Top cover bolt stress abnormity diagnosis method based on double-optical-fiber strain decoupling

Abstract

The invention discloses a top cover bolt stress abnormity diagnosis method based on double-optical-fiber strain decoupling, and relates to the technical field of bolt abnormity diagnosis, and the method comprises the steps: collecting the strain data of a top cover bolt through a double-optical-fiber strain sensor, and obtaining the double-optical-fiber strain data; carrying out strain decoupling processing based on the double-fiber strain data, and separating the axial stress and the shear stress of the bolt by adopting an orthogonally arranged fiber grating decoupling algorithm; and carrying out stress abnormity diagnosis according to the decoupled axial stress and shear stress, comparing an abnormity detection model based on a support vector machine with historical stress data, outputting an abnormity diagnosis result, and completing the top cover bolt stress abnormity diagnosis based on double-optical-fiber strain decoupling. According to the method, accurate separation and rapid anomaly identification of the axial stress and the shear stress of the top cover bolt under the complex load are realized, the measurement credibility and the diagnosis precision are improved, the anomaly type distinguishing capability is enhanced, the false alarm rate is effectively reduced, and online risk early warning and operation and maintenance decision are supported.

Description

Technical Field

[0001] This invention relates to the field of bolt anomaly diagnosis technology, and in particular to a method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling. Background Technology

[0002] As large-scale equipment, energy equipment, and complex structural systems develop towards higher reliability, higher load capacity, and longer operational cycles, bolted connections, as a key form of mechanical connection, have attracted widespread attention for their health status monitoring technology. Traditional bolt stress monitoring relies heavily on strain gauges, resistance strain gauges, or ultrasonic guided wave-based fastening force assessment methods. However, these methods are susceptible to factors such as temperature drift, electromagnetic interference, material aging, and contact instability under long-term operating conditions, making it difficult to achieve high-precision, long-term, continuous online monitoring.

[0003] In recent years, fiber optic sensing technology has been gradually introduced into the field of bolt health monitoring due to its advantages such as strong resistance to electromagnetic interference, long-term stable operation, and the ability to achieve quasi-distributed measurement. Among them, fiber Bragg grating (FBG) sensors, with their linear response to strain changes in wavelength drift, can be used to monitor the minute deformations that bolts undergo during operation. However, under complex load coupling, FBGs arranged in a single direction cannot effectively distinguish between multiple stress components such as axial stress, shear stress, and bending stress, making the monitoring results often unable to accurately reflect the true stress state of the bolt, thus limiting the accuracy of diagnosis.

[0004] Existing bolt stress monitoring technologies still have many shortcomings in multi-factor coupled environments. On the one hand, the bolt stress in operating structures has a significant multi-directional superposition effect, and traditional resistance and single-fiber strain measurement methods cannot distinguish the stress contributions in different directions, resulting in "aliasing" of monitoring information and making it difficult to support accurate diagnosis. On the other hand, existing methods usually adopt uniaxial or non-orthogonal arrangements, making it difficult to directly invert the mapping matrix between strain signals. This leads to unstable decoupling operations and insufficient robustness, failing to meet the engineering application requirements under complex load variation scenarios. Summary of the Invention

[0005] In view of the problems existing in the current method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling, this invention is proposed. Therefore, the problem to be solved by this invention is how to provide a method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling, which includes: acquiring strain data of top cover bolts through dual-fiber strain sensors to obtain dual-fiber strain data;

[0008] Strain decoupling processing is performed based on dual-fiber strain data, and an orthogonally arranged fiber optic grating decoupling algorithm is used to separate the axial stress and shear stress of the bolt.

[0009] Anomaly diagnosis is performed on the top cover bolts based on the decoupled axial stress and shear stress. The anomaly detection model based on support vector machine is compared with historical stress data to output the anomaly diagnosis results, thus completing the anomaly diagnosis of the top cover bolts based on dual fiber strain decoupling.

[0010] As a preferred embodiment of the method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling described in this invention, the dual-fiber strain sensors are attached to the surface of the top cover bolt head in an orthogonal direction, with the axis of sensor one parallel to the bolt axis and the axis of sensor two perpendicular to sensor one.

[0011] As a preferred embodiment of the method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling described in this invention, wherein: the separation of axial stress and shear stress in the bolts using an orthogonally arranged fiber optic grating decoupling algorithm includes:

[0012] Based on the fiber optic strain data, a strain observation vector is constructed. After inputting the strain observation vector into the decoupling algorithm, the true stress vector of the bolt is solved by linear matrix inversion.

[0013] The decoupling operation establishes a strain-stress mapping relationship based on the sensitive direction of the orthogonally arranged optical fibers. Axial stress and shear stress are obtained from the coupled strain, and the decoupling relationship is established, expressed as:

[0014] ;

[0015] In the formula, and These represent the strain data measured by the two orthogonal optical fibers, respectively. Indicates the axial stress of the bolt. Indicates the bolt shear stress. , , as well as Represents the coupling coefficient;

[0016] The system of equations is solved using matrix operations to obtain the axial stress and shear stress of the bolt at each moment.

[0017] As a preferred embodiment of the method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling described in this invention, the abnormal stress diagnosis includes:

[0018] The axial stress and shear stress obtained by decoupling are used as inputs to construct a pre-trained support vector machine model for real-time discrimination, and the results of anomaly probability and anomaly type are compared with the historical stress database.

[0019] The axial stress and shear stress are divided into sliding frames with a fixed window length, and the mean, root mean square and correlation coefficient of the axial stress and shear stress are calculated for each frame.

[0020] The mean, root mean square, and correlation coefficients of the obtained axial stress and shear stress are concatenated into a feature vector in a fixed order. The feature vector is then standardized and input into a pre-trained support vector machine model. A radial basis function kernel is used, and the decision function of the support vector machine model is expressed as:

[0021] ;

[0022] In the formula, This is the original score vector. For support vectors, The standardized feature vector, The coefficients obtained during training, The labels obtained during training, For bias terms, This represents the total number of support vectors.

[0023] The probability estimate is obtained by using Platt calibration for the score of each category. The model inference result is the probability of all categories, expressed as:

[0024] ;

[0025] In the formula, For the first Category probability, and These are the calibration parameters determined during training through cross-validation. For belonging to the first Category probability, For the first Category score;

[0026] Simultaneously compare the historical stress database, and use the covariance matrix and class center vector of each category calculated from the historical normal and abnormal sample sets during the training phase to calculate the distance between the current frame feature vector and the class center;

[0027] Based on the combined determination of the probability of each category and the distance from the category center, the abnormal force diagnosis result of the top cover bolt is output.

[0028] As a preferred embodiment of the method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling described in this invention, the formulas for calculating the mean, root mean square, and correlation coefficient of the axial stress and shear stress are as follows:

[0029] ;

[0030] ;

[0031] ;

[0032] In the formula, This represents the average value of the axial stress. This represents the average shear stress. The total number of frames. For the first Axial stress of the frame, For the first Shear stress of the frame; The root mean square of the axial stress is given. is the root mean square of the shear stress; This is the correlation coefficient between axial and shear stress.

[0033] As a preferred embodiment of the method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling described in this invention, wherein: calculating the distance between the current frame feature vector and the category center includes:

[0034] ;

[0035] In the formula, The current frame feature vector and the first Distance from the center of the category For the first The inverse matrix of class covariance, It is the transpose symbol. For the first Category center vector.

[0036] As a preferred embodiment of the method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling described in this invention, the output abnormal diagnosis result includes: if for a certain category, the category probability is greater than a predetermined probability and the distance is less than a predetermined distance, then it is determined to be abnormal; if the probability and distance of all categories are within the normal range, then it is determined to be normal, and finally the abnormal diagnosis result is output.

[0037] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling.

[0038] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements the steps of a method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling.

[0039] The beneficial effects of this invention are as follows: This invention enables accurate separation and rapid identification of anomalies in the axial and shear stress of top cover bolts under complex loads, improves measurement reliability and diagnostic accuracy, enhances the ability to distinguish anomaly types, effectively reduces false alarm rate, and supports online risk warning and operation and maintenance decision-making. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0041] Figure 1 This is a flowchart of a method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling. Detailed Implementation

[0042] To make the above-mentioned objects, features, and advantages of the present invention more readily understood, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0043] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0044] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0045] Reference Figure 1 This is the first embodiment of the present invention, which provides a method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling, including:

[0046] S1: The strain data of the top cover bolts is collected by dual fiber optic strain sensors and the signal is preprocessed to obtain dual fiber optic strain data.

[0047] S2: Based on dual-fiber strain data, strain decoupling processing is performed, and an orthogonally arranged fiber optic grating decoupling algorithm is used to separate the axial stress and shear stress of the bolt.

[0048] S3: Perform stress anomaly diagnosis based on the decoupled axial stress and shear stress. By comparing the anomaly detection model based on support vector machine with historical stress data, output the anomaly diagnosis results and complete the stress anomaly diagnosis of the top cover bolts based on dual fiber strain decoupling.

[0049] Specifically, two fiber optic strain sensors are securely attached to the surface of the top cover bolt head in an orthogonal direction (90 degrees to each other). The axis of sensor one is parallel to the bolt axis, and the axis of sensor two is perpendicular to sensor one. The two optical fibers are connected to an optical demodulation unit to convert the wavelength shift reflected by the optical fibers into a time series signal, thereby obtaining dual-fiber strain data.

[0050] The strain data of two optical fibers arranged in orthogonal directions are synchronized with a unified time scale to construct an original observation vector composed of two strain paths. Since the optical fibers are fixed on the bolt surface along the axial and tangential directions respectively, their strain response includes the combined effects of axial deformation and shear deformation. The two types of stress components are separated from the coupled strain through decoupling operations.

[0051] A strain response matrix is ​​established based on the spatial orientation of the two optical fibers. The two strain data are arranged in a fixed order to form a strain observation vector. A coefficient matrix containing the mapping relationship between the sensitive direction of the optical fiber and the stress components of the bolt is constructed. After the strain observation vector is input into the decoupling algorithm, the true stress vector of the bolt is solved by linear matrix inversion.

[0052] The decoupling operation utilizes the linear relationship between fiber Bragg wavelength drift and strain, and establishes a mapping relationship between strain and stress based on the sensitive direction of the orthogonally arranged fibers. Axial stress and shear stress are obtained from the coupled strain, establishing the following decoupling relationship, expressed as:

[0053] ;

[0054] In the formula, and These represent the strains measured by the two orthogonal optical fibers, respectively. Indicates the axial stress of the bolt. Indicates the bolt shear stress. , , , Represents the coupling coefficient;

[0055] Obtaining the coupling coefficients: First, two fiber Bragg gratings are orthogonally bonded to the bolt surface, and the assembly is calibrated in a loading fixture. Then, known pure axial tensile loads and pure shear loads are applied respectively, and the center wavelength drift of the two fiber Bragg gratings is recorded in real time, obtaining multiple sets of standard load and wavelength change data. Based on the stress response characteristics of the fiber Bragg gratings, a linear coupling matrix equation between wavelength change and axial stress and shear stress is established. The data from the loading test are substituted into this equation, and the four coupling coefficients corresponding to the fiber are solved using the least squares method. After obtaining the coupling coefficients, they are checked for error under other loading conditions. If the stress inversion error meets the preset accuracy requirements, the coupling coefficient is fixed for subsequent strain decoupling processing.

[0056] The system of equations is solved using matrix operations to obtain the axial stress and shear stress of the bolt at each moment.

[0057] By using a decoupling algorithm, the axial and shear stress components are separated without changing the physical installation structure, thus obtaining the actual bolt stress situation that changes over time.

[0058] Using the decoupled axial stress and shear stress as input, a pre-trained support vector machine (SVM) model is constructed for real-time discrimination, and the results of anomaly probability and anomaly type determination are compared with the historical stress database.

[0059] The pre-trained support vector machine (SVM) model comprises three parts: feature construction, training data preparation, and model training and parameter optimization. First, based on the historical axial and shear stresses of bolts during normal service, fundamental features including stress amplitude, stress change rate, and short-term stability indices are extracted to form a stress feature vector for training. Then, normal operating condition data is selected from long-term monitoring records, and a small number of typical abnormal operating conditions are manually labeled to construct a training set with category information. The training set is input into the SVM model, and the stress features are mapped to a high-dimensional space through a kernel function, making normal and abnormal samples separable. Simultaneously, cross-validation is used to optimize the kernel function type, penalty factor, and kernel parameters. After training, the model is validated offline to ensure that its sensitivity and generalization ability to different stress trends meet diagnostic requirements. Finally, the model is used as a discriminant for online stress anomaly detection.

[0060] Using a fixed window length, the axial stress and shear stress are divided into sliding frames. For each frame, the mean, root mean square, and correlation coefficient of the axial stress and shear stress are calculated, as follows:

[0061] ;

[0062] ;

[0063] ;

[0064] In the formula, This represents the average value of the axial stress. This represents the average shear stress. The total number of frames. For the first Axial stress of the frame, For the first Shear stress of the frame; The root mean square of the axial stress is given. is the root mean square of the shear stress; The correlation coefficient between axial and shear stress;

[0065] The feature vectors are concatenated in a fixed order, then standardized. The standardized feature vectors are then input into a pre-trained support vector machine (SVM) model using a radial basis function (RBF) kernel. The decision function of the SVM model is expressed as follows:

[0066] ;

[0067] In the formula, This is the original score vector. For support vectors, The standardized feature vector, The coefficients obtained during training, The labels obtained during training, For bias terms, This represents the total number of support vectors.

[0068] The probability estimate is obtained by using Platt calibration for the score of each category, as follows:

[0069] ;

[0070] In the formula, For the first Category probability, and These are the calibration parameters determined during training through cross-validation. For belonging to the first Category probability, For the first The scores for each category; the model's inference results are the probabilities for all categories.

[0071] Parallel to SVM discrimination, a comparison with the historical stress database is performed: using the class covariance matrices and class center vectors calculated from the historical normal and abnormal sample sets during the training phase, the distance between the current frame feature vector and the class center is calculated, expressed as:

[0072] ;

[0073] In the formula, The current frame feature vector and the first Distance from the center of the category For the first The inverse matrix of class covariance, It is the transpose symbol. For the first Category center vector;

[0074] A joint judgment is made based on the obtained category probabilities and distances: if a category simultaneously satisfies the condition that the category probability is greater than a predetermined probability and the distance is less than a predetermined distance, it is judged as abnormal; if all category probabilities and distances are within the normal range, it is judged as normal, and the final abnormal diagnosis result is output.

[0075] This embodiment also provides a computer device applicable to a method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement all or part of the steps of the method described in the above embodiments of the present invention.

[0076] This embodiment also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, it performs the method in any optional implementation of the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0077] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0078] In summary, this invention enables accurate separation and rapid identification of anomalies in the axial and shear stresses of top cover bolts under complex loads, improving measurement reliability and diagnostic accuracy, enhancing the ability to distinguish anomaly types, effectively reducing false alarm rates, and supporting online risk warning and operation and maintenance decisions.

[0079] It should be noted that 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 technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling, characterized in that: include, The strain data of the top cover bolts were collected by dual fiber optic strain sensors to obtain dual fiber optic strain data. Strain decoupling processing is performed based on dual-fiber strain data, and an orthogonally arranged fiber optic grating decoupling algorithm is used to separate the axial stress and shear stress of the bolt. Anomaly diagnosis is performed on the top cover bolts based on the decoupled axial stress and shear stress. The anomaly detection model based on support vector machine is compared with historical stress data to output the anomaly diagnosis results, thus completing the anomaly diagnosis of the top cover bolts based on dual fiber strain decoupling.

2. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 1, characterized in that: The dual fiber optic strain sensors are attached to the surface of the top cover bolt head in an orthogonal direction. The axis of sensor one is parallel to the bolt axis, and the axis of sensor two is perpendicular to sensor one.

3. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 1, characterized in that: The axial stress and shear stress of the bolt separated by the orthogonally arranged fiber optic grating decoupling algorithm include: Based on the fiber optic strain data, a strain observation vector is constructed. After inputting the strain observation vector into the decoupling algorithm, the true stress vector of the bolt is solved by linear matrix inversion. The decoupling operation establishes a strain-stress mapping relationship based on the sensitive direction of the orthogonally arranged optical fibers. Axial stress and shear stress are obtained from the coupled strain, and the decoupling relationship is established, expressed as: ； In the formula, and These represent the strain data measured by the two orthogonal optical fibers, respectively. Indicates the axial stress of the bolt. Indicates the bolt shear stress. , , as well as Represents the coupling coefficient; The system of equations is solved using matrix operations to obtain the axial stress and shear stress of the bolt at each moment.

4. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 1, characterized in that: The process of diagnosing abnormal stress includes: The axial stress and shear stress obtained by decoupling are used as inputs to construct a pre-trained support vector machine model for real-time discrimination, and the results of anomaly probability and anomaly type are compared with the historical stress database. The axial stress and shear stress are divided into sliding frames with a fixed window length, and the mean, root mean square and correlation coefficient of the axial stress and shear stress are calculated for each frame. The mean, root mean square, and correlation coefficients of the obtained axial and shear stresses are concatenated into a feature vector in a fixed order. The feature vector is then standardized and input into a pre-trained support vector machine (SVM) model. A radial basis function kernel is used, and the decision function of the SVM model is expressed as: ； In the formula, This is the original score vector. For support vectors, The standardized feature vector, The coefficients obtained during training, The labels obtained during training, For bias terms, This represents the total number of support vectors. The probability estimate is obtained by using Platt calibration for the score of each category. The model inference result is the probability of all categories, expressed as: ； In the formula, For the first Category probability, and These are the calibration parameters determined during training through cross-validation. For belonging to the first Category probability, For the first Category score; Simultaneously compare the historical stress database, and use the covariance matrix and class center vector of each category calculated from the historical normal and abnormal sample sets during the training phase to calculate the distance between the current frame feature vector and the class center; Based on the combined determination of the probability of each category and the distance from the category center, the abnormal force diagnosis result of the top cover bolt is output.

5. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 4, characterized in that: The formulas for calculating the mean, root mean square, and correlation coefficient of the axial stress and shear stress are as follows: ； ； ； In the formula, This represents the average value of the axial stress. This represents the average shear stress. The total number of frames. For the first Axial stress of the frame, For the first Shear stress of the frame; The root mean square of the axial stress is given. is the root mean square of the shear stress; This is the correlation coefficient between axial and shear stress.

6. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 4, characterized in that: The calculation of the distance between the current frame feature vector and the category center includes: ； In the formula, The current frame feature vector and the first Distance from the center of the category For the first The inverse matrix of class covariance, It is the transpose symbol. For the first Category center vector.

7. The method for diagnosing abnormal stress in top cover bolts based on dual-fiber strain decoupling as described in claim 6, characterized in that: The output of the abnormal diagnosis result includes: if a certain category simultaneously satisfies the condition that the category probability is greater than a predetermined probability and the distance is less than a predetermined distance, it is judged as abnormal; if the probability and distance of all categories are within the normal range, it is judged as normal, and the final abnormal diagnosis result is output.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for diagnosing abnormal stress on top cover bolts based on dual-fiber strain decoupling as described in any one of claims 1 to 7.

Images