CN114895647A - Small-sample ship part fault data-oriented diagnosis method and readable storage medium - Google Patents
Small-sample ship part fault data-oriented diagnosis method and readable storage medium Download PDFInfo
- Publication number
- CN114895647A CN114895647A CN202210383160.3A CN202210383160A CN114895647A CN 114895647 A CN114895647 A CN 114895647A CN 202210383160 A CN202210383160 A CN 202210383160A CN 114895647 A CN114895647 A CN 114895647A
- Authority
- CN
- China
- Prior art keywords
- data
- fault
- small
- sample
- diagnosis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a small-sample ship part fault data-oriented diagnosis method and a readable storage medium. Step 1: carrying out fault type marking on vibration acceleration signal data of the part in the running state process of the ship equipment; step 2: preprocessing the fault type labels in the step 1, and dividing a processed data set into a training set and a test set; and step 3: constructing a diagnosis model for fault data of the small-sample ship parts; and 4, step 4: using the training set in the step 2 for training the diagnostic model in the step 3 to obtain the generation data of the fault type; and 5: adopting a countermeasure generation idea based on the generated data in the step 4 to realize the fault data with the specified type of fault characteristics; step 6: and (5) realizing the diagnosis of the faults of the ship parts based on the combination of the generated data in the step (4) and the fault data in the step (5). The method is used for solving the problem of small sample class imbalance easily caused by the fact that the fault state is an accidental state in the prior art.
Description
Technical Field
The invention belongs to the field of fault diagnosis; in particular to a diagnosis method for fault data of small sample ship parts and a readable storage medium.
Background
With the continuous development of ship technology, maritime trading activities are more frequent. At the present stage, ship carriers with various functions are gradually popularized, and the accompanying problem of ship safety gradually comes into the visual field of people. The sinking of the luxury passenger ship 'tympany number' in the UK is almost a known catastrophic accident, and the 'maritime life safety convention' (SOLAS) is also made after the accident. Shipping is used as a main mode of the water transportation industry, the operation mode of the shipping is different from land transportation, and different accident environment backgrounds are generated, under the condition, the cost behind the accident occurrence is not only property loss, but also life threat, each accident should not be just a serious and weak thing, and experience teaching and training should be summarized, so that the system is perfected, equipment is improved, and management is enhanced to gradually improve the safety guarantee of the ship. And over 26,000 common marine accidents reported over the decade, over one third are caused by mechanical damage or failure, twice as many as the second leading cause of a marine collision. In 2019, 1044 accidents happen to ships with the weight of more than 100 total tons, and the mechanical faults account for more than one third of the total number of all accidents.
The effective ship part-level fault diagnosis method can improve the reliability of equipment, reduce the maintenance cost of a complex equipment system, improve the safety of ships and shipping industry and reduce corresponding risks. Meanwhile, fault diagnosis is realized on the parts through data change, and a manual complex processing mode is also avoided. Therefore, the research of the ship part-level fault diagnosis method has important theoretical significance and practical significance for realizing precision guarantee, improving guarantee efficiency and ship safety.
The deep learning aspect comprises more relatively mature basic model frameworks, and the current research situations of the method in fault diagnosis are mainly two types: the first is as a method of feature extraction and recognition, and the second is as a classification method. The deep learning method in any mode has wide application in the technical field and achieves certain research results. Nakamura et al, which adopts a deep learning-based diagnostic method, apply a long-term and short-term memory network to the faults of the equipment generator, thereby achieving a better diagnostic effect. Feng et al propose a rotating machine fault diagnosis method based on a deep learning network. The method combines the signal frequency spectrum with deep learning, and excavates information fed back in the signal frequency spectrum of the equipment through a deep network so as to realize the analysis of the equipment state. Wang et al used a particle swarm optimization algorithm for the parameter setting part in the model construction, and showed a superior effect in the part-level fault diagnosis.
In summary, although deep learning has achieved a better diagnosis effect in the field of fault diagnosis, it is important to use a small-sample ship component fault data diagnosis method, because the problem of small-sample type imbalance, which is likely to occur when a ship component is in a normal working state most of the time, and the fault state is in a sporadic state, directly affects the fault diagnosis result.
Disclosure of Invention
The invention provides a small-sample ship part fault data-oriented diagnosis method and a readable storage medium, which are used for solving the problem of small-sample unbalance easily caused by an accidental fault state in the prior art.
The invention is realized by the following technical scheme:
a small-sample ship part fault data oriented diagnosis method comprises the following steps:
step 1: carrying out fault type marking on vibration acceleration signal data of the part in the running state process of the ship equipment;
step 2: preprocessing the fault type labels in the step 1, and dividing a processed data set into a training set and a test set;
and step 3: constructing a diagnosis model for fault data of the small-sample ship parts;
and 4, step 4: using the training set in the step 2 for training the diagnostic model in the step 3 to obtain the generation data of the fault type;
and 5: adopting a countermeasure generation idea based on the generated data in the step 4 to realize the fault data with the specified type of fault characteristics;
step 6: and (5) realizing the diagnosis of the faults of the ship parts based on the combination of the generated data in the step (4) and the fault data in the step (5).
A small-sample ship part fault data oriented diagnosis method comprises the step 2 of dividing a preprocessed data set into a training set and a testing set according to the proportion of 7: 3.
A small-sample ship part fault data-oriented diagnosis method is disclosed, wherein the step 5 fault type generation data specifically comprises the following steps:
step 5.1: transmitting the real fault data as an initial input signal to a discriminator;
step 5.2: training the discriminator to enable the discriminator to master real data characteristics, so that real data can be distinguished from generated data;
step 5.3: the randomly generated noise is initially input to the generator,
step 5.4: causing the generator to generate random fault data and types;
step 5.5: and (4) carrying out countermeasure training on the discriminator trained in the step (5.2) and the generator in the step (5.4), and continuing training until similar generated fault data is obtained.
A diagnosis method for fault data of small sample ship parts is disclosed, wherein the step 5.2 is specifically that fault characteristics in the data and correlation between the fault characteristics and real data labels are obtained through a multilayer convolution neural network layer in a discriminator, and the fault characteristics and the labels are respectively output through a full connection layer; and comparing and judging fault data and tags mixed with the subsequent real data and the generated data.
5.4, enabling a generator to generate random fault data and types, namely, after the generator receives the data through an input layer, generating the fault data with the same size as the original data through four layers of one-dimensional convolutional neural network layers; the sizes of the data generated by the adjustment of the upper sampling layer among the convolutional neural networks of each layer gradually approach to the original data.
The step 5.5 is that the generator can continuously optimize and generate fault data which is more similar to real data according to the judgment of the discriminator, and the discriminator continuously improves the discrimination capability so as to find whether the data is the real data or the generated data.
A diagnosis method for fault data of small-sample ship parts is characterized in that a diagnosis part of a fault diagnosis model of the ship parts in step 4 specifically comprises the following steps:
step 6.1: combining the fault data generated in the step 5.4 with the real fault data in the training set in the step 2 as data input;
step 6.2: inputting data, namely combining the data in the step 6.1 to finish the feature extraction of the original data signal;
step 6.3: based on the feature extraction of the original data signal in the step 6.2, the diagnosis work of the equipment fault by the equipment fault signal is completed;
step 6.4: and 6.3, judging the fault type according to the extracted fault characteristics to obtain a stable fault diagnosis model.
6.2, extracting the characteristics of the original data signal, namely, completing the characteristic extraction of the original data signal by three one-dimensional convolutional neural network layers and corresponding maximum pooling layers and then connecting a long-term and short-term memory network layer; the initial multilayer one-dimensional convolutional neural network layer extracts the spatial characteristics of the fault signals, the temporal characteristics are extracted by the subsequent long-term and short-term memory network layer, and the characteristics of the fault signals are extracted more fully from multiple angles.
A diagnosis method for fault data of small-sample ship parts is characterized in that 6.3, the diagnosis of equipment faults through equipment fault signals is completed, namely, extracted features are received by a full connection layer and connected with the full connection layer responsible for classification.
A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method steps of the above.
The invention has the beneficial effects that:
in the invention, on the basis of basic fault diagnosis in the fault diagnosis of ship parts, the problems of excessive normal data volume and less fault data volume in actual production life are considered, and the countermeasure generation idea is utilized to capture one-dimensional fault type data characteristics to generate fault data highly similar to actual damage, so that a better fault diagnosis result can be obtained when the ship parts are oriented to small samples.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Figure 2 is a generator framework diagram of the present invention.
FIG. 3 is a diagram of the structure of the discriminator of the present invention.
FIG. 4 is a diagram showing a structure of a model diagnosis part of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention solves the problem that the occasional small sample fault affects the fault diagnosis result. As shown in figure 1 of the drawings, in which,
a small-sample ship part fault data oriented diagnosis method comprises the following steps:
step 1: carrying out fault type marking on vibration acceleration signal data of the part in the running state process of the ship equipment;
step 2: preprocessing the fault type labels in the step 1, and dividing a processed data set into a training set and a test set;
and step 3: constructing a diagnosis model for fault data of the small-sample ship parts;
and 4, step 4: using the training set in the step 2 for training the diagnostic model in the step 3 to obtain the generation data of the fault type;
and 5: adopting a countermeasure generation idea based on the generated data in the step 4 to realize the fault data with the specified type of fault characteristics;
step 6: and (5) realizing the diagnosis of the faults of the ship parts based on the combination of the generated data in the step (4) and the fault data in the step (5).
A small-sample ship part fault data oriented diagnosis method comprises the step 2 of dividing a preprocessed data set into a training set and a testing set according to the proportion of 7: 3.
A small-sample ship part fault data-oriented diagnosis method is disclosed, wherein the step 5 fault type generation data specifically comprises the following steps:
step 5.1: transmitting the real fault data as an initial input signal to a discriminator;
step 5.2: training the discriminator to enable the discriminator to master real data characteristics, so that real data can be distinguished from generated data;
step 5.3: the randomly generated noise is initially input to the generator,
step 5.4: causing the generator to generate random fault data and types;
step 5.5: and (4) carrying out countermeasure training on the discriminator trained in the step (5.2) and the generator in the step (5.4), and continuing training until similar generated fault data is obtained.
A diagnosis method for fault data of small-sample ship parts is provided, and the step 5.2 is specifically as shown in FIG. 2, obtaining fault characteristics in data and correlation between the fault characteristics and real data labels through a multilayer convolutional neural network layer in a discriminator, and respectively outputting the fault characteristics and the labels through a full connection layer; and comparing and judging fault data and tags mixed with the subsequent real data and the generated data.
5.4, enabling a generator to generate random fault data and types, namely, as shown in fig. 3, after the generator receives the data through an input layer, generating the fault data with the same size as the original data through four layers of one-dimensional convolutional neural network layers; the sizes of the data generated by the adjustment of the upper sampling layer among the convolutional neural networks are gradually approximated to the original data;
inputting the generated random noise n and randomly endowing the random noise n with a corresponding fault signal class label to a generator, and adding a generated data label (X) to the generated false data except the fault class label generate ＝G(label,n)。
The step 5.5 is that the generator can continuously optimize and generate fault data which is more similar to real data according to the judgment of the discriminator, and the discriminator continuously improves the discrimination capability so as to find whether the data is the real data or the generated data.
Preferably, the arbiter optimizes the loss function L of the true and false labels during the countermeasure training in step 5.5 S ＝E[logP(S＝real|X real )]+E[logP(S＝generate|X generate )]And fault data classification loss function L C ＝E[logP(C＝c|X real )]+E[logP(C＝c|X generate )]。
Wherein the loss functions of the arbiter and the generator for generating the countermeasure network in the model are respectively L Discriminator ＝L S +L C And L Generator ＝L S -L C And obtaining a better training model by maximizing the two.
A diagnosis method for fault data of small-sample ship parts is characterized in that a diagnosis part of a fault diagnosis model of the ship parts in step 4 specifically comprises the following steps:
step 6.1: combining the fault data generated in the step 5.4 with the real fault data in the training set in the step 2 as data input;
step 6.2: inputting data, namely combining the data in the step 6.1 to finish the feature extraction of the original data signal;
step 6.3: based on the feature extraction of the original data signal in the step 6.2, the diagnosis work of the equipment fault by the equipment fault signal is completed;
step 6.4: and 6.3, judging the fault type according to the extracted fault characteristics to obtain a stable fault diagnosis model.
A diagnosis method for fault data of small-sample ship parts is disclosed, as shown in FIG. 4, the step 6.2 of extracting the characteristics of original data signals specifically comprises the steps of passing through three one-dimensional convolutional neural network layers and corresponding maximum pooling layers, and then connecting a long-short term memory network layer to complete the characteristic extraction of the original data signals; the initial multilayer one-dimensional convolutional neural network layer extracts the spatial characteristics of the fault signals, the temporal characteristics are extracted by the subsequent long-term and short-term memory network layer, and the characteristics of the fault signals are extracted more fully from multiple angles.
A diagnosis method for fault data of small-sample ship parts is characterized in that 6.3, the diagnosis of equipment faults through equipment fault signals is completed, namely, extracted features are received by a full connection layer and connected with the full connection layer responsible for classification.
The step 6.4 is specifically that since the fault often includes multiple types, in the case of multiple classifications, if the data sample is (x, y), where x is the sample and y is the label of the sample, the prediction result is not between two values of {0, 1} but is a set of K classification label values, and the loss function is the set of K classification label values
And finally, training to obtain a better model.
A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method steps of the preceding description.
Claims (10)
1. A small-sample ship part fault data oriented diagnosis method is characterized by comprising the following steps:
step 1: carrying out fault type marking on vibration acceleration signal data of the part in the running state process of the ship equipment;
step 2: preprocessing the fault type labels in the step 1, and dividing a processed data set into a training set and a test set;
and step 3: constructing a diagnosis model for fault data of the small-sample ship parts;
and 4, step 4: using the training set in the step 2 for training the diagnostic model in the step 3 to obtain the generation data of the fault type;
and 5: adopting a countermeasure generation idea based on the generated data in the step 4 to realize the fault data with the specified type of fault characteristics;
step 6: and (5) realizing the diagnosis of the faults of the ship parts based on the combination of the generated data in the step (4) and the fault data in the step (5).
2. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 1, wherein the step 2 is to divide the preprocessed data set into a training set and a testing set according to a ratio of 7: 3.
3. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 1, wherein the step 5 fault type generation data specifically comprises the following steps:
step 5.1: transmitting the real fault data as an initial input signal to a discriminator;
step 5.2: training the discriminator to enable the discriminator to master real data characteristics, so that real data can be distinguished from generated data;
step 5.3: the randomly generated noise is initially input to the generator,
step 5.4: causing the generator to generate random fault data and types;
step 5.5: and (4) carrying out countermeasure training on the discriminator trained in the step (5.2) and the generator in the step (5.4), and continuing training until similar generated fault data is obtained.
4. The small-sample-oriented ship part fault data diagnosis method as claimed in claim 3, wherein the step 5.2 is specifically that fault features in the data and the correlation between the fault features and real data labels are obtained through a multilayer convolutional neural network layer in a discriminator, and the fault features and the labels are respectively output through a full connection layer; and comparing and judging fault data and tags mixed with the subsequent real data and the generated data.
5. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 3, wherein the step 5.4 is specifically that after the generator receives the data through an input layer, the generator generates the fault data with the same size as the original data through four layers of one-dimensional convolutional neural network layers; the sizes of the data generated by the adjustment of the upper sampling layer among the convolutional neural networks of each layer gradually approach to the original data.
6. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 3, wherein the step 5.5 is to optimize and generate the fault data more similar to the real data continuously according to the judgment of the discriminator, and the discriminator improves the discrimination continuously to find out whether the data is the real data or the generated data.
7. The method for diagnosing the fault data of the small-sample ship part as claimed in claim 3, wherein the step 4 of diagnosing the fault data of the ship part specifically comprises the following steps:
step 6.1: combining the fault data generated in the step 5.4 with the real fault data in the training set in the step 2 as data input;
step 6.2: inputting data, namely combining the data in the step 6.1 to finish the feature extraction of the original data signal;
step 6.3: based on the feature extraction of the original data signal in the step 6.2, the diagnosis work of the equipment fault by the equipment fault signal is completed;
step 6.4: and 6.3, judging the fault type according to the extracted fault characteristics to obtain a stable fault diagnosis model.
8. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 7, wherein the step 6.2 is to extract the features of the original data signals by connecting a long-term and short-term memory network layer after passing through three one-dimensional convolutional neural network layers and corresponding maximum pooling layers; the initial multilayer one-dimensional convolutional neural network layer extracts the spatial characteristics of the fault signals, the temporal characteristics are extracted by the subsequent long-term and short-term memory network layer, and the characteristics of the fault signals are extracted more fully from multiple angles.
9. The method for diagnosing the fault data of the small-sample ship parts as claimed in claim 7, wherein the step 6.3 is implemented for diagnosing the fault of the equipment by the fault signal of the equipment by receiving the extracted features from the full-connection layer and connecting the extracted features with the full-connection layer responsible for classification.
10. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210383160.3A CN114895647A (en) | 2022-04-13 | 2022-04-13 | Small-sample ship part fault data-oriented diagnosis method and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210383160.3A CN114895647A (en) | 2022-04-13 | 2022-04-13 | Small-sample ship part fault data-oriented diagnosis method and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114895647A true CN114895647A (en) | 2022-08-12 |
Family
ID=82717034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210383160.3A Pending CN114895647A (en) | 2022-04-13 | 2022-04-13 | Small-sample ship part fault data-oriented diagnosis method and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114895647A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115169506A (en) * | 2022-09-06 | 2022-10-11 | 中铁第四勘察设计院集团有限公司 | Method and system for rapidly diagnosing faults of power supply and transformation key equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112039687A (en) * | 2020-07-14 | 2020-12-04 | 南京邮电大学 | Small sample feature-oriented fault diagnosis method based on improved generation countermeasure network |
| CN112052902A (en) * | 2020-04-16 | 2020-12-08 | 北京信息科技大学 | Rolling bearing fault diagnosis method, system, computer program and storage medium |
| CN112396088A (en) * | 2020-10-19 | 2021-02-23 | 西安交通大学 | Intelligent diagnosis method for mechanical fault of implicit excitation countertraining under small sample |
| CN112649198A (en) * | 2021-01-05 | 2021-04-13 | 西交思创智能科技研究院(西安)有限公司 | Intelligent fault diagnosis method, system and equipment for quasi-unbalanced rolling bearing and application |
| CN114298267A (en) * | 2021-10-13 | 2022-04-08 | 华南理工大学 | Fault diagnosis method based on bidirectional attention generation countermeasure network and application |
-
2022
- 2022-04-13 CN CN202210383160.3A patent/CN114895647A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112052902A (en) * | 2020-04-16 | 2020-12-08 | 北京信息科技大学 | Rolling bearing fault diagnosis method, system, computer program and storage medium |
| CN112039687A (en) * | 2020-07-14 | 2020-12-04 | 南京邮电大学 | Small sample feature-oriented fault diagnosis method based on improved generation countermeasure network |
| CN112396088A (en) * | 2020-10-19 | 2021-02-23 | 西安交通大学 | Intelligent diagnosis method for mechanical fault of implicit excitation countertraining under small sample |
| CN112649198A (en) * | 2021-01-05 | 2021-04-13 | 西交思创智能科技研究院(西安)有限公司 | Intelligent fault diagnosis method, system and equipment for quasi-unbalanced rolling bearing and application |
| CN114298267A (en) * | 2021-10-13 | 2022-04-08 | 华南理工大学 | Fault diagnosis method based on bidirectional attention generation countermeasure network and application |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115169506A (en) * | 2022-09-06 | 2022-10-11 | 中铁第四勘察设计院集团有限公司 | Method and system for rapidly diagnosing faults of power supply and transformation key equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109615019B (en) | Abnormal behavior detection method based on space-time automatic encoder | |
| CN110298374B (en) | Driving track energy consumption analysis method and device based on deep learning | |
| CN106958484B (en) | Fleet's diesel engine remote state monitoring system and monitoring method based on Beidou satellite communication | |
| CN114895647A (en) | Small-sample ship part fault data-oriented diagnosis method and readable storage medium | |
| CN109712127A (en) | A kind of electric transmission line fault detection method for patrolling video flowing for machine | |
| CN103969052A (en) | Diesel engine fault diagnosis method based on outlier analysis | |
| CN112557826A (en) | Ship electric power system fault diagnosis method | |
| CN115114965B (en) | Wind turbine generator gearbox fault diagnosis method, device, equipment and storage medium | |
| CN113222108A (en) | Target detection processing method, device and equipment | |
| CN113641486B (en) | Intelligent turnout fault diagnosis method based on edge computing network architecture | |
| CN115061159A (en) | Satellite navigation positioning deception signal defense method | |
| CN105893941A (en) | Facial expression identifying method based on regional images | |
| CN113012107A (en) | Power grid defect detection method and system | |
| CN115099264A (en) | Ship part fault diagnosis method and device, computer and computer storage medium | |
| CN112948969B (en) | Ship rolling prediction method based on LSTMC hybrid network | |
| CN111340149B (en) | Excavator real-time detection method and system based on background difference method and deep learning | |
| Li et al. | MorseNet: a unified neural network for morse detection and recognition in spectrogram | |
| CN114818501A (en) | Light weight method for solar cell defect detection based on data enhancement | |
| CN113449618A (en) | Method for carrying out deep learning rolling bearing fault diagnosis based on feature fusion and mixed enhancement | |
| CN113554611A (en) | Insulator self-explosion defect detection method and device, terminal and storage medium | |
| CN115267462B (en) | Partial discharge type identification method based on self-adaptive label generation | |
| Arbon et al. | Anomaly detection in satellite communications networks using support vector machines | |
| Ma et al. | Bearing fault diagnosis based on attention mechanism and deep residual network | |
| CN117052607A (en) | Marine wind power generator fault detection system | |
| CN118051815A (en) | Method for detecting and restoring countermeasure sample for identifying modulation signal |
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 |