components according to the actual fault occurrence condition, and simulating the circuit faults under the actual condition to generate output waveforms;
step S12: collecting the electric signals of the measuring points by adopting a PCI-6229 collecting card;
step S13: and removing noise of the sampling signal through a simulink module, acquiring original sample data, and obtaining a sample with fault information.
Preferably, in this embodiment, when diagnosing the actually operating circuit, the extracted sample information is also recorded, and when the samples are accumulated to a certain number, the method may participate in constructing a new sample database to enrich the data of the sample database.
In this embodiment, step S3 specifically includes the following steps:
step S31: dividing a data sample library into a training sample and a test sample;
step S32: training the fault classifier by using a training sample as the input of the time convolution network fault classifier;
step S33: training the training samples by adopting different classifier functions, and using the classifier function with good effect through the training result.
Step S34: judging whether the training error meets the preset condition, if so, entering the step S35, otherwise, changing the classifier function for further adjustment;
step S35: acquiring a better classifier function and a hyper-parameter, wherein the hyper-parameter comprises a convolution kernel size k, an expansion coefficient d and a network depth n, and endowing the super-parameter to a time convolution network-based classifier, and testing the time convolution network fault classifier endowed with the better parameter by adopting a test sample;
step S36: and judging whether the testing accuracy meets the preset requirement, if so, taking the last selected better parameter as the optimal parameter and ending the process, otherwise, returning to the step S33.
In this embodiment, the Time Convolutional Network (TCN) fault classifier is a network structure capable of processing time series data, and infers new possible information according to an input sequence, and uses a judgment mechanism to evaluate the prediction effect, for example, an ordinary fully-connected layer uses MSE as a loss function.
The model structure diagram of TCN is shown in fig. 3, defining the model input sequence as: x is the number of0,x1,...,xtThe output sequence is as follows: y is0,y1,...,yt。
The largest difference between TCN and one-dimensional convolution is that it mainly uses dilation convolution to obtain global information of the whole sequence, so that each layer of hidden layer has the same size as the input sequence, and sets layer jump connection of residual convolution, and 1 × 1 convolution operation.
The general form of the dilated convolution kernel is:
in the formula, f represents a convolution kernel (filter), k represents a convolution kernel size (kernel size), d represents a dilation factor (scaling factor), and means the number of intervals of the convolution kernel, and s-d · i represents convolution only for a past state. As shown in FIG. 3, the higher the upper layers, the larger the convolution window, and the more "holes" in the convolution window, the larger the receptive field. Therefore, the advantage of the extended convolution is that the field of view can be enlarged without pooling lost information, so that each convolution output contains a larger range of information.
Residual concatenation is described using the formula:
o=Activation(x+F(x))
in the formula, Activation represents an Activation function, including functions such as ReLU, Sigmoid, tanh, and the like.
To obtain a larger receptive field, the network depth n has to be increased, so the residual units are constructed to train deeper networks. Residual convolution is to take the lower layer features to the higher layers to enhance accuracy.
The 1 x 1 convolution is used for dimensionality reduction. The TCN directly connects the feature map layer of the lower layer to the upper layer, and the feature map numbers (i.e., channel numbers) of each corresponding Cell are inconsistent, so that the addition operation of the feature map layer of the lower layer similar to Resnet cannot be directly performed. Therefore, in order to make the feature map quantity identical when two layers are added, 1 × 1 convolution is used for dimension reduction operation.
By adjusting the network depth n, the convolution kernel size k and the expansion coefficient d, the receptive field can be flexibly controlled, the calculated amount is reduced, and the method is suitable for different tasks.
Defining a probability density function ofWhere k is a training sample (k ═ 1, 2.., c), nkThe number of samples of the kth class training set.Representing the probability density function of the r-th sample of the kth class of samples.
When there are class c training samples, each class has a corresponding probability density functionConstructing a probability density function of all training samples asIt can be defined as:
from the probability density function of all training samples obtainedThen the specific gravity p for a single fault is defined as follows:
when specific gravity ρ(k)Is specific gravity ρ(1),ρ(2),...,ρ(c)And when the maximum value is the k-th fault type, the sample to be detected belongs to the k-th fault type. Thus, the TCN algorithm can be used to identify the type of the fault sample with known type.
Define β as the minimum value of specific gravity:
β=min{ρ}
where the beta value is the boundary between the known class fault sample and the unknown fault sample. When rho is larger than beta, the test sample belongs to a normal or known fault condition; when rho is less than beta, the test sample is a fault sample of unknown type, and the fault sample is output as an N + 1-th fault different from the known N-type faults, so that the system can safely operate when the unknown fault occurs. Thus, the TCN algorithm can be used for carrying out class distinction on the fault samples of the unknown classes.
Preferably, in this embodiment, by using the time convolution network, when an unknown fault occurs in a circuit, the difference between the fault state and the normal state can be distinguished, so as to implement self-checking, and the algorithm can reduce the time and complexity of sample training and reduce the dependence on fault samples.
Preferably, in the present embodiment, the expansion coefficient is designed to be a zigzag structure, for example, a cyclic structure of [1,2,5,1,2,5 ]. The properties of the saw-tooth shape can satisfy the requirements of small object and large object segmentation (the expansion coefficient of the small object is concerned with the short-distance information, and the expansion coefficient of the large object is concerned with the long-distance information) at the same time, and reduce the continuity of loss information.
Preferably, in this embodiment, a TCN fault classifier network is constructed, a classifier parameter selection rule is derived, the number of sampling points is changed according to the change of the actual working output frequency, and the same model constructed by the TCN is also used for classification and identification of faults, so as to realize detection under the self-adaptive frequency conversion condition; the single chip microcomputer is used for receiving data and sending the data to the cloud server, massive data are provided for a multi-scale and multi-level complex system, real-time online cloud processing is conducted on multi-machine fault data, dependence on historical data is reduced, real-time self-checking on a power electronic converter circuit is achieved, and high-efficiency and high-reliability fault identification and positioning are achieved. The wireless data transmission device has the advantages that wireless data transmission is used as a medium for transmitting data, data transmission is more convenient than wired data transmission, and the phenomenon that a large number of transmission lines are used in the same place and are too messy is avoided.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.
Claims (5)
1. A converter fault diagnosis method based on a time convolution network is characterized by comprising the following steps:
step S1: collecting electric signals of a measuring point and carrying out noise reduction treatment to obtain sample data with fault information;
step S2: carrying out dimension reduction treatment on sample data with fault information by adopting normalization, and establishing a data sample library by corresponding the obtained fault characteristics and fault types one by one;
step S3: constructing a fault classifier based on a time convolution network, and training and testing according to a data sample base to obtain an optimal network structure parameter;
step S4, reconstructing the fault classifier based on the time convolution network according to the optimal network structure parameters to obtain the fault classifier with the optimal parameters;
step S5: and writing the fault classifier network with the optimal parameters into simulink, and performing real-time fault diagnosis and positioning on the power electronic converter in actual operation.
2. The method for diagnosing the fault of the converter based on the time convolution network as claimed in claim 1, wherein the method comprises the following steps:
step S11: applying faults to circuit components according to the actual fault occurrence condition, and simulating the circuit faults under the actual condition to generate output waveforms;
step S12: collecting the electric signals of the measuring points by using a data acquisition card;
step S13: and removing noise of the sampling signal through a simulink module, acquiring original sample data, and obtaining a sample with fault information.
3. The method for diagnosing the fault of the converter based on the time convolution network as claimed in claim 2, characterized in that: the acquisition card adopts a PCI-6229 acquisition card.
4. The method for diagnosing the fault of the converter based on the time convolution network as claimed in claim 1, wherein the method comprises the following steps: step S3 specifically includes the following steps:
step S31: dividing a data sample library into a training sample and a test sample;
step S32: training the fault classifier by using a training sample as the input of the time convolution network fault classifier;
step S33: training the training samples by adopting different classifier functions, and using the classifier function with good effect according to the training result;
step S34: judging whether the training error meets the preset condition, if so, entering the step S35, otherwise, changing the classifier function for further adjustment;
step S35: acquiring a better classifier function and a hyper-parameter, assigning the better classifier function and the hyper-parameter to a time convolution network-based classifier, and testing the time convolution network fault classifier assigned with the better parameter by adopting a test sample;
step S36: and judging whether the testing accuracy meets the preset requirement, if so, taking the last selected better parameter as the optimal parameter and ending the process, otherwise, returning to the step S33.
5. The method for diagnosing the fault of the converter based on the time convolution network as claimed in claim 1, wherein the method comprises the following steps: the hyper-parameters include convolution kernel size k, dilation coefficient d, and network depth n.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911042913.9A CN110596506A (en) | 2019-10-30 | 2019-10-30 | Converter fault diagnosis method based on time convolution network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911042913.9A CN110596506A (en) | 2019-10-30 | 2019-10-30 | Converter fault diagnosis method based on time convolution network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110596506A true CN110596506A (en) | 2019-12-20 |
Family
ID=68851840
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911042913.9A Pending CN110596506A (en) | 2019-10-30 | 2019-10-30 | Converter fault diagnosis method based on time convolution network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110596506A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111325159A (en) * | 2020-02-25 | 2020-06-23 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Fault diagnosis method and device, computer equipment and storage medium |
CN111597996A (en) * | 2020-05-18 | 2020-08-28 | 中车永济电机有限公司 | Method for constructing wind turbine generator bearing fault identification model based on deep learning |
CN111798975A (en) * | 2020-06-01 | 2020-10-20 | 山东师范大学 | Disease diagnosis system, apparatus and medium based on cyclic time convolution network |
CN111803059A (en) * | 2020-06-30 | 2020-10-23 | 武汉中旗生物医疗电子有限公司 | Electrocardiosignal classification method and device based on time domain convolution network |
CN112016251A (en) * | 2020-09-02 | 2020-12-01 | 哈尔滨工程大学 | Nuclear power device fault diagnosis method and system |
CN112051481A (en) * | 2020-08-12 | 2020-12-08 | 华中科技大学 | Alternating current-direct current hybrid power grid fault area diagnosis method and system based on LSTM |
CN112348124A (en) * | 2021-01-05 | 2021-02-09 | 北京航空航天大学 | Data-driven micro fault diagnosis method and device |
CN112729529A (en) * | 2020-12-17 | 2021-04-30 | 苏州大学 | Motor defect detection method |
CN113962256A (en) * | 2021-09-29 | 2022-01-21 | 西安交通大学 | Intelligent fault diagnosis method and system for electromechanical actuator |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130208514A1 (en) * | 2010-06-18 | 2013-08-15 | ALSTOM Technlogy Ltd | Converter for hvdc transmission and reactive power compensation |
CN105606914A (en) * | 2015-09-06 | 2016-05-25 | 南京航空航天大学 | IWO-ELM-based Aviation power converter fault diagnosis method |
CN109116150A (en) * | 2018-08-03 | 2019-01-01 | 福州大学 | A kind of converters method for diagnosing faults based on Cerebellar Model Articulation Controller |
CN109828168A (en) * | 2019-01-31 | 2019-05-31 | 福州大学 | Converter method for diagnosing faults based on Density Estimator |
-
2019
- 2019-10-30 CN CN201911042913.9A patent/CN110596506A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130208514A1 (en) * | 2010-06-18 | 2013-08-15 | ALSTOM Technlogy Ltd | Converter for hvdc transmission and reactive power compensation |
CN105606914A (en) * | 2015-09-06 | 2016-05-25 | 南京航空航天大学 | IWO-ELM-based Aviation power converter fault diagnosis method |
CN109116150A (en) * | 2018-08-03 | 2019-01-01 | 福州大学 | A kind of converters method for diagnosing faults based on Cerebellar Model Articulation Controller |
CN109828168A (en) * | 2019-01-31 | 2019-05-31 | 福州大学 | Converter method for diagnosing faults based on Density Estimator |
Non-Patent Citations (2)
Title |
---|
南京航空航天大学科技部: "《南京航空航天大学论文集 2010年 第19册 自动化学院 第7分册》", 31 May 2011, 南京航空航天大学科技部 * |
阳景 等: "基于时间卷积网络的模拟电路故障诊断方法", 《电子测量技术》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111325159A (en) * | 2020-02-25 | 2020-06-23 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Fault diagnosis method and device, computer equipment and storage medium |
CN111325159B (en) * | 2020-02-25 | 2023-09-08 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Fault diagnosis method, device, computer equipment and storage medium |
CN111597996A (en) * | 2020-05-18 | 2020-08-28 | 中车永济电机有限公司 | Method for constructing wind turbine generator bearing fault identification model based on deep learning |
CN111597996B (en) * | 2020-05-18 | 2022-03-15 | 中车永济电机有限公司 | Method for constructing wind turbine generator bearing fault identification model based on deep learning |
CN111798975A (en) * | 2020-06-01 | 2020-10-20 | 山东师范大学 | Disease diagnosis system, apparatus and medium based on cyclic time convolution network |
CN111803059A (en) * | 2020-06-30 | 2020-10-23 | 武汉中旗生物医疗电子有限公司 | Electrocardiosignal classification method and device based on time domain convolution network |
CN112051481A (en) * | 2020-08-12 | 2020-12-08 | 华中科技大学 | Alternating current-direct current hybrid power grid fault area diagnosis method and system based on LSTM |
CN112016251B (en) * | 2020-09-02 | 2023-01-31 | 哈尔滨工程大学 | Nuclear power device fault diagnosis method and system |
CN112016251A (en) * | 2020-09-02 | 2020-12-01 | 哈尔滨工程大学 | Nuclear power device fault diagnosis method and system |
CN112729529A (en) * | 2020-12-17 | 2021-04-30 | 苏州大学 | Motor defect detection method |
CN112729529B (en) * | 2020-12-17 | 2023-02-03 | 苏州大学 | Motor defect detection method |
CN112348124A (en) * | 2021-01-05 | 2021-02-09 | 北京航空航天大学 | Data-driven micro fault diagnosis method and device |
CN112348124B (en) * | 2021-01-05 | 2021-04-20 | 北京航空航天大学 | Data-driven micro fault diagnosis method and device |
CN113962256A (en) * | 2021-09-29 | 2022-01-21 | 西安交通大学 | Intelligent fault diagnosis method and system for electromechanical actuator |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110596506A (en) | Converter fault diagnosis method based on time convolution network | |
CN108444708B (en) | Method for establishing rolling bearing intelligent diagnosis model based on convolutional neural network | |
CN106769052B (en) | A kind of mechanical system rolling bearing intelligent failure diagnosis method based on clustering | |
CN111914883B (en) | Spindle bearing state evaluation method and device based on deep fusion network | |
CN109460618A (en) | A kind of rolling bearing remaining life on-line prediction method and system | |
CN113870260B (en) | Welding defect real-time detection method and system based on high-frequency time sequence data | |
CN113158814B (en) | Bearing health state monitoring method based on convolution self-encoder | |
CN111562108A (en) | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC | |
CN110110768A (en) | Fault Diagnosis of Roller Bearings based on Concurrent Feature study and multi-categorizer | |
CN109389325B (en) | Method for evaluating state of electronic transformer of transformer substation based on wavelet neural network | |
CN109472226A (en) | A kind of sleep behavioral value method based on deep learning | |
CN109359665A (en) | A kind of family's electric load recognition methods and device based on support vector machines | |
CN112305388B (en) | On-line monitoring and diagnosing method for insulation partial discharge faults of generator stator winding | |
CN109828168A (en) | Converter method for diagnosing faults based on Density Estimator | |
CN112596016A (en) | Transformer fault diagnosis method based on integration of multiple one-dimensional convolutional neural networks | |
CN109670549A (en) | The data screening method, apparatus and computer equipment of fired power generating unit | |
CN115238785A (en) | Rotary machine fault diagnosis method and system based on image fusion and integrated network | |
CN114492642A (en) | Mechanical fault online diagnosis method for multi-scale element depth residual shrinkage network | |
CN114330486A (en) | Power system bad data identification method based on improved Wasserstein GAN | |
CN114048958A (en) | Cutter wear prediction method and system based on deep convolution residual shrinkage network | |
CN113642255A (en) | Photovoltaic power generation power prediction method based on multi-scale convolution cyclic neural network | |
CN117056865A (en) | Method and device for diagnosing operation faults of machine pump equipment based on feature fusion | |
CN110320802B (en) | Complex system signal time sequence identification method based on data visualization | |
CN117475191A (en) | Bearing fault diagnosis method for feature alignment domain antagonistic neural network | |
CN116244596A (en) | Industrial time sequence data anomaly detection method based on TCN and attention mechanism |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191220 |