Rotary machinery unsupervised anomaly detection method based on KAAE

Abstract

The invention discloses a KAAE-based unsupervised anomaly detection method for a rotating machine. The method comprises the following steps: 1, data acquisition: acquiring vibration signals of a key transmission part of the rotating machine in a normal state and an abnormal state; 2, data preprocessing: obtaining a sample set through sliding window sampling, normalization and fast Fourier transform, and dividing the sample set into a training set and a verification set only containing normal samples, and a test set containing normal samples and abnormal samples; 3, model building: in combination with the KAN and the auto-encoder AE, constructing a KAAE (Kolmogov-Arnod auto-encoder), and building a KAAE (Kolmogov-Arnod auto-encoder); 4, training, verifying and testing the model: training the model on the training set, obtaining an anomaly detection threshold value on the verification set, and evaluating the model performance on the test set; and 5, anomaly detection: carrying out anomaly detection on the rotating machinery by using the trained model. According to the method, the complex dependency relationship and the potential representation in the data are fully captured, so that the distribution of normal data is better fitted, and the method is used for unsupervised anomaly detection of the rotating machinery.

Description

Technical Field

[0001] The present invention relates to the technical field of rotating machinery anomaly detection, and in particular to an unsupervised anomaly detection method for rotating machinery based on KAAE. Background Art

[0002] With the rapid development of modern mechanical systems, rotating machinery has become indispensable in intelligent equipment, and academia and industry have paid great attention to its safety. Under high-intensity working conditions, various abnormalities will inevitably occur in the key transmission components of rotating machinery (such as bearings and gears). Abnormal transmission components directly affect the operational reliability of rotating machinery, which may lead to serious accidents, economic losses and even casualties. Therefore, it is of great research value to carry out abnormality detection of rotating machinery.

[0003] Traditional data-driven anomaly detection methods can be divided into three steps: signal acquisition, feature extraction, and state detection. The acquired signals can be vibration signals, current signals, acoustic signals, and temperature signals. Signal processing methods are widely used for feature extraction, such as empirical mode decomposition, Hilbert vibration decomposition, atomic decomposition, empirical wavelet transform, variational mode decomposition, and synchronous compression transform. The extracted features are used for state detection through machine learning algorithms, such as logistic regression, K nearest neighbor, naive Bayes, decision tree, support vector machine, and artificial neural network. However, machine learning algorithms are only suitable for small-scale data, and feature extraction through signal processing requires a high degree of reliance on expert prior knowledge and manual feature selection, which limits the generalization ability of these methods.

[0004] In recent years, deep learning technology has attracted widespread attention in the field of rotating machinery anomaly detection technology due to its adaptive feature extraction capabilities and end-to-end detection characteristics. Deep learning methods such as multi-layer perceptron, convolutional neural network, recurrent neural network, generative adversarial network and Transformer have shown good results in rotating machinery anomaly detection tasks. However, they need to rely on sufficient labeled data, which limits their application in industrial scenarios. Most rotating machinery in industrial scenarios is in normal state, and abnormal labeled data is very scarce or even unavailable. Therefore, unsupervised anomaly detection based only on normal state data can provide more flexibility and expand the capabilities of the above deep learning methods.

[0005] As a classic unsupervised deep learning model, autoencoders are widely used in data dimensionality reduction, feature extraction, and data reconstruction. In recent years, many scholars have explored and improved them to improve their unsupervised anomaly detection performance in different fields. However, traditional autoencoders are limited by the linear fitting of the fully connected layer, making it difficult to effectively reconstruct normal data and extract representative feature representations, resulting in insufficient detection capabilities for abnormal data.

[0006] Compared with the prior art, the differences are as follows:

[0007] Technical comparison with patent CN110132554A "A method for fault diagnosis of rotating machinery based on deep Laplace autoencoder"

[0008] Patent CN110132554A proposes a rotating machinery fault diagnosis method based on deep Laplace autoencoder, which belongs to the technical field of rotating machinery fault diagnosis; this patent proposes a rotating machinery unsupervised anomaly detection method based on KAAE, which belongs to the technical field of rotating machinery anomaly detection. There are essential differences between the technical fields involved in the two.

[0009] Patent CN110132554A constructs a deep Laplace autoencoder; this patent constructs a Kolmogorov-Arnold autoencoder. Both are improvements on traditional autoencoders, but their theoretical foundations are essentially different.

[0010] Patent CN110132554A constructs a deep Laplace autoencoder for rotating machinery fault diagnosis, which is a supervised learning algorithm; this patent constructs a Kolmogorov-Arnold autoencoder for unsupervised anomaly detection of rotating machinery, which is an unsupervised learning algorithm. There is an essential difference between the two at the algorithm level.

[0011] Technical comparison with patent CN110334764A "Intelligent fault diagnosis method for rotating machinery based on integrated deep autoencoder"

[0012] Patent CN110334764A proposes a rotating machinery intelligent fault diagnosis method based on an integrated deep autoencoder, which belongs to the technical field of rotating machinery fault diagnosis; this patent proposes a rotating machinery unsupervised anomaly detection method based on KAAE, which belongs to the technical field of rotating machinery anomaly detection. There are essential differences in the technical fields involved in the two.

[0013] Patent CN110334764A constructs an integrated deep autoencoder; this patent constructs a Kolmogorov-Arnold autoencoder. Both are improvements on traditional autoencoders, but their theoretical foundations are essentially different.

[0014] Patent CN110334764A constructs an integrated deep autoencoder for rotating machinery fault diagnosis, which is a supervised learning algorithm; this patent constructs a Kolmogorov-Arnold autoencoder for unsupervised anomaly detection of rotating machinery, which is an unsupervised learning algorithm. There is an essential difference between the two at the algorithm level. Summary of the invention

[0015] In order to solve the above technical problems, the present invention proposes a method for unsupervised anomaly detection of rotating machinery based on KAAE, and combines the Kolmogorov-Arnold network KAN with the autoencoder AE to construct a Kolmogorov-Arnold autoencoder KAAE. By utilizing the responsiveness and flexibility of KAN, the complex dependencies and potential representations in the data are fully captured, so as to better fit the distribution of normal data for unsupervised anomaly detection of rotating machinery.

[0016] To achieve the above object, the technical solution adopted by the present invention is:

[0017] A KAAE-based unsupervised anomaly detection method for rotating machinery, characterized by comprising the following steps:

[0018] Step 1: Data collection:

[0019] Collect vibration signals of key transmission components of rotating machinery in normal and abnormal states;

[0020] Step 2: Data preprocessing:

[0021] The sample set is obtained through sliding window sampling, normalization and fast Fourier transform, and is divided into a training set and a validation set containing only normal samples, and a test set containing both normal and abnormal samples;

[0022] Step 3: Model building:

[0023] Combine the Kolmogorov-Arnold network KAN and the autoencoder AE to construct the Kolmogorov-Arnold autoencoder KAAE;

[0024] Step 4: Model training, verification and testing:

[0025] Train the model on the training set, obtain the anomaly detection threshold on the validation set, and evaluate the model performance on the test set;

[0026] Step 5: Anomaly detection:

[0027] Use the trained model to perform anomaly detection on rotating machinery.

[0028] Furthermore, in the step 1 of data collection, one or more sensors are installed at different positions or directions near the rotating mechanical transmission component that needs abnormality detection, and the vibration signals of normal state and abnormal state are collected using data collection equipment.

[0029] Furthermore, in step 2, in order to avoid test leakage, when sliding window sampling, there is no overlap between adjacent windows and a certain interval is maintained; all samples are normalized and fast Fourier transformed.

[0030] Furthermore, in step 3, the overall structure of KAAE is similar to that of AE, with the main difference being that KAAE uses KAN layers to replace the fully connected layers in AE. The responsiveness and flexibility of KAN are utilized to fully capture the complex dependencies and potential representations in the data;

[0031] The KAAE is a symmetrical structure, including an encoding network and a decoding network, both of which have the same number of layers. The number of layers can be adjusted according to different task requirements;

[0032] In the KAAE, the encoding network compresses the high-dimensional input data into a low-dimensional feature representation through KAN, and each layer gradually reduces the dimension of the data; the decoding network converts the low-dimensional feature representation into a reconstructed form of the input data through KAN, and each layer gradually increases the dimension of the data. The goal is to reconstruct the input data to obtain a low-dimensional representation that best expresses the essential characteristics of the data, as shown in the following formula:

[0033] h=f(x)=KAN(x) (1)

[0034] z=g(h)=KAN′(h) (2)

[0035]

[0036] in, represents input data, f(·) represents the encoding network, represents the low-dimensional features obtained after encoding, g(·) represents the decoding network, Represents the reconstructed data obtained after decoding, L MSE Represents the mean square error reconstruction loss function, and the training goal of the model is to minimize it;

[0037] In the KAN, each layer is structured as a learnable activation function matrix, which is parameterized and optimized during the training process, as shown in the following formula:

[0038] Φ={φ q,p},p=1,...,n in ,q=1,...,n out (4)

[0039] Among them, n in and n out Represent the input feature dimension and output feature dimension respectively;

[0040] The KAN is implemented by stacking layers. The mapping transformation of each layer acts on the output of the previous layer to generate the input of the next layer. A KAN with L layers is shown in the following formula:

[0041] φ l,j,i ,l=0,...,L-1,j=1,...,nl+1 ,i=1,...,n l (5)

[0042]

[0043] Among them, φ l,j,i represents the learnable activation function connecting the i-th neuron in layer l and the j-th neuron in layer l+1, x l+1,j represents the activation value of the jth neuron in the l+1th layer, Φ l Represents the function matrix of the lth layer, containing n l ×n l+1 A learnable activation function;

[0044] The learnable activation function is a weighted sum of a spline function and a bias function, as shown in the following formula:

[0045] φ(x)=w s spline(x)+w b b(x) (9)

[0046]

[0047] Where spline(·) represents the linear combination of B-spline functions, c i represents the learnable combination coefficient, b(·) represents the bias function, and w s and w b Represents a learnable weight factor to better control the overall magnitude of the learnable activation function.

[0048] Furthermore, in step 4, the model training uses a training set containing only normal samples to train KAAE, and uses the Adam optimization algorithm to update the model parameters;

[0049] The model verification is performed by inputting a validation set containing only normal samples into the trained KAAE to obtain the reconstruction loss value of each validation sample. In order to avoid the difficulty of manually adjusting the threshold and selecting the percentage interval, the maximum value of the reconstruction loss value of all validation samples is selected as the threshold for detecting abnormal samples;

[0050] The model test inputs a test set containing normal samples and abnormal samples into the trained KAAE, obtains the reconstruction loss value of each test sample, determines it as normal or abnormal according to a threshold, and evaluates the model performance through multiple indicators.

[0051] Compared with the prior art, the advantages of the present invention are:

[0052] Combining the Kolmogorov-Arnold network (KAN) with the autoencoder (AE), a Kolmogorov-Arnold autoencoder (KAAE) is constructed. The responsiveness and flexibility of KAN are utilized to fully capture the complex dependencies and potential representations in the data, thereby better fitting the distribution of normal data for unsupervised anomaly detection of rotating machinery. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of the method of the present invention;

[0054] Figure 2 Schematic diagram of the structure of KAAE of the method of the present invention. DETAILED DESCRIPTION

[0055] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0056] In this embodiment, if Figure 1 As shown, a KAAE-based unsupervised anomaly detection method for rotating machinery includes the following steps: Step 1, data acquisition: collecting vibration signals of normal and abnormal states of key transmission components of rotating machinery (such as bearings and gears); Step 2, data preprocessing: obtaining a sample set through sliding window sampling, normalization and fast Fourier transform, and dividing it into a training set and a validation set containing only normal samples, and a test set containing normal samples and abnormal samples; Step 3, model construction: combining the Kolmogorov-Arnold network KAN and the autoencoder AE to construct the Kolmogorov-Arnold autoencoder KAAE; Step 4, model training, verification and testing: training the model on the training set, obtaining the anomaly detection threshold on the verification set, and evaluating the model performance on the test set; Step 5, anomaly detection: using the trained model to perform anomaly detection on rotating machinery.

[0057] In step one, one or more sensors are installed at different positions or directions near the rotating mechanical transmission parts (such as bearings, gears) that need to be detected abnormally, and the vibration signals of normal and abnormal states are collected using data acquisition equipment.

[0058] In this embodiment, two triaxial acceleration sensors are installed at reasonable positions near the high-speed aviation bearing, and the vibration signals of six channels are synchronously collected using a data acquisition device. During the collection process, the working conditions are changed by changing the speed and load of the bearing. The sampling frequency is 51200Hz, and a total of vibration signals of the normal state and six different abnormal states of the bearing are collected.

[0059] In this embodiment, in step 2, in order to avoid test leakage, when the sliding window is sampled, there is no overlap between adjacent windows and a certain interval is maintained. After sampling, each sample contains 1024 signal points; all samples are normalized and fast Fourier transformed, and each sample contains 512 signal points after data processing.

[0060] In this embodiment, the acquisition process includes three different working conditions, and the speed and load of the bearings in all working conditions are different. 500 normal samples are obtained for each working condition through data preprocessing, of which 200 samples are used for training, 100 samples are used for verification, and 200 samples are used for testing. 50 samples of each abnormal state are obtained for each working condition through data preprocessing, and 300 abnormal samples are used for testing, as shown in Table 1 below:

[0061]

[0062]

[0063] In step 3, the overall structure of KAAE is similar to AE, the main difference is that KAAE uses KAN layer to replace the fully connected layer in AE. The responsiveness and flexibility of KAN are used to fully capture the complex dependencies and potential representations in the data;

[0064] The KAAE is a symmetrical structure, including an encoding network and a decoding network, both of which have the same number of layers. The number of layers can be adjusted according to different task requirements;

[0065] In this embodiment, if Figure 2 As shown in the figure, in the KAAE, the encoding network compresses the high-dimensional input data into a low-dimensional feature representation through KAN, and each layer gradually reduces the dimension of the data; the decoding network converts the low-dimensional feature representation into a reconstructed form of the input data through KAN, and each layer gradually increases the dimension of the data. The goal is to reconstruct the input data to obtain a low-dimensional representation that best expresses the essential characteristics of the data, as shown in the following formula:

[0066] h=f(x)=KAN(x) (1)

[0067] z=g(h)=KAN′(h) (2)

[0068]

[0069] in, represents input data, f(·) represents the encoding network, represents the low-dimensional features obtained after encoding, g(·) represents the decoding network, Represents the reconstructed data obtained after decoding, L MSERepresents the mean square error reconstruction loss function, and the training goal of the model is to minimize it;

[0070] In the KAN, each layer is structured as a learnable activation function matrix, which is parameterized and optimized during the training process, as shown in the following formula:

[0071] Φ={φ q,p},p=1,...,n in ,q=1,...,n out (4)

[0072] Among them, n in and n out Represent the input feature dimension and output feature dimension respectively;

[0073] The KAN is implemented by stacking layers, where each layer's mapping transformation acts on the output of the previous layer to produce the input of the next layer. A KAN with L layers is shown below:

[0074] φ l,j,i ,l=0,...,L-1,j=1,...,n l+1 ,i=1,...,n l (5)

[0075]

[0076] Among them, φ l,j,i represents the learnable activation function connecting the i-th neuron in layer l and the j-th neuron in layer l+1, x l+1,j represents the activation value of the jth neuron in the l+1th layer, Φ l Represents the function matrix of the lth layer, containing n l ×n l+1 A learnable activation function;

[0077] The learnable activation function is a weighted sum of a spline function and a bias function, as shown in the following formula:

[0078] φ(x)=w s spline(x)+w b b(x) (9)

[0079]

[0080] Where spline(·) represents the linear combination of B-spline functions, c i represents the learnable combination coefficient, b(·) represents the bias function, and w s and w b represents a learnable weight factor to better control the overall magnitude of the learnable activation function;

[0081] In this embodiment, the encoding network and decoding network of the KAAE are both two-layer structures. Batch normalization BN is added to reduce the change of data distribution and stabilize the network training process. The linear rectifier unit ReLU is used as the activation function to improve the nonlinear expression ability of the network, alleviate the gradient disappearance and accelerate the network convergence. The detailed parameter configuration of the KAAE is as follows

[0082] As shown in Table 2:

[0083]

[0084] In step 4, the model is trained by using a training set containing only normal samples to train KAAE, and the Adam optimization algorithm is used to update the model parameters.

[0085] In this embodiment, the model training and hyperparameter configuration are as follows: the batch size is 128, the number of iterations is 300, the initial learning rate is 0.01, and it decays by 0.1 times at 20 and 200 iterations respectively.

[0086] In step 4, the model is verified by inputting the validation set containing only normal samples into the trained KAAE to obtain the reconstruction loss value of each validation sample. In order to avoid the difficulty of manually adjusting the threshold and selecting the percentage interval, the maximum value of the reconstruction loss value of all validation samples is selected as the threshold for detecting abnormal samples.

[0087] In step 4, the model is tested by inputting a test set containing normal samples and abnormal samples into the trained KAAE to obtain the reconstruction loss value of each test sample, determining whether it is normal or abnormal based on the threshold, and evaluating the model performance through multiple indicators.

[0088] In this embodiment, the calculation formulas of the four evaluation indicators (accuracy, precision, recall and F1 score) are:

[0089]

[0090] Among them, TP represents the number of correctly detected abnormal samples; TN represents the number of correctly detected normal samples; FP represents the number of incorrectly detected abnormal samples; FN represents the number of incorrectly detected normal samples;

[0091] In this example, in addition to comparing KAAE with traditional AE, we also selected the currently popular improved AE, including memory-enhanced autoencoder MAE, variational autoencoder VAE, sparse autoencoder SAE and denoising autoencoder DAE. It is worth noting that for a fair comparison, all the above models use the same input signal form, training strategy and hyperparameter configuration. In order to reduce the impact of randomness, each experiment was repeated five times, and the evaluation index was averaged as the final experimental result.

[0092] In this embodiment, the unsupervised anomaly detection results of all models are shown in Table 3 below:

[0093]

[0094] As can be seen from the table, the accuracy, precision, recall and F1 score of KAAE all exceeded 96%, reaching 97.84%, 99.77%, 96.62% and 98.13% respectively. Compared with the traditional AE, the evaluation indicators were improved by 14.71%, 0.02%, 24.53% and 15.41% respectively. Among them, the accuracy, recall and F1 score of the three evaluation indicators achieved the best among all models. The accuracy of SAE is 0.08% higher than that of KAAE, but the other three evaluation indicators are far behind KAAE.

[0095] It can be seen from the above embodiments that the present invention is a method for unsupervised anomaly detection of rotating machinery based on KAAE. Compared with the prior art, the advantages of the present invention are: combining the Kolmogorov-Arnold network KAN with the autoencoder AE to construct a Kolmogorov-Arnold autoencoder KAAE. By utilizing the responsiveness and flexibility of KAN, the complex dependencies and potential representations in the data are fully captured, so as to better fit the distribution of normal data for unsupervised anomaly detection of rotating machinery.

[0096] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. A KAAE-based unsupervised anomaly detection method for rotating machinery, characterized in that: The following steps are involved: Step 1: Data collection: Collect vibration signals of key transmission components of rotating machinery in normal and abnormal states; Step 2: Data preprocessing: The sample set is obtained through sliding window sampling, normalization and fast Fourier transform, and is divided into a training set and a validation set containing only normal samples, and a test set containing both normal and abnormal samples; Step 3: Model building: Combine the Kolmogorov-Arnold network KAN and the autoencoder AE to construct the Kolmogorov-Arnold autoencoder KAAE; Step 4: Model training, verification and testing: Train the model on the training set, obtain the anomaly detection threshold on the validation set, and evaluate the model performance on the test set; Step 5: Anomaly detection: Use the trained model to perform anomaly detection on rotating machinery.

2. The unsupervised anomaly detection method for rotating machinery based on KAAE according to claim 1 is characterized in that: In the step 1 of data collection, one or more sensors are installed at different positions or directions near the rotating mechanical transmission parts that need to be detected for abnormalities, and the vibration signals of normal and abnormal states are collected using data collection equipment.

3. The unsupervised anomaly detection method for rotating machinery based on KAAE according to claim 1 is characterized in that: In step 2, in order to avoid test leakage, when sliding window sampling, there is no overlap between adjacent windows and a certain interval is maintained; all samples are normalized and fast Fourier transformed.

4. The unsupervised anomaly detection method for rotating machinery based on KAAE according to claim 1 is characterized in that: In step 3, the overall structure of KAAE is similar to AE, the main difference is that KAAE uses KAN layer to replace the fully connected layer in AE. The responsiveness and flexibility of KAN are used to fully capture the complex dependencies and potential representations in the data; The KAAE is a symmetrical structure, including an encoding network and a decoding network, both of which have the same number of layers. The number of layers can be adjusted according to different task requirements; In the KAAE, the encoding network compresses the high-dimensional input data into a low-dimensional feature representation through KAN, and each layer gradually reduces the dimension of the data; the decoding network converts the low-dimensional feature representation into a reconstructed form of the input data through KAN, and each layer gradually increases the dimension of the data. The goal is to reconstruct the input data to obtain a low-dimensional representation that best expresses the essential characteristics of the data, as shown in the following formula: h=f(x)=KAN(x) (1) z=g(h)=KAN′(h) (2) in, represents input data, f(·) represents the encoding network, represents the low-dimensional features obtained after encoding, g(·) represents the decoding network, Represents the reconstructed data obtained after decoding, L MSE Represents the mean square error reconstruction loss function, and the training goal of the model is to minimize it; In the KAN, each layer is structured as a learnable activation function matrix, which is parameterized and optimized during the training process, as shown in the following formula: Φ={φ q,p },p=1,...,n in ,q=1,...,n out (4) Among them, n in and n out Represent the input feature dimension and output feature dimension respectively; The KAN is implemented by stacking layers. The mapping transformation of each layer acts on the output of the previous layer to generate the input of the next layer. A KAN with L layers is shown in the following formula: φ l,j,i ,l=0,...,L-1,j=1,...,n l+1 ,i=1,...,n l (5) Among them, φ l,j,i represents the learnable activation function connecting the i-th neuron in layer l and the j-th neuron in layer l+1, x l+1,j represents the activation value of the jth neuron in the l+1th layer, Φ l Represents the function matrix of the lth layer, containing n l ×n l+1 A learnable activation function; The learnable activation function is a weighted sum of a spline function and a bias function, as shown in the following formula: φ(x)=w s spline(x)+w b b(x) (9) Where spline(·) represents the linear combination of B-spline functions, c i represents the learnable combination coefficient, b(·) represents the bias function, and w s and w b Represents a learnable weight factor to better control the overall magnitude of the learnable activation function.

5. The unsupervised anomaly detection method for rotating machinery based on KAAE according to claim 1 is characterized in that: In step 4, the model training uses a training set containing only normal samples to train KAAE, and uses the Adam optimization algorithm to update the model parameters; The model verification is performed by inputting a validation set containing only normal samples into the trained KAAE to obtain the reconstruction loss value of each validation sample. In order to avoid the difficulty of manually adjusting the threshold and selecting the percentage interval, the maximum value of the reconstruction loss value of all validation samples is selected as the threshold for detecting abnormal samples; The model test inputs a test set containing normal samples and abnormal samples into the trained KAAE, obtains the reconstruction loss value of each test sample, determines it as normal or abnormal according to a threshold, and evaluates the model performance through multiple indicators.

Images