Trusted mechanical fault diagnosis method, system and device based on multi-modal data interpolation and fusion and medium
Through the combination of L2 space-time generation adversarial interpolation network and multi-input single-output autoencoder, the shortcomings of data interpolation and fusion in the existing technology are solved, high-quality data interpolation and fusion are achieved, and mechanical equipment is ensured to be able to Trust fault diagnosis.
Patent Information
- Application Number
- CN202410981340.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-22
AI Technical Summary
The existing data interpolation method cannot effectively capture the long-term dependencies of time series data, resulting in low quality of interpolation data, and data fusion method has data redundancy, increased dimensions, and low computing efficiency, making it impossible to achieve trusted fault diagnosis of mechanical equipment.
The L2 space-time generation adversarial interpolation network is used as the data interpolation model, and data fusion is combined with a multi-input single-output autoencoder. By constructing a spatiotemporal feature extraction module and a multi-head attention mechanism, the time and space characteristics of the data are considered to be used to realize the data. High-quality interpolation and fusion.
The quality of interpolated data is improved, the data distribution capture problem is solved, the data redundancy and inefficient computing efficiency are reduced, and the trusted fault diagnosis of mechanical equipment is realized.
Smart Images

Figure CN118940172A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of electrical digital data processing, and in particular to a reliable mechanical fault diagnosis method, system, device and medium based on multimodal data interpolation and fusion. Background Art
[0002] In practical engineering applications, the integrity and accuracy of sensor data are crucial to the operation monitoring and performance prediction of mechanical equipment systems. However, problems such as sensor failure, communication interruption, and environmental interference can easily lead to abnormal sensor data collection, among which the problem of sensor data loss is particularly prominent. Data loss refers to the lack or loss of some data, which makes the monitoring sensor data incomplete, which may affect the condition monitoring, fault diagnosis and prediction results of the mechanical equipment system, thereby reducing the reliability and performance of the equipment. Therefore, it is necessary to adopt effective data processing and interpolation methods to deal with the problem of sensor data loss in the process of mechanical equipment monitoring, improve the integrity and reliability of sensor detection data, and ensure the safe and stable operation of the equipment system.
[0003] Common data processing methods include deleting missing values and interpolation. Deleting missing values means directly deleting the missing values in the data. This method is suitable for situations where there is not much missing data and the impact of deleting missing values is small. Interpolation estimates missing data through known data. This method is suitable for situations where there is a lot of missing data and there is a certain correlation between missing values. Machine learning methods such as K-nearest neighbor and random forest use the statistical characteristics of time series in a complex way to achieve data interpolation.
[0004] Deep learning methods such as TCN (Temporal Convolutional Network) and GAIN (Generative Adversarial Imputation Network) extract data features and model missing values to interpolate missing values. Taking the Generative Adversarial Imputation Network as an example, the missing data, noise vector and mask vector are input into the generator, and the generator outputs the interpolated data. The interpolated data and the prompt mechanism are then input into the discriminator, and the discriminator outputs the probability of each component of the interpolated data being estimated to be true. After the two are trained adversarially, the generator can achieve high-quality interpolation of missing data. By introducing additional information for missing data, the Generative Adversarial Imputation Network can achieve data interpolation.
[0005] Data fusion methods such as data level fusion methods, feature level fusion methods, and decision level fusion methods can comprehensively utilize diverse information sources, improve the comprehensive ability and effect of data analysis, and thus provide more comprehensive support for fault diagnosis and prediction.
[0006] Based on the above description, although the data interpolation method has achieved certain results, it still has limitations. Directly deleting missing values will lead to a reduction in the amount of data on the one hand, and on the other hand, important data information will be lost, thus affecting the accuracy of the analysis results. Although the interpolation method can fill in missing values, it cannot consider the potential correlation and time dependency between data. Machine learning methods use the statistical characteristics of time series to interpolate data, but when the proportion of missing data is large, machine learning methods still cannot effectively capture the distribution of data. The generative adversarial interpolation network cannot take into account the long-term dependency of time series data, resulting in low quality of the generated interpolated data. In addition, the data-level fusion method has deficiencies such as data redundancy, increased dimensions, and low computational efficiency, which in turn makes it impossible to achieve reliable fault diagnosis of mechanical equipment. Summary of the invention
[0007] The purpose of this application is to provide a trusted mechanical fault diagnosis method, system, equipment and medium based on multimodal data interpolation and fusion, which can realize trusted fault diagnosis of mechanical equipment by considering long-term data dependencies and comprehensively utilizing data interpolation and fusion methods of multimodal information.
[0008] To achieve the above objectives, this application provides the following solutions.
[0009] In a first aspect, the present application provides a reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion, comprising:
[0010] Obtain vibration data and current data of the mechanical equipment to be diagnosed.
[0011] According to the operation cycle of the mechanical equipment to be diagnosed, a vibration signal is intercepted from the vibration data, and a current signal is intercepted from the current data.
[0012] Construct a data interpolation model and a data fusion model. The data interpolation model is a trained L2 spatiotemporal generative adversarial interpolation network. The data in the data set used when training the L2 spatiotemporal generative adversarial interpolation network are obtained by simulating the continuous loss of historical operating data under different mechanical health conditions. The historical operating data includes historical vibration data and historical current data. The data fusion model is a trained multi-input single-output autoencoder. The multi-input single-output autoencoder includes multiple encoders and a decoder.
[0013] The vibration signal and the current signal are respectively input into the data interpolation model to obtain vibration signal interpolation data and current signal interpolation data.
[0014] The vibration signal interpolation data and the current signal interpolation data are input into the data fusion model to obtain fused data.
[0015] A fault diagnosis model is constructed, and the fusion data is input into the fault diagnosis model to obtain a fault diagnosis result of the mechanical equipment to be diagnosed.
[0016] Optionally, constructing the data interpolation model and the data fusion model includes:
[0017] According to the operation cycle of the mechanical equipment, a historical vibration signal and a historical current signal are respectively intercepted from the historical vibration data and the historical current data.
[0018] Continuous value loss simulation is performed on the historical vibration signal and the historical current signal respectively to obtain a historical abnormal vibration signal and a historical abnormal current signal.
[0019] A first training set and a first test set are constructed based on the historical abnormal vibration signal and the historical abnormal current signal.
[0020] Construct an L2 spatiotemporal generative adversarial interpolation network. The L2 spatiotemporal generative adversarial interpolation network is a generative adversarial interpolation network including an L2 regularization function and a spatiotemporal feature extraction module.
[0021] The first training set is used to train the L2 space-time generative adversarial interpolation network until a first set condition is met, and the trained L2 space-time generative adversarial interpolation network is used as the data interpolation model.
[0022] The first test set is input into the data interpolation model to obtain historical vibration signal interpolation data and historical current signal interpolation data.
[0023] A second training set is constructed based on the historical vibration signal interpolation data and the historical current signal interpolation data.
[0024] Build a multi-input single-output autoencoder.
[0025] The multi-input single-output autoencoder is trained using the second training set until a second set condition is met, and the trained multi-input single-output autoencoder is used as the data fusion model.
[0026] Optionally, construct an L2 spatiotemporal generative adversarial interpolation network, specifically including:
[0027] The multi-head attention mechanism is added to the residual block of the temporal convolutional network to form a multi-head attention residual block.
[0028] A plurality of the multi-head attention residual blocks are connected in series to form the spatiotemporal feature extraction module.
[0029] The spatiotemporal feature extraction module is fed into the generator of the generative adversarial interpolation network to obtain an improved generator.
[0030] The L2 regularization function and the adversarial loss function of the original generator in the generative adversarial interpolation network are combined to obtain the loss function of the improved generator to complete the construction of the L2 spatiotemporal generative adversarial interpolation network.
[0031] Optionally, the first setting condition is to improve the generator and the discriminator to reach Nash equilibrium.
[0032] Optionally, the loss function of the improved generator is expressed as:
[0033] L G =αL adv_G +γL 2-norm .
[0034] Where, L adv_G represents the adversarial loss function, L 2-norm represents the L2 regularization function, α represents the weight coefficient of the adversarial loss function, γ represents the weight coefficient of the L2 regularization function, L G Represents the loss function of the improved generator.
[0035] Optionally, inputting the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data specifically includes:
[0036] The vibration signal interpolation data and the current signal interpolation data are respectively input into the encoder to obtain vibration data features and current data features.
[0037] The vibration data feature and the current data feature are fused to obtain a fused feature.
[0038] The decoder restores the dimension of the fused feature to the dimension of the vibration signal interpolation data and the current signal interpolation data to obtain the fused data.
[0039] Optionally, a fault diagnosis model is constructed, specifically including:
[0040] A second test set is constructed based on the historical vibration signal interpolation data and the historical current signal interpolation data.
[0041] The second test set is input into the data fusion model to obtain historical fusion data.
[0042] A third training set and a third test set are constructed based on the historical fusion data.
[0043] Construct a convolutional neural network model, and use the third training set and the third test set to train and test the convolutional neural network model respectively, until the third set condition is met, and use the trained convolutional neural network model as the fault diagnosis model.
[0044] In a second aspect, the present application provides a trusted mechanical fault diagnosis system based on multimodal data interpolation and fusion, and the trusted mechanical fault diagnosis system based on multimodal data interpolation and fusion is used to implement the steps of the trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion. The system includes:
[0045] The data acquisition module is used to acquire vibration data and current data of the mechanical equipment to be diagnosed.
[0046] A signal interception module is used to intercept a vibration signal from the vibration data and to intercept a current signal from the current data according to the operation cycle of the mechanical equipment to be diagnosed.
[0047] The interpolation fusion model construction module is used to construct a data interpolation model and a data fusion model. The data interpolation model is a trained L2 space-time generative adversarial interpolation network. The data in the data set used when training the L2 space-time generative adversarial interpolation network are obtained by continuous loss simulation of historical operating data in different mechanical health states. The historical operating data includes historical vibration data and historical current data. The data fusion model is a trained multi-input single-output autoencoder. The multi-input single-output autoencoder includes multiple encoders and a decoder.
[0048] The data interpolation module is used to input the vibration signal and the current signal into the data interpolation model respectively to obtain vibration signal interpolation data and current signal interpolation data.
[0049] The data fusion module is used to input the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data.
[0050] The fault diagnosis module is used to construct a fault diagnosis model and input the fusion data into the fault diagnosis model to obtain the fault diagnosis result of the mechanical equipment to be diagnosed.
[0051] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-mentioned trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion.
[0052] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion.
[0053] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion.
[0054] According to the specific embodiments provided in this application, this application discloses the following technical effects:
[0055] The present application provides a trusted mechanical fault diagnosis method, system, device and medium based on multimodal data interpolation and fusion. The method uses a trained L2 space-time generative adversarial interpolation network as a data interpolation model, and the data in the data set used when training the L2 space-time generative adversarial interpolation network are obtained by continuous loss simulation of historical operating data in different mechanical health states, taking into account the long-term dependency of time series data, solving the problem of not being able to effectively capture data distribution, and effectively improving the quality of interpolated data. The method uses a data fusion model to fuse interpolated data, and uses a fault diagnosis model to obtain the fault diagnosis result of the mechanical equipment to be diagnosed based on the fused data, which solves the problems of data redundancy, increased dimensionality, low computational efficiency, etc. in the existing data fusion method, and realizes trusted fault diagnosis of mechanical equipment. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0057] Figure 1 This is an application environment diagram of a trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion in one embodiment of the present application.
[0058] Figure 2 A flowchart of a trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion is provided in accordance with an embodiment of the present application.
[0059] Figure 3 A schematic diagram of the implementation process of a trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion provided in one embodiment of the present application.
[0060] Figure 4 This is an overall structural diagram of L2-TSGAIN provided in another embodiment of the present application.
[0061] Figure 5 This is an overall structural diagram of the TSFEM provided in one embodiment of the present application.
[0062] Figure 6This is an overall structural diagram of MISO-AE provided in one embodiment of the present application.
[0063] Figure 7 This is an overall structural diagram of a CNN provided in one embodiment of the present application.
[0064] Figure 8 A schematic diagram of the overall implementation flow of a trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion provided in another embodiment of the present application.
[0065] Fig. 9 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0066] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0067] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0068] The trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the vibration data and motor current data of the mechanical equipment to be diagnosed to the server 104. After the server 104 receives the vibration data and motor current data of the mechanical equipment to be diagnosed, for the vibration data and motor current data of the mechanical equipment to be diagnosed, the server 104 intercepts the vibration signal from the vibration data according to the operation cycle of the mechanical equipment to be diagnosed, and intercepts the current signal from the current data. The vibration signal and the current signal are respectively input into the constructed data interpolation model to obtain the vibration signal interpolation data and the current signal interpolation data. The vibration signal interpolation data and the current signal interpolation data are input into the constructed data fusion model to obtain the fusion data. The fusion data is input into the constructed fault diagnosis model to obtain the fault diagnosis result of the mechanical equipment to be diagnosed. The server 104 can feedback the obtained fault diagnosis result of the mechanical equipment to the terminal 102. In addition, in some embodiments, the trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly perform fault diagnosis on the vibration data and motor current data of the mechanical equipment to be diagnosed, or the server 104 can obtain the vibration data and motor current data of the mechanical equipment to be diagnosed from the data storage system and perform fault diagnosis on the vibration data and motor current data of the mechanical equipment to be diagnosed.
[0069] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.
[0070] In an exemplary embodiment, Figure 2 As shown, a reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 200 to 205. Among them:
[0071] Step 200: Obtain vibration data and current data of the mechanical equipment to be diagnosed, for example, obtain vibration data and current data of the planetary transmission system to be diagnosed.
[0072] Step 201: According to the operation cycle of the mechanical equipment to be diagnosed, a vibration signal is intercepted from the vibration data, and a current signal is intercepted from the current data.
[0073] Step 202: Construct a data interpolation model and a data fusion model. The data interpolation model is a trained L2 space-time generative adversarial interpolation network. The data in the data set used to train the L2 space-time generative adversarial interpolation network are obtained by simulating continuous loss of historical operating data under different mechanical health states. For example, 10%, 20%, and 30% of the overall sample points of the vibration signal and the current signal are randomly set to 0 to simulate the abnormal situation of sensor data with missing data. In this simulation process, the missing part of the abnormal data is continuously missing. The historical operating data includes historical vibration data and historical current data. The data fusion model is a trained multi-input single-output autoencoder. The multi-input single-output autoencoder includes multiple encoders and one decoder.
[0074] Step 203: Input the vibration signal and the current signal into the data interpolation model respectively to obtain vibration signal interpolation data and current signal interpolation data. For example, input the vibration data and current data of the planetary transmission system to be diagnosed into the trained L2-TSGAIN (i.e., the data interpolation model) to obtain interpolated vibration data and interpolated current data.
[0075] Step 204: Input the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data.
[0076] Step 205: construct a fault diagnosis model, and input the fused data into the fault diagnosis model to obtain the fault diagnosis result of the mechanical equipment to be diagnosed.
[0077] By implementing the above steps 200 to 205, the present application uses the trained L2 spatiotemporal generative adversarial interpolation network as a data interpolation model, and the data in the data set used when training the L2 spatiotemporal generative adversarial interpolation network is obtained by simulating continuous loss of historical operating data in different mechanical health states, taking into account the long-term dependency of time series data, solving the problem of not being able to effectively capture data distribution, and effectively improving the quality of interpolated data. By using a data fusion model to fuse interpolated data, and using a fault diagnosis model to obtain the fault diagnosis results of the mechanical equipment to be diagnosed based on the fused data, the problems of data redundancy, increased dimensions, and low computational efficiency existing in the existing data fusion methods are solved, and reliable fault diagnosis of mechanical equipment is achieved.
[0078] In another exemplary embodiment of the present application, taking the acquisition of historical vibration data and historical motor current data of the planetary transmission system under different health states as an example, the specific construction process of the data interpolation model and the data fusion model in this step is described as follows: Figure 3 As shown, specifically:
[0079] (1) According to the operation cycle of the mechanical equipment, the historical vibration signal and the historical current signal are respectively intercepted from the historical vibration data and the historical current data.
[0080] For example, the vibration data and current sensor data under different health states are obtained by installing an acceleration sensor and a motor current clamp on the planetary transmission system, and the vibration data and current sensor data under different health states are intercepted according to the operation cycle of the planetary transmission system to obtain several historical vibration signals. And historical current signal in, represents a single historical vibration signal, Represents a single current signal, i=1,2,...,n.
[0081] Based on this, a fault data set can be obtained, which mainly includes the vibration signal and current signal data of the planetary transmission system, and these data can reflect the different health states of the planetary transmission system. The different health states of the planetary transmission system include normal state, input shaft bearing inner ring fault, input shaft bearing outer ring fault, sun gear fault, outer gear ring fault, etc.
[0082] (2) Perform continuous value loss simulation on the historical vibration signal and the historical current signal respectively to obtain the historical abnormal vibration signal and the historical abnormal current signal.
[0083] For example, the historical vibration signal X v And the historical current signal X c 10%, 20%, and 30% of the overall sample points are randomly set to 0 to simulate the abnormal situation of sensor data with continuous missing data.
[0084] Based on the description of step (1) and step (2), in this embodiment, abnormal data of different health states are collected by sensors installed on the planetary transmission system.
[0085] (3) Construct a first training set and a first test set based on historical abnormal vibration signals and historical abnormal current signals.
[0086] Based on the description of steps (1) and (2) above, the training set and test set of the historical abnormal vibration signal finally obtained are respectively recorded as The training set and test set of historical abnormal current signals are denoted as in, represents a single historical abnormal vibration signal in the training set, represents a single historical abnormal vibration signal in the test set, represents a single historical abnormal current signal in the training set, Represents a single historical abnormal current signal in the test set, i=1,2,...,n.
[0087] (4) Construct an L2 temporal-spatial generative adversarial imputation network (L2-TSGAIN). The L2 temporal-spatial generative adversarial imputation network is a generative adversarial imputation network that includes an L2 regularization function and a spatiotemporal feature extraction module. The structure of the L2 temporal-spatial generative adversarial imputation network is as follows: Figure 4 As shown, the generator consists of two one-dimensional convolutional layers, a spatiotemporal feature extraction module and two fully connected layers, and the discriminator consists of two convolutional layers, two adaptive average pooling layers and two fully connected layers.
[0088] In actual application, the construction process of L2-TSGAIN can be described as:
[0089] (4-1) Temporal Convolutional Network (TCN) is introduced to focus on the long-term dependencies of time series data. The function expression of dilated causal convolution in TCN is:
[0090]
[0091] Where w[j] represents the weight of the kernel, x[i+j·d] is the input sequence, d is the expansion factor that controls the kernel spacing, y[i] represents the features extracted by the convolution layer, j represents the positional encoding of the features extracted by the convolution kernel at a time, j=1,2,...,k, k represents the convolution kernel size, i represents the positional encoding of the input sequence, i=1,2,...,n, n represents the total number of data.
[0092] By increasing the expansion factor d, L2-TSGAIN can capture more distant dependencies of the data.
[0093] (4-2) At the same time, a multi-head attention mechanism (MHA) is introduced to consider the spatial information between each sample point. For the single-head attention mechanism, there are:
[0094]
[0095] Where Q, K, and V represent the query matrix, key matrix, and value matrix of the input sequence, respectively. represents the scaling factor, Softmax(*) is the activation function, and Attention(*) represents the data features extracted by the single-head attention mechanism.
[0096] For the multi-head attention mechanism, there are:
[0097] MHA(x)=Concat(head1,head2,...,head h )W O .
[0098] In the formula, head p Indicates the head, p represents the index of the head, p = 1, 2, ..., h, h represents the total number of indexes, W Q , W K and W V are the weight matrices of Q, K, and V respectively, represents the weight matrix of the query vector of the p-th head, represents the weight matrix of the key vector of the p-th head, represents the weight matrix of the value vector of the p-th head, W O It is the linear projection of multi-head concatenation, MHA(x) represents the global features extracted by the multi-head attention mechanism, and Concat(*) represents the concatenation operation.
[0099] (4-3) Add MHA to the residual block of TCN to form a multi-head attention residual block (MHAR). Connect three MHARs in series to form a temporal-spatial feature extraction module (TSFEM), which is fed into the generator of the Generative Adversarial Imputation Network (GAIN). The specific structure of TSFEM is as follows: Figure 5 shown.
[0100] (4-4) The L2 regularization function is combined with the adversarial loss function of the generator to form an improved generator loss function, which is:
[0101] L G =αL adv_G +γL 2-norm .
[0102]
[0103] Where, L adv_G represents the adversarial loss function, L 2-normrepresents the L2 regularization function, α represents the weight coefficient of the adversarial loss function, γ represents the weight coefficient of the L2 regularization function, L G represents the loss function of the improved generator, D is the discriminator, is the interpolation data, w i is the weight coefficient of the generator, and λ is the weight coefficient used to control the regularization strength.
[0104] (5) Using the first training set to train the L2 space-time generative adversarial interpolation network until the first set condition is met, the trained L2 space-time generative adversarial interpolation network is used as the data interpolation model.
[0105] After L2-TSGAIN is constructed, it takes abnormal vibration signal or abnormal current signal as input and interpolation data as output. and training set Train L2-TSGAIN separately to obtain the trained L2-TSGAIN model.
[0106] In the actual training process, for the generator, the mask vector M is first defined to describe whether the data at each time point in the original data is missing, and its shape is the same as the missing data matrix. The elements of the mask vector M are expressed as:
[0107]
[0108] Among them, M ij is the element of the mask vector M, is the component in the abnormal data matrix, N is the sample length, and L is the number of samples.
[0109] Secondly, define the noise vector Z to obey Gaussian distribution, M, Z and abnormal data The same shape.
[0110] Then, M and Z are input into the generator of L2-TSGAIN, and the generator outputs synthetic data
[0111] Where G represents the generator and ⊙ represents element-by-element multiplication.
[0112] Finally, the abnormal data The missing parts are synthesized data Replace the sample points at the same position to obtain interpolated data
[0113]
[0114] For the discriminator, define the random variable B = (B1, B2, ..., Bn )∈{0,1} n , the prompt mechanism H is defined as:
[0115] H=B⊙M+0.5(1-B).
[0116] In the formula, B n Represents a random number, i=1,2,...,n.
[0117] The input to the discriminator is and H, the output is the estimation matrix P, which represents the probability of each component of the imputed data being predicted to be true, expressed as
[0118] After adversarial training, the generator and the discriminator reach Nash equilibrium, that is, the generator loss and the discriminator loss finally converge to a certain value. At this point, it can be considered that the generator has fully learned the data distribution of the original sample and obtained a good data interpolation effect. Save the parameters of L2-TSGAIN for subsequent interpolation data acquisition.
[0119] (6) Input the first test set into the data interpolation model to obtain historical vibration signal interpolation data and historical current signal interpolation data. and test set Input into L2-TSGAIN to obtain historical interpolated vibration signals and historical interpolated current signals in different health states.
[0120] In actual application, the training set and The interpolation data obtained by inputting into the trained L2-TSGAIN is and The test set and The interpolation data obtained by inputting into the trained L2-TSGAIN is and
[0121] (7) Construct a second training set based on historical vibration signal interpolation data and historical current signal interpolation data.
[0122] Based on the above description, the interpolation data used for data fusion is the interpolation data obtained in step (6) and For imputed data and Divide to get training data and Test Data and
[0123] (8) Construct a multi-input single-output autoencoder (MISO-AE). For example, MISO-AE has two encoders and one decoder. The two encoders are used to receive vibration and current data respectively. The specific structure of MISO-AE is as follows: Figure 6 As shown, each encoder consists of two fully connected layers, and the decoder structure is the same as the encoder.
[0124] (9) Using the second training set to train the multi-input single-output autoencoder until the second set condition is met, the trained multi-input single-output autoencoder is used as the data fusion model. and Train MISO-AE and obtain the trained MISO-AE.
[0125] Based on the above description, in actual application, the vibration signal interpolation data and the current signal interpolation data are input into the data fusion model, and the process of obtaining the fused data can be described as:
[0126] (1) The vibration signal interpolation data and the current signal interpolation data are respectively input into the encoder to obtain the vibration data features and the current data features. For example, based on the autoencoder, two encoders are set to extract the features of the vibration data and the current data respectively, that is, F q =ψ(X q ). Among them, ψ(·) represents the encoding process of MISO-AE, F q Represents the deep features extracted by the qth encoder, where q is the number of encoders, q=1,2.
[0127] (2) The vibration data features and the current data features are fused to obtain the fused features.
[0128] For example, the deep features of the extracted data of different modalities are fused, and the fusion processing formula is expressed as:
[0129] F=Concat(F1,F2).
[0130] Among them, F represents fusion features, Concat(*) represents concatenation operation, F1 represents the deep features extracted by the first encoder, and F2 represents the deep features extracted by the second encoder.
[0131] (3) The decoder restores the dimension of the fused feature to the dimension of the vibration signal interpolation data and the current signal interpolation data to obtain the fused data.
[0132] For example, the decoder restores the dimension of the fused feature to the dimension of the input data to obtain fused data. The fused data is expressed as: X' = ζ(F). In the formula, X' is the fused data, and ζ(·) represents the decoding process.
[0133] After constructing the trained MISO-AE (i.e., data fusion model), the test data Input the fusion samples of different health states obtained by MISO-AE. The fused data obtained by inputting into the trained MISO-AE is The test data The fused data obtained by inputting into the trained MISO-AE is
[0134] In another exemplary embodiment of the present application, after constructing a CNN (Convolutional Neural Network) fault classifier, a training set of fused data is used to train the CNN fault classifier to obtain a trained CNN fault classifier.
[0135] Specific:
[0136] (1) Construct a second test set based on historical vibration signal interpolation data and historical current signal interpolation data.
[0137] (2) Input the second test set into the data fusion model to obtain historical fusion data.
[0138] (3) Construct the third training set and the third test set based on the historical fusion data. Divide into training set and test set
[0139] (4) Construct a convolutional neural network model (CNN), and use the third training set and the third test set to train and test the CNN model respectively until the third set condition is met, and then use the trained CNN model as the fault diagnosis model. For example, the structure of the constructed CNN model is as follows: Figure 7 , which consists of four one-dimensional convolutional layers, four maximum pooling layers and two fully connected layers.
[0140] In actual application, the training set Input into the CNN fault classifier for training, and then the test set Input into the CNN fault classifier, the fault classifier will analyze the diagnosis results and obtain the test set The different health states included.
[0141] In summary, the reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion provided in this application has an overall process from model construction to fault detection. Figure 8 As shown, it can be summarized as:
[0142] Step 1: Acquire abnormal data.
[0143] Step 2: Data interpolation strategy based on L2-TSGAIN.
[0144] Step 3: Data fusion strategy based on MISO-AE.
[0145] Step 4: Intelligent fault diagnosis based on CNN.
[0146] This application first designs an L2-TSGAIN that introduces TSFEM and L2 regularization function, and inputs abnormal data of different modes into L2-TSGAIN respectively to realize the interpolation of abnormal data. Then a MISO-AE with a multi-encoder structure is designed, and the interpolation data of different modes are input into MISO-AE to realize the fusion of multimodal data. Finally, a CNN fault classifier is constructed, and the fused data is input into CNN to realize fault diagnosis. This application fully considers the impact of abnormal missing sensor data on fault diagnosis in industrial scenarios, and based on the L2-TSGAIN data interpolation strategy and the MISO-AE data fusion strategy, provides a new method for state monitoring and fault diagnosis of mechanical equipment, and provides important technical support for the safe and reliable operation of mechanical equipment.
[0147] In another exemplary embodiment of the present application, as shown in Table 1, method 1 in Table 1 indicates that only GAIN is used to interpolate abnormal data, and then the interpolated data is diagnosed using a CNN fault classifier. Method 2 indicates that GAIN with TSFEM is used to interpolate data, and then the interpolated data is diagnosed using a CNN fault classifier. Method 3 indicates that GAIN with TSFEM and L2 regularization function is used to interpolate data, and then the interpolated data is diagnosed using a CNN fault classifier. The proposed method indicates that the present application uses GAIN with TSFEM and L2 regularization function to interpolate data, then uses MISO-AE for data fusion, and finally uses a CNN fault classifier to diagnose the fused data. In this example, the fault diagnosis results of the planetary transmission system test data based on the diagnostic method of L2-TSGAIN data interpolation and MISO-AE data fusion show that the introduced TSFEM, L2 regularization function and MISO-AE data fusion methods can all improve the diagnostic accuracy, with the highest diagnostic accuracy reaching 99.68%. This demonstrates the effectiveness of the designed L2-TSGAIN data interpolation method and MISO-AE data fusion method in improving diagnostic accuracy.
[0148] Table 1 Missing rate results of various methods
[0149]
[0150] As shown in Table 2, KNN in Table 2 represents the K nearest neighbor interpolation method, GAIN represents the generative adversarial interpolation network, STA-GAN represents the spatiotemporal attention generative adversarial network, and ESM-GAIN represents the selective multi-generator ensemble generative adversarial interpolation network. The three quantitative indicators of mean absolute error (MAE), mean square error (MSE), and root mean square error (RMSE) are used to evaluate the effectiveness of different data interpolation methods. The comparison results show that the method proposed in this application achieves the best data interpolation effect under the condition of missing sensor data.
[0151] Table 2 Data interpolation effects of various methods
[0152]
[0153] As shown in Table 3, the diagnostic results of the fault diagnosis method based on L2-TSGAIN data interpolation and MISO-AE data fusion on the planetary transmission system test data show that compared with the traditional method and the existing advanced methods, the proposed method has the best results in terms of accuracy, precision, recall and F1 score, with the highest accuracy reaching 99.68%. Under the three missing ratios of 10%, 20% and 30%, this method can still achieve a diagnostic accuracy of more than 95%. This reflects the effectiveness of the multimodal data interpolation and fusion method of this application for trusted mechanical fault diagnosis.
[0154] Table 3 Diagnostic results of planetary transmission system test data
[0155]
[0156] Based on the above description, compared with the prior art, the method provided by the present application has the following advantages.
[0157] 1. The multimodal data interpolation and fusion method proposed in this application can extract the temporal and spatial characteristics of abnormal data and generate interpolated data, and can fuse interpolated data of different modes, providing an intelligent fault diagnosis method in the case of missing data.
[0158] 2. The data interpolation method based on L2-TSGAIN designed in this application constructs a TSFEM by introducing TCN and MHA to extract data features from the perspective of time and space, and integrates the L2 regularization function into the generator loss function to avoid overfitting. The designed L2-TSGAIN can fully learn the distribution of abnormal data samples and realize data interpolation.
[0159] 3. The data fusion method based on MISO-AE designed in this application realizes data fusion by deploying multiple encoders to extract features of different modal data respectively.
[0160] 4. This application can perform condition monitoring and fault diagnosis on mechanical equipment in the absence of data, can learn the distribution of abnormal data and interpolate abnormal data, can comprehensively utilize interpolated data of different modes to provide more powerful data support for fault diagnosis, and can provide technical support for condition monitoring, fault diagnosis and health management of many mechanical equipment.
[0161] Further, based on the same inventive concept, the embodiment of the present application also provides a trusted mechanical fault diagnosis system based on multimodal data interpolation and fusion for implementing the above-mentioned. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the trusted mechanical fault diagnosis system based on multimodal data interpolation and fusion provided below can refer to the limitations of the trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion in the above text, and will not be repeated here.
[0162] In an exemplary embodiment, the present application provides a reliable mechanical fault diagnosis system based on multimodal data interpolation and fusion, including:
[0163] The data acquisition module is used to acquire vibration data and current data of the mechanical equipment to be diagnosed.
[0164] The signal interception module is used to intercept the vibration signal from the vibration data and intercept the current signal from the current data according to the operation cycle of the mechanical equipment to be diagnosed.
[0165] The interpolation fusion model construction module is used to construct the data interpolation model and the data fusion model. The data interpolation model is a trained L2 spatiotemporal generative adversarial interpolation network. The data in the data set used to train the L2 spatiotemporal generative adversarial interpolation network are obtained by simulating the continuous loss of historical operating data under different mechanical health states. The historical operating data includes historical vibration data and historical current data. The data fusion model is a trained multi-input single-output autoencoder. The multi-input single-output autoencoder includes multiple encoders and one decoder.
[0166] The data interpolation module is used to input the vibration signal and the current signal into the data interpolation model respectively to obtain the vibration signal interpolation data and the current signal interpolation data.
[0167] The data fusion module is used to input the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data.
[0168] The fault diagnosis module is used to construct a fault diagnosis model and input the fusion data into the fault diagnosis model to obtain the fault diagnosis results of the mechanical equipment to be diagnosed.
[0169] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store video tag processing data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion is implemented.
[0170] Those skilled in the art will understand that Fig. 9The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.
[0171] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0172] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0173] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.
[0174] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0175] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.
[0176] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0177] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. A reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion, characterized in that: The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion includes: Obtain vibration data and current data of the mechanical equipment to be diagnosed; According to the operation cycle of the mechanical equipment to be diagnosed, a vibration signal is intercepted from the vibration data, and a current signal is intercepted from the current data; Constructing a data interpolation model and a data fusion model; the data interpolation model is a trained L2 spatiotemporal generative adversarial interpolation network; the data in the data set used when training the L2 spatiotemporal generative adversarial interpolation network are obtained by simulating the continuous loss of historical operating data under different mechanical health states; the historical operating data include historical vibration data and historical current data; the data fusion model is a trained multi-input single-output autoencoder; the multi-input single-output autoencoder includes multiple encoders and a decoder; Inputting the vibration signal and the current signal into the data interpolation model respectively to obtain vibration signal interpolation data and current signal interpolation data; Inputting the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data; A fault diagnosis model is constructed, and the fusion data is input into the fault diagnosis model to obtain a fault diagnosis result of the mechanical equipment to be diagnosed.
2. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 1 is characterized in that: Constructing the data interpolation model and the data fusion model includes: According to the operation cycle of the mechanical equipment, a historical vibration signal and a historical current signal are respectively intercepted from the historical vibration data and the historical current data; Performing continuous value loss simulation on the historical vibration signal and the historical current signal respectively to obtain a historical abnormal vibration signal and a historical abnormal current signal; Constructing a first training set and a first test set based on the historical abnormal vibration signal and the historical abnormal current signal; Constructing an L2 spatiotemporal generative adversarial interpolation network; the L2 spatiotemporal generative adversarial interpolation network is a generative adversarial interpolation network including an L2 regularization function and a spatiotemporal feature extraction module; Using the first training set to train the L2 spatiotemporal generative adversarial interpolation network until a first set condition is met, and using the trained L2 spatiotemporal generative adversarial interpolation network as the data interpolation model; Inputting the first test set into the data interpolation model to obtain historical vibration signal interpolation data and historical current signal interpolation data; Constructing a second training set based on the historical vibration signal interpolation data and the historical current signal interpolation data; Construct a multi-input single-output autoencoder; The multi-input single-output autoencoder is trained using the second training set until a second set condition is met, and the trained multi-input single-output autoencoder is used as the data fusion model.
3. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 2 is characterized in that: Construct an L2 spatiotemporal generative adversarial interpolation network, including: Add the multi-head attention mechanism to the residual block of the temporal convolutional network to form a multi-head attention residual block; Connecting multiple multi-head attention residual blocks in series to form the spatiotemporal feature extraction module; Feeding the spatiotemporal feature extraction module into the generator of the generative adversarial interpolation network to obtain an improved generator; The L2 regularization function and the adversarial loss function of the original generator in the generative adversarial interpolation network are combined to obtain the loss function of the improved generator to complete the construction of the L2 spatiotemporal generative adversarial interpolation network.
4. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 3 is characterized in that: The first setting condition is to improve the generator and the discriminator to achieve Nash equilibrium.
5. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 3 is characterized in that: The loss function of the improved generator is expressed as: L G =αL adv_G +γL 2-norm ; Where, L adv_G represents the adversarial loss function, L 2-norm represents the L2 regularization function, α represents the weight coefficient of the adversarial loss function, γ represents the weight coefficient of the L2 regularization function, L G Represents the loss function of the improved generator.
6. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 1 is characterized in that: Inputting the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data specifically includes: The vibration signal interpolation data and the current signal interpolation data are respectively input into the encoder to obtain vibration data features and current data features; fusing the vibration data feature and the current data feature to obtain a fused feature; The decoder restores the dimension of the fused feature to the dimension of the vibration signal interpolation data and the current signal interpolation data to obtain the fused data.
7. The reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion according to claim 2 is characterized in that: Construct a fault diagnosis model, including: Constructing a second test set based on the historical vibration signal interpolation data and the historical current signal interpolation data; Inputting the second test set into the data fusion model to obtain historical fusion data; Constructing a third training set and a third test set based on the historical fusion data; Construct a convolutional neural network model, and use the third training set and the third test set to train and test the convolutional neural network model respectively, until the third set condition is met, and use the trained convolutional neural network model as the fault diagnosis model.
8. A reliable mechanical fault diagnosis system based on multimodal data interpolation and fusion, characterized in that: The trusted mechanical fault diagnosis system based on multimodal data interpolation and fusion is used to implement the trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion as claimed in any one of claims 1 to 7; the system comprises: A data acquisition module, used to acquire vibration data and current data of the mechanical equipment to be diagnosed; A signal interception module, used to intercept a vibration signal from the vibration data and to intercept a current signal from the current data according to an operation cycle of the mechanical equipment to be diagnosed; An interpolation fusion model construction module is used to construct a data interpolation model and a data fusion model; the data interpolation model is a trained L2 spatiotemporal generative adversarial interpolation network; the data in the data set used when training the L2 spatiotemporal generative adversarial interpolation network are obtained by simulating the continuous loss of historical operating data under different mechanical health states; the historical operating data include historical vibration data and historical current data; the data fusion model is a trained multi-input single-output autoencoder; the multi-input single-output autoencoder includes multiple encoders and one decoder; A data interpolation module, used for inputting the vibration signal and the current signal into the data interpolation model respectively to obtain vibration signal interpolation data and current signal interpolation data; A data fusion module, used for inputting the vibration signal interpolation data and the current signal interpolation data into the data fusion model to obtain fused data; The fault diagnosis module is used to construct a fault diagnosis model and input the fusion data into the fault diagnosis model to obtain the fault diagnosis result of the mechanical equipment to be diagnosed.
9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the trusted mechanical fault diagnosis method based on multimodal data interpolation and fusion as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the reliable mechanical fault diagnosis method based on multimodal data interpolation and fusion described in any one of claims 1 to 7.
Citation Information
Patent Citations
Mechanical equipment fault diagnosis method based on small sample learning
CN116821697A
Urban road network checkpoint traffic data quality diagnosis and recovery method
CN117056693A
Air quality prediction method
CN118072873A
Battery simulator electromagnetic compatibility defect small sample cross-domain classification prediction method
CN118332424A
Cited By
Federal map learning-based power Internet of Things equipment anomaly detection method and system
CN120124674A
