Variable-speed bearing fault diagnosis method of wavelet-driven physical interpretable network

Abstract

The invention relates to the technical field of bearing faults, in particular to a variable-speed bearing fault diagnosis method based on a wavelet-driven physical interpretable network. Comprising the following steps: acquiring a bearing fault signal sample by using an acceleration sensor, and calculating a time-frequency diagram of a vibration signal by adopting short-time Fourier transform; extracting an instantaneous rotating speed from the time-frequency diagram of the vibration signal by adopting a local peak search algorithm; the time domain fault vibration signal is converted into a frequency domain signal by adopting fast Fourier transform (FFT); on the basis of the extracted instantaneous rotating speed information, constructing a learnable wavelet-driven physical interpretability network LWPI and training the learnable wavelet-driven physical interpretability network LWPI; fault diagnosis is realized through the trained LWPI network; by means of the signal processing technology, the convolutional neural network is endowed with certain physical significance, meanwhile, the rotating speed information is fused into construction of the convolutional neural network, the bearing fault diagnosis precision under the variable speed condition is improved, and the research and economic production efficiency is improved.

Description

Technical Field

[0001] The invention relates to the technical field of bearing failure, and in particular to a variable speed bearing failure diagnosis method of a wavelet-driven physically interpretable network. Background Art

[0002] With the development of Industry 4.0, the Predictive Health and Management (PHM) of rotating machinery has attracted wide attention from academia and industry, and the fault diagnosis of equipment plays an important role in its PHM. As one of the important components of rotating machinery, the health of rolling bearings is crucial to the safe and stable operation of the equipment. In addition, in actual engineering, bearings often operate under variable speed conditions. Therefore, bearing fault diagnosis under variable speed has attracted wide attention.

[0003] Vibration signals contain rich state signals of mechanical equipment. Fault diagnosis methods based on vibration signal analysis can reliably identify fault types according to fault mechanisms, so their diagnostic process has certain physical significance. Wavelet analysis, as a typical representative of traditional signal processing methods, has made remarkable achievements in bearing fault diagnosis. Li et al. proposed a mechanical fault diagnosis method based on empirical wavelet transform. Zhang et al. used orthogonal wavelet basis to transform vibration signals into time-scale domain, and performed envelope refinement spectrum analysis on wavelet coefficients in high-frequency scale domain to realize rolling bearing fault diagnosis. However, traditional wavelet analysis methods require various parameters to be set in advance according to diagnostic objectives and professional knowledge. Moreover, most methods are suitable for rotating machinery with fixed speed, and their diagnostic accuracy is greatly affected by speed changes. One of the more significant drawbacks is that traditional signal analysis methods are difficult to process massive fault data.

[0004] Data-driven deep learning methods can not only adaptively mine fault features from raw data, but also process massive amounts of data. However, deep learning methods extract local features through randomly initialized kernels in the convolution layer, and their internal decision-making process is a "black box" state, so it is difficult to explain the physical meaning of the extracted features. In order to give the neural network a certain physical meaning, Li et al. used a continuous wavelet convolution layer as the first convolution layer of the convolutional neural network and verified the robustness of their method using various gearbox data. Jiang et al. designed a multi-wavelet kernel convolution layer as the first layer of the neural network to locate and detect mechanical faults. Liu et al. used the bearing fault mechanism to demodulate the fault signal into multiple modal components, designed multiple filters, and combined the powerful feature extraction ability of the neural network to extract fault features and realize fault classification. The above work provides great research value for giving the convolutional neural network a certain physical meaning. However, the structure and working environment of rotating machinery are complex, and the vibration signal is often affected by the speed fluctuation of the rotating machinery and the interference of the vibration signals of other components, thereby reducing the performance of the fault diagnosis model. The above interpretability network ignores the influence of the running speed, which limits its ability in variable speed bearing fault diagnosis. Deepening the network layer may improve the impact of speed fluctuations and enhance diagnostic capabilities, but the higher the network layer, the more abstract and difficult to interpret the feature graph, and its internal decision-making process is more difficult to understand. Summary of the invention

[0005] The present invention discloses a bearing fault diagnosis method based on a learnable wavelet physics interpretability network, and the specific method is as follows:

[0006] The interpretability network is constructed with Morlet wavelet multi-wavelet filter layer as the first layer and CNN convolution block as the second layer;

[0007] Obtain historical time domain vibration signals when bearing fails;

[0008] The historical time-domain vibration signal is processed by Fourier transform to obtain the historical frequency-domain vibration signal;

[0009] Train an interpretability network with historical frequency-domain vibration signals;

[0010] The current time domain vibration signal of the bearing is collected by the acceleration sensor, and the variable speed bearing fault is diagnosed by the trained interpretable network.

[0011] Furthermore, the key parameters of the interpretability network are determined by the historical time-domain vibration signal as follows:

[0012] The historical time-domain vibration signal is processed by short-time Fourier transform to obtain the historical time-frequency diagram, and the historical speed information is extracted from the historical time-frequency diagram by using an improved local peak search algorithm.

[0013] The key parameters of the Morlet wavelet and multi-wavelet filter layers are fitted based on the historical speed information.

[0014] Furthermore, an improved local peak search algorithm is used to extract historical speed information from the historical time-frequency graph. The specific method is as follows:

[0015] The speed value μ(t1) at the first time point t1 in the historical time-frequency diagram is screened, and the frequency corresponding to the main spectrum peak is found according to the local maximum peak value. The frequency obtained from the smallest to the largest frequency is μ(t1)=[μ 11 ,μ 12 ,…,μ 1n ], taking into account the interference of noise and adjacent components in the time-frequency diagram, a penalty function is added to the local peak search algorithm:

[0016]

[0017] Among them, the local maximum peak of the time-frequency graph is:

[0018] P(t,μ(t))=|TFD(t,μ(t)| 2

[0019] The final penalty function of the local peak search algorithm is:

[0020]

[0021] Through the final penalty function of the local peak search algorithm, find the instantaneous rotation frequency corresponding to the minimum penalty function:

[0022]

[0023] in, is the instantaneous rotation frequency finally extracted at the i-th moment; is the lower limit for frequency search; is the upper limit of the frequency search; c1, c2 are constants, μ(t k ) is the frequency value at time point i within the frequency search range; μ(t j ) is the instantaneous frequency extracted from time i and all previous times.

[0024] Furthermore, the key parameters of the Morlet wavelet multi-wavelet filter layer are fitted according to the historical speed information. The specific method is as follows:

[0025] According to the extracted instantaneous rotation frequency, the Morlet wavelet multi-wavelet filter layer is designed. The frequency domain expression of the Morlet wavelet is:

[0026]

[0027] Among them, β is the bandwidth parameter, which controls the width of the wavelet window in the frequency domain; f0 is the center frequency, which determines the position of the wavelet window on the frequency axis; f is the set frequency range;

[0028] When constructing the Morlet wavelet multi-wavelet filter layer, the bandwidth parameter β i and the center frequency f i Obtained by fitting the extracted rotation frequency ridge:

[0029]

[0030] in, is the intermediate frequency of the ith time interval; n is the time interval divided when extracting the instantaneous rotation frequency; κ i0 , κ i1 , c i They are the bandwidth parameters β i and the center frequency f i The fitting coefficients of the Morlet wavelet and multi-wavelet filter layers are:

[0031]

[0032] The extracted feature expression of the i-th filter kernel is:

[0033] L i =ψ i (f)*F[x]

[0034] Where x is the original vibration signal and F[·] represents Fourier transform.

[0035] Furthermore, the CNN convolution block consists of a convolution layer, a BN layer, a global pooling layer GAP and an activation function ReLU, which is expressed as:

[0036] Z = GAP(BN(ReLU(W×L+b)))

[0037] Among them, W and b are the weight and bias of the convolution layer respectively, L is the fault feature extracted by the Morlet wavelet multi-wavelet filter layer; Z is the high-dimensional feature output by the convolution block;

[0038] The fully connected layer flattens the fault features extracted by the convolutional layer, and the expression is:

[0039] Z←flattening(Z)

[0040] q=ω z Z+b z

[0041] Among them, W z and b zare the weights and biases of the classifier. The softmax function is used to classify the input features. The expression is:

[0042]

[0043] Among them, S is the number of fault categories, o i is the predicted probability of the i-th category;

[0044] The cross-entropy loss is used to measure the distribution of predicted labels and true labels. Its expression is:

[0045]

[0046] Among them, r i is the true probability of the i fault category.

[0047] Furthermore, the interpretability network is trained with historical frequency domain vibration signals. The specific method is as follows:

[0048] The historical frequency domain vibration signals are randomly divided into a training set and a test set;

[0049] During the training process, according to the chain rule, the weights of the convolution kernels and the key coefficients of the Morlet wavelet multi-wavelet filter layer are optimized in the back-propagation process to obtain a trained interpretable network.

[0050] Due to the adoption of the above technical solution, the present invention has the following beneficial effects:

[0051] The present invention uses signal processing technology to give the convolutional neural network a certain physical meaning, and at the same time incorporates the speed information into the construction of the convolutional neural network, thereby improving the accuracy of bearing fault diagnosis under variable speed conditions and improving the efficiency of research and economic production.

[0052] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings of the present invention are as follows.

[0054] Figure 1 Construct graphs for the Morlet wavelet multiwavelet filter layer;

[0055] Figure 2 It is a schematic diagram of the overall process of the present invention;

[0056] Figure 3Diagnostic accuracy diagram of different diagnostic models on the self-made variable speed bearing dataset;

[0057] Figure 4 This is a diagram proving the physical meaning of the features extracted by the Morlt wavelet filter layer at a speed of 500 rpm;

[0058] Figure 5 This is a diagram proving the physical meaning of the features extracted by the Morlt wavelet filter layer at a speed of 1000 rpm;

[0059] Figure 6 This is a diagram proving the physical meaning of the features extracted by the Morlt wavelet filter layer at a speed of 1500rpm;

[0060] Figure 7 The rotation frequency extraction results of the local peak search algorithm for IA-2 under accelerated conditions before and after improvement.

[0061] Figure 8 The rotation frequency extraction result diagram of OB-2 under deceleration condition before and after the improvement of the local peak search algorithm. DETAILED DESCRIPTION

[0062] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0063] A bearing fault diagnosis method based on a learnable wavelet physics interpretability network, such as Figure 2 As shown, the specific steps are as follows:

[0064] S1. Obtain the historical time domain vibration signal when the bearing fails.

[0065] In step S1, a variable speed vibration signal is collected by an acceleration sensor.

[0066] S2. Use short-time Fourier transform to process the time-domain vibration signal to obtain a time-frequency diagram, and use a local peak search algorithm to extract speed information from the time-frequency diagram. The specific method is as follows:

[0067] The speed value μ(t1) at the first time point t1 in the time-frequency diagram (TFD) is screened, and the frequency corresponding to the main spectrum peak is found according to the local maximum peak value. The frequency obtained from the smallest to the largest frequency is μ(t1) = [μ 11 ,μ 12 ,…,μ 1n ]. Considering the interference of noise and adjacent components in the time-frequency diagram, a penalty function is added to the local peak search algorithm:

[0068]

[0069] Among them, the local maximum peak of the time-frequency graph is:

[0070] P(t,μ(t))=|TFD(t,μ(t)| 2 (2)

[0071] Combining formulas (1) and (2), the final penalty function of the local peak search algorithm is:

[0072]

[0073] From formula (1) and formula (3), the penalty function of the local peak search algorithm can be obtained as:

[0074]

[0075] in, is the instantaneous rotation frequency finally extracted at the i-th moment; is the lower limit for frequency search; is the upper limit of the frequency search; c1, c2 are constants, μ(t k ) is the frequency value at time point i within the frequency search range; μ(t j ) is the instantaneous frequency extracted from time i and all previous times.

[0076] S3. According to the historical speed information, the key parameters of Morlet wavelet are fitted to construct Figure 1 As shown, the specific method is as follows:

[0077] According to the extracted instantaneous rotation frequency, the Morlet wavelet multi-wavelet filter layer is designed. The frequency domain expression of the Morlet wavelet is:

[0078]

[0079] Among them, β is the bandwidth parameter, which controls the width of the wavelet window in the frequency domain; f0 is the center frequency, which determines the position of the wavelet window on the frequency axis; f is the set frequency range;

[0080] When constructing the Morlet wavelet multi-wavelet filter layer, the bandwidth parameter β i and the center frequency f i Obtained by fitting the extracted rotation frequency ridge:

[0081]

[0082] in, is the intermediate frequency of the ith time interval; n is the time interval divided when extracting the instantaneous rotation frequency; κ i0 , κ i1 , c i They are the bandwidth parameters β i and the center frequency f iThe fitting coefficients of the Morlet wavelet and multi-wavelet filter layers are:

[0083]

[0084] The extracted feature expression of the i-th filter kernel is:

[0085] L i =ψ i (f)*F[x]

[0086] Where x is the original vibration signal and F[·] represents Fourier transform.

[0087] S4. Use Fourier transform to process the historical time domain vibration signal to obtain the historical frequency domain vibration signal.

[0088] S5. The designed Morlet wavelet multi-wavelet filter layer is used as the first layer of LWPI, and the convolution block is used as the second layer. The CNN convolution block includes a convolution layer, a BN layer, a global pooling layer GAP and an activation function ReLU to form a convolution block, and its expression is:

[0089] Z = GAP(BN(ReLU(W×L+b)))

[0090] Among them, W and b are the weight and bias of the convolution layer respectively, L is the fault feature extracted by the Morlet wavelet multi-wavelet filter layer; Z is the high-dimensional feature output by the convolution block;

[0091] The fully connected layer flattens the fault features extracted by the convolutional layer, and the expression is:

[0092] Z←flattening(Z)

[0093] q=ω z Z+b z

[0094] Among them, W z and b z are the weights and biases of the classifier. The softmax function is used to classify the input features. The expression is:

[0095]

[0096] Among them, S is the number of fault categories, o i is the predicted probability of the i-th category;

[0097] The cross-entropy loss is used to measure the distribution of predicted labels and true labels. Its expression is:

[0098]

[0099] Among them, r i is the true probability of the i fault category.

[0100] The historical frequency domain vibration signal information is randomly divided into a training set and a test set;

[0101] During the training process, according to the chain rule, the weights of the convolution kernels and the key coefficients of the Morlet wavelet multi-wavelet filter are optimized in the back-propagation process to obtain the trained LWPI network model.

[0102] The test set is input into the trained LWPI network model to realize fault diagnosis.

[0103] S6. Collect the current time domain vibration signal of the bearing through the acceleration sensor; process the current time domain vibration signal by Fourier transform to obtain the current frequency domain vibration signal; process the current frequency domain vibration signal with the designed Morlet wavelet to extract the current preliminary diagnosis signal; identify the current preliminary diagnosis signal with the trained CNN convolution block to obtain the final diagnosis result.

[0104] To verify the effectiveness, the method was validated on two sets of data sets, including the homemade variable speed bearing data set and the Ottawa variable speed bearing data set. The homemade bearing data set contains two experimental data sets, one is conducted at time-varying speed, and the other is conducted at different constant speeds, including 500, 1000 and 1500rpm. In both experiments, there are 8 health types, including 1) ball fault (BF; 2) composite fault of inner raceway and ball IB; 3) inner raceway fault IF; 4) composite fault of inner and outer raceways IO; 5) normal NO; 6) composite fault of outer raceway and ball OB; 7) outer raceway fault OF; 8) composite fault of ball, inner raceway and outer raceway BIO. In the first experiment, the vibration signals of eight healthy types of bearings were collected in the acceleration T1 and deceleration T2 stages respectively. The sensor is used to collect the vibration signal, and the sampling frequency is set to 12kHz. For each health condition, the bearing vibration signal is divided into a total of 3656 samples by the sliding window technique, and the window width is 3072 points. These time domain signals are converted into frequency spectra by fast Fourier transform (FFT). Considering the conjugate property of FFT data, the first half of the FFT data points are used as input. The ratio of training samples to test samples is set to 7:3.

[0105] In order to further verify the superiority of the model, five classic and cutting-edge models are compared: MWA-CNN, GTFE-Net, Morlet-LeNet and Morlet-AlexNet and MA1DCNN. At the same time, two ablation experimental models are constructed. One is the Morlet wavelet in the Morlet wavelet filter layer without parameter learning function FF-CNN, and the other is the entire network structure without Morlet wavelet filter layer FL-CNN. The diagnostic results of different models under two variable speed conditions are shown in Figure 2. Figure 3 As shown in the figure, it can be seen that the proposed network structure LWPI diagnosis accuracy is higher than the five classic and cutting-edge models, and can effectively improve the accuracy of bearing fault diagnosis under variable speed. The physical meaning of the features extracted by the Morlet wavelet multi-wavelet filter layer is as follows Figures 4 to 6 As shown, Figures 4 to 6 The extraction of Morlet wavelet filter layer under different constant speeds is given respectively. It can be seen from the figure that the vibration signals extracted under different constant speeds have obvious fault characteristic frequencies, which shows that the features extracted by Morlet wavelet filter layer have significant physical significance. Therefore, the convolutional neural network can be endowed with certain physical significance.

[0106] In order to further verify the generalization of the model, further verification was carried out on the Ottawa variable speed bearing dataset. The vibration signals collected from the Ottawa variable speed bearing dataset under acceleration XA-2 and deceleration XB-2 were used. In this dataset, four types of bearings were used in the diagnostic experiment, including inner raceway fault, outer raceway fault, ball fault and normal. In each health state, the acceleration signal was collected for 10s at a sampling frequency of 20kHz. The signal was then split into 650 non-overlapping samples, each with a length of 3072. The ratio of the training set to the test set was also divided into 7:3.

[0107] The diagnostic accuracy of different models under two variable speed conditions is shown in the following table:

[0108]

[0109] The experimental results show that the proposed model has higher diagnostic accuracy than the five classical and cutting-edge models, further verifying the effectiveness and superiority of the proposed model. In addition, the local peak search algorithm before and after the improvement is also compared on this dataset, such as Figure 7 and Figure 8 As shown, Figure 7 and Figure 8The figures show the rotation frequency extraction results of the improved local peak search algorithm under the acceleration XA-2 and deceleration XB-2 speed conditions. It can be seen from the figure that the improved local peak search algorithm can effectively avoid the interference of noise and adjacent components and accurately extract the rotation frequency.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A variable speed bearing fault diagnosis method based on wavelet-driven physically interpretable network, characterized in that: The specific method is as follows: The interpretability network is constructed with Morlet wavelet multi-wavelet filter layer as the first layer and CNN convolution block as the second layer; Obtain historical time domain vibration signals when bearing fails; The historical time-domain vibration signal is processed by Fourier transform to obtain the historical frequency-domain vibration signal; Train an interpretability network with historical frequency-domain vibration signals; The current time domain vibration signal of the bearing is collected by the acceleration sensor, and the variable speed bearing fault is diagnosed by the trained interpretable network.

2. The variable speed bearing fault diagnosis method of wavelet driven physically interpretable network as claimed in claim 1 is characterized in that: The key parameters of the interpretability network are determined by the historical time-domain vibration signal as follows: The historical time-domain vibration signal is processed by short-time Fourier transform to obtain the historical time-frequency diagram, and the historical speed information is extracted from the historical time-frequency diagram by using an improved local peak search algorithm. The key parameters of the Morlet wavelet and multi-wavelet filter layers are fitted based on the historical speed information.

3. The variable speed bearing fault diagnosis method of wavelet driven physically interpretable network as claimed in claim 2 is characterized in that: The improved local peak search algorithm is used to extract historical speed information from the historical time-frequency diagram. The specific method is as follows: The speed value μ(t1) at the first time point t1 in the historical time-frequency diagram is screened, and the frequency corresponding to the main spectrum peak is found according to the local maximum peak value. The frequency obtained from the smallest to the largest frequency is μ(t1)=[μ 11 ,μ 12 ,…,μ 1n ], taking into account the interference of noise and adjacent components in the time-frequency diagram, a penalty function is added to the local peak search algorithm: Among them, the local maximum peak of the time-frequency graph is: P(t,μ(t))=|TFD(t,μ(t)| 2 The final penalty function of the local peak search algorithm is: Through the final penalty function of the local peak search algorithm, find the instantaneous rotation frequency corresponding to the minimum penalty function: in, is the instantaneous rotation frequency finally extracted at the i-th moment; is the lower limit for frequency search; is the upper limit of the frequency search; c1, c2 are constants, μ(t k ) is the frequency value at time point i within the frequency search range; μ(t j ) is the instantaneous frequency extracted from time i and all previous times.

4. The variable speed bearing fault diagnosis method of wavelet driven physically interpretable network as claimed in claim 3 is characterized in that: The key parameters of the Morlet wavelet and multi-wavelet filter layer are fitted according to the historical speed information. The specific method is as follows: According to the extracted instantaneous rotation frequency, the Morlet wavelet multi-wavelet filter layer is designed. The frequency domain expression of the Morlet wavelet is: Among them, β is the bandwidth parameter, which controls the width of the wavelet window in the frequency domain; f0 is the center frequency, which determines the position of the wavelet window on the frequency axis; f is the set frequency range; When constructing the Morlet wavelet multi-wavelet filter layer, the bandwidth parameter β i and the center frequency f i Obtained by fitting the extracted rotation frequency ridge: in, is the intermediate frequency of the ith time interval; n is the time interval divided when extracting the instantaneous rotation frequency; κ i0 , κ i1 , c i They are the bandwidth parameters β i and the center frequency f i The fitting coefficients of the Morlet wavelet and multi-wavelet filter layers are: The extracted feature expression of the i-th filter kernel is: L i =ψ i (f)*F[x] Where x is the original vibration signal and F[·] represents Fourier transform.

5. The variable speed bearing fault diagnosis method of wavelet driven physically interpretable network as claimed in claim 1 is characterized in that: The CNN convolution block consists of a convolution layer, a BN layer, a global pooling layer GAP and an activation function ReLU, which is expressed as: Z = GAP(BN(ReLU(W×L+b))) Among them, W and b are the weight and bias of the convolution layer respectively, L is the fault feature extracted by the Morlet wavelet multi-wavelet filter layer; Z is the high-dimensional feature output by the convolution block; The fully connected layer flattens the fault features extracted by the convolutional layer, and the expression is: Z←flattening(Z) q=ω z Z+b z Among them, W z and b z are the weights and biases of the classifier. The softmax function is used to classify the input features. The expression is: Among them, S is the number of fault categories, o i is the predicted probability of the i-th category; The cross-entropy loss is used to measure the distribution of predicted labels and true labels. Its expression is: Among them, r i is the true probability of the i fault category.

6. The variable speed bearing fault diagnosis method of wavelet driven physically interpretable network as claimed in claim 5 is characterized in that: The interpretability network is trained with historical frequency domain vibration signals. The specific method is as follows: The historical frequency domain vibration signals are randomly divided into a training set and a test set; During the training process, according to the chain rule, the weights of the convolution kernels and the key coefficients of the Morlet wavelet multi-wavelet filter layer are optimized in the back-propagation process to obtain a trained interpretable network.

Images