CN113935460A - Intelligent diagnosis method for mechanical fault under class imbalance data set - Google Patents
Intelligent diagnosis method for mechanical fault under class imbalance data set Download PDFInfo
- Publication number
- CN113935460A CN113935460A CN202111136682.5A CN202111136682A CN113935460A CN 113935460 A CN113935460 A CN 113935460A CN 202111136682 A CN202111136682 A CN 202111136682A CN 113935460 A CN113935460 A CN 113935460A
- Authority
- CN
- China
- Prior art keywords
- data
- fault
- model
- mechanical
- training
- 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
Images
Classifications
-
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- 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
- G06N3/047—Probabilistic or stochastic networks
-
- 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
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to an intelligent diagnosis method for mechanical faults under a class unbalance data set, which comprises the following steps: step (1), data preprocessing: converting the mechanical vibration signal into a frequency domain, and normalizing the amplitude value to a [0,1] range; step (2), model building: combining an automatic encoder with a generated countermeasure network to build a data generation model; step (3), model training: training the data generation model by utilizing fault data according to a preset loss function and an optimization algorithm; step (4), data generation: utilizing the low-dimensional characteristics of fault data learned by the data generation model in training to generate corresponding fault data after multiple times of interpolation and noise addition so as to realize the balance of various types of data; step (5), fault diagnosis: and training a preset fault diagnosis model by using the class balance data set, and intelligently diagnosing the mechanical fault by using the trained fault diagnosis model. And the mechanical fault diagnosis is realized by combining an automatic encoder and a generation countermeasure network.
Description
Technical Field
The invention relates to the field of intelligent fault diagnosis, in particular to an intelligent mechanical fault diagnosis method under a class-unbalanced data set.
Background
As rotary mechanical devices are continuously developed in the direction of intellectualization, precision and complication, the structures of the mechanical devices become more and more complex and compact. In the service process of the mechanical equipment, once a certain part fails, the operation of the whole mechanical equipment is influenced, and even safety accidents are caused. In order to ensure the healthy operation of mechanical equipment, deep learning theory is gradually applied to intelligent diagnosis of mechanical faults as a latest research result in the field of pattern recognition and machine learning. Compared with the traditional fault diagnosis method, the intelligent diagnosis model based on deep learning utilizes the deep network model to extract effective fault characteristics from the signals in a self-adaptive manner, has high diagnosis efficiency, does not depend on the signal processing experience of an operator, and is widely concerned.
Models commonly used in intelligent diagnosis of mechanical faults at present include Convolutional Neural Networks (CNN), deep confidence networks (DBN), residual error networks (ResNet), and the like. In the training process of the models, a large number of historical data sets are often required to be input as training samples, so that the corresponding relation between the data and the health state categories is established. Although a great potential safety hazard is brought to equipment operation when mechanical equipment fails, the failure is a sporadic event, and the equipment cannot operate for a long time in a failure state, so that the normal state data are more, the failure state data are less, and the problem of class imbalance of a data set is caused. The imbalance between the normal class and the fault class brings great difficulty and challenge to the identification of the mechanical health state, the imbalance class data set easily causes the performance reduction of a diagnosis model, namely the model is easy to over-fit normal signals with a large number of samples and under-fit fault signals with a small number of samples. In addition, as the fault samples are rare, the model can easily learn some redundant or even irrelevant features in the process of extracting the fault data features, and the features reduce the generalization capability of the model.
In order to solve the problem of performance degradation of the intelligent diagnosis model of mechanical faults caused by class imbalance, a dynamic weighting method and a data generation method are commonly used. The dynamic weighting method gives more attention to a smaller number of fault samples by adjusting the weighting parameters in the network, thereby improving the under-fitting problem of the fault samples. The data generation rule is that a small amount of fault data is utilized to generate new samples of the same category, the new samples are used for expanding fault samples, the fault data and normal data are balanced, and the balanced data set is used for training the intelligent diagnosis model. Conventional data generation methods include synthetic few-class upsampling (SMOTE), adaptive synthetic sampling (ADASYN), and the like.
The traditional technology has the following technical problems:
in the intelligent diagnosis of mechanical faults under the actual class of unbalanced data sets, the dynamic weight method needs to dynamically adjust the weight according to the unbalanced rate between normal and fault samples, so that the method is suitable for being applied to the condition that the unbalanced rate is known. And when extreme class unbalance occurs in data, the dynamic weight method is easily interfered by redundant features in a small number of fault samples, so that model overfitting is caused, and the accuracy of fault diagnosis is reduced. The data generation method enables various types in the data set to reach balance by performing upsampling on a small quantity of fault signals, and fundamentally solves the problem of class imbalance. However, the mechanical structure is complex and has a nonlinear characteristic, and the vibration signal of the vibration sensor usually has strong background noise under the actual working condition and shows an obvious non-stationary characteristic under the fault state. The traditional data generation method does not learn the distribution characteristics of data, directly generates signals in time domain signals through an interpolation technology, is easily interfered by measurement noise components, has low quality of generated data, and is also easy to cause the performance reduction of an intelligent diagnosis model.
Disclosure of Invention
The invention provides a mechanical fault intelligent diagnosis method under an unbalanced-class data set, which aims at the problems that the application scene of a dynamic weight method is limited, the traditional data generation method is easily interfered by noise and the quality of generated data is not high.
In order to solve the technical problem, the invention provides an intelligent diagnosis method for mechanical faults under a class-unbalanced data set, which comprises the following steps:
step (1), data preprocessing: converting the mechanical vibration signal into a frequency domain, and normalizing the amplitude value to a [0,1] range;
step (2), model building: combining an automatic encoder with a generated countermeasure network to build a data generation model;
step (3), model training: training the data generation model by utilizing fault data according to a preset loss function and an optimization algorithm;
step (4), data generation: utilizing the low-dimensional characteristics of fault data learned by the data generation model in training to generate corresponding fault data after multiple times of interpolation and noise addition so as to realize the balance of various types of data;
step (5), fault diagnosis: and training a preset fault diagnosis model by using the class balance data set, and intelligently diagnosing the mechanical fault by using the trained fault diagnosis model.
In one embodiment, in step (2), the automatic encoder is composed of an encoder and a decoder, the generation countermeasure network is composed of a generator and a discriminator, and the decoder is the generator; the automatic encoder learns the low-dimensional characteristics of input data, namely true data, through the encoder, and outputs generated data, namely false data, which is consistent with the distribution characteristics of the input data through the low-dimensional characteristics and class labels thereof through the decoder; and the discriminator in the generation countermeasure network carries out true and false discrimination and category classification on the input data and the generation data respectively.
In one embodiment, the encoder, the decoder and the discriminator are each constructed by one of a deep convolutional network, a deep belief network, and a residual network.
In one embodiment, in step (3), the preset loss function includes a mean square error loss function between the generator generated data and the encoder input data, a cross entropy classification loss function of the discriminator on true and false data, a Wasserstein distance or binary cross entropy loss function of the discriminator on true and false discrimination of data, and a mean square error loss function between the encoder output features and the intermediate implicit features of the discriminator.
In one embodiment, in step (3), the preset optimization algorithm includes, but is not limited to, one of a random gradient descent method (SGD), a random gradient descent of a Momentum (Momentum), a nertiov Momentum method, an adagadad algorithm, and an adaptive moment estimation method (Adam).
In one embodiment, in step (4), the interpolation is performed in different low-dimensional features of the same category of fault samples, the label of the category needs to be embedded before generating fault data, and the added noise is low-amplitude random noise.
In one embodiment, in step (5), the preset fault diagnosis model includes one of a support vector machine, a k-nearest neighbor algorithm, a random forest, a fuzzy system, or a deep neural network.
Based on the same inventive concept, the present application also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of any of the methods when executing the program.
Based on the same inventive concept, the present application also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of any of the methods.
Based on the same inventive concept, the present application further provides a processor for executing a program, wherein the program executes to perform any one of the methods.
The invention has the beneficial effects that:
compared with the prior art, the invention discloses an intelligent diagnosis method for mechanical faults under a class imbalance data set. The method is used for solving the problem of diagnosis precision reduction caused by imbalance of data sets in mechanical fault diagnosis, a new data generation method is provided, the data distribution characteristics of a small number of fault samples are learned by utilizing deep learning feature mining capability and an antagonistic training mechanism, new features are generated in a low-dimensional feature space of data by utilizing interpolation and noise addition, and the new fault samples are obtained through a generator after tags are embedded. The influence of measurement noise in signals can be eliminated by interpolation in a low-dimensional space, the diversity of generated samples can be improved by adding random noise, and the consistency of the generated samples and the sample data of the same type of fault can be ensured by embedding tags. Thus, the method has at least the following advantages: (1) the low-dimensional distribution characteristic of the data can be learned, and the interference of measurement noise is eliminated; (2) the generated data has consistency with the same type of fault data, and has certain diversity and high quality; (3) the accuracy rate of mechanical fault intelligent identification is high.
Drawings
FIG. 1 is a flow chart of the intelligent diagnosis method for mechanical failure under the unbalanced-like data set of the invention.
Fig. 2 is a comparison graph of generated data and real data of four fault types obtained in the intelligent diagnosis method for mechanical faults under the class imbalance data set, wherein the left column is the real data under four fault states, and the right column is the generated data corresponding to the real data.
Fig. 3 is a classification precision variation curve of the mechanical fault intelligent diagnosis method under the class unbalance data set and the traditional method under five unbalance rates.
Detailed Description
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
Fig. 1 shows a flow chart of a method for intelligently diagnosing a mechanical fault under a class imbalance data set, which specifically includes:
step 101: and (4) preprocessing data. The vibration signal is fourier transformed, the mechanical vibration signal is converted to the frequency domain, and the amplitude is normalized to the [0,1] range.
Step 102: and (5) building a model. And combining the automatic encoder with the generation countermeasure network to build a data generation model.
The automatic encoder consists of an encoder and a decoder, the generation countermeasure network consists of a generator and a discriminator, and the decoder is the generator. The automatic encoder learns the low-dimensional features of the input data (true data) through the encoder, and then the low-dimensional features and class labels thereof output the generated data (false data) consistent with the distribution characteristics of the input data through the decoder. The discriminator in the generation countermeasure network performs true and false discrimination and category classification on the input data and the generation data, respectively.
The encoder, decoder, and discriminator include, but are not limited to, being constructed by one of a deep convolutional network, a deep belief network, and a residual network.
Step 103: and (5) training a model. And training the data generation model according to a preset loss function and an optimization algorithm by using the fault data.
The loss function of the data generation model in the training process comprises the following steps:
1) the generator generates a mean square error loss function between the data and the encoder input data. Optimizing the loss function can ensure consistency of the generated data with the input data distribution characteristics.
2) And (4) carrying out cross entropy classification loss function on the true and false data by the discriminator. Optimizing the cross entropy classification loss function of the true data can improve the classification capability of the discriminator on the true data; the cross entropy classification loss function of the optimized false data can improve the classification discrimination capability of the discriminator on the generated data and the learning capability of the generator on the classification characteristics, and ensure the characteristic consistency among the generated data of the same type and the characteristic difference among the generated data of different types.
3) The Wasserstein distance or the binary cross entropy loss function of the discriminator for the true and false discrimination of the data. Optimizing the loss function may further improve the quality of the data generated by the generator and the discrimination ability of the discriminator.
4) The mean square error loss function between the encoder output features and the intermediate implicit features of the discriminators. Optimizing the loss function can improve the consistency of the extracted features of the generator and the discriminator on the data of the same category.
Through optimization of the loss functions, the generated data are close to the data distribution of the input data of the same type, the discriminator is difficult to discriminate the truth of the generated data and the truth of the input data of the same type, balance is achieved between the generator and the discriminator, and training of a data generation model is completed.
The preset optimization algorithm includes, but is not limited to, one of a random gradient descent method (SGD), a random gradient descent of Momentum (Momentum), a Nesterov Momentum method, an adagadad algorithm, and an adaptive moment estimation method (Adam).
Step 104: and (6) generating data. And (3) utilizing the low-dimensional characteristics of fault data learned by the data generation model in training, and generating corresponding fault data after multiple times of interpolation and noise addition to realize balance of various types of data.
Interpolation is carried out in different low-dimensional characteristics of fault samples of the same category, a label of the category needs to be embedded before fault data is generated, and the added noise is low-amplitude random noise.
And step 105, fault diagnosis. And training a preset fault diagnosis model by using the class balance data set, and intelligently diagnosing the mechanical fault by using the trained fault diagnosis model.
The preset fault diagnosis model includes but is not limited to one of a support vector machine, a k-nearest neighbor algorithm, a random forest, a fuzzy system and a deep neural network.
In order to more clearly understand the technical solution and the effects of the present invention, a detailed description is given below with reference to a specific embodiment.
Taking intelligent diagnosis of gear case faults as an example, a planetary gear case fault simulation test platform is built, and four fault states are respectively and manually set: broken teeth, missing teeth, cracked tooth root, worn tooth surface and normal state. An acceleration sensor is arranged on the planetary gear box to acquire a vibration signal of the gear box, and the sampling frequency is 5 kHz. Each health state comprises 2000 groups of signals, wherein 1000 groups of signals are not involved in training as test data, and each group of signals has a length of 2048 data points. In order to verify the effectiveness of the intelligent diagnosis method for mechanical faults under the class-unbalance data set, 5 unbalance rates are set in the example, namely the ratio of the number of the gearbox health samples to the number of each class of fault samples is 5:1, 10:1, 20:1, 50:1 and 100:1 respectively, and the data volume of the health samples under each unbalance rate is 1000.
The 5 groups of unbalanced class data sets are processed by the technology disclosed by the invention, the steps are shown in fig. 1, and the detailed information is as follows.
And (1) preprocessing data. The vibration signal is fourier transformed, the mechanical vibration signal is converted to the frequency domain, and the amplitude is normalized to the [0,1] range. The length of an original time domain signal is 2048 data points, and a frequency domain signal with 1024 lengths is taken as input data of the model after Fourier transform.
And (2) building a model. Combining an automatic encoder with a generated countermeasure network, and building a data generation model, wherein the specific embodiment is as follows:
an automatic encoder: comprising an encoder and a decoder, whose main role is to encode and decode the input data. The encoder adopts a four-layer one-dimensional convolution neural network structure, the dimension of each layer is 8, 16, 32 and 64 respectively, a convolution kernel with the length of 15 is adopted, a LeakyReLU activation function layer is connected among convolution layers, and a sample outputs a 64-dimensional potential feature vector after passing through the encoder. The decoder adopts a four-layer one-dimensional deconvolution neural network structure, the dimensionality of each layer is 64, 32, 16 and 8 respectively, a deconvolution kernel with the length of 15 is adopted, ReLU activation function layers are connected among deconvolution layers, the last layer of deconvolution of the decoder is connected with a Sigmoid activation function, and the amplitude of generated data is limited in the range of [0,1 ].
Generating a countermeasure network: including generators and discriminators. The generator is the decoder in the automatic encoder. The discriminator designs four one-dimensional convolution layers and two full-connection layers, the dimension of each layer is 8, 16, 32 and 64 respectively, a convolution kernel with the length of 15 is adopted, each convolution layer is connected with a LeakyRelu activation function layer, and the convolution layer finally outputs a feature vector with the length of 64 dimensions. The features are then separately input into two fully-connected layers, the first of which reduces the 64-dimensional feature vector to 1-dimensional for computing the Wasserstein distance between the generated data and the real data. The second fully connected layer reduces the feature vector of 64 dimensions to 4 dimensions (namely the number of fault categories needing to be upsampled) and is connected with the Softmax active layer to judge the category of the signal.
And (3) training a model. And training the data generation model according to a preset loss function and an optimization algorithm by using the fault data. In this embodiment there are 4 partial loss functions:
firstly, a generator generates a mean square error loss function between data and input data of an encoder;
a cross entropy classification loss function of the true data and the false data is performed by the discriminator;
the distance function of Wasserstein of the discriminator for identifying true and false of the data;
and fourthly, outputting a mean square error loss function between the characteristic and the intermediate implicit characteristic of the discriminator by the encoder.
And after the loss functions of all parts are added, back propagation is carried out through a root mean square transfer algorithm (RmsPorp), and the discriminator and the automatic encoder are optimized in sequence. And (5) repeatedly executing model training, after the model loss tends to be balanced after 2000 iterations, and finishing the network training.
And (4) generating data. Utilizing the low-dimensional characteristics of fault data learned by a data generation model in training to generate corresponding fault data after multiple times of interpolation and noise addition so as to realize the balance of various types of data;
training samples of the same category are input into an encoder, and potential feature vectors of input data are obtained. And then, similar feature vectors are selected for interpolation, in the embodiment, a K nearest neighbor method is adopted to select feature vectors, one feature vector is selected from low-dimensional feature vectors, 3 nearest neighbor vectors are found out, and one feature vector is selected from the nearest neighbor vectors for vector interpolation. After interpolation amplification, 0.02 times of standard white Gaussian noise is added to the newly obtained vector, and the label of the sample is embedded into the vector after noise addition, so that the amplification of the potential feature vector is realized. Finally, inputting the processed feature vector into a decoder to generate a new sample. Fig. 2 shows a comparison graph of the generated signals and the real signals of four fault types, and it can be seen that the generated signals obey the distribution rule of the real signals and have certain differences.
And (5) diagnosing faults. And training a preset fault diagnosis model by using the class balance data set, and intelligently diagnosing the mechanical fault by using the trained fault diagnosis model.
The fault diagnosis model adopts a support vector machine, and input data of the fault diagnosis model are 6 main characteristics of each data sample extracted by a main component analysis method. Firstly, a class balance data set is adopted to train the support vector machine, and then test set data (the data volume of each class is 1000) is adopted to test the classification accuracy of the trained support vector machine. Fig. 2 shows the classification accuracy obtained by training the support vector machine using the class-balanced data set obtained by the method of the present invention and the synthetic minority class upsampling technique, respectively, and also shows the classification accuracy obtained without using the data generation method. Under different unbalance rates, the method and the synthetic minority class upsampling technology provided by the invention can improve the classification accuracy of the classifier, and the method provided by the invention can obtain the highest classification accuracy, so that the quality of generated data obtained by the data generation method provided by the invention is high, and the improvement of the performance of the classifier is facilitated.
In summary, the invention combines the automatic encoder and the generation countermeasure network, and can learn the data distribution characteristics of a small number of fault samples by using the feature mining capability and the countermeasure training mechanism of deep learning. In addition, potential features are generated by utilizing interpolation and noise addition in a low-dimensional space, and data are generated by a decoder, so that the anti-interference capability and the data quality of measurement noise can be improved, and the intelligent diagnosis performance of mechanical faults is improved.
The above-mentioned embodiments are merely preferred embodiments for fully illustrating the present invention, and the scope of the present invention is not limited thereto. The equivalent substitution or change made by the technical personnel in the technical field on the basis of the invention is all within the protection scope of the invention. The protection scope of the invention is subject to the claims.
Claims (10)
1. An intelligent diagnosis method for mechanical faults under a class imbalance data set is characterized by comprising the following steps:
step (1), data preprocessing: converting the mechanical vibration signal into a frequency domain, and normalizing the amplitude value to a [0,1] range;
step (2), model building: combining an automatic encoder with a generated countermeasure network to build a data generation model;
step (3), model training: training the data generation model by utilizing fault data according to a preset loss function and an optimization algorithm;
step (4), data generation: utilizing the low-dimensional characteristics of fault data learned by the data generation model in training to generate corresponding fault data after multiple times of interpolation and noise addition so as to realize the balance of various types of data;
step (5), fault diagnosis: and training a preset fault diagnosis model by using the class balance data set, and intelligently diagnosing the mechanical fault by using the trained fault diagnosis model.
2. The intelligent diagnosis method for mechanical failure under the condition of imbalance-like data set of claim 1, wherein in the step (2), the automatic encoder is composed of an encoder and a decoder, the generation countermeasure network is composed of a generator and a discriminator, and the decoder is the generator; the automatic encoder learns the low-dimensional characteristics of input data, namely true data, through the encoder, and outputs generated data, namely false data, which is consistent with the distribution characteristics of the input data through the low-dimensional characteristics and class labels thereof through the decoder; and the discriminator in the generation countermeasure network carries out true and false discrimination and category classification on the input data and the generation data respectively.
3. The intelligent diagnosis method for mechanical failure under the unbalanced-like data set as claimed in claim 2, wherein the encoder, the decoder and the discriminator are all constructed by one of a deep convolutional network, a deep confidence network and a residual error network.
4. The intelligent diagnosis method for mechanical failure under the class of unbalanced data set as claimed in claim 1, wherein in step (3), the preset loss functions comprise a mean square error loss function between the generator generated data and the encoder input data, a cross entropy classification loss function of the discriminator on true and false data, a Wasserstein distance or a binary cross entropy loss function of the discriminator on true and false identification of the data, and a mean square error loss function between the encoder output features and the intermediate implicit features of the discriminator.
5. The intelligent diagnosis method for mechanical failure under the class of unbalanced data set as claimed in claim 1, wherein in the step (3), the preset optimization algorithm comprises one of a random gradient descent method, a random gradient descent with momentum, a Nesterov momentum method, an Adagarad algorithm and an adaptive moment estimation method.
6. The intelligent diagnosis method for mechanical failure under the condition of imbalance-like data set as claimed in claim 1, wherein in the step (4), the interpolation is carried out in different low-dimensional features of the same type of failure samples, the labels of the type are embedded before the failure data are generated, and the added noise is low-amplitude random noise.
7. The intelligent diagnosis method for mechanical failure under the imbalance-like data set of claim 1, wherein in the step (5), the preset failure diagnosis model comprises one of a support vector machine, a k-nearest neighbor algorithm, a random forest, a fuzzy system or a deep neural network.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 7 are implemented when the program is executed by the processor.
9. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
10. A processor, characterized in that the processor is configured to run a program, wherein the program when running performs the method of any of claims 1 to 7.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111136682.5A CN113935460B (en) | 2021-09-27 | 2021-09-27 | Intelligent diagnosis method for mechanical faults under unbalanced-like data set |
PCT/CN2021/123198 WO2023044979A1 (en) | 2021-09-27 | 2021-10-12 | Mechanical fault intelligent diagnosis method under class unbalanced dataset |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111136682.5A CN113935460B (en) | 2021-09-27 | 2021-09-27 | Intelligent diagnosis method for mechanical faults under unbalanced-like data set |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113935460A true CN113935460A (en) | 2022-01-14 |
CN113935460B CN113935460B (en) | 2023-08-11 |
Family
ID=79276976
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111136682.5A Active CN113935460B (en) | 2021-09-27 | 2021-09-27 | Intelligent diagnosis method for mechanical faults under unbalanced-like data set |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN113935460B (en) |
WO (1) | WO2023044979A1 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114057053A (en) * | 2022-01-18 | 2022-02-18 | 杭州浅水数字技术有限公司 | Method for monitoring fatigue degree of component of special machine |
CN114611233A (en) * | 2022-03-08 | 2022-06-10 | 湖南第一师范学院 | Rotating machinery fault unbalance data generation method and computer equipment |
CN114993677A (en) * | 2022-05-11 | 2022-09-02 | 山东大学 | Rolling bearing fault diagnosis method and system based on unbalanced small sample data |
CN116204786A (en) * | 2023-01-18 | 2023-06-02 | 北京控制工程研究所 | Method and device for generating designated fault trend data |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116401596B (en) * | 2023-06-08 | 2023-08-22 | 哈尔滨工业大学(威海) | Early fault diagnosis method based on depth index excitation network |
CN116432091B (en) * | 2023-06-15 | 2023-09-26 | 山东能源数智云科技有限公司 | Equipment fault diagnosis method based on small sample, construction method and device of model |
CN116993319B (en) * | 2023-07-14 | 2024-01-26 | 南京先维信息技术有限公司 | Remote equipment health monitoring method and device based on Internet of things |
CN116701948B (en) * | 2023-08-03 | 2024-01-23 | 东北石油大学三亚海洋油气研究院 | Pipeline fault diagnosis method and system, storage medium and pipeline fault diagnosis equipment |
CN116821697B (en) * | 2023-08-30 | 2024-05-28 | 聊城莱柯智能机器人有限公司 | Mechanical equipment fault diagnosis method based on small sample learning |
CN117056814B (en) * | 2023-10-11 | 2024-01-05 | 国网山东省电力公司日照供电公司 | Transformer voiceprint vibration fault diagnosis method |
CN117056734B (en) * | 2023-10-12 | 2024-02-06 | 山东能源数智云科技有限公司 | Method and device for constructing equipment fault diagnosis model based on data driving |
CN117076935B (en) * | 2023-10-16 | 2024-02-06 | 武汉理工大学 | Digital twin-assisted mechanical fault data lightweight generation method and system |
CN117593783B (en) * | 2023-11-20 | 2024-04-05 | 广州视景医疗软件有限公司 | Visual training scheme generation method and device based on self-adaptive SMOTE |
CN117332342B (en) * | 2023-11-29 | 2024-02-27 | 北京宝隆泓瑞科技有限公司 | Machine pump equipment operation fault classification method and device based on semi-supervised learning |
CN117725419A (en) * | 2023-12-22 | 2024-03-19 | 兰州理工大学 | Small sample unbalanced rotor fault diagnosis method and system |
CN117909652B (en) * | 2023-12-29 | 2024-06-25 | 广东电网有限责任公司江门供电局 | High-voltage circuit breaker fault diagnosis data processing method |
CN117610614B (en) * | 2024-01-11 | 2024-03-22 | 四川大学 | Attention-guided generation countermeasure network zero sample nuclear power seal detection method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109753998A (en) * | 2018-12-20 | 2019-05-14 | 山东科技大学 | The fault detection method and system, computer program of network are generated based on confrontation type |
CN113239991A (en) * | 2021-04-28 | 2021-08-10 | 浙江工业大学 | Flame image oxygen concentration prediction method based on regression generation countermeasure network |
CN113255078A (en) * | 2021-05-31 | 2021-08-13 | 南京信息工程大学 | Bearing fault detection method and device under unbalanced sample condition |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110428004B (en) * | 2019-07-31 | 2021-02-05 | 中南大学 | Mechanical part fault diagnosis method based on deep learning under data imbalance |
CN112396088B (en) * | 2020-10-19 | 2023-05-12 | 西安交通大学 | Mechanical fault intelligent diagnosis method for implicit excitation countermeasure training under small sample |
CN113298230B (en) * | 2021-05-14 | 2024-04-09 | 武汉嫦娥医学抗衰机器人股份有限公司 | Prediction method based on unbalanced data set generated against network |
-
2021
- 2021-09-27 CN CN202111136682.5A patent/CN113935460B/en active Active
- 2021-10-12 WO PCT/CN2021/123198 patent/WO2023044979A1/en unknown
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109753998A (en) * | 2018-12-20 | 2019-05-14 | 山东科技大学 | The fault detection method and system, computer program of network are generated based on confrontation type |
CN113239991A (en) * | 2021-04-28 | 2021-08-10 | 浙江工业大学 | Flame image oxygen concentration prediction method based on regression generation countermeasure network |
CN113255078A (en) * | 2021-05-31 | 2021-08-13 | 南京信息工程大学 | Bearing fault detection method and device under unbalanced sample condition |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114057053A (en) * | 2022-01-18 | 2022-02-18 | 杭州浅水数字技术有限公司 | Method for monitoring fatigue degree of component of special machine |
CN114057053B (en) * | 2022-01-18 | 2022-04-26 | 杭州浅水数字技术有限公司 | Method for monitoring fatigue degree of component of special machine |
CN114611233A (en) * | 2022-03-08 | 2022-06-10 | 湖南第一师范学院 | Rotating machinery fault unbalance data generation method and computer equipment |
CN114611233B (en) * | 2022-03-08 | 2022-11-11 | 湖南第一师范学院 | Rotating machinery fault imbalance data generation method and computer equipment |
CN114993677A (en) * | 2022-05-11 | 2022-09-02 | 山东大学 | Rolling bearing fault diagnosis method and system based on unbalanced small sample data |
CN116204786A (en) * | 2023-01-18 | 2023-06-02 | 北京控制工程研究所 | Method and device for generating designated fault trend data |
CN116204786B (en) * | 2023-01-18 | 2023-09-15 | 北京控制工程研究所 | Method and device for generating designated fault trend data |
Also Published As
Publication number | Publication date |
---|---|
WO2023044979A1 (en) | 2023-03-30 |
CN113935460B (en) | 2023-08-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113935460A (en) | Intelligent diagnosis method for mechanical fault under class imbalance data set | |
Han et al. | Multi-level wavelet packet fusion in dynamic ensemble convolutional neural network for fault diagnosis | |
CN103728551B (en) | A kind of analog-circuit fault diagnosis method based on cascade integrated classifier | |
CN109213121B (en) | Method for diagnosing clamping cylinder fault of fan braking system | |
CN110212528B (en) | Power distribution network measurement data missing reconstruction method | |
CN108062572A (en) | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on DdAE deep learning models | |
Zhang et al. | An intelligent fault diagnosis method of rolling bearings based on short-time Fourier transform and convolutional neural network | |
CN105678343B (en) | Hydropower Unit noise abnormality diagnostic method based on adaptive weighted group of sparse expression | |
CN105738109A (en) | Bearing fault classification diagnosis method based on sparse representation and ensemble learning | |
CN114778112A (en) | Audio identification and fault diagnosis method for mechanical fault of wind turbine generator system | |
CN105335698A (en) | Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network | |
CN111783531A (en) | Water turbine set fault diagnosis method based on SDAE-IELM | |
CN109214356A (en) | A kind of fan transmission system intelligent fault diagnosis method based on DCNN model | |
CN111275108A (en) | Method for performing sample expansion on partial discharge data based on generation countermeasure network | |
CN115409052A (en) | Fault diagnosis method and system for wind generating set bearing under data imbalance | |
CN113884844A (en) | Transformer partial discharge type identification method and system | |
CN114818806A (en) | Gearbox fault diagnosis method based on wavelet packet and depth self-encoder | |
CN112731137A (en) | Cage type asynchronous motor stator and rotor fault joint diagnosis method based on stack type self-coding and light gradient elevator algorithm | |
CN116628592A (en) | Dynamic equipment fault diagnosis method based on improved generation type countering network | |
CN115290326A (en) | Rolling bearing fault intelligent diagnosis method | |
CN115114965A (en) | Wind turbine generator gearbox fault diagnosis model, method, equipment and storage medium | |
Liang et al. | Multibranch and multiscale dynamic convolutional network for small sample fault diagnosis of rotating machinery | |
CN114897138A (en) | System fault diagnosis method based on attention mechanism and depth residual error network | |
CN114491823A (en) | Train bearing fault diagnosis method based on improved generation countermeasure network | |
CN114415018A (en) | Self-learning grating interference spectrum analysis technology for motor fault early warning |
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 |