A bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer
By combining variable pooling-multi-scale convolutional neural network with Transformer model, the problem of feature loss and overfitting in rolling bearing life prediction is solved, achieving more accurate life prediction and better generalization capabilities.
Patent Information
- Application Number
- CN202411645094.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-18
- Publication Date
- 2025-08-26
- Estimated Expiration
- 2044-11-18
AI Technical Summary
In the prior art, the MSCNN model has problems such as pooled layer dimensionality reduction leading to feature loss, predicted sequence degradation and overfitting in rolling bearing life prediction.
Combining the variable pooling-multi-scale convolutional neural network and the Transformer model, the data augmentation and variable pooling layer give different weights to feature and the Lable-Smoothing regularization technology is enhanced to enhance the feature extraction and model generalization capabilities, capture global and detailed features, and avoid overfitting.
It improves the accuracy and generalization ability of rolling bearing life prediction, effectively overcomes the problems of feature loss and sequence memory degradation, reduces the model's dependence on correct labels, and improves the prediction effect.
Smart Images
Figure CN119598287B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to related fields of bearing life, and in particular to a bearing life prediction method based on a combination of a variable pooling-multi-scale convolutional neural network and a Transformer. Background Art
[0002] Rolling bearings are key components of modern industrial production and equipment operation. Continuous wear during operation will reduce the service life of the equipment. Therefore, predicting the life of rolling bearings is of great significance and reference value for predicting the remaining service life of bearings in components such as engines.
[0003] like Figure 1 As shown, in the prior art, the method of establishing an MSCNN (1D-multiscale-CNN convolutional neural network) is generally used to predict the life of rolling bearings. It consists of four convolutional pooling layers, a fully connected layer, and a softmax layer. The multiscale layer (MS) is placed after the convolutional pooling layer. It aims to capture the different scale features of the input data through convolution kernels of different sizes. Smaller convolution kernels can capture detailed features, while larger convolution kernels can capture overall features, thereby predicting the life of rolling bearings. It can be seen that in the above prediction method, the pooling layer in the MSCNN only performs downsampling, reducing the dimensionality of the input raw acoustic emission data; however, this unmodified dimensionality reduction may lose some features with relatively large weights that are more important to the prediction task. At the same time, the original MSCNN also suffers from the problem of prediction sequence degradation. Moreover, the model is overconfident about the correct labels, which poses the risk of overfitting. Summary of the Invention
[0004] The purpose of the embodiments of the present invention is to provide a bearing life prediction method based on a variable pooling-multi-scale convolutional neural network combined with a Transformer, so as to alleviate the problem that the pooling layer in the MSCNN in the prior art only performs a downsampling operation and reduces the dimensionality of the input original acoustic emission data; however, this unmodified dimensionality reduction may lose some features with relatively large weights that are more important for the prediction task. At the same time, the original MSCNN also has the problem of prediction sequence degradation, and the model is overconfident about the correct labels, which has the technical problem of overfitting.
[0005] To solve the above technical problems, the embodiments of the present invention provide the following technical solutions:
[0006] A first aspect of the present invention provides a method for predicting the life of a bearing based on acoustic emission signals, comprising the following steps:
[0007] S10. The original bearing acoustic emission signal x *={x1,x2,…,x n} to collect;
[0008] Where x* represents the normalized data;
[0009] S20. Perform data enhancement on the original bearing acoustic emission signal, expand the data sample, and divide it into a training data set and a test data set;
[0010] S30.,establish a neural network that is sequentially equipped with a convolutional layer, a variable pooling layer, a multi-scale layer, and a Transformer model, and outputs the prediction result of the one-dimensional sequence through a fully connected layer and a Softmax function;
[0011] S40. Input the features output by the multi-scale layer in the variable pooling-multi-scale convolutional neural network into the Transformer model; after encoding and decoding, select the mean squared error as the loss function; and output the predicted value through the fully connected layer and softmax prediction output. The variable pooling layer changes the weight of the bearing signal features by assigning different weight coefficients to the channel domain features of the pooling layer. The variable pooling layer calculates the weight coefficients of the pooling layer channels and then combines the weight coefficients with the features of the pooling layer channels. By assigning weight coefficients to the pooling layer channels, feature information with a high predictive impact can be enhanced and feature information with a low predictive impact can be weakened.
[0012] Furthermore, in the multi-scale layer in step S30, a 1×1 convolution kernel is added before the 5×1 convolution kernel, and a 3×1 maximum pooling is set to perform feature extraction on the data.
[0013] Furthermore, the steps of establishing the variable pooling layer in step S30 are as follows:
[0014] S31. Combine the features generated by the convolutional layer and convert them into a feature vector, which is expressed as:
[0015]
[0016] Where H is the length of the pooling layer, u c represents the jth neuron in the cth pooling channel field of the pooling layer, z c Represents the average feature information of the c-th pooled channel domain;
[0017] S32. Calculate the weight coefficient of the channel domain feature of the pooling layer, the expression is:
[0018] S c =σ(W2δ(W1z c ))
[0019] Where δ is the Relu function, W1 is the weight of the Relu function; σ is the Sigmoid function, W2 is the weight of the Sigmoid function; the role of the Sigmoid function is to map the weight coefficient to between 0 and 1;
[0020] S33: Assignment of channel domain weight coefficients, using the weight coefficients s learned in step S20 c , multiplied by the channel domain features of the pooling layer, and different weight coefficients are assigned to the pooling channel domain features; the distribution expression of the weight coefficient is:
[0021] u c ′=s c *z c
[0022] Among them, u c ′ is the channel domain of the updated pooling layer.
[0023] Furthermore, during the prediction phase, the test dataset is imported into the trained variable pooling-multi-scale convolutional neural network, and the predicted remaining service life curve is compared with the actual bearing life curve;
[0024] The true bearing life curve is represented by the bearing reliability, and the reliability factor is then used as the remaining service life curve to construct the reliability factor. The reliability factor formula is as follows:
[0025]
[0026] Among them, R(i) is the reliability factor of the i-th sampling point, T i is the running time of the bearing at the i-th sampling point, T1 is the running time of the bearing at the initial fault point, T e is the total running time of the bearing.
[0027] Furthermore, in step S20, the data before the bearing device is at the initial fault point is labeled as 0; the data after the bearing device is at the initial fault point is labeled as 1, and the above data are used as label data of the training data set and the test data.
[0028] Furthermore, in step S20, the label-smoothing regularization technique is applied during the model training phase to add "noise" to the actual life label; that is, a small portion of ε = 0.1 is taken from the data with the bearing label of 1, and then assigned to the category with the label of 0 to reduce data overfitting.
[0029] Furthermore, the application of the label-smoothing regularization technique in the model training stage includes:
[0030] Convert hard labels into soft labels to make network optimization smoother; the original hard labels are
[0031]
[0032] The initial fault point is the sample point where the bearing is first detected to have a fault. After using the LabelSmoothing regularization technique, the updated label vector is as follows:
[0033]
[0034] Where K is the total number of classification categories, and K = 2, α is a small hyperparameter, generally set to 0.2.
[0035] The present invention proposes a method for predicting the life of bearings, combining VPMCNN and Transformer. It not only retains the ability of VPMCNN to capture global and detailed features, but also learns a rich representation of the sequence before and after the input signal through the self-attention mechanism of Transformer, which can better predict the remaining life of the bearing through the input rolling bearing acoustic emission signal; it solves the disadvantage of Transformer's inability to capture detailed features, and at the same time, the self-attention mechanism in the Transformer model can capture long-distance dependencies in the input signal, effectively overcoming the memory degradation problem of the prediction sequence. In order to reduce the model's overconfidence in the correct label, avoid overfitting, and improve the model's generalization ability, this patent embeds the Label-Smoothing regularization technology when labeling the training data to smooth the labels. The specific effects are as follows:
[0036] 1. In terms of feature extraction: A variable pooling-multi-scale convolutional neural network is proposed. Compared with the ordinary one-dimensional convolutional neural network, the variable pooling layer can assign higher weights to features that are more meaningful for life prediction, thereby better extracting features that are beneficial to bearing prediction and achieving better prediction results.
[0037] 2. This paper utilizes the Transformer model for improvement. The acoustic emission signal features extracted by the VPMCNN are first fed into the Transformer as input. A multi-head attention mechanism is then used to capture dependencies within the acoustic emission signal sequence, which, to a certain extent, mitigates the quadratic increase in computational speed as sequence length increases. Prediction output is then achieved through a fully connected layer and softmax. During training, a batch size of 128 is selected, and the Adam optimizer is used.
[0038] 3. To improve the generalization of model predictions and adapt to data noise, Label-Smoothing regularization technology is embedded when labeling the lifespan of training samples. During training, "hard" labels (one-hot encoded labels) are converted to "soft" labels (soft labels). Label Smoothing softens labels, making the model's predictions smoother and less dependent on a single feature or label. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments are briefly introduced below.
[0040] Figure 1 A flowchart of the MSCNN method in the prior art provided by the present invention;
[0041] Figure 2 A flow chart of a bearing life prediction method based on a variable pooling-multi-scale convolutional neural network combined with a Transformer provided by the present invention;
[0042] Figure 3 This is a structural principle diagram of the convolutional pooling layer in a bearing life prediction method based on a combination of variable pooling-multi-scale convolutional neural network and Transformer provided by the present invention;
[0043] Figure 4 A schematic diagram of the principle of the Transformer in a bearing life prediction method based on a variable pooling-multi-scale convolutional neural network combined with the Transformer provided by the present invention;
[0044] Figure 5 A schematic diagram of the principle of the reliability factor in a bearing life prediction method based on a variable pooling-multi-scale convolutional neural network combined with Transformer provided by the present invention;
[0045] Figure 6 Provides a schematic diagram of the LSTM structure in Example 2 of the present invention;
[0046] Figure 7 This is a VMD-AE flowchart of Example 2 of the bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer provided by the present invention. DETAILED DESCRIPTION
[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
[0049] Example 1
[0050] like Figure 2 As shown in the figure, a bearing life prediction method based on a variable pooling-multi-scale convolutional neural network combined with Transformer includes the following steps:
[0051] S10. The original bearing acoustic emission signal x * ={x1,x2,…,x n} to collect;
[0052] Where x* represents the normalized data;
[0053] S20. Data augmentation is performed on the original bearing acoustic emission signals, expanding the data samples and dividing them into training, validation, and test datasets. During the model training phase, label-smoothing regularization is applied to add "noise" to the true life labels. Specifically, a small portion of data after the initial bearing failure (labeled 1) with a value of ε = 0.1 is assigned to the class labeled 0 (data before the initial failure) to reduce data overfitting. In the test dataset, the model's output on the test dataset is directly used as the prediction result.
[0054] S30. Set the number of training epochs to N and establish a variable pooling-multiscale convolutional neural network that is sequentially provided with a convolutional pooling layer, a multi-scale layer, a Transformer model, a VP layer, a GAP layer, a softmax layer, and a Label-Smoothing model;
[0055] S40. Input the features output by the multi-scale layer in the variable pooling-multi-scale convolutional neural network into the Transformer model; after encoding and decoding, select the mean squared error as the loss function; and output the prediction through the fully connected layer and softmax.
[0056] In a multi-scale architecture, convolution kernels of different sizes (1*1, 3*1, and 5*1) are used within the same layer to extract features at different scales of the bearing signal. Finally, these extracted features are fused to achieve multi-scale feature extraction. When using convolution and pooling operations, padding is used to ensure that the output features are of the same size. The advantages of multi-scale blocks are as follows:
[0057] (1) Increased the depth and width of the network;
[0058] (2) Multi-scale reduces the parameters of the training network;
[0059] (3) Use convolution kernels of different sizes, 1*1, 3*1, and 5*1, to extract the multi-scale features of the original acoustic emission signal.
[0060] Finally, the multi-scale module extracts multiple features that can capture complementary and rich prediction information at different scales. This greatly improves the feature learning ability and thus improves the prediction performance.
[0061] S50. Embed Lable-Smoothing to perform label smoothing.
[0062] Combining VPMCNN and Transformer not only retains VPMCNN's ability to capture global and detailed features, but also allows Transformer to learn rich representations of the preceding and following sequences of input signals through its self-attention mechanism, enabling better prediction of the remaining life of bearings using the input acoustic emission signals of rolling bearings. This overcomes the Transformer's inability to capture detailed features, while the self-attention mechanism in the Transformer model can capture long-range dependencies in the input signal, effectively overcoming the memory degradation problem of the predicted sequence. To reduce the model's overconfidence in the correct label, avoid overfitting, and improve the model's generalization ability, this paper embeds Label-Smoothing after the VPMCNN-Transformer model to perform label smoothing. The specific effects are as follows:
[0063] In terms of feature extraction:
[0064] 1. A multiscale layer is introduced into the traditional 1D-CNN to extract local and full-view features;
[0065] 2. A variable pooling layer is proposed, which can give higher weights to features that are more meaningful for life prediction, thereby achieving better prediction results.
[0066] This paper uses the Transformer model to improve the method. The acoustic emission signal features extracted by the VPMCNN are first fed into the Transformer as input, which, to a certain extent, avoids the quadratic increase in the speed of calculation as the sequence length increases. Furthermore, a Label-Smoothing algorithm is embedded in the training data, converting "hard" labels into "soft" labels during training.
[0067] It should be noted that the Transformer model mainly consists of two parts: the encoder and the decoder. The encoder takes the features extracted by the variable pooling-multi-scale CNN as input and passes them into the self-attention layer in the Transformer. After that, it enters the feedforward neural network and is output to the next encoder. The encoder begins to process the input sequence. The output of the top encoder is then transformed into a set of attention vectors Key and Value. The attention vector will be used in the "encoder-decoder attention" layer of each decoder. If the acoustic emission data is directly input into the Transformer model without processing, the large number of samples will cause the number of model parameters to increase accordingly, and network training will be very difficult. This patent uses variable pooling-multi-scale CNN to first extract the time series feature vectors in the long-term degradation sequence of rolling bearing acoustic emissions to reduce the dimension of the input data.
[0068] Inputting the features output by the VPMCNN into the Transformer model allows for faster and better identification of time series relationships within rolling bearing acoustic emission signals, extracting information relevant to lifespan prediction. Furthermore, during model training, label-smoothing is applied to the labeling of the training dataset to smooth the output labels, aligning the extreme min and max values at both ends to the middle, improving generalization capabilities.
[0069] Furthermore, if Figure 3 As shown, in the multi-scale layer in step S10, a convolution kernel 1×1 is added before the convolution kernel 5×1, and a 3×1 maximum pooling is set to perform feature extraction on the data.
[0070] The 5×1 convolution kernel will increase the computational complexity of the model, so a 1×1 convolution kernel is added before the 5×1 convolution kernel to reduce the model parameters, and a 3×1 maximum pooling is introduced to further extract features from the data.
[0071] Furthermore, the steps for establishing the variable pooling layer in step S10 are as follows:
[0072] Step 1: Pooling layer channel domain feature combination
[0073] The features generated by the convolutional layer are combined and converted into a feature vector. The feature combination process is described as shown in Formula 1:
[0074]
[0075] In formula (1), H is the length of the pooling layer, u c represents the jth neuron in the cth pooling channel field of the pooling layer, z c Represents the average feature information of the c-th pooled channel domain.
[0076] Step 2: Calculate the weight coefficient of the channel domain feature of the pooling layer
[0077] The calculation process of the weight coefficient of the channel domain feature is shown in formula (2):
[0078] S c =σ(W2δ(W1z c )) (2)
[0079] In formula (2), δ is the Relu function, W1 is the weight of the Relu function, σ is the Sigmoid function, and W2 is the weight of the Sigmoid function. The role of the Sigmoid function is to map the weight coefficient to between 0 and 1.
[0080] Step 3: Assignment of channel-domain weight coefficients
[0081] Use the weight coefficient s learned in step 2 c , multiplied by the channel domain features of the pooling layer, and different weight coefficients are assigned to the pooled channel domain features. The distribution of weight coefficients is shown in formula (3).
[0082] u c ′=s c *z c (3)
[0083] In formula (3), u c ′ is the channel domain of the updated pooling layer.
[0084] Furthermore, during the prediction stage, the test dataset is imported into the trained variable pooling-multi-scale convolutional neural network, and the predicted RUL (remaining useful life) curve is compared with the actual bearing life curve.
[0085] like Figure 5 As shown, the current real bearing life curve is generally reflected by the reliability of the bearing, and then the reliability factor is constructed as RUL. The reliability factor formula and graphical representation are as follows:
[0086]
[0087] Where R(i) is the reliability factor of the i-th sampling point, T i is the running time of the bearing at the i-th sampling point, T1 is the running time of the bearing at the initial fault point, T e is the total running time of the bearing.
[0088] Before the acoustic emission signal is fed into the Transformer model for lifespan prediction, it undergoes feature extraction in the VPMCNN, assigning greater weight to variables that are meaningful to the prediction. Furthermore, the multi-scale layer extracts both local and global information from the data, facilitating lifespan prediction. To improve the generalization capability of the prediction model, a Label-Smoothing algorithm is embedded after the softmax during model training to smooth the labels and enhance the model's generalization capability. The specific steps are as follows:
[0089] (1) Offline modeling stage:
[0090] Step 1: Original bearing acoustic emission signal x * ={x1,x2,…,x n}, where x* represents the normalized data.
[0091] Step 2: Perform data enhancement on the original signal, expand the data samples, and divide them into training data set and test data set.
[0092] Step 3: To increase the generalization ability of the model, the Label-Smoothing regularization technology is embedded when labeling the training dataset. This can prevent the model from being overconfident about the correct labels and alleviate overfitting.
[0093] Step 4: Set the number of training iterations to N, the loss of the variable pooling-multi-scale convolutional neural network model to M, build the model, and perform feature extraction.
[0094] Step 5: Input the data processed by the variable pooling-multi-scale convolutional neural network model into the Transformer model, first encode it, then decode it, and select the root mean squared error (RMSE) as the loss function. Then, pass it through the fully connected layer and softmax to predict the output.
[0095] (2) Online prediction stage:
[0096] Step 1: Import the test data set into the trained variable pooling multi-scale-Transformer prediction model, and compare the predicted RUL curve with the true bearing life curve. The true bearing life curve is the reliability factor in the figure above. In this patent, the evaluation indicator root mean square error (RMSE) is used to evaluate the quality of the prediction results. The smaller the evaluation indicator value, the better the prediction effect. The root mean square error is the square root of the mean of the sum of squared errors. In order to reduce the error, the model will inevitably find a way to prioritize the sample with the largest deviation as close to the true value as possible. In other words, the greater the deviation of the sample, the greater the impact on the model. If this sample is noise, then this will have a significant negative impact on the accuracy of the model.
[0097]
[0098] Where the sampling time is N, i represents the i-th sampling point, RUL(i) pre represents the predicted RUL of the i-th sampling point, and RUL(i) real Represents the real RUL.
[0099] To better illustrate the algorithm used in the present invention, a brief supplementary explanation is given as follows:
[0100] like Figure 4 As shown, the Transformer model is a pre-trained model launched by GOOGLE in 2021. Some people have improved it and used it in life span prediction and achieved good results. However, the computational complexity of the Transformer model is relatively large, which is proportional to the square of the length. Therefore, corresponding processing can be performed before the data enters the Transformer. In the present invention, the feature extraction of VPMCNN is adopted, which can retain the multi-scale features of the data. At the same time, the proposed VP layer can assign different weights to channel variables, making the prediction results more accurate. The model is embedded in Lable-Smoothing, which increases the generalization performance of the model.
[0101] Lable-Smoothing is introduced as follows:
[0102] Label smoothing is a regularization method in the field of machine learning. It is usually used to label the classification problem of training data in diagnosis and prediction. The purpose is to prevent the model from overconfidently predicting labels during training and improve the problem of poor generalization ability. It converts hard labels into soft labels to make network optimization smoother. The original hard label is
[0103]
[0104] The initial fault point is the sample point where the bearing is first detected to have a fault. After using the LabelSmoothing regularization technique, the updated label vector is as follows:
[0105]
[0106] Where K is the total number of classification categories. In this patent, K=2, and α is a small hyperparameter (generally 0.2).
[0107] Example 2
[0108] This patent application's method analyzes bearing fatigue acoustic emission signals. To further demonstrate its effectiveness, the proposed method is compared with a traditional VMD denoising method that extracts degradation features and then incorporates an autoencoder (AE) network (VMD-AE). The proposed algorithm extracts features from the raw signals collected from bearings in the dataset, achieving better results than methods such as VMD, EMD, EEMD, and EWT. This method avoids the difficulties of excessive data dimensionality and high computational complexity.
[0109] The experiment grouped bearing data with similar failure modes into one group (similar failure modes refer to bearing data with similar degradation trends. Such data sets often contain failures caused by the same type of fault). Data with different failure modes were grouped into different groups (different failure modes refer to bearing data with very low or even completely different degradation trends. Such data sets often contain failures caused by different types and degrees of faults). The VMD-AE flow chart is shown in Figure 7:
[0110] The method of this application is also compared with LSTM prediction. The LSTM structure is as follows Figure 6 As shown in Figure 2, the prediction effect is better than that of LSTM, and the root mean square error of the proposed method is much smaller than the corresponding value of LSTM.
[0111] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.
Claims
1. A bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer, characterized in that: The method comprises the following steps: S10. The original bearing acoustic emission signal x * ={x1,x2,...,x n } to collect; Where x* represents the normalized data; S20. Perform data enhancement on the original bearing acoustic emission signal, expand the data sample, and divide it into a training data set and a test data set; In step S20, the label-smoothing regularization technique is applied during the model training phase to add "noise" to the actual life labels. That is, a small portion of ε = 0.1 is taken from the data with the bearing label 1 and then assigned to the class with the label 0 to reduce data overfitting. S30. Establish a neural network that sequentially includes a convolutional layer, a variable pooling layer, a multi-scale layer, and a Transformer model, and outputs a prediction result of a one-dimensional sequence through a fully connected layer and a Softmax function; In the multi-scale layer in step S30, a 1×1 convolution kernel is added before the 5×1 convolution kernel, and a 3×1 maximum pooling is set to perform feature extraction on the data; The variable pooling layer changes the proportion of bearing signal features by assigning different weight coefficients to the channel domain features of the pooling layer; The variable pooling layer calculates the weight coefficient of the pooling layer channel, and then combines the weight coefficient with the characteristics of the pooling layer channel; by assigning weight coefficients to the pooling layer channels, the feature information with great predictive effect can be enhanced, and the feature information with small predictive effect can be weakened; During the prediction phase, the test data set is imported into the trained variable pooling-multi-scale convolutional neural network, and the predicted remaining service life curve is compared with the actual bearing life curve; The true bearing life curve is represented by the bearing reliability, and the reliability factor is then used as the remaining service life curve to construct the reliability factor. The reliability factor formula is as follows: Among them, R(i) is the reliability factor of the i-th sampling point, T i is the running time of the bearing at the i-th sampling point, T1 is the running time of the bearing at the initial fault point, T e is the total running time of the bearing; S40. Input the features of the multi-scale layer output in the variable pooling-multi-scale convolutional neural network into the Transformer model; after encoding and decoding, select the mean square error as the loss function; after passing through the fully connected layer and softmax prediction output, output the predicted value.
2. The bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer according to claim 1 is characterized in that: The steps for establishing the variable pooling layer in step S30 are as follows: S31. Combine the features generated by the convolutional layer and convert them into a feature vector, which is expressed as: Where H is the length of the pooling layer, u c represents the jth neuron in the cth pooling channel field of the pooling layer, z c Represents the average feature information of the c-th pooled channel domain; S32. Calculate the weight coefficient of the channel domain feature of the pooling layer, the expression is: S c =σ(W2δ(W1z c )) Where δ is the Relu function, W1 is the weight of the Relu function; σ is the Sigmoid function, W2 is the weight of the Sigmoid function; the role of the Sigmoid function is to map the weight coefficient to between 0 and 1; S33: Assignment of channel domain weight coefficients, using the weight coefficients s learned in step S32 c , multiplied by the channel domain features of the pooling layer, and different weight coefficients are assigned to the pooling channel domain features; the distribution expression of the weight coefficient is: u c ′=s c *z c Among them, u c ′ is the channel domain of the updated pooling layer.
3. The bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer according to claim 1 is characterized in that: In step S20, the data before the bearing device is at the initial fault point is labeled as 0; the data after the bearing device is at the initial fault point is labeled as 1, and the above data are used as label data of the training data set and the test data.
4. The bearing life prediction method based on variable pooling-multi-scale convolutional neural network combined with Transformer according to claim 1 is characterized in that: The application of Lable-smoothing regularization technology in the model training stage includes: Convert hard labels into soft labels to make network optimization smoother; the original hard labels are The initial fault point is the sample point where the bearing is first detected to have a fault. After using the Label Smoothing regularization technique, the updated label vector is as follows: Where K is the total number of classification categories, and K = 2, α is a small hyperparameter, generally set to 0.2.