An aviation tubing fitting leak level classification method, system, and apparatus

Abstract

The application discloses a kind of aviation pipeline joint leakage level classification method, system and equipment, it is related to aviation pipeline system leakproofness detection technical field, the method includes: carrying out multiple groups of pipeline leakage experiment, and the ultrasonic signal of leaked gas under different leakage levels is collected;The ultrasonic signal is carried out time-frequency segmentation two-dimensional feature fusion;Leak rate classification model is constructed;The leak rate classification model is trained based on the two-dimensional feature;The ultrasonic signal of aviation pipeline joint leakage is classified by the trained leak rate classification model, and the leakage level is determined.The application can realize the classification of aviation pipeline leakage ultrasonic signal under multiple different leakage levels, so as to determine the leakage level of aviation pipeline.

Description

Technical Field

[0001] This invention relates to the field of airtightness testing technology for aviation piping systems, and particularly to a method, system, and equipment for classifying leakage levels of aviation piping joints. Background Technology

[0002] Aircraft component piping systems are characterized by their numerous, scattered, long, and complex nature, containing thousands of connecting pipes exceeding 1000 meters in length, with intricate spatial structures and diverse connection methods. Under the combined effects of various airworthiness loads, these piping systems are prone to leaks. Traditional methods for detecting leaks in aviation piping, such as visual inspection and pressure testing, while commonly used, have limitations when dealing with the complex piping systems of modern aerospace engineering. The most significant drawback is the difficulty in quantitatively classifying leak points. Visual inspection can only detect obvious external signs of leaks, such as drips or corrosion marks, but cannot quantify the severity of the leak. Similarly, pressure testing can only indicate the presence of a leak, but cannot provide detailed information on the leak rate or leakage magnitude. This lack of information hinders the determination of repair and maintenance priorities, potentially leading to wasted resources or overlooking potentially hazardous leaks.

[0003] The internal pressure of aircraft piping is higher than the external air pressure. If a leak occurs, the ejected medium often has a high Reynolds number due to the small size of the leak, thus creating turbulence. The sound waves generated by this turbulence near the leak have frequencies greater than 20kHz, and these ultrasonic signals can be captured by ultrasonic sensors. Existing techniques have provided mechanistic analyses of the sound generation principles and source characteristics during leaks and performed aeroacoustic simulations of the sound wave field, but no method has been found to quantitatively determine or automatically classify the relationship between the leakage rate and the ultrasonic signal. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, and equipment for classifying the leakage level of aviation pipeline joints, which can quickly classify the leakage level by locating the leak point through non-contact measurement.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A method for classifying leakage levels of aviation pipeline joints, comprising:

[0007] Multiple sets of pipeline leakage experiments were conducted, and ultrasonic signals of leaked gas under different leakage levels were collected;

[0008] The ultrasonic signal is subjected to time-frequency segmented two-dimensional feature fusion;

[0009] Construct a leak rate classification model; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module;

[0010] The leakage rate classification model is trained based on the two-dimensional features;

[0011] The ultrasonic signals of leaks in aviation pipeline joints are classified using a trained leak rate classification model to determine the leak level.

[0012] The present invention also provides the following solutions:

[0013] A leakage classification system for aviation piping joints includes:

[0014] The ultrasonic signal acquisition module is used to conduct multiple pipeline leakage experiments and acquire ultrasonic signals of leaked gas under different leakage levels.

[0015] The segmented fusion module is used to perform time-frequency segmented two-dimensional feature fusion on the ultrasonic signal;

[0016] The model building module is used to build a leak rate classification model; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module.

[0017] The training module is used to train the leakage rate classification model based on the two-dimensional features;

[0018] The grading module is used to classify the ultrasonic signals of leaks in aviation pipeline joints using a trained leak rate grading model, and to determine the leak level.

[0019] The present invention also provides the following solutions:

[0020] An electronic device includes a memory and a processor, the memory storing a computer program, and the processor running the computer program to cause the electronic device to perform the above-described method for classifying leakage levels of aviation pipeline joints.

[0021] The present invention also provides the following solutions:

[0022] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for classifying leakage levels of aviation pipeline joints.

[0023] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0024] This invention performs time-frequency segmented two-dimensional feature fusion on ultrasonic signals of pipeline leaks, and establishes a leak rate classification model including a Transformer feature extraction module and a multi-sphere depth support vector data description and classification module. This enables the classification of ultrasonic signals of aviation pipeline leaks under various different leak levels, thereby determining the leak level of aviation pipelines. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart of the method for classifying leakage levels of aviation pipeline joints provided by the present invention;

[0027] Figure 2 This is a schematic diagram of a leakage rate classification model;

[0028] Figure 3 This is a schematic diagram illustrating the working process of the Transformer feature extraction module.

[0029] Figure 4 A schematic diagram illustrating the working process of the multi-sphere depth support vector data description classification module;

[0030] Figure 5 This is a schematic diagram of the aircraft pipeline leakage test bench provided by the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] To overcome the deficiency of existing pipeline testing methods in lacking leakage rate classification functionality, this invention provides a method, system, and equipment for classifying leakage levels of aviation pipeline joints.

[0033] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] Example 1

[0035] like Figure 1 As shown, the leakage level classification method for aviation pipeline joints provided by the present invention includes the following steps:

[0036] S1: Conduct multiple sets of pipeline leakage experiments and collect ultrasonic signals of leaked gas under different leakage levels.

[0037] To obtain ultrasonic signals of leaked gas at different leakage levels, this invention discloses an aircraft pipeline leak test bench, which includes a leaking pipeline, gas source, air filter, pressure regulator, and pressure gauge. Figure 5 Assemble as shown. Place the SoundCam Ultra acoustic camera horizontally directly in front of the leak hole, 0.2m away. The acoustic camera uses a high-performance ultrasonic sensor with a sensitivity of -45dB and a frequency range of 20–20000Hz. The sampling frequency of the acquisition system is 44kHz, and the single sampling time is 5s.

[0038] By manually adjusting the size of the standard leak hole, the leakage rate under different leakage conditions is simulated. By gradually reducing the leakage rate, different stages of pipeline leakage are simulated, and corresponding ultrasonic signals are collected.

[0039] Because 72 microphone sensors record data simultaneously during the sampling process, 72 sets of data are collected in a single experiment for the same leak.

[0040] S2: Perform time-frequency segmented two-dimensional feature fusion on the ultrasonic signal.

[0041] To better uncover leakage features in ultrasonic signals and improve classification accuracy, this invention first constructs a time-frequency segmented two-dimensional feature fusion method, which stacks the time-domain signal with the squared envelope spectrum to form a two-dimensional input.

[0042] Step S2 specifically includes:

[0043] S21: In ultrasound signal x i The length f is extracted from (t)(i∈N). s / 2, where f s The sampling frequency.

[0044] S22: Construct a bandpass filter h(n) based on a fourth-order Butterworth filter, with a lower cutoff frequency f. l =50kHz, upper cutoff frequency f l =70KHz.

[0045] S23: Using h(n) to analyze the ultrasonic signal x i (t) is filtered, and the filtered signal x' is calculated. i SES (square envelope spectrum of t) i (f), the formula is as follows:

[0046]

[0047] Where H represents the Hilbert transform, F represents the Fourier transform, and n is the number of sampling points.

[0048] S24: Perform two-dimensional concatenation of the time-domain and squared envelope spectra of N ultrasound signals to obtain a two-dimensional matrix M.i .

[0049] This invention can simultaneously capture different aspects of ultrasound signals in both the time and frequency domains. The time-domain representation reflects the characteristics of the signal as it changes over time, while the squared envelope spectrum representation provides the signal's components at different frequencies. Combining these two representations provides more diverse and comprehensive information, helping the grading model better understand the signal. Furthermore, the time-domain representation may be more sensitive to noise, while the squared envelope spectrum representation can distinguish between signal and noise at different frequency bands. By considering both representations simultaneously, the leak rate grading model can more effectively reduce the impact of noise and improve robustness.

[0050] S3: Construct a leak rate classification model; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module.

[0051] To learn the mapping relationship from the time-frequency segmented 2D feature fusion matrix to the final task output in an end-to-end manner, a leak rate hierarchical model is built, including a Transformer feature extraction module and a multi-sphere support vector data description classification module, such as... Figure 2 As shown, the model can automatically learn how to extract useful features from the raw signal without requiring manual design of a feature extractor.

[0052] like Figure 3 As shown, the Transformer feature extraction module adopts a ViT structure, which includes four parts:

[0053] 1) Linear Projective Layer Transformation. Preferably, the linear projective layer transformation divides the input two-dimensional matrix into 128×128 patches, each patch having dimensions of 128×128×1=16384. Each image will generate (N×f) s / 2) / (128×128) patches, the input after passing through the linear projection layer has a dimension of (N×f) s There are 128×128)×16384 tags, each with a dimension of 16384. Additionally, a leakage rate level label is added, so the final dimension is (N×f) / (128×128)×16384. s / 2) / (128×128)×16385.

[0054] 2) Location information embedding. Preferably, the location information embedding layer needs to add location encoding to the output features, and the encoding is added through the SUM function.

[0055] 3) Multi-head attention mechanism. Preferably, the multi-head self-attention layer considers the relationship between different positions within the input sequence when calculating attention weights, thus forming a self-attention mechanism.

[0056] 4) Multilayer perceptron. Preferably, the multilayer perceptron has 5 layers, and the activation function is ReLU. The number of neurons in the output layer is 256.

[0057] To optimize feature extraction performance, this invention utilizes a ViT structure to construct a Transformer feature extraction module. The ViT structure extracts and organizes features through a series of layers of processing, including embedding, attention, and feedforward neural networks. Through a self-attention mechanism, it captures the dependencies between features globally.

[0058] The Transformer feature extraction module uses a self-attention mechanism to calculate the attention weights between each position in the input sequence and other positions. This allows the model to focus on the correlations between different positions in the input sequence, thereby capturing the correlations between signals in the time and frequency domains. This is crucial for understanding long-range dependencies in ultrasound signal time-series data.

[0059] Generally, increasing model depth can improve the model's expressive power, but it also increases computation and training difficulty; the ViT layer count is 3. Multi-head attention allows the model to focus on different feature subspaces, thereby improving the model's representational power; the ViT attention head count is 16. Embedding dimension defines the feature dimension of the input matrix in the model, which is determined according to the size of the input matrix. Batch size determines the number of images processed by the model in one iteration; the batch size is 128.

[0060] The multi-sphere support vector data description and classification module takes the features extracted by the Transformer feature extraction module as input, and includes a fully connected layer and n hyperspheres, where n corresponds to the number of leakage levels. Preferably, the number of leakage levels is n = 4. The multi-sphere support vector data description and classification module uses the k-means algorithm to determine the center position of each hypersphere. During the training of the leakage rate grading model, the hyperparameters of the Transformer feature extraction module and the multi-sphere support vector data description and classification module are adjusted using the cross-entropy loss function. Preferably, the ultrasound data and the output data of the leakage rate grading model are visualized using t-SNE.

[0061] The output of the Transformer feature extraction module is input into the multi-sphere deep support vector data description module. This module combines support vector data description and the multi-sphere method to reduce the adverse effects of many redundant and irrelevant features on classification learning when dealing with classification problems in high-dimensional data. It learns a compact representation of the data using the k-means algorithm, thereby highlighting key features in the signal. Classification is achieved by mapping data points to different spheres. The multi-sphere model exhibits better adaptability on different datasets, improving the model's generalization performance under complex data conditions.

[0062] like Figure 4 As shown, the features are first activated by the ReLU function and then passed through a max pooling layer. Using the center positions of each class in the k-means function as the centers of the hyperspheres, the volume of the entire hypersphere is calculated, and the optimization objective is to minimize this volume.

[0063] S4: Train the leakage rate classification model based on the two-dimensional features.

[0064] This invention, through a thorough analysis of the pipeline leakage process, selects ultrasonic signal data to indirectly reflect the leakage state, further establishes a mapping relationship between the data and the leakage level, and trains the model based on this.

[0065] S5: The ultrasonic signals of leaks in aviation pipeline joints are classified using a trained leak rate classification model to determine the leak level.

[0066] The model is deployed on an edge computing device, using real-time ultrasonic signals measured by an acoustic camera as input, and outputting the pipeline leakage level at that location, typically from level one to four, i.e., 5 × 10⁻⁶. -3 MPa (Level 1), 9×10 -3 MPa (Level II), 5×10 -2 MPa (Level III), 5×10 -1 MPa (Level 4) represents four leakage levels, ranging from minor to severe.

[0067] After obtaining the classification results, appropriate actions are taken based on the severity of the leak. For Level 1 or 2 leaks, preventative maintenance measures can be taken, while for Level 4 or 5 leaks, immediate emergency repair measures are required to prevent potential aircraft safety risks.

[0068] Compared with existing technologies, the leak rate classification model provided by this invention can automatically learn the leak-related features in ultrasonic monitoring data and significantly improve the accuracy of different leak rate classifications.

[0069] Another beneficial effect of this invention is that, in response to the problems of lack of pipeline leakage monitoring methods and unclear leakage status assessment standards in aircraft pipeline sealing tests, this invention can divide the leakage point into different leakage stages according to the leakage pressure, obtain ultrasonic data of different stages to form a pipeline leakage ultrasonic signal dataset, establish a quantitative leakage assessment criterion, and then reflect the leakage status through ultrasonic data of different leakage levels, providing data support for the training and testing of leakage rate classification models.

[0070] Another beneficial effect of the present invention is that, through the monitoring indicators provided by the present invention, namely ultrasonic data and ultrasonic signal acquisition method, and combined with the leak rate classification model of the present invention, continuous and real-time monitoring of the sealing performance of the entire aircraft piping can be achieved.

[0071] Example 2

[0072] In order to implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, an aviation pipeline joint leakage level classification system is provided below.

[0073] The system includes:

[0074] The ultrasonic signal acquisition module is used to conduct multiple pipeline leakage experiments and acquire ultrasonic signals of leaked gas under different leakage levels.

[0075] The segmented fusion module is used to perform time-frequency segmented two-dimensional feature fusion on the ultrasonic signal.

[0076] The model building module is used to build a leak rate classification model; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module.

[0077] The training module is used to train the leak rate classification model based on the two-dimensional features.

[0078] The grading module is used to classify the ultrasonic signals of leaks in aviation pipeline joints using a trained leak rate grading model, and to determine the leak level.

[0079] Furthermore, the segmented fusion module specifically includes:

[0080] The interception unit is used to intercept the ultrasonic signal according to the sampling frequency to obtain the time domain of the ultrasonic signal.

[0081] The square envelope spectrum calculation unit is used to filter the ultrasonic signal through a bandpass filter and calculate the square envelope spectrum of the filtered ultrasonic signal; the bandpass filter is a bandpass filter based on a fourth-order Butterworth filter.

[0082] A two-dimensional splicing unit is used to splice the time domain and the squared envelope spectrum of the ultrasound signal in two dimensions to obtain two-dimensional features.

[0083] Example 3

[0084] Embodiment 3 of the present invention provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the aviation pipeline joint leakage level classification method provided in Embodiment 1.

[0085] In practical applications, the aforementioned electronic devices can be servers.

[0086] In practical applications, electronic devices include: at least one processor, memory, bus, and communication interface.

[0087] The processor, communication interface, and memory communicate with each other via a communication bus.

[0088] A communication interface is used to communicate with other devices.

[0089] The processor is used to execute programs, specifically the methods described in the above embodiments.

[0090] Specifically, the program may include program code, which includes computer operation instructions.

[0091] The processor may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The electronic device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0092] Memory is used to store programs. Memory may include high-speed RAM, and may also include non-volatile memory, such as at least one disk drive.

[0093] Example 4

[0094] Based on the description of Embodiment 3, Embodiment 4 of the present invention provides a storage medium storing a computer program thereon, which can be executed by a processor to implement the aviation pipeline joint leakage level classification method of Embodiment 1.

[0095] The aviation pipeline joint leakage level classification system provided in Embodiment 2 of this invention exists in various forms, including but not limited to:

[0096] (1) Mobile communication devices: These devices are characterized by their mobile communication capabilities and primarily aim to provide voice and data communication. These terminals include smartphones, multimedia phones, feature phones, and low-end phones.

[0097] (2) Ultra-mobile personal computer devices: These devices fall under the category of personal computers, possessing computing and processing capabilities, and generally also have mobile internet access capabilities. These terminals include PDAs, MIDs, and UMPCs, etc.

[0098] (3) Portable entertainment devices: These devices can display and play multimedia content. This category includes audio and video players, handheld game consoles, e-book readers, as well as smart toys and portable car navigation devices.

[0099] (4) Other electronic devices with data interaction functions.

[0100] Specific embodiments of the subject matter have now been described. Other embodiments are within the scope of the appended claims. In some cases, the actions described in the claims can be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing can be advantageous.

[0101] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0102] For ease of description, the above apparatus is described by dividing it into various functional units. Of course, in implementing this invention, the functions of each unit can be implemented in one or more software and / or hardware components. Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take 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 code.

[0103] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0104] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0105] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0106] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0107] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0108] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined in this invention, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0109] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0110] This invention can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules.

[0111] Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific transactions or implement specific abstract data types. This invention can also be practiced in distributed computing environments where transactions are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0113] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for classifying the leakage levels of aviation pipeline joints, characterized in that, include: Multiple sets of pipeline leakage experiments were conducted, and ultrasonic signals of leaked gas under different leakage levels were collected. Specifically, pipeline leakage experiments were conducted using an aircraft pipeline leakage test bench, which includes a pipeline under test with a leak, a gas source, an air filter, a pressure regulating valve, and a pressure gauge. The ultrasonic signal is subjected to time-frequency segmentation and two-dimensional feature fusion; specifically, this includes: truncating the ultrasonic signal according to the sampling frequency to obtain the time domain of the ultrasonic signal; filtering the ultrasonic signal through a bandpass filter and calculating the square envelope spectrum of the filtered ultrasonic signal; the bandpass filter is a bandpass filter based on a fourth-order Butterworth filter; and concatenating the time domain and the square envelope spectrum of the ultrasonic signal in two dimensions to obtain two-dimensional features. A leak rate classification model is constructed; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module; the Transformer feature extraction module adopts a ViT structure, including linear projection layer transformation, position information embedding, multi-head attention mechanism and multilayer perceptron; The leakage rate classification model is trained based on the two-dimensional features; The ultrasonic signals of leaks in aviation pipeline joints are classified using a trained leak rate classification model to determine the leak level.

2. The method for classifying leakage levels of aviation pipeline joints according to claim 1, characterized in that, The multi-sphere support vector data description classification module includes a fully connected layer and multiple hyperspheres, with the number of hyperspheres corresponding to the number of leakage levels.

3. The method for classifying leakage levels of aviation pipeline joints according to claim 2, characterized in that, The multi-sphere support vector data description and classification module uses the k-means algorithm to determine the center position of each hypersphere.

4. The method for classifying leakage levels of aviation pipeline joints according to claim 1, characterized in that, During training, the leakage rate classification model adjusts the hyperparameters of the Transformer feature extraction module and the multi-sphere support vector data description classification module using the cross-entropy loss function.

5. A leakage level classification system for aviation pipeline joints, characterized in that, include: An ultrasonic signal acquisition module is used to conduct multiple sets of pipeline leakage experiments and acquire ultrasonic signals of leaked gas under different leakage levels. Pipeline leakage experiments are conducted using an aircraft pipeline leakage test bench, which includes a pipeline under test with a leak, a gas source, an air filter, a pressure regulating valve, and a pressure gauge. The segmented fusion module is used to perform time-frequency segmented two-dimensional feature fusion on the ultrasonic signal; specifically, it includes: a truncation unit, used to truncate the ultrasonic signal according to the sampling frequency to obtain the time domain of the ultrasonic signal; a square envelope spectrum calculation unit, used to filter the ultrasonic signal through a bandpass filter and calculate the square envelope spectrum of the filtered ultrasonic signal; the bandpass filter is a bandpass filter based on a fourth-order Butterworth filter; and a two-dimensional splicing unit, used to splice the time domain of the ultrasonic signal and the square envelope spectrum in two dimensions to obtain two-dimensional features; The model building module is used to build a leak rate classification model; the leak rate classification model includes a Transformer feature extraction module and a multi-sphere support vector data description classification module; the Transformer feature extraction module adopts a ViT structure, including linear projection layer transformation, position information embedding, multi-head attention mechanism and multilayer perceptron; The training module is used to train the leakage rate classification model based on the two-dimensional features; The grading module is used to classify the ultrasonic signals of leaks in aviation pipeline joints using a trained leak rate grading model, and to determine the leak level.

6. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the aviation pipeline joint leakage level classification method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the aviation pipeline joint leakage level classification method as described in any one of claims 1-4.

Images