Aero-engine fault diagnosis method and device, storage medium and equipment

Abstract

The invention discloses an aero-engine fault diagnosis method and device, a storage medium and equipment, and belongs to the technical field of aero-engine fault diagnosis, and the method comprises the steps: obtaining a vibration signal of a set position on an aero-engine, and carrying out the preprocessing of the vibration signal, and obtaining a preprocessed vibration signal; inputting the preprocessed vibration signal into the trained improved binary neural network to obtain a fault diagnosis result of the position; the improvement of the improved binary neural network relative to the binary neural network comprises the following steps: carrying out binarization processing on weights of all convolution kernels and full connection layers; the operation executed by the depth separable convolution module is improved as follows: the preprocessed vibration signal is converted into a frequency domain and then deep convolution is carried out, and then the output of the deep convolution is converted back to a time domain for point-by-point convolution; and shielding a part of input channels of the sparse residual block module through a set sparse mask in each residual block of the sparse residual block module. According to the invention, efficient fault diagnosis is realized by improving the model.

Description

Technical Field

[0001] The present invention relates to an aero-engine fault diagnosis method, device, storage medium and equipment, belonging to the technical field of aero-engine fault diagnosis. Background Art

[0002] The main power source of aircraft is the aircraft engine, and its performance and safety directly determine the reliability of the flight mission. Due to long-term exposure to extreme conditions of high temperature, high pressure and high vibration, the key components of the aircraft engine are prone to wear, fatigue or failure. In order to ensure flight safety and improve maintenance efficiency, establishing an efficient and accurate fault diagnosis system has become an important direction in aircraft engine research.

[0003] Traditional aircraft engine fault diagnosis mainly relies on the time domain feature analysis of vibration signals, such as root mean square value, peak factor, etc., to classify the fault type. However, aircraft engine vibration signals often show strong nonlinearity and non-stationarity, and traditional methods have difficulty capturing subtle differences in fault modes. In addition, complex data analysis increases the computational burden, especially on resource-constrained edge devices, where real-time performance is significantly affected.

[0004] In recent years, deep learning technology, especially convolutional neural networks (CNN), has made significant progress in the field of pattern recognition. CNN has powerful feature extraction capabilities and provides a new perspective for the analysis of complex signals. Since its introduction, the residual network (ResNet) has become one of the important architectures in deep neural networks. Its core feature is that by introducing residual blocks, that is, introducing skip connections in each layer of the network, the network can better transfer gradients during training, solving the problem of gradient disappearance.

[0005] However, although deep neural networks work very well on many tasks, their computational and storage requirements are extremely high, especially on complex models and large-scale datasets. Therefore, many studies have begun to try to significantly reduce these requirements by binarizing the weights and activation values ​​of neural networks. Binary neural networks came into being, which limit weights and activation values ​​to two states, -1 and +1 or 0 and +1, thereby greatly reducing the accuracy of parameter representation and reducing computational overhead.

[0006] However, the advantages of binary neural networks are not obvious on small data sets. Binary neural networks usually require a large amount of training data to fully play their role. When the amount of data is small, the binary network cannot effectively learn enough features when the amount of data is insufficient, resulting in insignificant efficiency improvement during training. On the contrary, due to its special training method, the computational burden is increased, and the speed of model training and reasoning may even be slower than that of traditional networks.

[0007] In summary, the prior art still has the problems of low fault diagnosis efficiency and heavy computational burden when the amount of data is small. Summary of the invention

[0008] The purpose of the present invention is to provide an aircraft engine fault diagnosis method, device, storage medium and equipment to solve the problems of low diagnostic efficiency and heavy calculation burden in the prior art.

[0009] To achieve the above objectives, the present invention is implemented by adopting the following technical solutions: In a first aspect, the present invention provides an aircraft engine fault diagnosis method, comprising: Acquire a vibration signal of a set position on the aircraft engine, and obtain a preprocessed vibration signal after preprocessing; Inputting the preprocessed vibration signal into the trained improved binary neural network to obtain the fault diagnosis result of the position; Among them, the improvements of the improved binary neural network relative to the binary neural network include: binarizing the weights of all convolution kernels and fully connected layers; improving the operation performed by the depthwise separable convolution module to: converting the preprocessed vibration signal into the frequency domain and then performing deep convolution, and then converting the output of the deep convolution back into the time domain for point-by-point convolution; and masking part of the input channels of the sparse residual block module by a set sparse mask in each residual block of the sparse residual block module.

[0010] Further, the preprocessing includes spectral kurtosis data enhancement; The spectral kurtosis data enhancement comprises the following steps: The spectral kurtosis is calculated by the following formula: ; ; ; in, It is a vibration signal. is the window function, τ represents the time integral variable, f is the frequency variable, t is the time variable, j represents the imaginary unit, represents the power spectral density at time t and frequency f, is the mean of the power spectral density, E[ ] represents the expected value calculation, represents the spectral kurtosis; According to the spectral kurtosis, the vibration signal is enhanced by the following formula: ; in, is the i-th sample, each sample consists of multiple vibration signals, is the i-th sample after enhancement, is the disturbance coefficient; The vibration signal and the enhanced vibration signal are combined to obtain a preprocessed vibration signal.

[0011] Furthermore, the weights of all convolution kernels and fully connected layers are binarized using the following formula: ; in, Indicates that the weight The elements in are binarized to 1 or -1, W b Represents the binarization weight.

[0012] Furthermore, the preprocessed vibration signal is converted into the frequency domain and then deep convolution is performed, and then the output of the deep convolution is converted back into the time domain for point-by-point convolution, which is performed by the following formula: ; ; ; in, represents the element-wise product in the frequency domain, represents the output of the preprocessed vibration signal after Fourier transform, Represents the binary deep convolution kernel The output of the Fourier transform is express and The result of element-wise multiplication, y d represents the output of the deep convolution in the time domain, represents the inverse Fourier transform, e represents the natural exponential, j represents the imaginary unit, is the point-by-point convolution kernel after binarization. is the output of point-wise convolution in the time domain.

[0013] Furthermore, the expression of the sparse residual block module is: ; Among them, y m represents the output of the sparse residual block module, represents the depthwise separable convolutional layer, y sparse Represents the input of the sparse residual block module with the sparse mask added.

[0014] Furthermore, the sparse mask is added to the input of the sparse residual block module by the following formula: ; in, is a sparse mask, is the input of the sparse residual block module. The sparse mask is a matrix containing 0 and 1. When an element in is 0, The corresponding channels in will be masked, and the elements in the sparse mask and The channels correspond one to one.

[0015] Furthermore, the improved binary neural network is verified by accuracy after training. If the accuracy is greater than a preset value, it means that the improved binary neural network has been trained. Otherwise, the improved binary neural network is retrained until the accuracy is greater than the preset value. The accuracy is calculated by the following formula: ; in, is the true label of the setting The probability of the class, The model predicts The probability of the class, is the number of categories, is the indicator function, when and The value is 1 if they are equal, otherwise 0.

[0016] In a second aspect, the present invention provides an aircraft engine fault diagnosis device, comprising: The vibration signal acquisition module is configured to: acquire a vibration signal at a set position on the aircraft engine, and obtain a preprocessed vibration signal after preprocessing; The fault diagnosis module is configured to: input the preprocessed vibration signal into the trained improved binary neural network to obtain a fault diagnosis result of the position; Among them, the improvements of the improved binary neural network relative to the binary neural network include: binarizing the weights of all convolution kernels and fully connected layers; improving the operation performed by the depthwise separable convolution module to: converting the preprocessed vibration signal into the frequency domain and then performing deep convolution, and then converting the output of the deep convolution back into the time domain for point-by-point convolution; and masking part of the input channels of the sparse residual block module by a set sparse mask in each residual block of the sparse residual block module.

[0017] In a third aspect, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the steps of the aircraft engine fault diagnosis method described in any one of the first aspects.

[0018] In a fourth aspect, the present invention provides a computer system, comprising: Memory, for storing computer programs / instructions; A processor is used to execute the computer program / instructions to implement the steps of the aircraft engine fault diagnosis method according to any one of claims 1 to 7.

[0019] Compared with the prior art, the beneficial effects achieved by the present invention are: The present invention provides an aircraft engine fault diagnosis method, device, storage medium and equipment. By improving the binary neural network, that is, binarizing the weights of all convolution kernels and fully connected layers, the computational complexity and data storage are reduced, and the amount of calculation is reduced. By improving the operation in the depthwise separable convolution module to converting to the frequency domain before convolution, the training time is optimized, the training efficiency and the accuracy of the model are improved, and by introducing sparse mask processing in the residual block, the data storage is reduced, and efficient processing of vibration signals and efficient diagnosis of faults are achieved.

[0020] At the same time, for the environment of small data sets, the data enhancement strategy effectively solves the problems of heavy computational burden and insufficient performance of binary neural networks when the amount of data is small, further enhancing the generalization ability of the model.

[0021] Therefore, the present invention can not only quickly and accurately identify the type of aircraft engine faults, but also realize efficient and immediate fault diagnosis in an edge computing environment with limited resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of an aircraft engine fault diagnosis method provided by an embodiment of the present invention; Figure 2 is a schematic diagram of the structure of a sparse residual block provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the protection scope of the present invention.

[0024] Embodiment 1, as Figure 1 As shown, this embodiment provides an aircraft engine fault diagnosis method, including: Acquire a vibration signal of a set position on the aircraft engine, and obtain a preprocessed vibration signal after preprocessing; Inputting the preprocessed vibration signal into the trained improved binary neural network to obtain the fault diagnosis result of the position; Among them, the improvements of the improved binary neural network relative to the binary neural network include: binarizing the weights of all convolution kernels and fully connected layers; improving the operation performed by the depthwise separable convolution module to: converting the preprocessed vibration signal into the frequency domain and then performing deep convolution, and then converting the output of the deep convolution back into the time domain for point-by-point convolution; and masking part of the input channels of the sparse residual block module by a set sparse mask in each residual block of the sparse residual block module.

[0025] The present invention reduces computational complexity and data storage, and reduces the amount of calculation by improving the binary neural network, that is, binarizing the weights of all convolution kernels and fully connected layers. The present invention also optimizes the training time, improves the training efficiency and the accuracy of the model, and reduces data storage by introducing sparse mask processing in the residual block, thereby achieving efficient processing of vibration signals and efficient diagnosis of faults.

[0026] Embodiment 2 further illustrates the implementation of the present invention in conjunction with the accompanying drawings. Figure 1 is a network flow chart of an aircraft engine fault diagnosis method provided by the present invention, such as Figure 1 As shown, the method includes the following seven steps. The improved binary neural network used is specifically a binary sparse residual network. The specific process is as follows.

[0027] The first step is vibration signal input and preprocessing.

[0028] Get the vibration signal of the aircraft engine, denoted as X, , represents a space of dimension A×N, where is the number of samples of the vibration signal obtained, is the number of data points per sample, and N is the sum of the number of data points and the number of labels per sample.

[0029] Construct N-1 vibration signals into a sample, put the true label of the sample into the last position of the sample, and get a complete sample; the i-th sample is recorded as x i , , where x i1 represents the first vibration signal in the i-th sample, x i2 represents the second vibration signal in the i-th sample, x i3 represents the third vibration signal in the i-th sample, xiN-1 The N-1th vibration signal in the ith sample and the mth vibration signal in the ith sample can be expressed by the general formula x im Indicates that m=1,2,3,…,N-1, , represents the true label of the i-th sample.

[0030] In order to solve the problem of insufficient performance of binary neural networks when the amount of data is small, data enhancement is performed in the first step. For this, the spectral kurtosis data enhancement method based on short-time Fourier transform (STFT) is used. Spectral kurtosis is a quantitative statistic of the sharpness of the signal spectrum. The calculation method is as follows: First, the spectrum of the signal is calculated by short-time Fourier transform STFT. Given the vibration signal , spectrum It can be expressed as: ; in, is the window function, which is used to control the duration of each Fourier transform, τ represents the time integral variable, is a frequency variable, is the time variable and j represents the imaginary unit.

[0031] Next, calculate the power spectral density (PSD), which is the square of the STFT calculation result (the spectrum of the signal): ; in, represents the power spectral density at time t and frequency f.

[0032] Finally, calculate the spectral kurtosis , the spectral kurtosis is calculated by the following formula: ; in, represents the spectral kurtosis, is the mean of the power spectral density, and E[ ] represents the expected value calculation.

[0033] For each sample, set multiple perturbation coefficients (such as 0.05, 0.1, 0.15, etc.) for enhancement processing. The specific method is: ; in, is the i-th sample, is the i-th sample after enhancement, is the disturbance coefficient, is the spectral kurtosis of the signal.

[0034] Then, the original signal and the enhanced signal are combined to obtain the preprocessed vibration signal: , the signal will be transmitted to the subsequent convolution module.

[0035] A high spectral kurtosis value indicates that the spectrum is more concentrated and the signal is sharper; a low spectral kurtosis value indicates that the signal is smoother. Through this method, the signal data can be enhanced, the diversity of training data can be increased, and the problem of insufficient performance of binary neural networks when the amount of data is small can be solved.

[0036] The second step is to binarize the convolution kernel and the fully connected layer.

[0037] In order to improve the efficiency of convolution operations, weight binarization technology is introduced. The weights of the convolution kernel and the fully connected layer in the binary neural network will be binarized and their values ​​will be limited to +1 and -1, thereby reducing computational complexity and data storage.

[0038] The preprocessed vibration signal is input into the binary neural network. Before entering the convolution operation, the binary neural network performs binarization on the weights of all convolution kernels and fully connected layers, that is, converting the original weight W into a binary weight. , indicating the dimension space, where is the number of input channels of the convolution kernel, is the number of output channels of the convolution kernel, is the size of the convolution kernel.

[0039] The original weight W is converted into a binary weight using the following formula: ; in, Indicates that the weight The elements in are binarized to ±1, which significantly reduces the amount of calculation. b Represents the binarization weight.

[0040] The third step is to process the signal using the depthwise separable frequency domain convolution module.

[0041] Preprocessed vibration signal Feature extraction is performed through the depthwise separable frequency domain convolution module. Depthwise separable convolution is implemented in two steps: depthwise convolution and pointwise convolution. Depthwise convolution performs convolution operations independently for each input channel, while pointwise convolution integrates the output results of depthwise convolution into new output channels with the help of 1×1 convolution kernel. In order to further improve the efficiency of convolution operation, frequency domain convolution is added to the depthwise separable convolution to obtain the depthwise separable frequency domain convolution module. In the depthwise separable frequency domain convolution module, the input signal and convolution kernel are first converted to the frequency domain through Fourier transform, depthwise convolution is performed in the frequency domain, and then restored back to the time domain through inverse Fourier transform, and finally pointwise convolution is performed.

[0042] Assume that the preprocessed vibration signal The output of the Fourier transform is , the binary deep convolution kernel The output of the Fourier transform is , , Representation Dimension The convolution operation can be expressed in the frequency domain as: ; in, represents the element-wise product in the frequency domain, represents the output of the convolution operation in the frequency domain, which is also and The result of performing the element-wise multiplication.

[0043] The final output of the deep convolution in the time domain is: ; Among them, y d represents the output of the deep convolution in the time domain, represents inverse Fourier transform, e represents natural exponent, and j represents imaginary unit; The output of point-wise convolution in the time domain is: ; in, is the point-by-point convolution kernel after binarization. , Representation Dimension space, is the output of point-wise convolution in the time domain.

[0044] The frequency domain convolution result is converted back to the time domain signal through inverse Fourier transform. Through frequency domain convolution, the frequency domain information of the signal can be extracted more effectively, thereby enhancing the fault diagnosis ability of the network.

[0045] The fourth step is to process the signal using the sparse residual block module.

[0046] The output of the depth-separable frequency domain convolution module is passed through the activation function and the maximum pooling layer to obtain the input of the sparse residual block module. ,Will Enter the sparse residual block module. In the design of this module, static sparse masks are introduced to improve computational efficiency. Specifically, in each residual block, a custom sparse mask is used to randomly mask The partial channels of the sparse residual block are obtained to reduce the amount of calculation. Figure 2 shown.

[0047] The operation process of each residual block in the sparse residual block module can be expressed as: ; in, represents a depthwise separable convolutional layer, is the input of the sparse residual block module, It is the output of the residual block after the residual connection.

[0048] In order to introduce sparsity, the input of the sparse residual block module A sparse mask is applied on , the formula is: ; Among them, y sparse represents the input of the sparse residual block module with the sparse mask added, is a sparse mask, which is a matrix containing 0s and 1s used to mask When part of the channel When an element in is 0, The corresponding channels in will be masked, thereby reducing the computational complexity, and the elements and The channels correspond one to one.

[0049] Finally, the expression of the sparse residual block module is: ; Among them, y m represents the output of the sparse residual block module, Represents a depthwise separable convolutional layer.

[0050] Step 5: Fully connected layer and classification.

[0051] Output of the sparse residual block module The global average pooling layer is used for dimensionality reduction to extract global features. The features after the global average pooling layer are input into a binary fully connected layer for final fault classification. Through the softmax function, the network outputs the probability of each fault category, and finally obtains the model prediction value. .

[0052] The overall process of the network is: vibration signal After preprocessing, the preprocessed vibration signal is obtained , the preprocessed signal Enter the depth-separable frequency domain convolution module, and obtain it after deep frequency domain convolution and point-by-point convolution. , obtained through activation function and maximum pooling layer ,Will Input to the sparse residual block module, masked by sparse mask Some channels of the network are connected through residual connections to obtain the final output. ,Finally, after the fully connected layer and classification processing, ,the network outputs the fault category label to achieve fault diagnosis.

[0053] The sixth step is to train the network.

[0054] In the training phase, the AdamW (Adam with Weight Decay Fix) optimizer is used to optimize the network performance with the help of the cross entropy loss function (CrossEntropyLoss). The cross entropy loss function is used for classification tasks and is defined as: ; Among them, L is the value of the cross entropy loss function, is the true label of the setting The probability of the class, The model predicts The probability of the class, is the number of categories.

[0055] The gradient is calculated through back propagation and the network parameters are updated, ultimately minimizing the loss function L and improving the overall performance of the network.

[0056] Step 7: Fault diagnosis.

[0057] After the model training is completed, the model is evaluated through the test set. The test data is input into the network, the prediction results are obtained, and the results are compared with the true labels to calculate the accuracy.

[0058] The accuracy calculation formula is: ; in is the indicator function, when the predicted value With the true label The value is 1 if they are equal, otherwise 0.

[0059] Through this process, the model can quickly and accurately diagnose faults in different situations and provide a basis for equipment maintenance.

[0060] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0061] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0062] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0063] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0064] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for diagnosing aircraft engine faults, characterized in that: include: Acquire a vibration signal of a set position on the aircraft engine, and obtain a preprocessed vibration signal after preprocessing; Inputting the preprocessed vibration signal into the trained improved binary neural network to obtain the fault diagnosis result of the position; Among them, the improvements of the improved binary neural network relative to the binary neural network include: binarizing the weights of all convolution kernels and fully connected layers; improving the operation performed by the depthwise separable convolution module to: converting the preprocessed vibration signal into the frequency domain and then performing deep convolution, and then converting the output of the deep convolution back into the time domain for point-by-point convolution; and masking part of the input channels of the sparse residual block module by a set sparse mask in each residual block of the sparse residual block module.

2. The method for diagnosing aircraft engine faults according to claim 1, characterized in that: The preprocessing includes spectral kurtosis data enhancement; The spectral kurtosis data enhancement comprises the following steps: The spectral kurtosis is calculated by the following formula: ； ； ； in, It is a vibration signal. is the window function, τ represents the time integral variable, f is the frequency variable, t is the time variable, j represents the imaginary unit, represents the power spectral density at time t and frequency f, is the mean of the power spectral density, E[ ] represents the expected value calculation, represents the spectral kurtosis; According to the spectral kurtosis, the vibration signal is enhanced by the following formula: ； in, is the i-th sample, each sample consists of multiple vibration signals, is the i-th sample after enhancement, is the disturbance coefficient; The vibration signal and the enhanced vibration signal are combined to obtain a preprocessed vibration signal.

3. The method for diagnosing aircraft engine faults according to claim 1, characterized in that: The weights of all convolution kernels and fully connected layers are binarized using the following formula: ； in, Indicates that the weight The elements in are binarized to 1 or -1, W b Represents the binarization weight.

4. The method for diagnosing aircraft engine faults according to claim 1, characterized in that: The preprocessed vibration signal is converted into the frequency domain and then deep convolution is performed, and then the output of the deep convolution is converted back to the time domain for point-by-point convolution, which is performed by the following formula: ； ； ； in, represents the element-wise product in the frequency domain, represents the output of the preprocessed vibration signal after Fourier transform, Represents the binary deep convolution kernel The output of the Fourier transform is express and The result of element-wise multiplication, y d represents the output of the deep convolution in the time domain, represents the inverse Fourier transform, e represents the natural exponential, j represents the imaginary unit, is the point-by-point convolution kernel after binarization. is the output of point-wise convolution in the time domain.

5. The method for diagnosing aircraft engine faults according to claim 1, characterized in that: The expression of the sparse residual block module is: ; Among them, y m represents the output of the sparse residual block module, represents the depthwise separable convolutional layer, y sparse Represents the input of the sparse residual block module with the sparse mask added.

6. The method for diagnosing aircraft engine faults according to claim 5, characterized in that: The sparse mask is added to the input of the sparse residual block module by the following formula: ; in, is a sparse mask, is the input of the sparse residual block module. The sparse mask is a matrix containing 0 and 1. When an element in is 0, The corresponding channels in will be masked, and the elements in the sparse mask and The channels correspond one to one.

7. The method for diagnosing aircraft engine faults according to claim 1, characterized in that: After the training is completed, the improved binary neural network is also verified by the accuracy rate. If the accuracy rate is greater than a preset value, it means that the improved binary neural network has been trained. Otherwise, the improved binary neural network is retrained until the accuracy rate is greater than the preset value. The accuracy is calculated by the following formula: ； in, is the true label of the setting The probability of the class, The model predicts The probability of the class, is the number of categories, is the indicator function, when and The value is 1 if they are equal, otherwise 0.

8. An aircraft engine fault diagnosis device, characterized in that: include: The vibration signal acquisition module is configured to: acquire a vibration signal at a set position on the aircraft engine, and obtain a preprocessed vibration signal after preprocessing; The fault diagnosis module is configured to: input the preprocessed vibration signal into the trained improved binary neural network to obtain a fault diagnosis result of the position; Among them, the improvements of the improved binary neural network relative to the binary neural network include: binarizing the weights of all convolution kernels and fully connected layers; improving the operation performed by the depthwise separable convolution module to: converting the preprocessed vibration signal into the frequency domain and then performing deep convolution, and then converting the output of the deep convolution back into the time domain for point-by-point convolution; and masking part of the input channels of the sparse residual block module by a set sparse mask in each residual block of the sparse residual block module.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the steps of the aircraft engine fault diagnosis method described in any one of claims 1 to 7 are implemented.

10. A computer system, characterized in that: include: Memory, for storing computer programs / instructions; A processor is used to execute the computer program / instructions to implement the steps of the aircraft engine fault diagnosis method according to any one of claims 1 to 7.

Images