Gearbox compound fault diagnosis method and storage medium

Abstract

The application discloses a kind of gear box compound fault diagnosis method, storage medium, it is related to fault diagnosis technical field.Gear box compound fault diagnosis method includes: using improved feature mode decomposition algorithm respectively to the gear box vibration signal under multiple working conditions is mode decomposition, obtains data sample, wherein, data sample includes source domain data with label and target domain data without label;Using source domain data and its corresponding label pre-training fault diagnosis model, wherein, fault diagnosis model includes multi-scale attention enhancement network and Transformer network;Based on feature library and pre-trained fault diagnosis model, target domain data is carried out double migration learning, realizes the fault diagnosis of target domain data, wherein, feature library includes the source domain feature that pre-trained fault diagnosis model is carried out to source domain data Feature extraction obtains.The method can realize the accurate diagnosis of gear box compound fault.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fault diagnosis, and in particular to a gearbox compound fault diagnosis method and a storage medium. BACKGROUND

[0002] As a core component of mechanical transmission system, gearbox is widely used in wind power, rail transportation, aerospace and many other fields, and its running state is directly related to the safety and reliability of the equipment. Under complex working conditions such as variable load, high speed rotation and high temperature, gearbox is prone to compound faults such as gear wear and bearing crack. With the advancement of equipment intelligence, it has become a research focus in industry and academia to realize the diagnosis of new or rare faults of gearbox.

[0003] Currently, traditional fault diagnosis methods face many difficulties: on the one hand, the gearbox vibration signal has nonlinear and non-stationary characteristics, and traditional signal processing methods such as wavelet transform and Fourier transform cannot effectively extract the weak fault features. On the other hand, deep learning-based methods highly depend on large-scale labeled data for training, and their diagnosis ability is obviously insufficient when facing unseen fault types. Existing zero-shot diagnosis techniques, such as attribute-based transfer learning, rely on expert-defined fault attributes, and their feature extraction is not optimized for gearbox compound faults, which cannot meet the actual engineering requirements. In addition, most existing methods are based on single working condition data, and the utilization of multi-source information is not sufficient, which limits the generalization ability of the model. SUMMARY

[0004] The purpose of the present application is to provide a gearbox compound fault diagnosis method and a storage medium to realize accurate diagnosis of gearbox compound faults.

[0005] In a first aspect, the present application provides a gearbox compound fault diagnosis method, which comprises: using an improved feature mode decomposition algorithm to decompose gearbox vibration signals under multiple working conditions to obtain data samples, wherein the data samples include source domain data with labels and target domain data without labels; pre-training a fault diagnosis model using the source domain data and its corresponding labels, wherein the fault diagnosis model includes a multi-scale attention enhancement network and a Transformer network; performing double transfer learning on the target domain data based on a feature library and the pre-trained fault diagnosis model to realize fault diagnosis of the target domain data, wherein the feature library includes source domain features obtained by the pre-trained fault diagnosis model performing feature extraction on the source domain data.

[0006] In some embodiments of the present application, the method further comprises: training the pre-trained fault diagnosis model using the target domain data and its corresponding fault diagnosis result to obtain a trained fault diagnosis model; and using the trained fault diagnosis model to diagnose the compound fault of the gearbox.

[0007] In some embodiments of the present application, the number of decomposition modes and the number of frequency band segments used in the improved characteristic mode decomposition algorithm are obtained based on genetic algorithm optimization.

[0008] In some embodiments of the present application, when the number of decomposition modes and the number of frequency band segments are optimized based on genetic algorithm, chromosomes are formed by binary coding parameters, the correlation kurtosis is used as the fitness function, and the population is iteratively updated by roulette selection, crossover and mutation until the fitness converges, to obtain the optimized number of decomposition modes and the number of frequency band segments.

[0009] In some embodiments of the present application, the multi-scale attention enhancement network first strengthens the key channel features through the channel attention module, then fuses the spatial position correlation information through the position attention module, and fuses the output features of the channel attention module and the position attention module to realize multi-scale feature enhancement.

[0010] In some embodiments of the present application, the output features of the channel attention module and the position attention module are fused by the following formula:

[0011] ;

[0012] wherein, represents the fusion result, represents the output feature of the channel attention module, represents the output feature of the position attention module, represents a scaling factor.

[0013] In some embodiments of the present application, the feature library is further optimized by the following objective function:

[0014] ;

[0015] wherein, represents the source domain feature, D represents the feature library, represents a reconstruction error term, represents a sparse coding coefficient matrix and represents a regularization term.

[0016] In some embodiments of the present invention, the dual transfer learning of the target domain data based on the feature library and the pre-trained fault diagnosis model includes: sequentially performing DBSCAN clustering and k-means optimization on the target domain data to generate pseudo-labels; using the pre-trained fault diagnosis model to extract features from the target domain data to obtain target domain features; performing transfer learning on the target domain features based on the features in the feature library; using the transferred-learned target domain features and the pseudo-labels as feedback, performing k-means optimization again to update the pseudo-labels and the feature library, and returning to the step of performing transfer learning on the target domain features based on the features in the feature library, until the fault diagnosis result of the target domain data is obtained.

[0017] In some embodiments of the present invention, k-means optimization is performed using the following formula:

[0018] ;

[0019] in, This represents the center vector of the j-th cluster. This represents the number of data points in the j-th cluster. This indicates that it belongs to the j-th cluster. The i-th data point in This indicates that the summation operation is performed on all data points in the j-th cluster.

[0020] Secondly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the gearbox composite fault diagnosis method described in the first aspect of the embodiments above.

[0021] The gearbox composite fault diagnosis method and storage medium of this invention employ an improved feature pattern decomposition algorithm to perform pattern decomposition on gearbox vibration signals under multiple operating conditions. It also uses source domain data and its corresponding labels to pre-train a fault diagnosis model including a multi-scale attention enhancement network and a Transformer network. Based on the feature library and the pre-trained fault diagnosis model, it performs dual transfer learning on the target domain data, which can achieve accurate diagnosis of unknown faults in the target domain data. Attached Figure Description

[0022] Figure 1 This is a flowchart of a gearbox composite fault diagnosis method according to an embodiment of the present invention;

[0023] Figure 2 This is a flowchart of a gearbox composite fault diagnosis method according to an example of the present invention;

[0024] Figure 3is a flow chart of an improved feature mode decomposition algorithm of an embodiment of the present application;

[0025] Figure 4 is a structural schematic diagram of a channel attention module of an example of the present application;

[0026] Figure 5 is a structural schematic diagram of a position attention module of an example of the present application;

[0027] Figure 6 is a structural schematic diagram of a double transfer clustering diagnosis module of an embodiment of the present application;

[0028] Figure 7 is a structural schematic diagram of a Transformer network of an example of the present application;

[0029] Figure 8 is a schematic diagram of clustering of an example of the present application. DETAILED DESCRIPTION

[0030] Embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numbers represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are intended to explain the present application, and cannot be understood as limiting the present application.

[0031] A gearbox composite fault diagnosis method and a storage medium of an embodiment of the present application are described below with reference to the accompanying drawings.

[0032] The conventional gearbox fault diagnosis method has difficulty in effectively extracting weak fault features when processing nonlinear and non-stationary vibration signals, and is highly dependent on a large amount of labeled data, and has insufficient diagnosis capability when facing unobserved fault types. At the same time, the existing zero-shot diagnosis technology relies on expert-defined fault attributes, and the feature extraction is not optimized for gearbox composite faults, which cannot meet the actual engineering requirements.

[0033] Therefore, the present application proposes a gearbox composite fault diagnosis method to break through the dependence of the conventional technology on labeled data and significantly improve the diagnosis accuracy and robustness of gearbox composite faults under complex working conditions. The method uses an IFMD (Improved Feature Mode Decomposition) algorithm to preprocess the vibration signal, suppresses noise interference and enhances fault features; with the multi-scale attention mechanism of the multi-scale attention enhancement network, it adaptively captures key frequency band features and realizes feature enhancement; through a double transfer clustering diagnosis method, it completes knowledge transfer and pseudo-label mining, and then realizes effective diagnosis of unobserved fault types of the gearbox.

[0034] Figure 1is a flowchart of a gearbox composite fault diagnosis method of an embodiment of the present application.

[0035] As shown in Figure 1 , the gearbox composite fault diagnosis method comprises:

[0036] S11, using an improved feature mode decomposition algorithm to respectively perform mode decomposition on gearbox vibration signals under multiple working conditions to obtain data samples, wherein the data samples include source domain data with labels and target domain data without labels.

[0037] As shown in Figure 2 , the gearbox vibration signals under multiple working conditions can be collected by vibration displacement sensors x1, x2, …, xm deployed at the input end of the gearbox and vibration displacement sensors y1, y2, …, ym deployed at the output end of the gearbox.

[0038] Due to the large amount of labeling work, in order to reduce the workload, part of the gearbox vibration signals can be labeled for fault and fault type to obtain labeled gearbox vibration signals, and the other part is not labeled, which is unlabeled gearbox vibration signal. Then, the labeled gearbox vibration signals and the unlabeled gearbox vibration signals can be randomly divided into training set and test set according to a certain proportion. For example, the number of labeled gearbox vibration signals is N1, and the number of unlabeled gearbox vibration signals is N2. N1x7 / 10 labeled gearbox vibration signals and N2x7 / 10 unlabeled gearbox vibration signals are used as the training set, and the rest is used as the test set.

[0039] In order to suppress noise and highlight fault features, the IFMD algorithm can be used to preprocess the gearbox vibration signals under multiple working conditions to remove interference information, and the processing result is converted into a frequency domain signal. As shown in Figure 2 , the IFMD algorithm can decompose each gearbox vibration signal into n groups of signals, providing input for the multi-scale attention enhancement network in the subsequent fault diagnosis model.

[0040] As an implementation, the number of decomposition modes and the number of frequency band segments used in the improved feature mode decomposition algorithm are optimized based on a genetic algorithm.

[0041] Exemplarily, when the number of decomposition modes and the number of frequency band segments are optimized based on a genetic algorithm, a chromosome is formed by binary coding parameters, the correlation kurtosis is used as the fitness function, the roulette selection, crossover and mutation are iteratively updated to update the population until the fitness converges, and the optimized number of decomposition modes and the number of frequency band segments are obtained.

[0042] Specifically, the traditional FMD algorithm has important significance in the field of rotating machinery fault diagnosis, but it has certain limitations in parameter selection. The FMD algorithm mainly decomposes the signal by means of an adaptive FIR (‌Finite Impulse Response, finite impulse response) filter, in order to more accurately evaluate the decomposition mode quality and effectively extract the fault feature signal, a CK (Cumulant Kurtosis, correlation kurtosis) is used as the objective function, and its calculation formula is as follows (1):

[0043] (1)

[0044] Wherein, represents the value of the kth point in the nth mode component, represents the value of the kth point in the nth mode component, , N represents the length of the signal, M represents the correlation order set in advance, represents the fault period. This method initializes the FIR filter set through the Hanning window, divides the signal frequency band into K segments, and the cutoff frequency of the kth filter is as follows (2): (2) Wherein, fs represents the sampling frequency,

[0045] , respectively represent the lower cutoff frequency and the upper cutoff frequency of the kth FIR filter.

[0046] However, the decomposition performance of the traditional FMD algorithm is highly dependent on the artificial setting of the mode number n (i.e. the number of mode components) and the frequency band segmentation number K, and has problems such as strong parameter sensitivity and insufficient modal aliasing suppression. In order to solve the parameter dependence problem in the traditional FMD algorithm, the IFMD algorithm in the present application introduces a genetic algorithm to optimize the decomposition mode number n and the frequency band segmentation number K. The specific optimization process is as shown in The parameter combination (n, K) is formed into a chromosome by binary coding, the correlation kurtosis is used as the fitness function, and the roulette selection, crossover and mutation iteration are performed to output the optimal parameter combination (n, K).

[0047] After obtaining (n, K), (n, K) is obtained.

[0048] Figure 3

[0049] Referring to Figure 3 , after obtaining (n, K), (n, K) is obtained. ​​​​​​​​) for feature mode decomposition, so that the IMF (Intrinsic Mode Function) component reflecting the fault feature more accurately can be obtained, thereby improving the quality of the input signal of the subsequent diagnosis process.

[0050] By introducing a genetic algorithm based on a traditional feature mode decomposition algorithm, the number of decomposition layers and filter parameters can be intelligently screened, thereby more efficiently processing the nonlinear and non-stationary vibration signals of the gearbox. The algorithm decomposes the original signal (i.e., the gearbox vibration signal) into multiple feature mode components, each component carrying different scale fault feature information, forming a hierarchical feature mode set, and realizing the transformation from the original signal to the fault feature signal, thereby providing a more valuable data basis for subsequent fault diagnosis.

[0051] As another implementation, the number of decomposition modes and the number of frequency band segments used in the improved feature mode decomposition algorithm are obtained based on an artificial bee algorithm.

[0052] S12, pre-training a fault diagnosis model using source domain data and its corresponding labels, wherein the fault diagnosis model comprises a multi-scale attention enhancement network and a Transformer network.

[0053] In some embodiments of the present application, the multi-scale attention enhancement network first strengthens the key channel features through the channel attention module, then fuses the spatial position correlation information through the position attention module, and fuses the output features of the channel attention module and the position attention module, thereby realizing multi-scale feature enhancement.

[0054] For example, the output features of the channel attention module and the position attention module are fused by the following formula (3):

[0055] (3)

[0056] wherein, denotes the fusion result, denotes the output feature of the channel attention module, denotes the output feature of the position attention module, denotes a scaling factor.

[0057] Specifically, in the self-attention mechanism, after the linear transformation of the input sequence, the query vector Q, the key vector K and the value vector V are generated respectively. Here, is a scalar, which is usually equal to the dimension number of the key vector K, and mainly serves to scale the attention score to reduce the variance. The specific calculation process is as follows: first, the dot product operation of the query vector Q and the key vector K is performed, so that the correlation degree score between all elements can be obtained. Then, divide these scores by The scale is then applied, and the softmax function is used to convert it into probability weights. Finally, these probability weights are weighted and summed with the value vector V to obtain the final output result. The self-attention formula is shown in formula (4).

[0058] (4)

[0059] Next, we will introduce in detail the positional attention module and the channel attention module used in the multi-scale attention enhancement network. The channel attention module aims to select key channel features and achieves feature enhancement along the channel dimension through a systematic process, as shown in the flowchart below. Figure 4 As shown. First, for the input features Perform average pooling (AvgPool) and max pooling (MaxPool) respectively to obtain... and Subsequently, the two pooling results are fused using a shared multilayer perceptron (MLP): ,in, It is the Sigmoid activation function. b1 is the dimensionality-reduced weight matrix (r is the compression ratio, e.g., r=16), and b1 is the bias term. Finally, the generated channel attention weights... Input features Multiplication is used to achieve weighted enhancement of key channels, as shown in equation 5 below.

[0060] (5)

[0061] in, This indicates a channel-wise multiplication operation, outputting features. Channel information, which is more discriminative for fault diagnosis, has been retained.

[0062] After the channel attention module completes the filtering of key channel information for the input features, the output feature A then enters the position attention module to initiate spatial dimension feature optimization. The position attention module focuses on mining spatial positional dependencies; its specific flowchart is shown below. Figure 5 As shown. See also Figure 5 For the output feature A of the channel attention module, features B, C, and D are generated through three sets of convolutions, and then reshaped into... , , Then, calculate the spatial attention matrix. ( (as feature dimension), then through Integrating spatial correlation features ( The learnable coefficient, That is Figure 5E) to achieve the weighted aggregation of each position feature and enhance the discriminability of fault features in the spatial dimension.

[0063] After the channel attention module and the position attention module are processed, a parallel fusion strategy is used to realize multi-dimensional feature enhancement. Specifically, the features A output by the channel attention are added to the features E output by the position attention according to elements, that is, , to retain the key band information of the channel dimension and the position dependence of the spatial dimension. The fusion process can further adjust the weights of the two by a scaling factor , as shown in equation (3) above, to finally generate a dual attention feature containing channel screening and spatial correlation, and obtain a multi-scale discriminative feature to provide a more discriminative input representation for subsequent feature classification or clustering.

[0064] By integrating the relevant features between all channel mappings through the channel attention module, the channel mappings that exist in mutual dependence can be selectively emphasized, effectively highlighting the key channel information; by weighting and aggregating the features at all positions through the position attention module, the features at each position are selectively aggregated to mine the spatial position dependence; by combining the two to realize multi-scale feature enhancement, the obtained fault features are more comprehensive and accurate, and the discriminability of the fault diagnosis model is improved.

[0065] As shown in Figure 6 , the fault labels of the source domain data are input into the Transformer network for feature extraction after classification, and the specific process is shown in Figure 7 . The source sequence (i.e., the sequence output by the source domain data after processing by the multi-scale attention enhancement network) is first mapped to a high dimension by an embedding layer and position encoding . The encoder layer calculates: , , where the multi-head self-attention mechanism captures the long-range dependence within the sequence, and the feed-forward network (FFN) enhances the nonlinear expression ability. The decoder layer processes the target sequence (i.e., the label corresponding to the source domain data) , , , where the masked multi-head self-attention prevents information leakage and cross-attention fusion of encoder and decoder features. Finally, the output is obtained as the prediction result.

[0066] When pre-training the fault diagnosis model, the loss function can be calculated based on the prediction result and the corresponding true label, and the parameters of the fault diagnosis model can be updated based on the loss function.

[0067] S13, based on the feature library and the pre-trained fault diagnosis model, double migration learning is performed on the target domain data to realize fault diagnosis of the target domain data, wherein the feature library comprises source domain features obtained by the pre-trained fault diagnosis model performing feature extraction on the source domain data.

[0068] In some embodiments of the present application, the feature library is further optimized by a target function of formula (6) as follows:

[0069] (6)

[0070] wherein, represents the source domain features, D represents the feature library, represents a reconstruction error term, represents a sparse coding coefficient matrix and represents a regularization term.

[0071] Specifically, the source domain features extracted through the Transformer network will enter the feature library. In order to make the feature library store typical features of various faults in the source domain data, the source domain features in the feature library can be further optimized and processed, including: screening, integrating and learning the extracted source domain features, removing possible noise and redundant information in the source domain features, so that the source domain features in the feature library are more refined and representative. Finally, these optimized source domain features will be stored in the feature library. The target function for optimization processing is formula (6) as follows, The sparse coding coefficient matrix restricts the linear representation ability of D on the source domain features, ensuring the integrity of the fault features; and respectively restrict the redundancy of D and the sparsity of the coding to extract the most discriminative fault mode basis vector.

[0072] In some embodiments of the present application, based on the feature library and the pre-trained fault diagnosis model, double migration learning is performed on the target domain data, including: generating pseudo labels by sequentially performing DBSCAN clustering and k-means optimization on the target domain data; obtaining target domain features by performing feature extraction on the target domain data using the pre-trained fault diagnosis model; performing migration learning on the target domain features based on the features in the feature library; taking the migration-learned target domain features and the pseudo labels as feedback, and performing k-means optimization again to update the pseudo labels and the feature library, and returning to the step of performing migration learning on the target domain features based on the features in the feature library, until the fault diagnosis result of the target domain data is obtained.

[0073] Exemplarily, k-means optimization is performed by formula (7) as follows:

[0074] (7)

[0075] where, denotes the center vector of the jth cluster, denotes the number of data points in the jth cluster, denotes the ith data point belonging to the jth cluster denotes the summation operation on all data points in the jth cluster.

[0076] In this embodiment, Figure 2 the double migration cluster diagnosis module in the source domain knowledge migration and target domain unsupervised learning collaborative construction of zero sample diagnosis, the specific flow chart as Figure 6 shown. The labeled source domain data is extracted by the encoder in the pre-trained fault diagnosis model of the Transformer network Deep fault features (i.e. source domain features) are stored in the feature library; the unlabeled target domain data is first filtered by DBSCAN algorithm and initial clustering, and then the pseudo label is generated by k-means iterative optimization. Subsequently, the target domain data realizes cross-domain knowledge alignment through double migration, combines with the pseudo label feedback to optimize the clustering result, and adjusts the deep layer through the feature library, forms a "feature extraction-knowledge transfer-clustering optimization-iterative learning" closed loop, realizes zero sample fault recognition.

[0077] Specifically, the target domain data is different from the source domain data, which lacks explicit fault labels. To solve this problem, first, the pre-trained fault diagnosis model of the source domain data is used and the target domain data is preprocessed using the DBSCAN algorithm. Among them, DBSCAN is an unsupervised clustering algorithm based on data density, which can divide data into different clusters according to the density relationship between data points, and the DBSCAN point classification diagram is shown in Figure 8 . In the target domain data, different fault modes may exhibit different data distribution characteristics, and the DBSCAN algorithm can discover the rules of these data distribution and preliminarily divide the cluster structure of the data. However, the clustering result obtained by the DBSCAN algorithm may not be compact and accurate enough, so the k-means algorithm will be used to optimize the clustering result, and its calculation formula is formula (7) above.

[0078] ​The k-means algorithm assigns data points to different clusters through iteration, so that the points in the cluster are as close to the cluster center as possible to improve the clustering quality. The pseudo-labels are generated by DBSCAN and k-means processing, and a weak supervision signal is constructed for the target domain data. Then, the target domain data is processed through double transfer learning: on the one hand, the pre-trained fault diagnosis model of the source domain data is reused, and the rich fault feature extraction ability learned by the model is used to quickly extract cross-domain general features, avoiding the tediousness and data insufficiency of training from scratch; on the other hand, the source domain features in the feature library are called, and the typical fault features of the source domain are associated with the target domain features, so that the target domain features learn to approach the source domain knowledge system, and the label missing problem is solved.

[0079] After the transfer learning of the target domain features, the pseudo-labels are fed back to the k-means again, and the clustering distribution is further optimized to accurately reflect the fault mode. The optimized features enter the feature library, and deep learning is performed based on the source domain knowledge and the clustering results, and the feature representation is continuously adjusted to adapt to the fault characteristics of the target domain. Through the iterative closed loop of "feature extraction-knowledge transfer-feature optimization", the source domain knowledge is finally used to effectively diagnose the unknown faults of the target domain in the zero sample scene.

[0080] Through the double transfer clustering diagnosis method, the double collaborative mechanism of parameter transfer and feature knowledge transfer of the pre-trained fault diagnosis model of the source domain data is realized, and the diagnosis difficulty of the unknown faults of the gearbox in the target domain data without labels is effectively solved.

[0081] In some embodiments of the present application, the gearbox compound fault diagnosis method further comprises: training the pre-trained fault diagnosis model using the target domain data and the corresponding fault diagnosis results of the target domain data to obtain a trained fault diagnosis model; and diagnosing the compound fault of the gearbox by using the trained fault diagnosis model.

[0082] Specifically, after obtaining the fault diagnosis results of the target domain data, the target domain data can obtain labels, and at this time, the pre-trained fault diagnosis model can be further trained by using the target domain data and the corresponding labels, so that the trained fault diagnosis model can adapt to the diagnosis of more fault types of the gearbox. In actual use, the trained fault diagnosis model can be directly used to diagnose the compound fault of the gearbox, which is simple and efficient.

[0083] Exemplarily, when the pre-training and / or training of the fault diagnosis model is completed, the test set data can be sequentially passed through the IFMD, the multi-scale attention enhancement network and the double transfer clustering diagnosis module, and the output results of the predicted classification are used to complete the diagnosis of the compound fault of the gearbox.

[0084] For example, after the pre-training of the fault diagnosis model is completed, the pre-trained fault diagnosis model is tested by using a test set corresponding to the labeled source domain data. For another example, after the training of the fault diagnosis model is completed, the pre-trained fault diagnosis model is tested by using a test set corresponding to the labeled source domain data. For another example, after the pre-training of the fault diagnosis model is completed, a first prediction result is obtained by performing double migration learning on the pre-trained fault diagnosis model by using a test set corresponding to the unlabeled source domain data, and after the training of the fault diagnosis model is completed, a second prediction result is obtained by testing the trained fault diagnosis model by using a test set corresponding to the unlabeled source domain data, and the test is realized by comparing the two prediction results.

[0085] Based on the gearbox composite fault diagnosis method of the above-mentioned embodiments, the application provides a computer-readable storage medium.

[0086] In the embodiments of the application, a computer program is stored on a computer-readable storage medium, and the computer program is executed by a processor to implement the gearbox composite fault diagnosis method of the above-mentioned embodiments.

[0087] To sum up, the gearbox composite fault diagnosis method and the storage medium of the embodiments of the application can efficiently diagnose the composite fault of the gearbox, have good stability, are strong in adaptability to complex working conditions, and have excellent generalization ability. Specifically, first, with the help of the improved feature mode decomposition algorithm, the decomposition parameters are optimized by using the genetic algorithm, the nonlinear non-stationary vibration signal is adaptively processed, the fault mode is separated, the noise interference is suppressed, and the decomposition result containing the fault feature is obtained, so that the quality and efficiency of fault feature extraction are greatly improved. Then, the decomposition result is converted into a frequency domain feature matrix, which is input into a multi-scale attention enhancement network, the key frequency band features are dynamically captured through the parallel structure of channel and position attention, multi-scale feature enhancement is realized, and the fault feature is more comprehensive and accurate. Finally, the enhanced features are input into a double migration clustering diagnosis module, the source domain knowledge migration and target domain unsupervised learning are cooperated, the feature library migration and pseudo-label mining mechanism are used, the dependence on labeled data is broken, the diagnosis of the gearbox fault type that has not been seen is realized, the diagnosis precision and robustness of the gearbox composite fault under complex working conditions are effectively improved, the fault diagnosis demand of actual engineering is met, and under the multiple actions of fault feature extraction, enhancement and knowledge migration, the composite fault of the gearbox can be more accurately diagnosed.

[0088] It is to be appreciated that the above description and the examples that follow are intended to be illustrative only and that changes can be made to the description, as represented by the above listed elements, by the steps recited in the flow charts, and by the examples that follow, without departing from the spirit of the application. Accordingly, the scope of the present application is intended to be defined only by the appended claims.

[0089] It should be understood that aspects of the application can be implemented in hardware, software, firmware or a combination thereof. In the above embodiments, various steps or methods can be implemented in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, can be used: a hybrid of the above technologies, a combination of any of the above technologies, etc.

[0090] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Also, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in an appropriate manner.

[0091] In the description of the application, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the application.

[0092] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise explicitly specified and limited.

[0093] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements, unless otherwise explicitly limited. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0094] In the present application, unless otherwise explicitly specified and limited, the first feature "on" or "under" the second feature can be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, the first feature "above", "above" and "above" the second feature can be that the first feature is directly above or obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "below" and "below" the second feature can be that the first feature is directly below or obliquely below the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature.

[0095] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as a limitation on the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A gear box compound fault diagnosis method, characterized in that, The method comprises the following steps: The improved feature mode decomposition algorithm is used to decompose the vibration signals of the gearbox under multiple working conditions to obtain data samples, wherein the data samples include source domain data with labels and target domain data without labels; A pre-trained fault diagnosis model is trained using the source domain data and its corresponding labels, wherein the fault diagnosis model comprises a multi-scale attention enhancement network and a Transformer network; Double transfer learning is performed on the target domain data based on a feature library and the pre-trained fault diagnosis model to realize fault diagnosis of the target domain data, wherein the feature library comprises source domain features obtained by the pre-trained fault diagnosis model performing feature extraction on the source domain data; The double transfer learning of the target domain data based on the feature library and the pre-trained fault diagnosis model comprises the following steps: DBSCAN clustering and k-means optimization are sequentially performed on the target domain data to generate pseudo labels; target domain features are extracted from the target domain data by using the pre-trained fault diagnosis model; transfer learning is performed on the target domain features based on the features in the feature library; the pseudo labels and the target domain features after transfer learning are used as feedback to perform k-means optimization again to update the pseudo labels and the feature library, and the step of performing transfer learning on the target domain features based on the features in the feature library is returned until the fault diagnosis result of the target domain data is obtained.

2. The gear box compound fault diagnostic method of claim 1, wherein, The method further comprises the following steps: The pre-trained fault diagnosis model is trained using the target domain data and its corresponding fault diagnosis result to obtain a trained fault diagnosis model; The trained fault diagnosis model is used to diagnose the compound fault of the gearbox.

3. The gear box compound fault diagnostic method of claim 1, wherein, The number of decomposition modes and the number of frequency band segments used in the improved feature mode decomposition algorithm are obtained based on a genetic algorithm.

4. The gear box compound fault diagnostic method of claim 3, wherein, When the number of decomposition modes and the number of frequency band segments are optimized based on the genetic algorithm, chromosomes are formed by binary coding parameters, the correlation kurtosis is used as the fitness function, the roulette wheel selection, crossover and mutation are iteratively updated to update the population until the fitness converges, and the optimized number of decomposition modes and the number of frequency band segments are obtained.

5. The gear box compound fault diagnostic method of claim 1, wherein, The multi-scale attention enhancement network first strengthens the key channel features through a channel attention module, then fuses the spatial position correlation information through a position attention module, and finally fuses the output features of the channel attention module and the position attention module to realize multi-scale feature enhancement.

6. The gear box compound fault diagnostic method of claim 5, wherein, The output features of the channel attention module and the position attention module are fused by the following formula: ； wherein, denotes the fusion result, denotes the output feature of the channel attention module, denotes the output feature of the position attention module, denotes a scaling factor.

7. The gear box compound fault diagnostic method of claim 1, wherein, The feature library is further optimized by the following objective function: ； wherein, denotes the source domain feature, D denotes the feature library, denotes a reconstruction error term, denotes a sparse coding coefficient matrix and denotes a regularization term.

8. The gear box compound fault diagnostic method of claim 1, wherein, k-means optimization is performed by the following formula: ； in, This represents the center vector of the j-th cluster. This represents the number of data points in the j-th cluster. This indicates that it belongs to the j-th cluster. The i-th data point in This indicates that the summation operation is performed on all data points in the j-th cluster.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, the gearbox compound fault diagnosis method according to any one of claims 1-8 is realized.

Images