Civil aircraft hydraulic system data enhancement method based on WGAN-GP and attention mechanism

Abstract

The invention discloses a civil aircraft hydraulic system data enhancement method based on WGAN-GP and an attention mechanism, and the method comprises the steps: constructing a group of convolutional neural networks and deconvolutional networks based on a Wharisstein generative adversarial network with gradient penalty, deeply extracting the high-dimensional features of a sample, and enhancing the authenticity of the generated sample; a self-attention mechanism is added, so that more attention is added to features with greater influence on data by the network, and a sample with higher quality is generated; an instance standardization layer is added after each layer of convolution to replace batch standardization so as to adapt to the condition of small samples, and when batch comparison is small, independence of each generated sample is kept, and model convergence is accelerated; and a filtering method based on an independent forest is used to remove generated samples which are greatly different from real samples, low-quality samples mixed in the generated samples are deleted, and the quality of the generated samples is improved. The existing data enhancement method is improved from the perspective of data generation, and the method can be used for data enhancement under the condition that fault samples of the civil aircraft hydraulic system are scarce.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent fault diagnosis of aircraft systems, and in particular to a data enhancement method for a civil aircraft hydraulic system based on WGAN-GP and an attention mechanism. Background Art

[0002] The hydraulic system is an important component of the aircraft's electromechanical system, including hydraulic pumps, hydraulic valves, hydraulic cylinders and other components. Among them, key components such as hydraulic pumps and hydraulic cylinders are prone to failures that affect aircraft control and may cause downtime or even casualties. As a complex nonlinear system, the maintenance and troubleshooting diagnosis method based on threshold monitoring can only roughly estimate the fault and cannot quickly locate the fault. It is urgent to introduce intelligent fault diagnosis algorithms to improve the accuracy and efficiency of diagnosis.

[0003] With the rapid development of science and technology, the amount of available aircraft onboard sensor data has exploded. The cross-linking relationship of the hydraulic system is complex, and the signal types and data volume are huge. It is necessary to clarify the aircraft fault characterization information from a large amount of high-dimensional and heterogeneous data. The data-driven deep learning method can perform deep information mining and feature dimensionality reduction extraction on high-dimensional data, and can be used for the study of civil aircraft fault feature characterization methods. Due to the characteristics of high safety, high reliability, and multi-tasking, the aircraft can obtain a large amount of normal operating status data, but it is difficult to collect fault data, which leads to insufficient fault samples and the problem that intelligent methods cannot correctly identify faults. Common deep learning models belong to supervised learning, which relies on a large number of labeled samples for training to achieve classification diagnosis of faults, and have limitations for data with fewer samples. Therefore, it is necessary to introduce data enhancement methods while avoiding overfitting to solve the problem that supervised models cannot perform intelligent diagnosis of small sample faults. Summary of the invention

[0004] Purpose of the invention: The present invention provides a data enhancement method for a civil aircraft hydraulic system based on WGAN-GP and attention mechanism, which improves the existing data enhancement method from the perspective of data generation and can be used for data enhancement in the case of scarce fault samples of civil aircraft hydraulic system.

[0005] Technical solution: The data enhancement method for a civil aircraft hydraulic system based on WGAN-GP and attention mechanism described in the present invention comprises the following steps:

[0006] Step 1: Signal acquisition: by placing corresponding sensors at key measuring points, one or more sensor signals containing information that can reflect the fault status are obtained, and the signals are classified according to different fault modes;

[0007] Step 2: Data preprocessing: normalize the data so that the features can be better learned;

[0008] Step 3: Construct a Wasserstein generative adversarial network based on a gradient penalty mechanism, where the generator contains four layers of deconvolution, three layers of instance normalization, and one self-attention layer, and the discriminator contains four layers of convolution, four layers of instance normalization, and one self-attention mechanism layer. The loss function uses Wasserstein loss.

[0009] Step 4: Iterative parameter update: input the processed data into the model in step 3, randomly extract a fixed number of data from the training set to participate in the training process, iteratively optimize the model parameters, and stop training and save the model when the set number of iterations is met or the Nash equilibrium is met;

[0010] Step 5: Input random noise into the trained generator to obtain generated samples, use the independent forest-based filtering method to remove generated samples that are too different from real samples, delete low-quality samples mixed in the generated samples, and improve the quality of generated samples;

[0011] Step 6: Implement fault diagnosis. Mix the generated data with real data to build a training data set. Use real data to build a test set. Use a common convolutional neural network for training to implement fault diagnosis. Use real data for testing. Finally, output the diagnosis results to complete the fault diagnosis process.

[0012] Furthermore, in step 2, the data normalization refers to scaling all data to the interval (-1, 1) by using the maximum and minimum standardization (MinMaxScaler) method on the original samples.

[0013] Furthermore, in step 3, a Wasserstein generative adversarial network based on a gradient penalty mechanism is constructed, in which the generator includes four layers of deconvolution, three layers of instance normalization layers, and one self-attention layer, and the discriminator includes four layers of convolution, four layers of instance normalization layers, and one self-attention mechanism layer. The loss function uses Wasserstein loss and specifically includes the following steps:

[0014] Step 31: The input of the SA layer is represented as x∈R C×N , where N and C represent the length of the sequence and the number of channels of the sample, respectively. Subsequently, x is transformed into three different feature spaces, namely f, g, and h, which represent the query, key, and value, respectively. The calculation formula is shown in the following formula:

[0015] f(x)=xW f

[0016] g(x)=xW g

[0017] h(x)=xW h

[0018] in and are all weight matrices that are continuously updated. Represents the number of channels output by the self-attention mechanism. Here, the number of channels of the sample is set equal to the number of channels output by the attention mechanism;

[0019] Step 32: Calculate the attention score s i,j , the attention score is obtained by multiplying f and g, and then the softmax function is used to regularize the attention score, denoted as α i,j , the attention score and the regularized score are expressed as follows:

[0020] s i,j =f T g

[0021]

[0022] Among them, f represents the query of x transformation, g represents the key of x transformation, T represents the transposition calculation, and the subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region;

[0023] Step 33: Use the obtained α i,j and h calculate the output vector o of the SA layer corresponding to each input vector x = (o 1 ,o 2 ,…,o j ,…,o N )∈R C×N It can be expressed as follows:

[0024] v(x i )=W v x i

[0025]

[0026] Where, subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region, W v Represents the weight matrix updated during the training process;

[0027] Step 34. In order to make the model learn gradually from simple to complex, the output of the self-attention layer is further multiplied by a constant γ and added to the input, which is expressed as follows:

[0028] y i =γo i +x i

[0029] Among them, γ is a constant, which can be learned and its initial value is 0;

[0030] Step 35. Add an instance normalization (IN) layer after each convolution layer. Most of the normalization layers used in previous GANs are batch normalization (BN), which is not suitable for fault sample enhancement under small samples. This is because when the number of learnable samples is very scarce, the batch size of the model is too small each time it is trained, and the BN layer cannot estimate the mean and variance of the entire data set. First, the mean μ and variance σ of the feature dimension data of each sample are calculated as follows:

[0031]

[0032] In the formula, t represents the sequence number of the feature, i represents the signal of the sample channel, N represents the feature dimension, and M represents the number of features in a batch.

[0033] Then, normalize the samples as follows:

[0034]

[0035] Finally, two learnable variables, scaling γ and translation β, are added to obtain the final output;

[0036] y i =γo i +x i

[0037] When defining the loss function of the improved GAN, the Wasserstein distance is selected to define the distance between distributions, and its calculation formula is:

[0038]

[0039] where |f| L is the Lipschitz constraint of the function f(x). The gradient penalty strategy is used to satisfy the Lipschitz constraint of the Wasserstein distance, P data represents the actual distribution of data, P g represents the distribution of generated data, k represents the constant of normalization scale, It represents finding the function that satisfies the K-Lipschitz condition and maximizes the expected difference. x stands for mathematical expectation;

[0040] The final loss function of the model is defined as follows:

[0041]

[0042] Among them, D(x) represents the discriminator, λ represents the gradient penalty coefficient, Represents the gradient of the discriminator.

[0043] Furthermore, in step 4, the parameters are iteratively updated, the processed data is input into the model of step 3, a fixed number of data in the training set is randomly extracted to participate in the training process, the model parameters are iteratively optimized, and the training is stopped and the model is saved when the set number of iterations is met or the Nash equilibrium is met. Specifically, the steps include:

[0044] Step 41: The constructed generator uses the ReLU activation function, and the constructed discriminator uses the LeakyReLU activation function. The formula of the LeakyReLU activation function is as follows:

[0045] f(x)=max{αx,x}

[0046] When x < 0, the function value is f(x) = αx, and when x > 0, the function value is f(x) = x;

[0047] Step 42: During the training process, the generator is trained once and the discriminator is trained five times. The Adam optimizer is used for training, and the learning rate is set to 0.0001. At the same time, the real sample label is -0.5 and the generated sample label is 0.5.

[0048] Furthermore, in step 5, random noise is input into the trained generator to obtain generated samples, and the generated samples that differ too much from the real samples are removed using the independent forest-based filtering method, and the low-quality samples mixed in the generated samples are deleted. The quality of the generated samples is improved, and the specific steps include the following:

[0049] Step 51: For each data point, calculate its path length on all trees in the forest. The path length is the number of split nodes passed by the path from the root node to the leaf node, and the formula is as follows:

[0050]

[0051] Where t is the number of trees in the forest, is the path length of point x in the i-th tree;

[0052] Step 52: Calculate the anomaly score based on the path length. The shorter the path length, the higher the score, indicating that the point is more likely to be an anomaly. The anomaly score formula is as follows:

[0053]

[0054] Where c(n) is an adjustment factor, which represents the average path length when the dataset size is n. The formula is as follows:

[0055]

[0056] Step 53: Determine the outlier points. If the outlier score of a point is close to 1, then the point is very likely to be an outlier point; if the outlier score is close to 0.5, it means that the point is a normal point.

[0057] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: (1) The original batch normalization in the Wasserstein generative adversarial network with gradient penalty is changed to instance normalization, which better adapts to the learning of data by the neural network when the fault sample is small, and improves the quality of generated samples; (2) A self-attention mechanism is added to the Wasserstein generative adversarial network with gradient penalty, so that the network can pay more attention to the features that have a greater impact on the system, thereby improving the quality of generated samples; (3) An independent forest strategy is used to filter a part of the generated samples, and then the real samples are mixed into them, which improves the utilization rate of the samples and facilitates classification. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the method flow of the present invention.

[0059] Figure 2 It is a schematic diagram of the network model structure of the present invention.

[0060] Figure 3 The spatial distribution diagram of the samples generated by the present invention and the real samples for a certain fault is shown.

[0061] Figure 4 This is a confusion matrix diagram of a certain experiment of the present invention. DETAILED DESCRIPTION

[0062] like Figure 1 and Figure 2 As shown in FIG. 1 , a data enhancement method for a civil aircraft hydraulic system based on WGAN-GP and an attention mechanism includes the following steps:

[0063] Step 1: Signal acquisition: by placing corresponding sensors at key measuring points, one or more sensor signals containing information that can reflect the fault status are obtained, and the signals are classified according to different fault modes;

[0064] Step 2: Data preprocessing: normalize the data so that the features can be better learned;

[0065] Step 3: Construct a Wasserstein generative adversarial network based on a gradient penalty mechanism, where the generator contains four layers of deconvolution, three layers of instance normalization, and one self-attention layer, and the discriminator contains four layers of convolution, four layers of instance normalization, and one self-attention mechanism layer. The loss function uses Wasserstein loss.

[0066] Step 4: Iterative parameter update: input the processed data into the model in step 3, randomly extract a fixed number of data from the training set to participate in the training process, iteratively optimize the model parameters, and stop training and save the model when the set number of iterations is met or the Nash equilibrium is met;

[0067] Step 5: Input random noise into the trained generator to obtain generated samples, use the independent forest-based filtering method to remove generated samples that are too different from real samples, delete low-quality samples mixed in the generated samples, and improve the quality of generated samples;

[0068] Step 6: Implement fault diagnosis. Mix the generated data with real data to build a training data set. Use real data to build a test set. Use a common convolutional neural network for training to implement fault diagnosis. Use real data for testing. Finally, output the diagnosis results to complete the fault diagnosis process.

[0069] The data normalization described in step 2 refers to scaling all data to the (-1, 1) interval using the MinMaxScaler method for the original samples.

[0070] The specific calculation steps of the self-attention mechanism layer described in step 3 are as follows:

[0071] S1: The input of the SA layer is represented as x∈R C×N , where N and C represent the length of the sequence and the number of channels of the sample, respectively. Subsequently, x is transformed into three different feature spaces, namely f, g, and h, which represent the query, key, and value, respectively. The calculation formula is shown below:

[0072] f(x)=xW f

[0073] g(x)=xW g

[0074] h(x)=xW h

[0075] in and are all weight matrices that are continuously updated. Represents the number of channels output by the self-attention mechanism. Here, the number of channels of the sample is set equal to the number of channels output by the attention mechanism.

[0076] S2: Calculate the attention score s i,j , the attention score is obtained by multiplying f and g, and then the softmax function is used to regularize the attention score, denoted as α i,j , the attention score and the regularized score can be expressed as follows:

[0077] s i,j=f T g

[0078]

[0079] Where f represents the query of x transformation, g represents the key of x transformation, T represents the transposition calculation, and the subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region;

[0080] S3: Using the obtained α i,j and h calculate the output vector o of the SA layer corresponding to each input vector x = (o 1 ,o 2 ,...,o j ,...,o N )∈R C×N It can be expressed as follows:

[0081] v(x i )=W v x i

[0082]

[0083] Where, subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region, W v Represents the weight matrix updated during the training process;

[0084] S4: In order to enable the model to learn gradually from simple to complex, the output of the self-attention layer is further multiplied by a constant γ and added to the input, expressed as follows:

[0085] y i =γo i +x i

[0086] Among them, γ is a constant, which can be learned and its initial value is 0;

[0087] S5: Add an Instance Normalization (IN) layer after each convolution layer. Most of the normalization layers used in previous GANs are batch normalization (BN), which is not suitable for fault sample enhancement under small samples. This is because when the number of learnable samples is very scarce, the batch size of the model is too small each time it is trained, and the BN layer cannot estimate the mean and variance of the entire dataset. First, the mean μ and variance σ of the feature dimension data of each sample are calculated as follows:

[0088]

[0089]

[0090] In the formula, t represents the sequence number of the feature, i represents the signal of the sample channel, N represents the feature dimension, and M represents the number of features in a batch.

[0091] Then, normalize the sample and calculate it as follows:

[0092]

[0093] Finally, two learnable variables, scaling γ and translation β, are added to obtain the final output;

[0094] y i =γo i +x i

[0095] When defining the loss function of the improved GAN, the Wasserstein distance is selected to define the distance between distributions, and its calculation formula is:

[0096]

[0097] where |f| L is the Lipschitz constraint of the function f(x). The gradient penalty strategy is used to satisfy the Lipschitz constraint of the Wasserstein distance, P data represents the actual distribution of data, P g represents the distribution of generated data, k represents the constant of normalization scale, It represents finding the function that satisfies the K-Lipschitz condition and maximizes the expected difference. x stands for mathematical expectation.

[0098] The final loss function of the model is defined as follows:

[0099]

[0100] Among them, D(x) represents the discriminator, λ represents the gradient penalty coefficient, Represents the gradient of the discriminator.

[0101] The training of the parameter update process described in step 4 is as follows:

[0102] S1: The constructed generator uses the ReLU activation function, and the constructed discriminator uses the LeakyReLU activation function. The formula of the LeakyReLU activation function is as follows:

[0103] f(x)=max{αx,x}

[0104] When x < 0, the function value is f(x) = αx, and when x > 0, the function value is f(x) = x;

[0105] S2: During the training process, the generator is trained once and the discriminator is trained five times. The Adam optimizer is used for training, and the learning rate is set to 0.0001. At the same time, the real sample label is -0.5 and the generated sample label is 0.5.

[0106] The filtering strategy of the independent forest described in step 5 is as follows:

[0107] S1: For each data point, calculate its path length on all trees in the forest. The path length is the number of split nodes passed by the path from the root node to the leaf node, and the formula is as follows:

[0108]

[0109] Where t is the number of trees in the forest, is the path length of point x in the i-th tree. ;

[0110] S2: Calculate the anomaly score based on the path length. The shorter the path length, the higher the score, which means that the point is more likely to be an anomaly. The anomaly score formula is as follows:

[0111]

[0112] Where c(n) is an adjustment factor, which represents the average path length when the dataset size is n. The formula is as follows:

[0113]

[0114] S3: Determine the outliers. If the outlier score of a point is close to 1, then the point is very likely to be an outlier. If the outlier score is close to 0.5, then the point is a normal point.

[0115] Implementation case: In order to verify the performance of the method of the present invention, a certain type of aircraft hydraulic actuation system is simulated and modeled and fault injection is implemented. The typical fault modes are shown in Table 1. Due to the complex coupling relationship between the components of the hydraulic actuation system, multiple parameter measurement points are selected to reflect the system status. The 16 characteristic parameters are shown in Table 2. It is assumed that normal samples are sufficient and fault samples are lacking. No data enhancement is performed on the normal state, and only data enhancement is performed on five fault states. 100 samples are randomly selected for each fault for data expansion. A basic convolutional neural network is trained using the expanded data, and then tested using real data.

[0116] The model structure parameters and training parameters of the proposed inventive method are shown in Table 3 and Table 4 respectively.

[0117] Table 1 Typical failure modes of hydraulic actuation system

[0118]

[0119]

[0120] Table 2 Characteristic parameters of aircraft hydraulic actuation system

[0121]

[0122] Table 3 Model structure parameters

[0123]

[0124]

[0125] Table 4 Model training parameter settings

[0126]

[0127] In order to explore the performance of the proposed method, FID and MMD are used as evaluation indicators. The lower the value, the better the quality of the generated data, as shown in Tables 5 and 6. At the same time, the generated samples and real samples of one of the fault modes are used to generate the TSNE diagram, as shown in Figure 3 As shown in the figure, it can be seen that the spatial distribution of the real data is very similar to that of the generated data.

[0128] Table 5 FID data generated by each fault mode

[0129]

[0130] Table 6 Data MMD generated by each fault mode

[0131]

[0132] The convolutional neural network is used to classify the data before and after data enhancement and using real samples. The classification accuracy of the test set is shown in Table 7, and the confusion matrix of the classification test is shown in Figure 4 shown.

[0133] Table 7 Classification accuracy of test set

[0134]

[0135] It can be seen that the data quality after sample enhancement has basically reached the level of real data, and the classification accuracy after training with generated data is similar to that after training with real data. In the confusion matrix, 0 represents normal, 1-5 represents fault 1-fault 5, and the classification effect of fault 5 is the lowest. The reason is that fault 5 is a fire shut-off valve fault. Before the set fault occurs, the signal is consistent with the normal situation, which leads to easy misclassification.

Claims

1. A data enhancement method for civil aircraft hydraulic system based on WGAN-GP and attention mechanism, characterized in that: The steps include: Step 1: Signal acquisition: by placing corresponding sensors at key measuring points, one or more sensor signals containing information that can reflect the fault status are obtained, and the signals are classified according to different fault modes; Step 2: Data preprocessing: normalize the data so that the features can be better learned; Step 3: Construct a Wasserstein generative adversarial network based on a gradient penalty mechanism, where the generator contains four layers of deconvolution, three layers of instance normalization, and one self-attention layer, and the discriminator contains four layers of convolution, four layers of instance normalization, and one self-attention mechanism layer. The loss function uses Wasserstein loss. Step 4: Iterative parameter update: input the processed data into the model in step 3, randomly extract a fixed number of data from the training set to participate in the training process, iteratively optimize the model parameters, and stop training and save the model when the set number of iterations is met or the Nash equilibrium is met; Step 5: Input random noise into the trained generator to obtain generated samples, use the independent forest-based filtering method to remove generated samples that are too different from real samples, delete low-quality samples mixed in the generated samples, and improve the quality of generated samples; Step 6: Implement fault diagnosis. Mix the generated data with real data to build a training data set. Use real data to build a test set. Use a common convolutional neural network for training to implement fault diagnosis. Use real data for testing. Finally, output the diagnosis results to complete the fault diagnosis process.

2. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: In step 2, the data normalization refers to scaling all data to the (-1, 1) interval by using the maximum and minimum normalization method for the original samples.

3. The civil aircraft hydraulic system data enhancement method based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: In step 3, a Wasserstein generative adversarial network based on a gradient penalty mechanism is constructed, in which the generator includes four layers of deconvolution, three layers of instance normalization, and one self-attention layer, and the discriminator includes four layers of convolution, four layers of instance normalization, and one self-attention mechanism layer. The loss function uses Wasserstein loss and specifically includes the following steps: Step 31: The input of the SA layer is represented as x∈R C×N , where N and C represent the length of the sequence and the number of channels of the sample, respectively. Subsequently, x is converted into three different feature spaces, namely f, g, and h, representing query, key, and value, respectively. The calculation formula is shown below: f(x)=xW f g(x)=xW g h(x)=xW h in and are all weight matrices that are continuously updated. Represents the number of channels output by the self-attention mechanism. Here, the number of channels of the sample is set equal to the number of channels output by the attention mechanism; Step 32: Calculate the attention score s i,j , the attention score is obtained by multiplying f and g, and then the softmax function is used to regularize the attention score, denoted as α i,j , the attention score and the regularized score are expressed as follows: s i,j =f T g Among them, f represents the query of x transformation, g represents the key of x transformation, T represents the transposition calculation, and the subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region; Step 33: Use the obtained α i,j and h calculate the output vector o of the SA layer corresponding to each input vector x = (o1, o2, ..., o j ,…,o N )∈R C×N It is expressed as follows: v(x i )=W v x i Where, subscripts i and j represent the attention score of the model on the jth point when synthesizing the i-th region, W v Represents the weight matrix updated during the training process; Step 34. In order to make the model learn gradually from simple to complex, the output of the self-attention layer is further multiplied by a constant γ and added to the input, which is expressed as follows: y i =γo i +x i Among them, γ is a constant, which can be learned and its initial value is 0; Step 35. Add an instance normalization IN layer after each convolution layer, and calculate the mean μ and variance σ of the feature dimension data of each sample as follows: In the formula, t represents the sequence number of the feature, i represents the signal of the sample channel, N represents the feature dimension, and M represents the number of features in a batch; Then, normalize the samples as follows: Finally, two learnable variables, scaling γ and translation β, are added to obtain the final output; y i =γo i +x i When defining the loss function of the improved GAN, the Wasserstein distance is selected to define the distance between distributions, and its calculation formula is: where |f| L is the Lipschitz limit of the function f(x), and the gradient penalty strategy is used to satisfy the Lipschitz limit of the Wasserstein distance, P data represents the actual distribution of data, P g represents the distribution of generated data, k represents the constant of normalization scale, It represents finding the function that satisfies the K-Lipschitz condition and maximizes the expected difference. x stands for mathematical expectation; The final loss function of the model is defined as follows: Among them, D(x) represents the discriminator, λ represents the gradient penalty coefficient, represents the gradient of the discriminator.

4. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: When defining the loss function of the improved GAN, the Wasserstein distance is selected to define the distance between distributions, and its calculation formula is: where |f| L is the Lipschitz limit of the function f(x), and the gradient penalty strategy is used to satisfy the Lipschitz limit of the Wasserstein distance. The final loss function of the model is defined as follows:

5. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: In step 4, the parameters are iteratively updated. The processed data is input into the model of step 3, a fixed number of data from the training set is randomly selected to participate in the training process, and the model parameters are iteratively optimized. When the set number of iterations is met or the Nash equilibrium is met, the training is stopped and the model is saved. The specific steps include the following: Step 41: The constructed generator uses the ReLU activation function, and the constructed discriminator uses the LeakyReLU activation function. The formula of the LeakyReLU activation function is as follows: f(x)=max{αx,x} When x < 0, the function value is f(x) = αx, and when x > 0, the function value is f(x) = x; Step 42: During the training process, the generator is trained once and the discriminator is trained five times, using the Adam optimizer for training.

6. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: In step 42, the learning rate is set to 0.0001, while the true sample label is -0.5 and the generated sample label is 0.

5.

7. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 1, characterized in that: In step 5, random noise is input into the trained generator to obtain generated samples. The generated samples that differ too much from the real samples are removed using the independent forest-based filtering method, and the low-quality samples mixed in the generated samples are deleted. The specific steps to improve the quality of the generated samples are as follows: Step 51: For each data point, calculate its path length on all trees in the forest. The path length is the number of split nodes passed by the path from the root node to the leaf node. The formula is as follows: Where t is the number of trees in the forest, is the path length of point x in the i-th tree; Step 52: Calculate the anomaly score based on the path length. The shorter the path length, the higher the score, indicating that the point is more likely to be an anomaly. The anomaly score formula is as follows: Where c(n) is an adjustment factor, which represents the average path length when the dataset size is n. The formula is as follows: Step 53: Determine the abnormal point.

8. The method for data enhancement of a civil aircraft hydraulic system based on WGAN-GP and attention mechanism as claimed in claim 7, characterized in that: In step 53, if the anomaly score of a point is close to 1, then the point is very likely to be an anomaly point; if the anomaly score is close to 0.5, it means that the point is a normal point.

Images