FBG (fiber bragg grating) sensor fault positioning method and system

Abstract

The invention relates to the technical field of fault positioning, and particularly discloses a fiber grating sensor fault positioning method and system, and the method comprises the steps: firstly obtaining first signal data of a fiber grating sensor, and carrying out the empirical mode decomposition of the first signal data, further introducing a signal processing technology based on deep learning to carry out feature extraction on each sensor signal component, and carrying out feature selection based on feature distribution guidance of each sensor signal component so as to filter noise and redundant information and realize high-efficiency expression of signal key features; and on the basis, intelligent identification of the fault type of the fiber grating sensor is carried out. In this way, noise interference can be effectively avoided, and therefore the accuracy and stability of fault positioning of the fiber grating sensor are improved.

Description

Technical Field

[0001] This application relates to the field of fault location technology, and more specifically, to a method and system for fault location of fiber Bragg grating sensors. Background Art

[0002] Fiber Bragg grating (FBG) sensors, as advanced optical measurement devices, are widely used in structural health monitoring, communication networks, and oil and gas pipeline monitoring due to their high precision, long-distance monitoring capabilities, and good adaptability to harsh environments. However, because they operate in complex and changing environments for extended periods, FBG sensors are susceptible to various factors such as temperature fluctuations, mechanical vibrations, and electromagnetic interference, leading to performance degradation or malfunctions. Therefore, fault location is a crucial step in ensuring the stable operation and timely maintenance of FBG sensors.

[0003] In the prior art, the invention patent with publication number CN118035855A discloses a method for fault location of fiber optic grating sensors. It processes signal data through a preset signal decomposition algorithm to obtain sensor data components, and then encodes fault categories based on the signal characteristics of each sensor data component to construct a feature sample set. This set is used to train a preset machine learning model, and then the sensor data to be analyzed is analyzed according to the trained target fault location model to obtain the fault location result.

[0004] While the methods described above improve the automation and accuracy of fiber Bragg grating sensor fault identification to some extent, some challenges remain in practical applications. For example, signal decomposition algorithms decompose noise and interference components in the sensor signal, resulting in sensor data components containing a lot of redundant information. Existing technologies mainly rely on statistics or simple mathematical transformations for feature extraction, which makes it difficult to capture the deep correlations and subtle differences between different signal components. This may lead to the omission of key features or interference from noise and redundant information, introducing unnecessary errors and reducing the accuracy and stability of fault location.

[0005] Therefore, an optimized method and system for fault location of fiber Bragg grating sensors is needed. Summary of the Invention

[0006] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a method and system for fault location of a fiber Bragg grating sensor. First, it acquires first signal data from the fiber Bragg grating sensor. After performing empirical mode decomposition on the first signal data, it further introduces deep learning-based signal processing techniques to extract features from each sensor signal component. Based on the feature distribution of each sensor signal component, feature selection is performed to filter out noise and redundant information, achieving efficient expression of key signal features. On this basis, intelligent identification of the fiber Bragg grating sensor fault type is performed. This approach effectively avoids noise interference, thereby improving the accuracy and stability of fiber Bragg grating sensor fault location.

[0007] According to one aspect of this application, a method for fault location of a fiber Bragg grating sensor is provided, comprising: Acquire the first signal data from the fiber Bragg grating sensor; The first signal data is decomposed using an empirical mode decomposition algorithm to obtain multiple sensor data components. Feature extraction is performed on each of the multiple sensor data components to obtain a set of sensor signal component encoding feature vectors; The set of sensor signal component coding feature vectors is subjected to feature distribution-guided feature selection to obtain a sparse set of sensor signal component coding feature vectors; The fault type of the fiber Bragg grating sensor is determined based on the set of sparse sensor signal component encoded feature vectors.

[0008] According to another aspect of this application, a fiber Bragg grating sensor fault location system is provided, comprising: The first signal data acquisition module is used to acquire the first signal data of the fiber optic grating sensor. The signal data decomposition module is used to perform signal decomposition on the first signal data based on the empirical mode decomposition algorithm to obtain multiple sensor data components. The feature extraction module is used to extract features from each of the multiple sensor data components to obtain a set of sensor signal component encoded feature vectors. The feature selection module is used to perform feature distribution-guided feature selection on the set of sensor signal component encoded feature vectors to obtain a sparse set of sensor signal component encoded feature vectors. The fault type identification module is used to determine the fault type of the fiber Bragg grating sensor based on the set of sparse sensor signal component encoded feature vectors.

[0009] Compared with existing technologies, the fiber Bragg grating sensor fault location method and system provided in this application first acquires the first signal data of the fiber Bragg grating sensor. After performing empirical mode decomposition on the first signal data, it further introduces deep learning-based signal processing technology to extract features from each sensor signal component. Based on the feature distribution of each sensor signal component, feature selection is performed to filter out noise and redundant information, achieving efficient expression of key signal features. On this basis, intelligent identification of the fiber Bragg grating sensor fault type is then performed. This approach effectively avoids noise interference, thereby improving the accuracy and stability of fiber Bragg grating sensor fault location. Attached Figure Description

[0010] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 This is a flowchart of a fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0012] Figure 2 This is a schematic diagram of the data flow in the fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0013] Figure 3 This is a flowchart of sub-step S4 of the fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0014] Figure 4 This is a flowchart of sub-step S41 of the fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0015] Figure 5 This is a flowchart of sub-step S42 of the fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0016] Figure 6 This is a flowchart of sub-step S5 of the fiber Bragg grating sensor fault location method according to an embodiment of this application.

[0017] Figure 7 This is a block diagram of a fiber Bragg grating sensor fault location system according to an embodiment of this application. Detailed Implementation

[0018] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0019] While this application makes various references to certain modules of the systems according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0020] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0021] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0022] It is worth noting that all data acquisition actions in this application were carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization granted by the owner of the relevant device.

[0023] As mentioned in the background section above, patent CN118035855A proposes a method for fault location of fiber optic grating sensors. This method processes signal data using a preset signal decomposition algorithm to obtain sensor data components. Then, based on the signal characteristics of each sensor data component, it encodes the fault category and constructs a feature sample set. This set is used to train a preset machine learning model. Finally, based on the trained target fault location model, the method analyzes the sensor data to be analyzed and obtains the fault location result.

[0024] While the aforementioned methods have improved the automation and accuracy of fiber Bragg grating sensor fault identification to some extent, several challenges remain in practical applications. For example, signal decomposition techniques, while analyzing sensor data, also break down accompanying noise and interference factors, resulting in a significant amount of redundant information in the final data components. Existing technologies primarily rely on statistical analysis or basic mathematical transformations to extract features, making it difficult to capture the complex relationships and subtle differences between various signal components. This may lead to overlooking key features or being affected by noise and redundant data, causing additional misjudgments and reducing the accuracy and reliability of fault location.

[0025] To address the aforementioned technical problems, this application proposes an optimized method for fault location of fiber Bragg grating (FBG) sensors. First, it acquires the first signal data from the FBG sensor. After performing empirical mode decomposition (EMD) on the first signal data, it further introduces deep learning-based signal processing techniques to extract features from each sensor signal component. Based on the feature distribution of each sensor signal component, feature selection is performed to filter out noise and redundant information, achieving efficient expression of key signal features. On this basis, intelligent identification of the FBG sensor fault type is then performed. This approach effectively avoids noise interference, thereby improving the accuracy and stability of FBG sensor fault location.

[0026] Figure 1 This is a flowchart of a fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 2 This is a schematic diagram of the data flow in a fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 1 and Figure 2 As shown, the fiber Bragg grating sensor fault location method includes the following steps: S1, acquiring first signal data from the fiber Bragg grating sensor; S2, performing signal decomposition on the first signal data based on an empirical mode decomposition algorithm to obtain multiple sensor data components; S3, extracting features from each of the multiple sensor data components to obtain a set of sensor signal component encoded feature vectors; S4, performing feature selection based on feature distribution guidance on the set of sensor signal component encoded feature vectors to obtain a sparse set of sensor signal component encoded feature vectors; S5, determining the fault type of the fiber Bragg grating sensor based on the set of sparse sensor signal component encoded feature vectors.

[0027] In the aforementioned fiber Bragg grating sensor fault location method, step S1 involves acquiring the first signal data of the fiber Bragg grating sensor. It should be understood that the fiber Bragg grating sensor operates based on the grating principle. When the external physical quantity being measured changes, it causes a change in the propagation path or phase of light within the fiber, thereby altering the interference mode. The fiber Bragg grating sensor analyzes the interference light, converting the physical quantity information into changes in parameters such as light signal intensity and wavelength, thus achieving the measurement of the physical quantity being measured. In practical applications, fiber Bragg grating sensors are affected by various factors, such as environmental changes, structural deformation, and equipment aging, which may lead to problems such as chirping, fiber optic head damage, base detachment, and cable breakage, resulting in changes in the measurement data. Therefore, this application aims to capture abnormal changes in the measured physical quantity by analyzing the first signal data acquired by the fiber Bragg grating sensor, thereby achieving accurate identification of the fault type of the fiber Bragg grating sensor.

[0028] Specifically, in the hardware preparation phase, it is crucial to ensure that the fiber Bragg grating sensor is correctly installed on the object to be monitored. The installation method will vary depending on the application scenario. For example, in structural health monitoring, the sensor may need to be embedded inside a bridge or building; while in pipeline monitoring, it is laid along the pipe. In either case, care must be taken to avoid applying external forces to the sensor, such as pulling or squeezing, to prevent affecting its sensitivity and accuracy. Furthermore, the fiber Bragg grating sensor needs to be connected to reading devices, such as wavelength demodulators or spectrometers, which can analyze the light signal reflected from the fiber Bragg grating and convert it into data that can be used for subsequent processing. Before initial use, the entire system must be calibrated to ensure data accuracy. The calibration process typically involves comparison and adjustment using a standard light source and a reference sample with known reflectivity to ensure the system's measurement accuracy.

[0029] After completing the hardware preparation, the next step is to focus on setting up the measurement environment. To ensure data consistency and reliability, the impact of external factors on the measurement results should be minimized. This includes maintaining stable temperature and humidity conditions, as temperature changes can cause wavelength drift in the fiber Bragg grating sensor, thus affecting the measurement results. Simultaneously, strong sources of electromagnetic interference should be avoided, and mechanical vibrations should be prevented from being transmitted to the sensor, as these factors can introduce additional noise and degrade data quality. Furthermore, appropriate safety precautions must be taken when working in certain special environments. For example, when working near high-voltage electric fields, relevant safety procedures should be followed; in flammable and explosive environments, explosion-proof requirements must be observed to ensure the safety of personnel and equipment.

[0030] When setting data acquisition parameters, it's crucial to select an appropriate sampling frequency based on the characteristics of the phenomenon to be monitored. For rapidly changing processes, such as vibration monitoring, a higher sampling rate is necessary to capture transient events; while for slower changes, such as long-term structural health monitoring, a lower sampling rate can be used to save storage space and reduce computational load. Besides the sampling frequency, defining the duration of each measurement (i.e., the time window) is also important. Longer time windows help capture periodic or trending changes but increase the data volume; shorter time windows are more suitable for real-time monitoring of transient events. Setting trigger conditions is equally important. For example, sometimes it may be desirable to start recording data only under specific conditions, requiring the setting of trigger conditions. Trigger conditions can be based on time (timed triggering), external signals (such as switch states), or other physical quantities (such as stress or strain reaching a certain threshold) to improve the effectiveness and relevance of the data.

[0031] Finally, after everything is ready, the reading device can be started and the predetermined parameter configuration file loaded, all components initialized, and the system status indicators checked to confirm normal operation. Then, the measurement process is initiated by pressing the "Start" button or sending a command through the software interface. At this time, the reading device will collect reflectance spectral information from the fiber Bragg grating sensor according to the predetermined parameters and convert it into a digital signal. Since the raw acquired data usually contains some noise and other unwanted components, preliminary data cleaning is often required before saving or further processing. This may involve removing outliers, filtering, or applying other data preprocessing techniques to improve the quality of subsequent analysis. In addition, considering the uncertainties and environmental changes in practical applications, adaptive data acquisition strategies can be designed to enable the system to automatically adjust parameters to a certain extent to meet the measurement needs under different conditions.

[0032] In the aforementioned fiber Bragg grating sensor fault location method, step S2 involves performing signal decomposition on the first signal data using an empirical mode decomposition algorithm to obtain multiple sensor data components. Specifically, this application considers that the signal data of fiber Bragg grating sensors is typically nonlinear and non-stationary, containing multiple frequency components, noise interference, and abnormal fluctuations caused by sensor faults. Therefore, to effectively extract key features from the sensor signal, this application employs an empirical mode decomposition algorithm to decompose the first signal data, adaptively decomposing the complex signal into multiple intrinsic mode function (IMF) components and residual components, thereby obtaining multiple sensor data components. It should be understood that the empirical mode decomposition algorithm is based on the local characteristic time scale of the signal. By continuously searching for local maxima and minima of the signal, it constructs upper and lower envelopes, calculates their mean curves, and then subtracts the mean curve from the original signal to obtain a new signal. This process is repeated until a specific stopping criterion is met, thereby decomposing the signal into a series of IMF components of different frequencies. Each IMF component represents the oscillation mode of the signal in different frequency bands. This decomposition process can separate the first signal data from the complex mixed state, allowing subsequent feature extraction to focus on each component with independent physical meaning, more effectively mining fault feature information, avoiding mutual influence between different frequency components and interference factors, and improving the accuracy of fault diagnosis.

[0033] Specifically, for fiber optic grating (FBG) sensors, their reflection spectrum varies with changes in external physical quantities such as temperature and stress. However, real-world monitoring environments often contain various interference factors, making the acquired first signal data complex and non-stationary. Therefore, the EMD algorithm has become one of the effective tools for extracting useful information from such complex signals. FBG sensors are widely used in structural health monitoring, communication networks, and other fields due to their high sensitivity and strong resistance to electromagnetic interference. However, due to long-term exposure to complex natural environments, the signals received by FBG sensors are inevitably affected by various noises and interferences, such as environmental noise (including external factors such as mechanical vibration and wind disturbance), temperature fluctuations (temperature changes cause the fiber itself to expand and contract, thus affecting the position of the FBG's reflection wavelength), electromagnetic interference (although the fiber itself has good resistance to electromagnetic interference, interference may still occur in certain special situations), and multi-physics coupling effects (when multiple physical quantities act on the same FBG simultaneously, the physical quantities may interact, increasing the difficulty of signal analysis). To overcome these challenges, advanced signal processing techniques are needed to improve data quality and analytical accuracy. Among them, the EMD algorithm has become an ideal choice because it can automatically adapt to the characteristics of the input signal and effectively separate different frequency components.

[0034] Before performing EMD decomposition, the original signal typically requires preprocessing to reduce interference from irrelevant information, such as removing DC bias and smoothing filters. For fiber Bragg grating sensors, this means ensuring the acquired initial signal data is as pure as possible so that subsequent steps can more accurately capture useful features. Next, we move to the core of EMD—constructing the IMF sequence. First, all the locations of the maximum and minimum values ​​of the signal within a given time period are identified. Then, a smooth upper envelope and a lower envelope are fitted based on these extreme points. The average value between the upper and lower envelopes is calculated and subtracted from the original signal to obtain a new residual signal. Next, it is checked whether the newly obtained residual signal meets the definition of an IMF. If it does, it is output as an IMF; otherwise, the above steps are repeated until the condition is met. Finally, the current IMF is subtracted from the original signal to obtain a new remainder, which is then used as the object for the next decomposition step. This process is repeated until the remainder no longer contains any effective IMF components, i.e., it becomes a trend term or approaches a constant.

[0035] As EMD decomposition deepens, the original complex signal is peeled away layer by layer, forming multiple distinct IMF components. Each IMF represents the oscillating behavior of the original signal at a specific time scale. These IMFs not only retain the main characteristics of the original signal but also, because each IMF corresponds to a different frequency band, can more clearly reveal the information hidden within the signal. This is particularly beneficial for devices like fiber Bragg grating sensors, which are extremely sensitive to environmental changes. The IMFs obtained through EMD decomposition can help better understand the specific contributions of different physical quantities to the sensor response, thus providing a solid foundation for further feature extraction and fault diagnosis.

[0036] In the aforementioned fiber Bragg grating sensor fault location method, step S3 involves extracting features from each of the multiple sensor data components to obtain a set of sensor signal component encoded feature vectors. In a specific example of this application, step S3 includes: using a one-dimensional convolutional layer-based signal feature extractor to extract features from each of the multiple sensor data components to obtain a set of sensor signal component encoded feature vectors. Here, considering that each sensor data component represents an oscillation mode of different frequency bands decomposed from the original signal, reflecting different fault characteristic information, this application introduces deep learning-based signal processing technology to accurately capture and efficiently express the features of each sensor data component. It utilizes a one-dimensional convolutional layer-based signal feature extractor to perform feature learning on each sensor data component, automatically extracting deep features of each sensor data component to obtain a set of sensor signal component encoded feature vectors. It should be understood that a one-dimensional convolutional layer has powerful local feature extraction capabilities, extracting feature patterns in local regions, such as local frequency changes and amplitude abrupt changes, by sliding the convolution kernel along the time dimension of the sensor data components. As convolutional layers are stacked and parameters are learned and updated, the signal feature extractor can gradually extract higher-level feature representations, transforming the original sensor data components into more abstract and representative feature vectors, thereby providing more discriminative feature representations for subsequent fault diagnosis and improving the accuracy of fault location.

[0037] In the aforementioned fiber Bragg grating sensor fault location method, step S4 involves performing feature selection based on feature distribution guidance on the set of sensor signal component encoded feature vectors to obtain a sparser set of sensor signal component encoded feature vectors. It should be understood that since the sensor data components obtained after signal decomposition are numerous and contain a certain degree of redundancy and correlation, directly using them for fault identification may increase computational complexity and introduce unnecessary noise. Therefore, to improve the robustness and computational efficiency of feature representation, this application further performs feature selection on the set of sensor signal component encoded feature vectors. Based on the feature correlation distribution between the various sensor signal component encoded feature vectors, the set of sensor signal component encoded feature vectors is filtered, retaining feature representations that contribute significantly to fault identification and have low redundancy, while removing redundant and noisy features. This achieves dimensionality reduction and sparsity of the feature space, reducing computational burden and improving the accuracy and speed of fault identification. Figure 3 This is a flowchart of sub-step S4 of the fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 3As shown, step S4 includes the following steps: S41, based on the feature distribution hub of the set of sensor signal component coded feature vectors, calculating the fine-grained ablation factor of each sensor signal component coded feature vector in the set of sensor signal component coded feature vectors to obtain a set of fine-grained ablation factors for sensor signal components; S42, based on the set of fine-grained ablation factors for sensor signal components, performing feature sparsification processing on the set of sensor signal component coded feature vectors to obtain the sparsified set of sensor signal component coded feature vectors.

[0038] Figure 4 This is a flowchart of sub-step S41 of the fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 4 As shown, step S41 includes the following steps: S411, performing a feature distribution hub search on the set of sensor signal component encoded feature vectors to obtain sensor signal component feature distribution hub encoded vectors; S412, calculating the fine-grained ablation factor of each sensor signal component encoded feature vector in the set of sensor signal component encoded feature vectors relative to the sensor signal component feature distribution hub encoded vector to obtain the set of sensor signal component fine-grained ablation factors.

[0039] More specifically, in a specific example of this application, step S411 includes: first, calculating the feature difference factor of each sensor signal component encoded feature vector in the set of sensor signal component encoded feature vectors relative to the set of sensor signal component encoded feature vectors to obtain a set of sensor signal component feature difference factors, expressed by the formula:

[0040]

[0041] in, This represents the set of coded feature vectors of the sensor signal components. , , , and These respectively represent the first, second, and third eigenvectors in the set of encoded feature vectors of the sensor signal components. , The and the first Each sensor signal component is encoded as a feature vector. The number of feature vectors encoded for the sensor signal components. Represents the transpose of a vector. For the and stated The covariance matrix between them For the The sensor signal component characteristic difference factor.

[0042] In other words, by performing a feature distribution hub search on the set of encoded feature vectors of the sensor signal components, key feature points located at the feature distribution center or with significant distinguishability are identified to characterize the main feature distribution trend of the first signal data. Specifically, during the search process, the average distance of each sensor signal component's encoded feature vector relative to the overall features of the set is first calculated to understand the cooperation patterns and interaction effects between different component features, thereby revealing the connections and hierarchical structure between the various sensor data components and helping to initially assess the importance of each sensor data component feature. In this way, the essential characteristics of the original signal can be reflected more accurately, providing a solid foundation for subsequent analysis, effectively selecting the most representative features, reducing redundant information, and improving the efficiency and accuracy of the fault location model.

[0043] Then, after taking the negative of each sensor signal component feature difference factor in the set of sensor signal component feature difference factors, normalization based on the Softmax function is performed to obtain the set of sensor signal component hub feature correlation factors; finally, using the set of sensor signal component hub feature correlation factors as a weight distribution, the set of sensor signal component encoded feature vectors is weighted and aggregated to obtain the sensor signal component feature distribution hub encoding vector, expressed by the formula:

[0044] in, It represents exponential operations with the natural constant as the base. This represents the hub encoding vector representing the characteristic distribution of sensor signal components.

[0045] In other words, firstly, normalization is used to determine the weight distribution of the features of each sensor data component. Then, a weighted aggregation method is used to construct the feature distribution hub of the set of encoded feature vectors of the sensor signal components, i.e., the sensor signal component feature distribution hub encoding vector. This serves as a holistic description and general expression of the core feature distribution of the first signal data, providing a key reference standard for subsequent feature selection and sparsity processing. This approach helps to extract the most representative information, thus laying the foundation for subsequent data analysis steps.

[0046] More specifically, step S412 is expressed by the formula:

[0047] in, The norm of a vector This indicates the operation of retrieving the maximum value. It is a very small positive number, used to prevent the denominator from being zero. Indicates the Fine-grained ablation factor for sensor signal components.

[0048] In other words, using the hub encoding vector of the sensor signal component feature distribution as a benchmark, a fine-grained ablation factor is calculated between the encoded feature vectors of each sensor signal component and the hub encoding vector of the sensor signal component feature distribution using contrastive learning techniques. This assesses the impact of removing features from each sensor signal component on the overall expression of the hub feature. In this way, the semantic correlation strength between each sensor data component and the core features of the first signal data can be revealed, thereby quantifying the importance of the encoded feature vectors of each sensor signal component.

[0049] Figure 5 This is a flowchart of sub-step S42 of the fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 5 As shown, step S42 includes the following steps: S421, inputting each fine-grained ablation factor of the sensor signal component in the set of fine-grained ablation factors of the sensor signal component into a sparsification module based on a gate function to obtain a set of sparsified fine-grained ablation factors of the sensor signal component; S422, using the set of sparsified fine-grained ablation factors of the sensor signal component as a weight distribution, performing weighted modulation on the set of encoded feature vectors of the sensor signal component to obtain the set of sparsified encoded feature vectors of the sensor signal component.

[0050] More specifically, step S421 is expressed by the formula:

[0051] in, Represents the gate function. Indicates the gate threshold. Indicates the Fine-grained ablation factor for sparsified sensor signal components.

[0052] More specifically, step S422 is expressed by the formula:

[0053] in, Indicates the The corresponding sparse sensor signal component encoded feature vector, This represents the set of coded feature vectors of the sparsed sensor signal components.

[0054] In other words, by further utilizing a gating mechanism, each fine-grained ablation factor is screened, and the sensor signal component feature vectors corresponding to lower-fine-grained ablation factors are treated as redundant or noise features and filtered out. Simultaneously, sensor signal component features that significantly influence the overall expression of the core features of the first signal data are retained. This achieves effective sparsification of the set of sensor signal component encoded feature vectors. In this way, the subsequent fault identification process can focus more on key features, thereby reducing the risk of false alarms and false negatives, and enhancing the reliability and stability of the diagnosis. Furthermore, sparsification of the feature space can reduce computational resource consumption and improve the algorithm's execution efficiency.

[0055] In the above-described fiber Bragg grating sensor fault location method, step S5 determines the fault type of the fiber Bragg grating sensor based on the set of sparsed sensor signal component encoded feature vectors. Wherein, Figure 6 This is a flowchart of sub-step S5 of the fiber Bragg grating sensor fault location method according to an embodiment of this application. Figure 6 As shown, step S5 includes the following steps: S51, performing feature concatenation on the set of sparse sensor signal component encoding feature vectors to obtain a global feature encoding vector for the sensor signal; S52, inputting the global feature encoding vector for the sensor signal into the trained target fault location model to obtain a fault location result, wherein the fault location result is a fault type label for the fiber Bragg grating sensor.

[0056] Specifically, in step S51, the set of sparse sensor signal component encoded feature vectors is concatenated to obtain a global feature encoded vector for the sensor signal. It should be understood that although the sensor signal component encoded feature vectors after feature selection each carry signal features from different frequency bands in the first signal data, they lack overall coherence. Therefore, to achieve a global feature description of the first signal data, this application further concatenates the set of sparse sensor signal component encoded feature vectors to integrate the scattered local feature representations into a global feature encoded vector for the sensor signal, comprehensively reflecting the overall characteristics and fault modes of the first signal data.

[0057] Specifically, in step S52, the global feature encoding vector of the sensor signal is input into the trained target fault location model to obtain a fault location result, which is the fault type label of the fiber Bragg grating sensor. In a specific example of this application, the target fault location model is constructed based on a support vector machine (SVM). During the training phase, the target fault location model learns from a large amount of known fault types and corresponding sensor feature data, continuously adjusting its parameters and structure to optimize its classification ability for different fault types. During the fault location process, the model matches and compares the input global feature encoding vector of the sensor signal with various fault modes learned during training, and determines the fault type label that best matches the input features according to the maximum margin classification principle in SVM, thereby achieving fault location of the fiber Bragg grating sensor.

[0058] Preferably, when each sensor signal component encoding feature vector in the set of sensor signal component encoding feature vectors represents a sensor signal component feature, the process of performing feature sparsification based on feature distribution hubs may result in the accidental deletion of effective information. This leads to insufficient long-distance effective information encoding representation of the global feature encoding vector of the sensor signal, thereby reducing the expressive effect of the global feature encoding vector of the sensor signal relative to the trained target fault location model and affecting the accuracy of the fault location result obtained by inputting it into the trained target fault location model.

[0059] Therefore, in one example, when the global feature encoding vector of the sensor signal is input into the trained target fault location model, the global feature encoding vector of the sensor signal is optimized. The optimization includes the following steps: The implicit parameter topology framework is derived from the target fault location model; The global feature encoding vector of the sensor signal is embedded into the differential manifold of the implicit parameter topology framework to generate the sensor signal mimicry criterion mapping encoding;

[0060] in, This represents the global feature encoding vector of the sensor signal. Represents an implicitly parameterized topology framework. This represents the mapping encoding of sensor signal mimicry criteria; A nonlinear gain operation is applied to the sensor signal mimicry criterion mapping code to output the sensor signal scaling convergence mapping code;

[0061] in, This represents the sensor signal scaling convergence mapping code; A multi-scale convergent coding tensor for sensor signals is generated based on the sensor signal scaling convergent mapping code and the sensor signal mimicry criterion mapping code.

[0062] in, This indicates the transpose operation. This represents vector multiplication. This indicates the length of the sensor signal scaling convergence mapping code. This represents the multi-scale convergent coding tensor of the sensor signal; Entropy regularization is performed on the multi-scale convergent coding tensor of the sensor signal to generate a steady-state retrieval architecture for the sensor signal;

[0063] in, This represents the steady-state retrieval architecture for sensor signals. The global feature encoding vector of the sensor signal is reparameterized to the Riemann space of the steady-state retrieval architecture of the sensor signal to obtain the incremental representation of the sensor signal criteria;

[0064] in, This represents the incremental characterization of sensor signal criteria; By coupling the global feature encoding vector of the sensor signal with the incremental representation of the sensor signal criterion, the adjusted global feature encoding vector of the sensor signal is output.

[0065] in, and Indicates the weighted hyperparameter. This represents the adjusted global feature encoding vector of the sensor signal. Finally, the adjusted global feature encoding vector of the sensor signal is input into the trained target fault location model to obtain the fault location result.

[0066] Here, relying on the set of orthogonal quantities of the separating hyperplane anchored by the implicit cognitive graph primitives of the model, the original feature vectors are driven to evolve into a higher-order separability space, and the correlation dimension with the category criterion is significantly enhanced through a nonlinear gain mapper. Subsequently, a multi-scale convergent retrieval field for classification features is constructed as an information distillation field to capture incremental representations that contribute to criterion decision-making, and a parameter sensitivity adjustment mechanism is used to fuse these incremental criterion representations. The optimized classification features are better suited to the elastic scaling convergence of the implicit parameter topology of the target fault location model, improving the accuracy of the fault location results obtained by inputting them into the trained target fault location model.

[0067] After initially identifying the fault type of the fiber Bragg grating sensor, a more in-depth fault analysis is required. This step includes a physical inspection of the sensor and its related components to identify any visible damage or anomalies. For example, checking for obvious mechanical damage to the sensor or broken connecting cables. Simultaneously, by combining historical and currently acquired data, an attempt is made to pinpoint the root cause of the fault. By comparing the data characteristics during normal operation with those under fault conditions, a better understanding of the specific manifestations and scope of the fault can be achieved. Furthermore, simulation tools can be used to model sensor behavior under different conditions, helping engineers to more accurately locate the problem.

[0068] After identifying the type of fault, the next step is to assess its impact on the overall system performance. In critical applications such as aerospace and oil and gas pipeline monitoring, even minor faults can have serious consequences. Therefore, it is essential to quantify the negative impact of the fault on monitoring accuracy, response time, and other aspects to determine whether immediate action is necessary. Furthermore, it is crucial to consider whether the fault will trigger other potential problems, thus informing subsequent decision-making. For example, if a sensor exhibits decreased temperature sensitivity, it could affect the accuracy of data relying on that sensor throughout the monitoring network, thereby impacting all decisions made based on that data. Therefore, a comprehensive assessment of the fault's impact is necessary to determine the most appropriate response.

[0069] Based on the above analysis, the next step is to develop a specific repair plan. This process typically includes selecting appropriate repair methods (such as hardware replacement, software updates, parameter adjustments, etc.) and planning an implementation schedule. If the failure is caused by external factors, such as performance degradation due to environmental changes, then in addition to repairing the sensor itself, measures need to be taken to improve the installation environment to prevent similar problems from recurring. For internal failures, such as aging or failure of sensor components, it may be necessary to replace the entire sensor unit or specific components. At this stage, cost-effectiveness should also be considered to select the optimal solution that solves the problem without causing unnecessary waste of resources.

[0070] Once the repair plan is finalized, the execution phase begins. This phase includes both actual repair operations and related coordination. For on-site work, ensuring sufficient technical support personnel are available is crucial, as they can quickly and effectively handle any unexpected situations. Simultaneously, to avoid disrupting normal business operations, repairs are often scheduled during off-peak hours, minimizing downtime. Furthermore, if collaboration between multiple departments is involved, advance communication and planning are essential to ensure all parties clearly understand their tasks and timelines.

[0071] Finally, after all repair work is completed, a comprehensive performance verification must be performed to ensure that the system has returned to its expected operating state. This step can be done by recalibrating sensors, collecting new data, and comparing it with previous benchmarks. Performance verification is not only a test of the repair work but also a guarantee of future system stability. By setting a series of rigorous test standards, any residual problems or new hidden dangers can be effectively detected. Only when all indicators meet the requirements can the fault be officially declared completely resolved, and the system can be put back into normal use.

[0072] In summary, the fiber Bragg grating sensor fault location method based on the embodiments of this application is explained. It first acquires the first signal data of the fiber Bragg grating sensor, performs empirical mode decomposition on the first signal data, and then introduces deep learning-based signal processing technology to extract features from each sensor signal component. Based on the feature distribution of each sensor signal component, feature selection is performed to filter out noise and redundant information, achieving efficient expression of key signal features. On this basis, intelligent identification of the fiber Bragg grating sensor fault type is then performed. This approach effectively avoids noise interference, thereby improving the accuracy and stability of fiber Bragg grating sensor fault location.

[0073] Furthermore, a fiber optic grating sensor fault location system is also provided.

[0074] Figure 7 This is a block diagram of a fiber Bragg grating sensor fault location system according to an embodiment of this application. Figure 7 As shown, the fiber Bragg grating sensor fault location system 100 according to an embodiment of this application includes: a first signal data acquisition module 110, used to acquire first signal data of the fiber Bragg grating sensor; a signal data decomposition module 120, used to perform signal decomposition on the first signal data based on an empirical mode decomposition algorithm to obtain multiple sensor data components; a feature extraction module 130, used to extract features from each of the multiple sensor data components to obtain a set of sensor signal component encoded feature vectors; a feature selection module 140, used to perform feature selection based on feature distribution guidance on the set of sensor signal component encoded feature vectors to obtain a sparse set of sensor signal component encoded feature vectors; and a fault type identification module 150, used to determine the fault type of the fiber Bragg grating sensor based on the sparse set of sensor signal component encoded feature vectors.

[0075] Here, those skilled in the art will understand that the specific operation of each module in the above-described fiber Bragg grating sensor fault location system has been referenced above. Figures 1 to 6The method for fault location of fiber Bragg grating sensors has been described in detail, and therefore, its repeated description will be omitted.

[0076] The basic principles of the present invention have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in the present invention are merely examples and not limitations, and should not be considered as essential features of each embodiment of the present invention. Furthermore, the specific details of the above embodiments are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the present invention to the necessity of employing the specific details described above.

[0077] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. In the several embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for example, the unit division is only a logical functional division, and other division methods may exist in actual implementation. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the present invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0079] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in a system claim may also be implemented by a single unit through software or hardware.

[0080] Finally, it should be noted that the above description has been given for illustrative and descriptive purposes. Furthermore, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention.

Claims

1. A method for locating a fiber grating sensor fault, characterized in that: include: Acquiring first signal data of the fiber grating sensor; Performing signal decomposition based on an empirical mode decomposition algorithm on the first signal data to obtain a plurality of sensor data components; Performing feature extraction on each of the plurality of sensor data components to obtain a set of sensor signal component encoding feature vectors; Performing feature selection based on feature distribution guidance on the set of sensor signal component encoding feature vectors to obtain a sparse set of sensor signal component encoding feature vectors; The fault type of the fiber grating sensor is determined based on the set of sparse sensor signal component encoding feature vectors.

2. The fiber grating sensor fault location method according to claim 1, characterized in that: Extracting features from each of the plurality of sensor data components to obtain a set of sensor signal component encoding feature vectors includes: A signal feature extractor based on a one-dimensional convolutional layer is used to perform feature extraction on each of the multiple sensor data components to obtain a set of encoded feature vectors of the sensor signal components.

3. The fiber grating sensor fault location method according to claim 2, characterized in that: Performing feature selection on the set of sensor signal component encoding feature vectors to obtain a sparse set of sensor signal component encoding feature vectors, comprising: Based on the feature distribution hub of the set of sensor signal component encoding feature vectors, calculating the fine-grained ablation factor of each sensor signal component encoding feature vector in the set of sensor signal component encoding feature vectors to obtain a set of sensor signal component fine-grained ablation factors; Based on the set of fine-grained ablation factors of the sensor signal components, feature thinning processing is performed on the set of encoded feature vectors of the sensor signal components to obtain the set of thinned encoded feature vectors of the sensor signal components.

4. The fiber Bragg grating sensor fault location method according to claim 3, characterized in that: Based on the feature distribution hub of the set of sensor signal component encoding feature vectors, calculating the fine-grained ablation factor of each sensor signal component encoding feature vector in the set of sensor signal component encoding feature vectors to obtain a set of sensor signal component fine-grained ablation factors, including: Performing a feature distribution hub search on the set of sensor signal component encoding feature vectors to obtain a sensor signal component feature distribution hub encoding vector; The fine-grained ablation factors of each sensor signal component encoding feature vector in the set of sensor signal component encoding feature vectors relative to the sensor signal component feature distribution hub encoding vector are respectively calculated to obtain the set of sensor signal component fine-grained ablation factors.

5. The fiber grating sensor fault location method according to claim 4, characterized in that: Performing a feature distribution hub search on the set of sensor signal component encoding feature vectors to obtain a sensor signal component feature distribution hub encoding vector, comprising: Calculating a characteristic difference factor of each sensor signal component encoding feature vector in the set of sensor signal component encoding feature vectors relative to the set of sensor signal component encoding feature vectors to obtain a set of sensor signal component characteristic difference factors; Taking the inverse of each sensor signal component feature difference factor in the set of sensor signal component feature difference factors and performing normalization processing based on a Softmax function to obtain a set of sensor signal component hub feature correlation factors; The set of sensor signal component hub feature correlation factors is used as a weight distribution, and the set of sensor signal component encoding feature vectors is weightedly aggregated to obtain the sensor signal component feature distribution hub encoding vector.

6. The fiber grating sensor fault location method according to claim 5, characterized in that: Based on the set of fine-grained ablation factors of the sensor signal components, performing feature sparse processing on the set of encoded feature vectors of the sensor signal components to obtain the set of sparsely encoded feature vectors of the sensor signal components, comprising: Inputting each sensor signal component fine-grained ablation factor in the set of sensor signal component fine-grained ablation factors into a sparse module based on a gating function to obtain a set of sparse sensor signal component fine-grained ablation factors; The set of the sparse sensor signal component fine-grained ablation factors is used as a weight distribution, and the set of the sensor signal component encoding feature vectors is weighted modulated to obtain the set of the sparse sensor signal component encoding feature vectors.

7. The fiber grating sensor fault location method according to claim 6, characterized in that: Determining the fault type of the fiber grating sensor based on the set of sparsely encoded feature vectors of the sensor signal components comprises: Performing feature concatenation on the set of the sparsely encoded feature vectors of the sensor signal components to obtain a sensor signal global feature encoding vector; The sensor signal global feature encoding vector is input into the trained target fault location model to obtain a fault location result, and the fault location result is a fault type label of the fiber grating sensor.

8. A fiber grating sensor fault location system, characterized in that: include: A first signal data acquisition module, used to acquire first signal data of the fiber grating sensor; A signal data decomposition module, configured to perform signal decomposition based on an empirical mode decomposition algorithm on the first signal data to obtain a plurality of sensor data components; A feature extraction module, configured to extract features from each of the plurality of sensor data components to obtain a set of sensor signal component encoding feature vectors; A feature selection module, configured to perform feature selection based on feature distribution guidance on the set of sensor signal component encoding feature vectors to obtain a sparse set of sensor signal component encoding feature vectors; The fault type identification module is used to determine the fault type of the fiber Bragg grating sensor based on the set of sparse sensor signal component encoding feature vectors.

Images