A method and system for monitoring optical fiber stress

Abstract

The application relates to the technical field of stress monitoring, in particular to a kind of optical fiber stress monitoring method and system.It includes: preprocessing original optical signal data to form initial signal sequence;Based on frequency separation, extract polarization state fluctuation characteristics, determine polarization response mode;According to polarization response mode analysis coupling effect, extract low-frequency component to obtain purified signal sequence, calculate polarization elliptical parameter change rate to determine the principal stress direction coordinate of target material;Combining cluster analysis to identify optical signal cluster, quantize irreversible change between clusters to determine the plastic deformation starting point;Based on phase shift integral value, establish deformation quantization model, calculate plastic deformation degree index;Update polarization response mode and correct coupling effect parameter, output real-time deformation monitoring result of target material.The application solves the problem that signal coupling interference is large and plastic deformation is difficult to accurately identify in the existing optical fiber sensing technology, and realizes high sensitivity, high reliability and real-time monitoring of material stress.

Description

A fiber optic stress monitoring method and system Technical Field

[0001] This invention relates to the field of stress monitoring technology, and in particular to a fiber optic stress monitoring method and system. Background Technology

[0002] In recent years, with the increasing demands for structural safety and performance monitoring in fields such as aerospace, bridge engineering, and precision manufacturing, real-time monitoring technology for material stress and deformation has become a key component of structural health monitoring (SHM) systems. Traditional stress detection methods mainly include strain gauges, resistive sensors, and ultrasonic testing. While these methods can achieve a certain degree of stress measurement, they suffer from limitations such as limited measurement points, weak anti-interference capabilities, and the inability to perform long-term online monitoring. These limitations make it difficult to meet the high-sensitivity and high-resolution monitoring requirements of complex structures under multi-load environments. Especially in aerospace composite materials and large metal structures, minute deformations within the material are often difficult for mechanical sensors to capture in the early stages, thus increasing the potential risk of failure.

[0003] Fiber optic sensing technology has gradually become an important technical approach for stress and deformation monitoring due to its advantages such as resistance to electromagnetic interference, high temperature resistance, high sensitivity, and distributed measurement capability. However, existing fiber optic sensing methods still have several shortcomings: on the one hand, multipath interference and external vibrations can easily cause optical signal coupling, making polarization state and phase information severely affected by noise; on the other hand, traditional signal analysis often relies on fixed models or single feature extraction, lacking adaptive analytical capabilities for the dynamic response of light wave polarization under complex stress fields, resulting in limited monitoring accuracy and stability.

[0004] To address the aforementioned problems, this invention proposes a fiber optic stress monitoring method based on a fiber optic sensor array. By deploying a fiber optic array on the material surface, it acquires real-time information on the polarization state and intensity phase of light waves. Combining frequency domain separation, signal purification, polarization response mode recognition, and deformation quantization model construction, it achieves a complete analysis from the original optical signal to stress state parameters. This method, through the fusion processing of polarization response mode parameters and purified signal sequences, can effectively distinguish changes in optical characteristics caused by internal material stress, external coupling, or vibration, thereby achieving accurate identification of the initiation point of plastic deformation and quantitative assessment of the degree of deformation. Compared with existing technologies, this invention utilizes the high sensitivity of the optical domain and multi-dimensional signal modeling techniques to significantly improve the accuracy, real-time performance, and environmental adaptability of stress monitoring, making it particularly suitable for long-term online monitoring scenarios in aerospace, civil engineering, and high-reliability structures. Summary of the Invention

[0005] This invention provides a fiber optic stress monitoring method and system, which enables high-precision real-time monitoring of material stress state through dynamic extraction of optical signals, spectrum analysis, polarization response modeling, and data closed-loop correction.

[0006] In a first aspect, the present invention provides a method for monitoring fiber optic stress, the method comprising:

[0007] Step S1: Acquire light wave polarization state and light intensity phase information by deploying an optical fiber sensor array on the surface of the target material; obtain raw optical signal data based on the light wave polarization state and light intensity phase information, and perform preprocessing to obtain an initial signal sequence; separate frequency components according to the initial signal sequence, extract polarization state fluctuation characteristics, and determine the polarization response mode;

[0008] Step S2: Determine the coupling effect of the signal based on the polarization response mode, obtain a purified signal sequence based on the low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode; calculate the rate of change of polarization ellipse parameters using the purified signal sequence, identify the potential stress direction vector of the target material, and determine the principal stress direction coordinates of the target material.

[0009] Step S3: Based on the principal stress direction coordinates and the purified signal sequence, obtain optical signal clusters through cluster analysis, classify the signal clusters, quantify the degree of irreversible change between different clusters, and determine the plastic deformation initiation point of the target material;

[0010] Step S4: Based on the plastic deformation initiation point of the target material, extract the phase shift integral value from the purified signal sequence, establish a deformation quantification model, and obtain the plastic deformation degree index of the target material;

[0011] Step S5: When the plastic deformation degree index is greater than the set value, the polarization response mode is updated by the deformation quantification model, the updated polarization response mode is compared with the initial signal sequence, the coupling influence parameters are corrected, and the deformation monitoring output result of the target material is obtained.

[0012] As a preferred embodiment of the present invention, step S1, obtaining the initial signal sequence, includes:

[0013] The polarization state and intensity phase information of light waves on the surface of the target material are acquired by an optical fiber sensor array; for the polarization state of the light waves, the polarization direction angle and degree of polarization of the light waves are determined, and a polarization state sequence is generated; for the intensity phase information of the light waves, phase change features are extracted, and a phase sequence is generated.

[0014] Based on the polarization state sequence and the phase sequence, an initial signal sequence containing time and spatial dimensions is synthesized, wherein the initial signal sequence characterizes the optical response characteristics of the target material surface; the initial signal sequence is preprocessed to remove noise interference, resulting in a standardized initial signal sequence.

[0015] As a preferred embodiment of the present invention, step S1, determining the polarization response mode, includes:

[0016] Fourier transform is applied to the initial signal sequence to separate the high-frequency and low-frequency components of the light wave; for the high-frequency components, the fluctuation frequency and amplitude characteristics of the polarization state are extracted; based on the fluctuation frequency and amplitude characteristics, a polarization state fluctuation feature vector is constructed; using the polarization state fluctuation feature vector, cluster analysis is used to determine the polarization response mode, wherein the polarization response mode characterizes the dynamic response law of the optical signal on the surface of the target material; and polarization response mode parameters are generated based on the polarization response mode.

[0017] As a preferred embodiment of the present invention, step S2, obtaining the purified signal sequence, includes:

[0018] Analyze the frequency distribution characteristics of the polarization response mode to determine whether it exhibits a multi-peak distribution; if the polarization response mode exhibits a multi-peak distribution, then coupling influence is determined to be the dominant factor; for the low-frequency components in the initial signal sequence, a filtering algorithm is applied to remove high-frequency noise and extract the low-frequency signal components; based on the low-frequency signal components, a purified signal sequence is generated, wherein the purified signal sequence represents the main optical signal after removing coupling influence; the purified signal sequence is normalized to obtain a standardized purified signal sequence.

[0019] As a preferred embodiment of the present invention, step S2, determining the principal stress direction coordinates, includes:

[0020] Based on the purified signal sequence, the ellipticity and azimuth angle of the polarization ellipse are calculated to generate a polarization ellipse parameter sequence. For the polarization ellipse parameter sequence, the rate of change in the time dimension is calculated to obtain the polarization ellipse parameter change rate. Based on the polarization ellipse parameter change rate, the portion of the change rate exceeding a preset threshold is determined. For the portion of the change rate exceeding the preset threshold, the corresponding polarization state features are extracted to generate the potential stress direction vector of the target material. Based on the potential stress direction vector, the principal stress direction coordinates of the target material are determined.

[0021] As a preferred embodiment of the present invention, step S3, determining the starting point of plastic deformation of the target material, includes:

[0022] Based on the principal stress direction coordinates of the target material and the purified signal sequence, an optical signal feature matrix is ​​constructed; the optical signal feature matrix is ​​classified using a support vector machine to obtain multiple optical signal clusters; for each optical signal cluster, the inter-cluster distance is calculated to quantify the degree of irreversible change; based on the degree of irreversible change, it is determined whether there is a change exceeding a preset threshold; if there is a change exceeding the preset threshold, the plastic deformation initiation point is determined, wherein the plastic deformation initiation point characterizes the initial moment when the material undergoes irreversible deformation.

[0023] As a preferred embodiment of the present invention, in step S4, a deformation quantification model is established to obtain an index of the degree of plastic deformation, including:

[0024] Based on the plastic deformation initiation point, a corresponding time window is determined in the purified signal sequence; for the purified signal sequence within the time window, phase shift features are extracted; the phase shift features are integrated to obtain the phase shift integral value; based on the phase shift integral value and the principal stress direction coordinates, a deformation quantization model is constructed, wherein the deformation quantization model characterizes the mapping relationship between phase shift and stress direction; and the degree of plastic deformation is calculated using the deformation quantization model.

[0025] As a preferred embodiment of the present invention, step S5, obtaining the monitoring output, includes:

[0026] When the plastic deformation degree index is greater than a set value, an updated polarization response mode parameter is generated according to the deformation quantization model; the updated polarization response mode parameter is compared with the initial signal sequence to determine the deviation of the coupling effect parameter; based on the deviation, the coupling effect parameter is corrected to generate the corrected coupling effect parameter.

[0027] Based on the corrected coupling effect parameters, the polarization response mode is adjusted to obtain the final polarization response mode; through the final polarization response mode, an accurate monitoring output is generated, wherein the monitoring output result characterizes the real-time state of material deformation.

[0028] The present invention also provides an optical fiber stress monitoring system for implementing the above method, the system comprising:

[0029] The signal acquisition and processing unit is used to acquire light wave polarization state and light intensity phase information through an optical fiber sensor array deployed on the surface of the target material, acquire raw optical signal data based on the light wave polarization state and light intensity phase information, and perform preprocessing to obtain an initial signal sequence; separate frequency components according to the initial signal sequence, extract polarization state fluctuation characteristics, and determine the polarization response mode;

[0030] The stress identification unit is used to determine the coupling effect of the signal based on the polarization response mode, obtain a purified signal sequence based on the low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode, calculate the rate of change of polarization ellipse parameters through the purified signal sequence, identify the potential stress direction vector of the target material, and determine the principal stress direction coordinates of the target material.

[0031] The clustering analysis unit is used to obtain optical signal clusters based on the principal stress direction coordinates and the purified signal sequence through clustering analysis, classify the signal clusters, quantify the degree of irreversible change between different clusters, and determine the plastic deformation initiation point of the target material.

[0032] The deformation modeling unit is used to extract the phase shift integral value from the purified signal sequence based on the plastic deformation initiation point of the target material, establish a deformation quantification model, and obtain the plastic deformation degree index of the target material.

[0033] The deformation monitoring unit is used to update the polarization response mode through the deformation quantification model when the plastic deformation degree index is greater than a set value, compare the updated polarization response mode with the initial signal sequence, correct the coupling influence parameters, and obtain the deformation monitoring output result of the target material.

[0034] The present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0035] The beneficial effects of this invention are as follows:

[0036] This invention uses the initial signal sequence acquired by a fiber optic sensor array as a foundation. Through frequency domain analysis, noise and coupling interference are separated, and low-frequency components are extracted to form a purified signal sequence, ensuring that only optical response features directly related to stress changes are retained. The polarization ellipticity and azimuth angle are calculated from the purified signal sequence to generate a polarization elliptic parameter sequence. Through time-dimensional rate of change analysis, segments with significant polarization state changes are identified, and potential stress direction vectors are extracted and principal stress direction coordinates are determined, achieving spatial positioning of the stress direction. The principal stress direction coordinates are fused with the purified signal sequence to construct an optical signal feature matrix. A support vector machine model is used to classify the matrix data, dividing the optical signals into different clusters representing the optical response modes of the material under different stress states. The degree of irreversible change is quantified by calculating the inter-cluster distance to determine the starting point of plastic deformation, providing a basis for subsequent deformation quantification. Using this starting point as a time window, the purified signal sequence is analyzed. Phase shift features are extracted from the signal and integral values ​​are calculated. A deformation quantification model is established in conjunction with the principal stress direction to quantify the degree of plastic deformation, thereby realizing a numerical expression of the relationship between material stress and deformation. Based on the deformation quantification model, updated polarization response mode parameters are generated and compared with the initial signal sequence to identify coupling deviations. The optimized response parameters are obtained through least squares weighted correction. The above correction results are used to dynamically adjust the polarization response mode, realizing closed-loop feedback between the signal level and the model level, so that the system maintains stable accuracy under different environmental and load conditions. Through the cooperation of the above technical solutions, and through the multi-level cooperation of fiber optic signal frequency domain decomposition, polarization modeling, machine learning classification and model self-correction, an adaptive, scalable and highly sensitive fiber optic stress monitoring system is constructed, which can achieve continuous and reliable stress state identification and plastic deformation early warning in complex structures such as aerospace composite materials and bridge steel structures. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 is a flowchart of an optical fiber stress monitoring method in this embodiment;

[0039] Figure 2 is a flowchart of the method for determining the principal stress direction coordinates in this embodiment;

[0040] Figure 3 is a structural diagram of an optical fiber stress monitoring system in this embodiment. Detailed Implementation

[0041] This invention provides a fiber optic stress monitoring method and system. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0042] For ease of understanding, the specific process of an embodiment of the present invention is described below. As shown in Figure 1, an embodiment of the present invention provides a fiber optic stress monitoring method, including:

[0043] Step S1: Acquire light wave polarization state and light intensity phase information by deploying an optical fiber sensor array on the surface of the target material; obtain raw optical signal data based on the light wave polarization state and light intensity phase information, and perform preprocessing to obtain an initial signal sequence; separate frequency components according to the initial signal sequence, extract polarization state fluctuation characteristics, and determine the polarization response mode;

[0044] In step S1, the initial signal sequence is obtained, including:

[0045] The polarization state and intensity phase information of light waves on the surface of the target material are acquired by an optical fiber sensor array; for the polarization state of the light waves, the polarization direction angle and degree of polarization of the light waves are determined, and a polarization state sequence is generated; for the intensity phase information of the light waves, phase change features are extracted, and a phase sequence is generated.

[0046] Based on the polarization state sequence and the phase sequence, an initial signal sequence containing time and spatial dimensions is synthesized, wherein the initial signal sequence characterizes the optical response characteristics of the target material surface; the initial signal sequence is preprocessed to remove noise interference, resulting in a standardized initial signal sequence.

[0047] Specifically, existing material stress and deformation monitoring technologies mostly rely on resistance strain gauges or single-point fiber optic sensors, which have limited measurement accuracy and anti-interference capabilities. These limitations make them unsuitable for real-time monitoring of complex structures such as aircraft, bridges, and pipelines under multiaxial stress and dynamic load conditions. Traditional methods often fail to effectively separate coupling noise caused by temperature, vibration, and structural inhomogeneity in optical signals, and cannot accurately identify the principal stress direction and the starting point of plastic deformation, leading to delayed or misjudged monitoring results. To overcome these shortcomings, this invention deploys a fiber optic sensor array on the target material surface, combining multidimensional feature extraction of polarization state and phase signals, spectral separation, polarization ellipse analysis, and clustering recognition algorithms to establish a complete monitoring system from the original optical signal to a deformation quantization model. This achieves high-precision identification and dynamic tracking of stress distribution and material deformation state, significantly improving the reliability and intelligence level of structural health monitoring.

[0048] In one embodiment, real-time monitoring and deformation identification of the stress state of the target material surface are achieved by acquiring and processing the optical signals corresponding to the fiber optic sensor array on the target material surface. First, the fiber optic sensor array deployed on the target material surface collects light wave polarization state and intensity phase information. This fiber optic sensor array is used to convert the minute optical changes caused by the force on the target material surface into quantifiable optical signals. Furthermore, the sensor array is arranged along the main stress direction of the material to enhance polarization response sensitivity and ensure spatial resolution of the signal. The polarization state parameters of the incident and reflected light, as well as the intensity and phase information contained in the light wave signal, constitute the raw optical signal data. For the aforementioned polarization states of light waves, the polarization direction angle and degree of polarization of the light waves are determined by polarization calculation algorithms to form a polarization state sequence. Specifically, the polarization state is first represented by the Jones vector, and the polarization direction angle is obtained by calculating the ratio of the electric field components of the incident light and the reflected light. The polarization direction angle refers to the angle between the vibration plane of the light wave and the reference axis, usually expressed in radians. The ratio of the intensity of the fully polarized light to the total intensity of the light wave is used as the degree of polarization to characterize the purity of the light wave polarization. By temporalizing and serializing the above polarization direction angle and degree of polarization, a polarization state sequence is formed. The above sequence can continuously reflect the polarization change trend of the target material during the stress deformation process and can be used for subsequent stress characteristic analysis.

[0049] For the aforementioned light intensity and phase information, a phase demodulation algorithm is used to extract phase change features and generate a phase sequence. The demodulation process analyzes the changes in the phase components of the light intensity interference signal, separating and quantifying the phase shift caused by the stress of the target material. The demodulated phase sequence characterizes the strain evolution behavior of the material in the time dimension, forming a complementary relationship with the polarization state sequence. The polarization state sequence reflects changes in stress direction and properties, while the phase sequence reflects strain amplitude and dynamic response. Combining the two allows for a complete mapping from optical signals to mechanical parameters. The polarization state sequence and the phase sequence are then synchronized in time and aligned in spatial coordinates to synthesize an initial signal sequence containing both time and spatial dimensions. Specifically, firstly, the sampling points of the two types of sequences are interpolated and resampled on the time axis to ensure a correspondence at the same time step. Then, based on the numbering information of the sensor array in its spatial distribution, the... The sequence is matrixed to form a two-dimensional or three-dimensional matrix structure, where rows correspond to time nodes and columns correspond to spatial location or depth information. This matrix structure enables a spatiotemporal integrated expression of the optical response characteristics of the material surface. After synthesizing the initial signal sequence, it is preprocessed to remove noise interference, resulting in a standardized initial signal sequence. Specifically, wavelet transform or adaptive filtering algorithms are used to identify and filter out high-frequency random noise and systematic error components, thereby retaining the effective signal reflecting the true stress response of the target material. The preprocessing process includes normalization, baseline correction, and time window smoothing to ensure that the signal amplitudes of each sensor node are on a uniform scale, avoiding pseudo-signal deviations caused by differences in sensitivity between arrays. The standardized initial signal sequence has a high signal-to-noise ratio and spatiotemporal consistency, and can serve as the input data basis for subsequent polarization response mode extraction, stress identification, and deformation modeling.

[0050] The above technical solution realizes a complete processing path for data acquired by fiber optic sensor arrays, from raw optical signals to standardized initial signal sequences. This path, with polarization and phase joint analysis as its core, constructs a mapping relationship between multidimensional optical features and material mechanical responses, making the physical meaning of optical signals correspond to stress characteristics. It also realizes the logical association between polarization state sequences, phase sequences, and spatial distribution matrices at the data level, thereby providing highly reliable data support for subsequent stress pattern recognition and plastic deformation judgment.

[0051] Further, in step S1, determining the polarization response mode includes:

[0052] Fourier transform is applied to the initial signal sequence to separate the high-frequency and low-frequency components of the light wave; for the high-frequency components, the fluctuation frequency and amplitude characteristics of the polarization state are extracted; based on the fluctuation frequency and amplitude characteristics, a polarization state fluctuation feature vector is constructed; through the polarization state fluctuation feature vector, cluster analysis is used to determine the polarization response mode, wherein the polarization response mode characterizes the dynamic response law of the optical signal on the surface of the target material.

[0053] Specifically, a Fourier transform is applied to the initial signal sequence to convert the time-domain signal into a frequency-domain signal, thereby distinguishing the optical changes at different time scales during the material's stress process. Specifically, a Fast Fourier Transform (FFT) algorithm is used to analyze the signal's frequency spectrum, classifying components above a preset frequency as high-frequency components and those below as low-frequency components. This separates different frequency components reflecting dynamic fluctuations and overall trends. The high-frequency components primarily correspond to the minute strain and vibration responses generated by the material at the moment of stress, while the low-frequency components represent the slow evolution of the material's overall stress state. After frequency separation, the high-frequency components are further processed to extract the polarization state fluctuation frequencies. The amplitude characteristics are obtained by first using Hilbert transform to obtain the envelope of the high-frequency polarized signal and then calculating the fluctuation frequency to capture the rhythm of the change in the polarization state of the light wave in the time dimension. The instantaneous frequency can reflect the dynamic disturbance characteristics of the micro-regions on the material surface and is an important parameter for identifying early stress concentration and local deformation. At the same time, the amplitude characteristics are obtained by calculating the difference between the peak and valley values ​​of the high-frequency polarized signal envelope, which is used to characterize the intensity and stability of the fluctuation of the polarization state of the light wave. The combined characteristics of the fluctuation frequency and amplitude reveal the energy distribution law and response intensity of the optical signal. The two form a corresponding relationship in time, that is, the frequency change and amplitude change at each moment together constitute the transient characteristics of the optical response.

[0054] The extracted frequency and amplitude features are structurally combined to construct a polarization-state wave feature vector. This feature vector, expressed as a matrix of multidimensional optical parameters, forms a high-dimensional feature space containing multiple pairs of frequency and amplitude features. This space comprehensively characterizes the wave patterns of optical signals under dynamic stress environments. The polarization-state wave feature vector not only preserves the correlation between the signal in the time and frequency domains but also strengthens the correspondence between polarization features and the material's stress state through multidimensional mapping, providing a mathematical foundation for subsequent cluster analysis. After constructing the polarization-state wave feature vector, a cluster analysis algorithm is used to determine the polarization response mode. Specifically, the K-means clustering method is employed to analyze the polarization-state wave feature vector. Data points in the quantity space are grouped; K-means clustering minimizes the intra-cluster variance by iteratively calculating the distance between data points and the centers of each cluster to achieve cluster convergence. The clustering process involves multiple rounds of updates to gradually converge the feature vectors to the optimal distribution, thereby identifying signal clusters corresponding to different response modes. Each signal cluster reflects the stable mode of the optical signal under different stress states. The position of the cluster center represents the typical polarization response characteristics, while the distance and density distribution between clusters reflect the coupling degree and stability of signal changes. If the clustering result shows a multi-peak distribution, it indicates that the polarization response mode is affected by multiple stress factors and is a coupling-dominated mode; if the clustering result shows a single-peak concentration, it indicates that the target material is in a stable stress state.

[0055] Through the above technical solution, the determination of the polarization response mode enables a logical correlation between the polarization fluctuation characteristics and the stress response of the target material, ensuring that the polarization response mode can truly reflect the dynamic optical behavior of the target material under complex stress environment, and providing accurate data support for subsequent principal stress direction identification and plastic deformation detection.

[0056] Step S2: Determine the coupling effect of the signal based on the polarization response mode, obtain a purified signal sequence based on the low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode; calculate the rate of change of polarization ellipse parameters using the purified signal sequence, identify the potential stress direction vector of the target material, and determine the principal stress direction coordinates of the target material.

[0057] In step S2, the purified signal sequence is obtained, including:

[0058] The frequency distribution characteristics of the polarization response mode are analyzed based on the polarization response parameters to determine whether it exhibits a multi-peak distribution. If the polarization response mode exhibits a multi-peak distribution, coupling influence is determined to be the dominant factor. For the low-frequency components in the initial signal sequence, a filtering algorithm is applied to remove high-frequency noise and extract low-frequency signal components. Based on the low-frequency signal components, a purified signal sequence is generated, wherein the purified signal sequence represents the main optical signal after removing coupling influence. The purified signal sequence is normalized to obtain a standardized purified signal sequence.

[0059] Specifically, based on the polarization response parameters obtained from the polarization response mode, its frequency distribution characteristics are statistically analyzed. This involves converting the time-domain signal of the polarization response mode into a frequency-domain signal using Fourier transform, identifying the amplitude and position of each frequency peak in the spectrum, smoothing the spectral lines to reduce high-frequency noise interference, and then using a peak detection algorithm to identify local maxima. If two or more peaks are detected, and the frequency difference between adjacent peaks exceeds a preset threshold, the polarization response mode is determined to exhibit a multi-peak distribution. This multi-peak distribution indicates a complex coupling effect in the optical signal caused by internal stress, temperature gradient, or external mechanical vibration of the target material. Therefore, in this case, the coupling effect is determined to be the dominant factor, requiring further analysis of the signal. The purification process involves a step-by-step approach. After identifying the dominant coupling mode, a low-pass filtering algorithm is used to remove high-frequency noise from the initial signal sequence, thereby extracting the low-frequency signal components that reflect the true stress characteristics of the material. The filtering algorithm can employ a Butterworth low-pass filter or an adaptive filtering method. By setting the cutoff frequency and combining it with transition band smoothing, high-frequency noise is suppressed while low-frequency signals are preserved. The low-frequency signal components represent the slow optical response generated on the material surface under stress loading, and their phase changes are mainly related to the internal stress state of the material, rather than external disturbance signals. The above process ensures the effectiveness of signal extraction through joint time-frequency domain analysis, so that the subsequently generated purified signal sequence retains only the main components reflecting the stress response.

[0060] After the low-frequency signal components are extracted, a purified signal sequence is generated based on these components. The generation process includes restoring the low-frequency signal components to the time domain using an inverse Fourier transform, obtaining an optical signal in time-series form. This time-domain reconstruction ensures that the purified signal remains synchronized with the original signal sequence in the time dimension, while eliminating high-frequency noise and coupling interference in amplitude and phase. This results in a purified signal sequence characterizing the true stress behavior of the material surface. In scenarios where fiber optic sensor arrays are used for monitoring bridge steel structures, this purified signal can clearly reflect low-frequency stress fluctuations caused by vehicle loads without being affected by environmental vibrations or temperature disturbances. In the monitoring of composite materials or aerospace structures, this process helps distinguish between stresses caused by the material itself and those caused by other factors. The differences between internal coupling and actual force response caused by different structural orientations are addressed to improve data reliability. Furthermore, to enhance the comparability of signals across different measurement points and improve the stability of subsequent algorithm processing, the generated purified signal sequence is normalized. The normalization process employs a minimum-maximum normalization method, mapping the value range of the purified signal sequence to a standard interval. This ensures that the signal amplitudes between different sensor nodes have uniform dimensions, avoiding amplitude deviations caused by differences in array sensitivity. The normalized purified signal sequence not only maintains the dynamic change characteristics of the original signal in the time dimension but also achieves consistency in the numerical scale, thus facilitating subsequent calculation of polarization ellipse parameters, identification of principal stress directions, and quantitative analysis of plastic deformation.

[0061] The above technical solution realizes a complete process from multi-peak pattern recognition to low-frequency signal extraction, then to purified signal reconstruction and standardization. Starting with frequency domain analysis, it identifies coupling characteristics in polarization response modes, achieves physical purification of signals through filtering and reconstruction algorithms, and then constructs a uniform-scale optical response sequence through normalization operations. This effectively overcomes the problems of high noise, strong coupling and low resolution in traditional fiber optic signal monitoring, realizes accurate mapping of optical signals to real stress responses, and improves the stability and accuracy of target material deformation monitoring.

[0062] Further, in step S2, determining the principal stress direction coordinates, as shown in Figure 2, includes:

[0063] Based on the purified signal sequence, the ellipticity and azimuth angle of the polarization ellipse are calculated to generate a polarization ellipse parameter sequence. For the polarization ellipse parameter sequence, the rate of change in the time dimension is calculated to obtain the polarization ellipse parameter change rate. Based on the polarization ellipse parameter change rate, the portion of the change rate exceeding a preset threshold is determined. For the portion of the change rate exceeding the preset threshold, the corresponding polarization state features are extracted to generate the potential stress direction vector of the target material. Based on the potential stress direction vector, the principal stress direction coordinates of the target material are determined.

[0064] Specifically, in this embodiment, the process of determining the principal stress direction coordinates includes calculating and dynamically analyzing the polarization ellipse parameters based on the purified signal sequence to extract the potential stress direction vector and map it to the material coordinate system, ultimately determining the principal stress direction. Specifically, firstly, based on the light wave polarization state and intensity phase information contained in the purified signal sequence, the ellipticity and azimuth angle of the polarization ellipse are obtained using the standard elliptic polarization calculation method. The ellipticity describes the ratio of the major and minor axes of the polarization ellipse, and the azimuth angle characterizes the deflection angle of the major axis of the polarization ellipse relative to the reference coordinate axis. Through continuous sampling operations in the time dimension, a time-varying polarization ellipse is generated. The parameter sequence described above reflects the dynamic process of the optical response of a material surface changing with stress. The material coordinate system described above is a set of "reference coordinates" used to describe the orientation of the material itself. It is not fixed in space like geographical coordinates, but follows the structural orientation of the material itself. For example, for a metal plate or composite material, the material coordinate system usually consists of three mutually perpendicular directions: one along the main extension direction of the material, such as the rolling or fiber direction, one in the transverse direction, and one in the thickness direction. In this way, when the material is subjected to stress or deformation, we can use this set of coordinates to accurately indicate in which direction the stress or deformation occurs.

[0065] For the polarization ellipse parameter sequence, its rate of change in the time dimension is calculated to quantify the dynamic amplitude and rate of change of the polarization parameters. This calculation can be achieved through differential operations, i.e., by subtracting the ellipticity and azimuth angle at adjacent time points and dividing by the sampling time interval, a polarization ellipse parameter rate of change sequence representing the rate of change of optical response is obtained. A larger rate of change indicates a more drastic change in polarization state within that time period, corresponding to a more significant change in the stress or strain on the material surface. A preset rate of change threshold is used to filter out signal intervals reflecting significant stress changes. Specifically, when the rate of change exceeds the preset threshold, it indicates that stress concentration or microscopic plastic slip may occur in a local area of ​​the material, and this time period is marked as a key analysis object to ensure that subsequent stress direction identification is based solely on the actual stress response rather than random noise disturbances. For the aforementioned rate of change exceeding... For the portion with a preset rate of change threshold, polarization state characteristic parameters corresponding to the time period are extracted from the purified signal sequence, including the amplitude and phase shift of polarization fluctuations. These features together constitute a parameter set reflecting the directional change of optical response. To further reveal the dominant change trend of the optical signal in three-dimensional space, principal component analysis is performed on the extracted multidimensional polarization feature data. Multiple feature components with the largest variance contribution are identified as dominant fluctuation direction components and combined to form a potential stress direction vector. The aforementioned potential stress direction vector comprehensively characterizes the mapping relationship between the polarization change of light waves and the internal stress direction of the material, representing the potential principal stress direction of the material under the current stress state. Through the above potential stress direction vectorization processing, the complex optical signal characteristics are transformed into stress direction information with clear physical meaning, ensuring the traceability and mechanical consistency of signal interpretation.

[0066] After obtaining the potential stress direction vector, the potential stress direction vector is mapped to the three-dimensional coordinate system of the material to determine the principal stress direction coordinates of the target material. The mapping process is achieved by calculating the projection of the vector on the x, y, and z axes. First, the vector is normalized to eliminate the influence of dimensions. Then, the dot product of each axis and the vector is calculated to obtain its component values ​​in each coordinate axis direction. The dominant component of the stress direction is determined according to the magnitude of the component values. The direction with the largest projection value is marked as the principal stress direction coordinate. The above principal stress direction coordinates represent the spatial orientation of the principal stress in the form of angles and can be used to quantitatively reflect the relationship between the stress distribution direction and the stress state of the structure.

[0067] The system also verifies the principal stress direction coordinates by combining the phase information of the purified signal sequence. By comparing the strain trend obtained from the phase shift integral with the distribution of the principal stress direction vector, if the difference between the two is within the preset error range, the accuracy of the principal stress direction coordinates is confirmed. The verification process can use the minimum deviation matching algorithm to ensure that the stress direction identification result is consistent with the actual deformation trend of the material, thereby improving the robustness of the monitoring system under dynamic loads and complex stress environments. For example, in the monitoring scenario of bridge steel structures, the rate of change of polarization ellipse parameters calculated from the purified signal shows a periodic peak distribution, corresponding to the stress pulsation under vehicle load. Based on this, the system generates a potential stress direction vector and maps it to obtain the principal stress direction coordinates, successfully capturing the principal stress concentration trend in the mid-span region and identifying structural fatigue risks in advance. In the monitoring of aerospace composite materials, the system adjusts the threshold parameters according to the anisotropic characteristics of the material, so that the potential stress direction vector can accurately reflect the stress transmission direction between fiber layers, thereby supporting high-precision monitoring under multi-axis stress conditions.

[0068] The above technical solution realizes a multi-level correlation between optical signal parameters and material stress information, transforms polarization state changes into stress direction indicators, not only achieves high-precision determination of principal stress directions, but also provides accurate data input for subsequent plastic deformation quantification and structural health assessment, significantly improving the intelligence level and engineering application reliability of the fiber optic sensing monitoring system.

[0069] Step S3: Based on the principal stress direction coordinates and the purified signal sequence, optical signal clusters are obtained through cluster analysis, and these clusters are classified. The degree of irreversible change between different clusters is quantified to determine the starting point of plastic deformation of the target material; specifically including:

[0070] Based on the principal stress direction coordinates of the target material and the purified signal sequence, an optical signal feature matrix is ​​constructed; the optical signal feature matrix is ​​classified using a support vector machine to obtain multiple optical signal clusters; for each optical signal cluster, the inter-cluster distance is calculated to quantify the degree of irreversible change; based on the degree of irreversible change, it is determined whether there is a change exceeding a preset threshold; if there is a change exceeding the preset threshold, the plastic deformation initiation point is determined, wherein the plastic deformation initiation point characterizes the initial moment when the material undergoes irreversible deformation.

[0071] Specifically, by synergistically fusing the principal stress direction coordinates and the purified signal sequence, an optical signal feature matrix is ​​constructed to achieve a unified modeling of the optical response and stress distribution on the material surface. Here, the principal stress direction coordinates are vectors calculated using the rate of change of the polarization ellipse parameters, representing the potential stress distribution on the material surface. First, the principal stress direction coordinate vectors calculated using the rate of change of the polarization ellipse parameters are registered with the corresponding light intensity and phase information in the purified signal sequence. This forms the aforementioned optical signal feature matrix through matrix mapping. In this initial matrix, rows represent stress direction components, columns represent the optical response time series, and matrix elements reflect the dynamic amplitude and phase shift of the optical signal under different stress directions. Normalization is performed on the initial matrix to map the values ​​of each matrix element to a unified range, eliminating the influence of different dimensions on subsequent analysis, thus forming a standardized optical signal feature matrix. This matrix not only preserves the correlation between stress direction and optical response but also realizes a structured expression of multidimensional features of the optical signal, providing a data foundation for nonlinear pattern recognition. The optical signal is then analyzed using a support vector machine (SVM) model. The SVM model is classified using a feature matrix to identify optical signal clusters corresponding to different stress response modes. Based on the maximum margin principle, the optimal segmentation hyperplane is determined during training to distinguish between different categories of signal samples. During model construction, the radial basis function (RBF) is preferred as the kernel function to map nonlinear features to a high-dimensional space, making the complex optical signal distribution in the original data linearly separable in the high-dimensional space. Cross-validation is used during model training to determine kernel parameters and penalty factors to prevent overfitting. A hold-out method is used on some samples to evaluate classification accuracy, ensuring the stability and repeatability of signal cluster division. After classification, multiple optical signal clusters are output, each representing the optical response category of the target material under specific stress, such as elastic strain response clusters, plastic deformation response clusters, and high-frequency coupling interference clusters. This achieves patterned identification of the material's stress state. The training data for the SVM model consists of historical optical monitoring data of the target material and its corresponding stress state labels. The historical optical monitoring data includes historical principal stress direction coordinate vectors and corresponding purified optical signals.

[0072] After classification, the inter-cluster distance is calculated for each optical signal cluster to quantify the degree of irreversible change between different clusters. Specifically, Euclidean distance is selected as the metric function of the feature space, and the distance between the centers of each cluster is calculated. Euclidean distance is the square root of the straight-line distance between two points and is used to measure the cluster separation. The larger the value, the more obvious the separation between the two signal clusters, that is, the greater the response difference between different deformation stages. The above distance values ​​are then weighted and averaged. The larger the number of samples in a cluster, the greater the weight of the cluster. The weighted average represents the overall degree of change index. The above quantitative index reflects the overall degree of separation between different clusters. The larger the value, the more obvious the difference between signal clusters, and the more irreversible the internal stress change of the corresponding material. To facilitate threshold determination, the degree of change value is normalized to a standardized scale to give it a unified physical meaning, which can be directly mapped to the deformation ratio or irreversibility index. It shows good versatility in scenarios such as composite materials, metal structures, and pipeline monitoring: when the inter-cluster distance increases significantly, it is judged that the stress state has changed from the reversible elastic stage to the irreversible plastic stage.

[0073] Based on the quantified degree of irreversible change, the system determines whether a change exceeds a preset threshold. If the change exceeds the threshold, the material is marked as entering the plastic deformation stage. The time point at which the degree of change first exceeds the threshold is taken as the starting point of plastic deformation. This time point characterizes the initial moment when the material undergoes irreversible deformation, corresponding to the abrupt inflection point of the deformation response in the optical signal sequence. Through the above determination, signal feature analysis can be directly correlated with structural deformation events, forming a self-correcting closed loop, enabling the monitoring output to have time-localization and event recognition capabilities. In practical applications, such as in the monitoring of bridge steel structures or aerospace composite materials, the system can track the dynamic changes between signal clusters in real time. Once a change rate is detected to be continuously higher than the threshold, a plastic deformation initiation alarm is output, achieving proactive identification of structural fatigue and instability risks. Overall, this technical solution effectively transforms the high-dimensional features of optical signals into measurable indicators of material stress state through a continuous logical path of feature matrix construction, pattern classification, distance quantization, and threshold judgment. This ensures both the accuracy and real-time performance of the monitoring process and improves the system's robustness and engineering applicability under complex stress environments.

[0074] Step S4: Based on the plastic deformation initiation point of the target material, extract the phase shift integral value from the purified signal sequence, establish a deformation quantification model, and obtain the plastic deformation degree index of the target material; specifically including:

[0075] Based on the plastic deformation initiation point, a corresponding time window is determined in the purified signal sequence; for the purified signal sequence within the time window, phase shift features are extracted; the phase shift features are integrated to obtain the phase shift integral value; based on the phase shift integral value and the principal stress direction coordinates, a deformation quantization model is constructed, wherein the deformation quantization model characterizes the mapping relationship between phase shift and stress direction; and the degree of plastic deformation is calculated using the deformation quantization model.

[0076] Specifically, in a preferred embodiment, based on the plastic deformation initiation point, a corresponding time window is first determined in the purified signal sequence to limit the time domain range of subsequent analysis. Specifically, the timestamp position of the plastic deformation initiation point in the purified signal sequence is identified and used as the starting boundary of the time window. This boundary is then automatically extended backward according to a preset duration to form a time window covering the initial stage of material deformation, thereby ensuring that the extracted data reflects the true dynamic process of plastic deformation. For the purified signal sequence within the aforementioned time window, phase shift features directly related to deformation are extracted. Periodic jumps, such as 2π periodic jumps, are removed using signal decomposition and phase unwinding algorithms. To maintain the continuity of phase information, wavelet transform is used to extract high-frequency local features at different scales, highlighting the phase perturbations caused by changes in microstructure. Furthermore, by calculating the statistics of the phase shift sequence, such as the mean and variance, a set of phase shift features characterizing the dynamic properties of deformation is formed. These phase shift features are then integrated to obtain the phase shift integral value, which reflects the cumulative optical response intensity during deformation and is crucial for subsequent deformation quantification analysis. The integration calculation employs a numerical integration method to ensure that signals at different time resolutions can be processed uniformly, outputting an integral quantity with a consistent scale, thereby establishing a foundation for comparable deformation features across different scenarios.

[0077] After obtaining the phase shift integral value, a deformation quantification model is constructed based on the integral value and the principal stress direction coordinates. This deformation quantification model reveals the quantitative correspondence mechanism between optical response and mechanical behavior by establishing a mapping relationship between phase shift and stress direction. Specifically, the phase shift integral value is used as the input variable, and the principal stress direction coordinate vector is used as the mapping direction parameter. The least squares method is used to fit the linear relationship between the two to obtain a preliminary mapping function. The gradient descent algorithm is used to optimize the function parameters, so that the model error gradually converges in multiple iterations, thereby ensuring the stability and accuracy of the mapping relationship under different stress levels. Through the above fitting and optimization process, the deformation quantification model can reflect the inherent coupling characteristics of the material surface optical signal change and stress direction distribution, and realize a unified characterization of different types of deformation response. The training data of the above deformation quantification model are the historical phase shift integral value, principal stress direction coordinate vector and corresponding plastic deformation degree index corresponding to the historical purified signal sequence of the target material. The historical purified signal sequence is located within the above time window.

[0078] After the model is built, the phase offset integral value and principal stress direction coordinates at the current moment are input to perform model mapping calculations, obtaining a plastic deformation degree index. This plastic deformation degree index is used as the final output to quantitatively describe the degree of plastic deformation of the target material under a specific stress state. The larger the value, the higher the degree of deformation. This achieves cross-domain quantitative conversion from optical signals to mechanical deformation, enabling fiber optic sensing monitoring to not only detect stress change trends but also accurately quantify the degree of plastic deformation. For example, when this method is applied to bridge steel structures or aerospace aluminum alloy components, by identifying the starting point of plastic deformation and extracting the phase offset integral value, the deformation degree index can be output in real time, providing a continuous and quantifiable assessment basis for structural safety monitoring, thereby achieving high-precision dynamic monitoring of material deformation processes under complex stress environments.

[0079] Step S5: When the plastic deformation degree index is greater than a set value, the polarization response mode is updated through the deformation quantification model. The updated polarization response mode is compared with the initial signal sequence to correct the coupling influence parameters and obtain the deformation monitoring output result of the target material; specifically including:

[0080] When the plastic deformation degree index is greater than a set value, an updated polarization response mode is generated according to the deformation quantization model; the updated polarization response mode parameters are compared with the initial signal sequence to determine the deviation of the coupling effect parameters; based on the deviation, the coupling effect parameters are corrected to generate corrected coupling effect parameters.

[0081] Based on the corrected coupling effect parameters, the polarization response mode is adjusted to obtain the final polarization response mode; through the final polarization response mode, an accurate monitoring output is generated, wherein the monitoring output result characterizes the real-time state of material deformation.

[0082] Specifically, in a preferred embodiment, when the aforementioned plastic deformation degree index exceeds a set value, it indicates that the target material may have undergone significant plastic deformation. To obtain more accurate monitoring results, the deformation quantization model is used to calculate the rate of change of polarization ellipse parameters by continuously extracting the phase shift integral value and principal stress direction coordinates from the purified signal sequence. This rate of change is then matched with the polarization response modes corresponding to the aforementioned parameters stored within the deformation quantization model to generate updated polarization response modes. These updated parameters reflect the dynamic changes in the optical response characteristics of the material under stress over time and are used to describe the evolution of the polarization state of the material under stress. Specifically, the deformation quantization model is used to fit and analyze the time-series data of the optical response, calculate the rate of change of the major and minor axis ratios and azimuth angles of the polarization ellipse at each measuring point over time, and thus obtain a new set of parameter vectors. These parameter vectors can quantify the microscopic deformation characteristics of the material under different load conditions. The response mode parameters include principal axis direction, ellipticity distribution, response sensitivity coefficient, etc. For example, in the monitoring scenario of aerospace composite materials, the updated polarization response mode parameters can capture the optical response of small interlayer slippage or local stress concentration, thereby significantly improving monitoring accuracy and sensitivity.

[0083] The updated polarization response mode parameters are compared point-by-point with the polarization state and intensity phase information of the light wave in the initial signal sequence to determine the deviation of the coupling effect parameters. The coupling effect parameters include at least polarization state fluctuation and phase shift. The comparison process analyzes the difference between the high-frequency components of the polarization fluctuation under the updated polarization response mode parameters and the corresponding frequency components in the initial signal in the frequency domain, and constructs a deviation vector to quantify the response difference between the two. If the comparison result shows that the deviation exceeds a set threshold, it indicates that the coupling effect of the external environment or material structure has a significant impact on the signal. The above method achieves the continuity and integrity of data comparison by retaining the initial signal sequence as an optical reference, and avoids information loss caused by multiple data conversions.

[0084] For the identified deviations, the least squares method is used to fit and correct the coupling effect parameters, generating corrected coupling effect parameters. In this process, a linear relationship is established between the deviation vector and the parameter adjustment coefficients, and the sum of squared errors is calculated and gradually minimized to achieve automatic parameter optimization. For significant deviations, a weighting factor based on the intensity of low-frequency purified signals is introduced for weighted correction, thereby improving the response balance of the deformation quantization model to different signal intensity ranges. The optimized parameters not only reflect the true optical response of the material surface but also effectively filter out errors caused by sensor coupling, temperature fluctuations, or environmental noise. For example, in bridge steel... In long-term stress monitoring of structures, this correction step can reduce environmental interference deviations to an extremely low level, improving the accuracy and stability of identifying the initiation point of plastic deformation. After parameter correction, the polarization response mode is adjusted according to the corrected coupling influence parameters to modify its frequency distribution and characteristic shape, thereby obtaining the final polarization response mode. The above mode achieves accurate restoration of optical signal characteristics by fine-tuning the multi-peak and single-peak characteristics, making the response mode consistent with the actual stress state of the material. In the final mode, the dynamic distribution of polarization characteristics can clearly distinguish the different stages of the material in elastic deformation, plastic deformation, or coupling interference.

[0085] Based on the final polarization response mode, a precise monitoring output is generated. This monitoring output reflects the current deformation state of the material in the form of real-time data and can further output comprehensive indicators such as stress direction vector, degree of plastic deformation, and deformation rate. By converting the dynamic evolution results of the polarization response mode into quantitative parameters, the mapping of optical signals to structural mechanical information is realized. For example, in the monitoring of industrial pipelines or aerospace structures, the deformation degree index is output in real time according to the final polarization response mode, and a safety warning is automatically triggered when the index exceeds a preset threshold, thereby effectively preventing potential structural failures and ensuring that the monitoring system has a high sensitivity and high reliability continuous monitoring capability.

[0086] The present invention also provides an optical fiber stress monitoring system for implementing the above-described method, as shown in Figure 3, the system comprising:

[0087] The signal acquisition and processing unit is used to acquire light wave polarization state and light intensity phase information through an optical fiber sensor array deployed on the surface of the target material, acquire raw optical signal data based on the light wave polarization state and light intensity phase information, and perform preprocessing to obtain an initial signal sequence; separate frequency components according to the initial signal sequence, extract polarization state fluctuation characteristics, and determine the polarization response mode;

[0088] The stress identification unit is used to determine the coupling effect of the signal based on the polarization response mode, obtain a purified signal sequence based on the low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode, calculate the rate of change of polarization ellipse parameters through the purified signal sequence, identify the potential stress direction vector of the target material, and determine the principal stress direction coordinates of the target material.

[0089] The clustering analysis unit is used to obtain optical signal clusters through clustering analysis based on the principal stress direction coordinates and the purified signal sequence, classify the signal clusters, quantify the degree of irreversible change between different clusters, and determine the plastic deformation initiation point of the target material.

[0090] The deformation modeling unit is used to extract the phase shift integral value from the purified signal sequence based on the plastic deformation initiation point of the target material, establish a deformation quantification model, and obtain the plastic deformation degree index of the target material.

[0091] The deformation monitoring unit is used to update the polarization response mode through the deformation quantification model when the plastic deformation degree index is greater than a set value, compare the updated polarization response mode with the initial signal sequence, correct the coupling influence parameters, and obtain the deformation monitoring output result of the target material.

[0092] The present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0093] In summary, this invention uses the initial signal sequence acquired by a fiber optic sensor array as a foundation. Through frequency domain analysis, noise and coupling interference are separated, and low-frequency components are extracted to form a purified signal sequence, ensuring that only optical response features directly related to stress changes are retained in the signal. The polarization ellipticity and azimuth angle are calculated from the purified signal sequence to generate a polarization elliptic parameter sequence. Through time-dimensional rate of change analysis, segments with significant polarization state changes are identified, and potential stress direction vectors are extracted and principal stress direction coordinates are determined, achieving spatial positioning of the stress direction. The principal stress direction coordinates are fused with the purified signal sequence to construct an optical signal feature matrix. A support vector machine model is used to classify the matrix data, dividing the optical signals into different clusters representing the optical response modes of the material under different stress states. The degree of irreversible change is quantified by calculating the inter-cluster distance, determining the starting point of plastic deformation and providing a basis for subsequent deformation quantification. Using this starting point as a time window, from the pure... Phase shift features are extracted from the signal and integral values ​​are calculated. A deformation quantification model is established in conjunction with the principal stress direction to quantify the degree of plastic deformation, thereby achieving a numerical expression of the relationship between material stress and deformation. Based on the deformation quantification model, updated polarization response mode parameters are generated and compared with the initial signal sequence to identify coupling deviations. The optimized response parameters are obtained through least squares weighted correction. The above correction results are used to dynamically adjust the polarization response mode, realizing closed-loop feedback between the signal level and the model level, so that the system maintains stable accuracy under different environmental and load conditions. Through the cooperation of the above technical solutions, and through multi-level cooperation of fiber optic signal frequency domain decomposition, polarization modeling, machine learning classification, and model self-correction, an adaptive, scalable, and highly sensitive fiber optic stress monitoring system is constructed, which can achieve continuous and reliable stress state identification and plastic deformation early warning in complex structures such as aerospace composite materials and bridge steel structures.

[0094] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0095] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] The above-described 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fiber optic stress monitoring method, characterized in that, The method includes: Step S1: Acquiring light wave polarization state and intensity phase information through an optical fiber sensor array deployed on the surface of the target material; obtaining raw optical signal data based on the light wave polarization state and intensity phase information, and performing preprocessing to obtain an initial signal sequence; separating frequency components according to the initial signal sequence, extracting polarization state fluctuation characteristics, and determining the polarization response mode; Step S2: Judging the coupling effect of the signal according to the polarization response mode; obtaining a purified signal sequence based on low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode; calculating the rate of change of polarization ellipse parameters through the purified signal sequence, identifying the potential stress direction vector of the target material, and determining the principal stress direction coordinates of the target material; Step S3: Obtaining optical signal clusters through cluster analysis based on the principal stress direction coordinates and the purified signal sequence, classifying the signal clusters, quantifying the degree of irreversible change between different clusters, and determining the plastic deformation initiation point of the target material; Step S4: Extracting the phase shift integral value from the purified signal sequence based on the plastic deformation initiation point of the target material, establishing a deformation quantification model, and obtaining the target material's... The plastic deformation degree index of the material; Step S5: When the plastic deformation degree index is greater than a set value, the polarization response mode is updated through the deformation quantification model, the updated polarization response mode is compared with the initial signal sequence, the coupling influence parameters are corrected, and the deformation monitoring output result of the target material is obtained; wherein, determining the polarization response mode includes: applying Fourier transform to the initial signal sequence to separate the high-frequency and low-frequency components of the light wave; extracting the fluctuation frequency and amplitude characteristics of the polarization state for the high-frequency components; constructing a polarization state fluctuation feature vector based on the fluctuation frequency and amplitude characteristics; determining the polarization response mode through cluster analysis using the polarization state fluctuation feature vector, the polarization response mode characterizing the dynamic response law of the optical signal on the surface of the target material; wherein, the K-means clustering method is used to group the data points in the polarization state fluctuation feature vector space; K-means clustering achieves cluster convergence by iteratively calculating the distance between the data points and the center of each cluster and minimizing the intra-cluster variance; wherein, the clustering process gradually converges the feature vector to the optimal distribution through multiple rounds of updates, thereby identifying the signal clusters corresponding to different response modes.

2. The method as described in claim 1, characterized in that, Step S1 involves obtaining an initial signal sequence, including: acquiring the polarization state and intensity phase information of light waves on the surface of the target material using an optical fiber sensor array; determining the polarization direction angle and degree of polarization of the light waves based on the polarization state, and generating a polarization state sequence; extracting phase change features based on the intensity phase information of the light waves, and generating a phase sequence; synthesizing an initial signal sequence containing time and spatial dimensions based on the polarization state sequence and the phase sequence, wherein the initial signal sequence characterizes the optical response characteristics of the target material surface; and preprocessing the initial signal sequence to remove noise interference, resulting in a standardized initial signal sequence.

3. The method as described in claim 1, characterized in that, In step S2, obtaining the purified signal sequence includes: analyzing the frequency distribution characteristics of the polarization response mode to determine whether it exhibits a multi-peak distribution; if the polarization response mode exhibits a multi-peak distribution, then determining that coupling influence is the dominant factor; applying a filtering algorithm to remove high-frequency noise and extract low-frequency signal components from the low-frequency components in the initial signal sequence; generating a purified signal sequence based on the low-frequency signal components, wherein the purified signal sequence represents the main optical signal after removing coupling influence; and normalizing the purified signal sequence to obtain a standardized purified signal sequence.

4. The method as described in claim 3, characterized in that, In step S2, determining the principal stress direction coordinates includes: calculating the ellipticity and azimuth of the polarization ellipse based on the purified signal sequence to generate a polarization ellipse parameter sequence; calculating the rate of change in the time dimension for the polarization ellipse parameter sequence to obtain the polarization ellipse parameter change rate; determining the portion of the change rate exceeding a preset threshold based on the polarization ellipse parameter change rate; extracting the corresponding polarization state features for the portion of the change rate exceeding the preset threshold to generate a potential stress direction vector for the target material; and determining the principal stress direction coordinates of the target material based on the potential stress direction vector.

5. The method as described in claim 1, characterized in that, In step S3, determining the starting point of plastic deformation of the target material includes: constructing an optical signal feature matrix based on the principal stress direction coordinates of the target material and the purified signal sequence; classifying the optical signal feature matrix using a support vector machine to obtain multiple optical signal clusters; calculating the inter-cluster distance for each optical signal cluster to quantify the degree of irreversible change; determining whether there is a change exceeding a preset threshold based on the degree of irreversible change; if there is a change exceeding the preset threshold, determining the starting point of plastic deformation, wherein the starting point of plastic deformation characterizes the initial moment when the material undergoes irreversible deformation.

6. The method as described in claim 5, characterized in that, In step S4, a deformation quantification model is established to obtain an index of the degree of plastic deformation, including: determining the time window corresponding to the purified signal sequence based on the starting point of plastic deformation; extracting phase shift features for the purified signal sequence within the time window; integrating the phase shift features to obtain a phase shift integral value; constructing a deformation quantification model based on the phase shift integral value and the principal stress direction coordinates, wherein the deformation quantification model characterizes the mapping relationship between phase shift and stress direction; and calculating the index of the degree of plastic deformation using the deformation quantification model.

7. The method as described in claim 1, characterized in that, In step S5, obtaining the monitoring output includes: when the plastic deformation degree index is greater than a set value, generating updated polarization response mode parameters according to the deformation quantification model; comparing the updated polarization response mode parameters with the initial signal sequence to determine the deviation of the coupling influence parameters; correcting the coupling influence parameters according to the deviation to generate corrected coupling influence parameters; adjusting the polarization response mode according to the corrected coupling influence parameters to obtain the final polarization response mode; and generating an accurate monitoring output through the final polarization response mode, wherein the monitoring output result characterizes the real-time state of material deformation.

8. A fiber optic stress monitoring system for implementing the method as described in any one of claims 1-7, characterized in that, The system includes: a signal acquisition and processing unit, used to acquire light wave polarization state and intensity phase information through an optical fiber sensor array deployed on the surface of the target material; acquire raw optical signal data based on the light wave polarization state and intensity phase information; preprocess the data to obtain an initial signal sequence; separate frequency components based on the initial signal sequence; extract polarization state fluctuation characteristics; and determine the polarization response mode; a stress identification unit, used to determine the coupling effect of the signal based on the polarization response mode; acquire and obtain a purified signal sequence based on low-frequency components by analyzing the frequency distribution characteristics of the polarization response mode; calculate the rate of change of polarization ellipse parameters using the purified signal sequence; identify the potential stress direction vector of the target material; and determine the principal stress direction coordinates of the target material; and clustering. The analysis unit is used to obtain optical signal clusters through cluster analysis based on the principal stress direction coordinates and the purified signal sequence, classify the signal clusters, quantify the degree of irreversible change between different clusters, and determine the plastic deformation initiation point of the target material. The deformation modeling unit is used to extract the phase shift integral value from the purified signal sequence based on the plastic deformation initiation point of the target material, establish a deformation quantification model, and obtain the plastic deformation degree index of the target material. The deformation monitoring unit is used to update the polarization response mode through the deformation quantification model when the plastic deformation degree index is greater than a set value, compare the updated polarization response mode with the initial signal sequence, correct the coupling influence parameters, and obtain the deformation monitoring output result of the target material.

9. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the method as described in any one of claims 1-7.

Images