CN114818904A - Fan fault detection method based on Stack-GANs model and storage medium - Google Patents
Fan fault detection method based on Stack-GANs model and storage medium Download PDFInfo
- Publication number
- CN114818904A CN114818904A CN202210422554.5A CN202210422554A CN114818904A CN 114818904 A CN114818904 A CN 114818904A CN 202210422554 A CN202210422554 A CN 202210422554A CN 114818904 A CN114818904 A CN 114818904A
- Authority
- CN
- China
- Prior art keywords
- data
- generator
- formula
- stack
- fan
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 208000009119 Giant Axonal Neuropathy Diseases 0.000 title claims abstract description 28
- 238000001514 detection method Methods 0.000 title claims abstract description 28
- 201000003382 giant axonal neuropathy 1 Diseases 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 15
- 238000007637 random forest analysis Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 abstract description 12
- 238000012545 processing Methods 0.000 abstract description 7
- 230000000694 effects Effects 0.000 abstract description 4
- 238000004804 winding Methods 0.000 description 10
- 238000003066 decision tree Methods 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 6
- 238000001816 cooling Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000010720 hydraulic oil Substances 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Economics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a fan fault detection method based on a Stack-GANs model, which comprises the steps of forming a fan data set by working condition parameters of a wind turbine generator, processing the fan data set by the Stack-GANs model, and inputting the processed fan data set into a fan fault detection model to obtain a detection result. The invention utilizes the effectiveness of the Stack-GANs algorithm in processing the problem of unbalanced fan data, reduces the negative effect of unbalanced data set categories on fan fault detection, and improves the efficiency and accuracy of fan fault detection.
Description
Technical Field
The invention relates to the technical field of computer and fan fault detection, in particular to a fan fault detection method and a storage medium based on a Stack-GANs model.
Background
The data processing and detection of the wind turbine generator faults have great significance for improving the reliability of the wind turbine generator in power generation and reducing the operation and maintenance cost. In addition to the traditional detection method, an artificial intelligence method is also used at present, and the method is characterized in that various data during the operation of the fan are collected to be manufactured into a data set, and the data set is used for training a model to finally achieve the purpose of fan fault detection. The method comprises the following steps of firstly, acquiring a fault data set of the wind turbine generator, and then, acquiring a fault data set of the wind turbine generator, wherein the fault data set of the wind turbine generator is a very critical step, and in the actual data set, the proportion of the number of the normal operation data and the number of the fault data of the wind turbine generator is very unbalanced, so that the next model training can be adversely affected, and further, the detection efficiency of the fault of the wind turbine generator is influenced.
Disclosure of Invention
The invention provides a fan fault detection method based on a Stack-GANs model, which utilizes the effectiveness of a Stack-GANs algorithm in processing the problem of fan data imbalance, reduces the negative effect of the imbalance of data set categories on fan fault detection, and improves the efficiency and accuracy of fan fault detection.
The method comprises a fan data set formed by working condition parameters of a wind turbine generator, wherein the fan data set usually takes part or all of 18 parameters such as U2 winding current, U3 winding current, U1 winding voltage, U2 winding voltage, U3 winding voltage, variable pitch distance, hydraulic oil temperature, generator cooling temperature, generator slip ring temperature, generator rotating speed, impeller rotating speed, gearbox blade temperature, gearbox filter pressure, wind speed 1, wind speed 2, wind direction 1, wind direction 2, cabinet temperature and the like as elements, and the elements are processed by a Stack-GANs model and then input into a general fan fault detection model to obtain a detection result.
The construction of the Stack-GANs model comprises the following steps:
(1) screening the characteristics in the fan data set according to the importance degree of the working condition parameters of the wind turbine generator by using a random forest method to obtain a new characteristic subset, and extracting a minority of characteristic data sets P (P) { P } from the new characteristic subset 1 ,p 2 ,…,p i ,…,p m }。
(2) Dividing the minority class feature data set P by using a Pearson correlation coefficient method and an MIC analysis method to respectively obtain a Group1 feature data subsetAnd Group2 feature data subsets
(3) Using the Group1 and Group2 feature data subsets for training GANs1 and GANs2, respectively, and training a Discrimatoror and a Generator by alternately using formula (a) and formula (b) using random noise as input in the training process; wherein the content of the first and second substances,
in the formula: p (x) represents a subset of feature data P 1 And P 2 Distribution of (2), x t Representing a subset of feature data P 1 And P 2 Real sample data at time t, p (z) represents the distribution of random noise input into the Generator, z t Denotes random noise data input into the Generator at time t, D denotes a Discriminator, G denotes the Generator,andrepresenting the hidden state values at the last time of RNN in D and G, respectively.
(4) After the model training of step (3) is finished, two discriminators are removed, and the corresponding generor 1 and generor 2 are reserved.
(5) Generating corresponding grouped data groups 3 and 4 by using the Generator1 and the Generator2, and splicing the grouped data groups 3 and 4 to form a data Group which is coarse sample data.
(6) Training a Discriminator and a Generator by alternately using formula (c) and formula (d) with the coarse sample data as input; wherein the content of the first and second substances,
the formula (c) is:
in the formula: g1 and G2 represent Generator1 and Generator2, p, respectively, retained in step (4) A (x) Representing the distribution, x, of a few classes of feature data sets P t A Representing the true sample data at time t in the minority class feature data set P, P G1 (z) and p G2 (z) represents the distribution of random noise input into G1 and G2, respectively, z t G1 And z t G2 Random noise data in G1 and G2 at time t are shown, respectively, D denotes a Discriminator, G denotes a Generator,andrepresenting the hidden state values at the last time of RNN in D and G, respectively.
(7) After the training of step (6) is finished, the Discrimidator is removed and the corresponding Generator3 is reserved.
(8) The construction of the Stack-GANs model is completed based on reserved Generator1, Generator2 and Generator 3.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the above described method of fan fault detection.
Drawings
FIG. 1 is a flow chart of feature screening using random forests in accordance with the present invention;
FIG. 2 is a diagram of a Stack-GANs model method construction provided by the present invention;
FIG. 3 is a network structure diagram of a Discrimatoror and Generator in the present invention;
fig. 4a and 4b are comparison graphs of minority class samples generated by the present invention and real minority class samples.
Detailed Description
A specific implementation example is given below to explain the technical solution of the present invention in detail, the data set used in this embodiment is derived from historical operating data of a certain wind farm in Yunnan, and the data set is divided, where a training set includes 315600 pieces of data, a test set includes 20190 pieces of data, and the entire data set includes 30 features.
(1) In order to reduce the difficulty of subsequent model construction and avoid generating unnecessary data characteristics, the invention firstly carries out characteristic importance sequencing and screening on a data set by adopting a Random Forest (RF) method, wherein the RF is an algorithm comprising a plurality of decision trees and has the advantages of high accuracy and good stability. The specific flow of feature importance selection by RF is shown in fig. 1. The detailed steps are as follows:
first, N new data sets may be formed by randomly taking a certain number of data samples from the original data set with a put-back sample to form a new data set, and repeating the process N times. And randomly selecting a plurality of features from each new data set to form feature subsets, constructing a corresponding decision tree by using each feature subset, and regarding each decision tree, remaining a part of feature data subsets which do not participate in the construction of the decision tree, and referring the feature subsets which are not utilized as the data outside the bag of the corresponding decision tree.
Secondly, when calculating the importance score for the feature x, all decision trees constructed by the feature x are used as a base model, and the corresponding data outside the bag is used as test data to calculate an Error value Error 1.
Random noise is then added to the features x of all samples in the out-of-bag data, and an Error value Error2 is calculated again for the noisy out-of-bag data using the corresponding decision tree.
Finally, the importance score of the feature x is calculated by the formula (e).
Wherein N represents the number of decision trees in which the features x participate.
After the data set is subjected to RF feature importance sorting and screening, 18 important features are reserved, including: the method comprises the following steps of U2 winding current, U3 winding current, U1 winding voltage, U2 winding voltage, U3 winding voltage, pitch distance, hydraulic oil temperature, generator cooling temperature, generator slip ring temperature, generator rotating speed, impeller rotating speed, gearbox blade temperature, gearbox filter pressure, wind speed 1, wind speed 2, wind direction 1, wind direction 2 and cabinet temperature, wherein specific conditions are shown in Table 1.
TABLE 1 Fan characteristics after RF Algorithm screening
And (3) performing correlation analysis on the data set after the characteristic screening by using a Pearson correlation coefficient method and an MIC analysis method, and finally selecting the threshold values of the Pearson correlation coefficient method and the MIC analysis method as 0.6 and 0.7 respectively through multiple experiments. In order to make both linear and non-linear relationships between features strongly correlated, features are grouped together when both the Pearson coefficient value and the MIC value between the features are greater than the above threshold values. The final features are divided into two groups, Group1 and Group2, as shown in table 2.
TABLE 2 Fan characteristic grouping
(2) The invention is based on the idea of the GANs, a Stack-GANs model is built to generate a few types of sample data, so that the problem of data imbalance is solved, the overall model construction is shown in figure 2, the main steps of the model construction are described in the invention content, more specifically, for the Generator and the Discrimator for constructing the GANs, the time sequence characteristics in the data are mined by adopting an RNN network structure model considering that the wind turbine data has time sequence, and the specific structure is shown in figure 3.
(3) Referring to fig. 4a and 4b, the comparison between the mean value and the standard deviation of 18 characteristic variables in the minority class data generated by the Stack-GANs model and the real minority class data can be seen, and the situation shows that: in most features, the mean and the standard deviation of the generated few types of sample data are relatively close to those of the real few types of sample data. This shows that the minority class sample data generated by the Stack-GANs method has similarity with the real minority class sample data, and the method and the model provided by the invention can generate the minority class sample data with high quality.
(4) Furthermore, a comparative experiment is carried out on the Stack-GANs model-based processing method provided by the invention and other data imbalance processing methods. In the experiment, in order to eliminate the influence of random factors on each fan fault detection model, 10 experiments are carried out for each algorithm, the average values of F1-Score, G-mean and AUC are calculated, and the experiment results of each data imbalance processing algorithm are shown in Table 3.
TABLE 3 comparison of F1, G-mean, and AUC values of the data imbalance processing algorithms (unit:%)
The results of the experiments in the analysis table can be concluded as follows: compared with an unprocessed data set, the data set processed by the Stack-GANs method can improve the comprehensive performance of the fan fault detection model, and compared with other algorithms, the Stack-GAN method obtains the optimal value of the evaluation index on each fan fault detection model. Secondly, in each fan fault detection model, the BSMOTE-Sequence algorithm has a slightly poorer improvement effect on the model than Stack-GANs, but is better than other algorithms, and probably because the BSMOTE-Sequence algorithm considers the time Sequence characteristics among data but does not consider the correlation reasons among characteristics when synthesizing a few types of samples. The model trained by using the data set sampled by the TomekLink has poor comprehensive effect because the TomekLink is an undersampling technology, and certain important information is lost when a sample is deleted, so that the performance of the model is reduced.
The effectiveness of the Stack-GANs method in processing the fan data imbalance problem is proved through the experimental results, the correlation and the time sequence characteristics of the data characteristics are comprehensively considered when a few new samples are synthesized, the generated data are more real, the model can better learn the distribution of the fault characteristics, and therefore the overall performance of the model is improved.
The invention has the technical characteristics and beneficial effects that: for the Stack-GANS model in the invention, important characteristics in fault data are screened out by using an RF algorithm, and the complexity of constructing the model is reduced; secondly, according to the idea of the GANs, the RNN is used for constructing the Discriminator and the Generator at each stage, and the Discriminator and the Generator can be used for capturing the time sequence characteristics among data; and finally, a Stack-GANS model is established by a progressive method, so that strong and weak correlation among characteristics can be considered, a high-quality fan fault data sample is generated, the problem of imbalance in fan fault data concentration is solved beneficially, and the efficiency and accuracy of fan fault detection are improved.
Claims (2)
1. A fan fault detection method based on a Stack-GANs model comprises a fan data set formed by working condition parameters of a wind turbine generator, and the fan data set is processed by the Stack-GANs model and then input into a fan fault detection model to obtain a detection result, and is characterized in that the construction of the Stack-GANs model comprises the following steps:
(1) screening the characteristics in the fan data set according to the importance degree of the working condition parameters of the wind turbine generator by using a random forest method to obtain a new characteristic subset, and extracting a minority of characteristic data sets P (P) { P } from the new characteristic subset 1 ,p 2 ,…,p i ,…,p m };
(2) Dividing the minority class feature data set P by using a Pearson correlation coefficient method and an MIC analysis method to respectively obtain a Group1 feature data subsetAnd Group2 feature data subsets
(3) Using the Group1 and Group2 feature data subsets for training GANs1 and GANs2, respectively, and training a Discrimatoror and a Generator by alternately using formula (a) and formula (b) using random noise as input in the training process; wherein the content of the first and second substances,
in the formula: p (x) represents a subset of feature data P 1 And P 2 Distribution of (2), x t Representing a subset of feature data P 1 And P 2 Real sample data at time t, p (z) represents the distribution of random noise input into the Generator, z t Denotes random noise data input into the Generator at time t, D denotes a Discriminator, G denotes the Generator,andrespectively representing hidden state values of RNN at the previous moment in D and G;
(4) after the model training in the step (3) is finished, removing two Discrimators, and reserving the corresponding Generator1 and Generator 2;
(5) generating corresponding grouped data groups 3 and 4 by using the Generator1 and the Generator2, and splicing the grouped data groups 3 and 4 to form a data Group which is coarse sample data;
(6) training a Discriminator and a Generator by alternately using formula (c) and formula (d) with the coarse sample data as input; wherein the content of the first and second substances,
the formula (c) is:
in the formula: g1 and G2 represent Generator1 and Generator2, p, respectively, retained in step (4) A (x) Representing the distribution, x, of a few classes of feature data sets P t A Representing the true sample data at time t in the minority class feature data set P, P G1 (z) and p G2 (z) represents the distribution of random noise input into G1 and G2, respectively, z t G1 And z t G2 Random noise data in G1 and G2 at time t are shown, respectively, D denotes a Discriminator, G denotes a Generator,andrespectively representing hidden state values of RNN at the previous moment in D and G;
(7) after the training in the step (6) is finished, removing the Discrimatoror and reserving the corresponding Generator 3;
(8) the construction of the Stack-GANs model is completed based on reserved Generator1, Generator2 and Generator 3.
2. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the steps of the Stack-GANs model based wind turbine fault detection method of claim 1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210422554.5A CN114818904B (en) | 2022-04-21 | 2022-04-21 | Fan fault detection method and storage medium based on Stack-GANs model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210422554.5A CN114818904B (en) | 2022-04-21 | 2022-04-21 | Fan fault detection method and storage medium based on Stack-GANs model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114818904A true CN114818904A (en) | 2022-07-29 |
CN114818904B CN114818904B (en) | 2024-03-15 |
Family
ID=82505516
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210422554.5A Active CN114818904B (en) | 2022-04-21 | 2022-04-21 | Fan fault detection method and storage medium based on Stack-GANs model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114818904B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110766175A (en) * | 2019-10-25 | 2020-02-07 | 长沙理工大学 | Pitch system fault detection method and device based on optimal interval distribution machine |
EP3620983A1 (en) * | 2018-09-05 | 2020-03-11 | Sartorius Stedim Data Analytics AB | Computer-implemented method, computer program product and system for data analysis |
US20200202221A1 (en) * | 2018-12-20 | 2020-06-25 | Shandong University Of Science And Technology | Fault detection method and system based on generative adversarial network and computer program |
CN111337243A (en) * | 2020-02-27 | 2020-06-26 | 上海电力大学 | ACGAN-based wind turbine generator planet wheel gearbox fault diagnosis method |
CN112633317A (en) * | 2020-11-02 | 2021-04-09 | 国能信控互联技术有限公司 | CNN-LSTM fan fault prediction method and system based on attention mechanism |
CN112801151A (en) * | 2021-01-18 | 2021-05-14 | 桂林电子科技大学 | Wind power equipment fault detection method based on improved BSMOTE-Sequence algorithm |
-
2022
- 2022-04-21 CN CN202210422554.5A patent/CN114818904B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3620983A1 (en) * | 2018-09-05 | 2020-03-11 | Sartorius Stedim Data Analytics AB | Computer-implemented method, computer program product and system for data analysis |
US20200202221A1 (en) * | 2018-12-20 | 2020-06-25 | Shandong University Of Science And Technology | Fault detection method and system based on generative adversarial network and computer program |
CN110766175A (en) * | 2019-10-25 | 2020-02-07 | 长沙理工大学 | Pitch system fault detection method and device based on optimal interval distribution machine |
CN111337243A (en) * | 2020-02-27 | 2020-06-26 | 上海电力大学 | ACGAN-based wind turbine generator planet wheel gearbox fault diagnosis method |
CN112633317A (en) * | 2020-11-02 | 2021-04-09 | 国能信控互联技术有限公司 | CNN-LSTM fan fault prediction method and system based on attention mechanism |
CN112801151A (en) * | 2021-01-18 | 2021-05-14 | 桂林电子科技大学 | Wind power equipment fault detection method based on improved BSMOTE-Sequence algorithm |
Non-Patent Citations (2)
Title |
---|
柳青秀;马红占;褚学宁;马斌彬;王峥;: "基于长短时记忆―自编码神经网络的风电机组性能评估及异常检测", 计算机集成制造系统, no. 12, 15 December 2019 (2019-12-15), pages 233 - 243 * |
陈宇韬;唐明珠;吴华伟;赵琪;匡子杰;: "基于极端随机森林的大型风电机组发电机故障检测", 湖南电力, no. 06, 25 December 2019 (2019-12-25), pages 49 - 55 * |
Also Published As
Publication number | Publication date |
---|---|
CN114818904B (en) | 2024-03-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111460956B (en) | Unbalanced electrocardiogram sample classification method based on data enhancement and loss weighting | |
CN110147760B (en) | Novel efficient electric energy quality disturbance image feature extraction and identification method | |
CN110570346A (en) | Method for performing style migration on calligraphy based on cyclic generation countermeasure network | |
CN114742097B (en) | Optimization method for automatically determining variation modal decomposition parameters based on bearing vibration signals | |
Zhang et al. | A method for modulation recognition based on entropy features and random forest | |
CN114548190B (en) | Wind turbine fault diagnosis method and device based on self-adaptive residual error neural network | |
CN114441173B (en) | Rolling bearing fault diagnosis method based on improved depth residual error shrinkage network | |
CN113466634A (en) | Ground fault waveform identification method based on fault indicator | |
CN116619136A (en) | Multi-working-condition multi-source data cutter abrasion prediction method | |
CN114004091A (en) | CEEMDAN-BNs-based wind variable pitch system fault diagnosis method | |
CN116071352A (en) | Method for generating surface defect image of electric power safety tool | |
CN117171544B (en) | Motor vibration fault diagnosis method based on multichannel fusion convolutional neural network | |
CN114818904B (en) | Fan fault detection method and storage medium based on Stack-GANs model | |
CN109635428A (en) | A kind of GIS mechanical failure diagnostic method based on the analysis of machine performance signal | |
CN114298136A (en) | Wind speed prediction method based on local mean decomposition and deep learning neural network | |
CN116735198A (en) | Fan gear box fault diagnosis method and system based on speEMD and double index screening | |
CN116304836A (en) | Small-sample semi-supervised bearing fault diagnosis method based on domain knowledge guidance | |
CN107784015B (en) | Data reduction method based on online historical data of power system | |
CN112464708B (en) | Double-fed asynchronous fan power quality abnormal fault diagnosis method | |
CN115017445A (en) | Propeller state identification method based on CEEMDAN combined approximate entropy | |
CN114301719A (en) | Malicious update detection method and model based on variational self-encoder | |
CN114781244A (en) | Grouping and parameter optimization method in wind power plant | |
CN109902007B (en) | Test case generation method based on point dyeing model | |
Ceylan et al. | GraphDCA--a Framework for Node Distribution Comparison in Real and Synthetic Graphs | |
CN116245022A (en) | Efficient construction method for health index of rolling bearing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |