A buried pipeline stress inversion and correction method based on pattern matching

Abstract

The present application relates to the technical field of oil and gas storage and transportation facility safety monitoring and evaluation, in particular to a buried pipeline stress inversion and correction method based on pattern matching, which first peels off temperature and pressure load to construct a pipeline reference internal force model, then inverses the internal force component of the monitoring section surface displacement through multi-source measured data, then realizes the whole line internal force deduction with pattern matching as the constraint, finally superimposes and corrects to obtain the whole line real stress field and completes risk identification, each step is progressive and interlocking, the previous step provides data basis and technical support for the next step, the next step iterates and applies the results of the previous step, forming a complete technical closed loop from model construction-single point inversion-whole line deduction-stress correction-risk identification, the present application realizes the physical level decoupling of internal force under different loads, lays a foundation for the accurate analysis of surface displacement additional stress, and greatly improves the accuracy of surface displacement internal force component inversion.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring and assessment technology for oil and gas storage and transportation facilities, specifically to a method for stress inversion and correction of buried pipelines based on pattern matching. Background Technology

[0002] Long-distance oil and gas pipelines often traverse complex geological areas, which are prone to geological disasters such as landslides and subsidence, causing surface displacement. Surface displacement will generate additional stress on buried pipelines, changing the original mechanical state of the pipelines. If the additional stress accumulates over a long period of time, it will seriously threaten the normal operation and safety of the pipelines, and may even cause safety accidents such as pipeline rupture and medium leakage. Therefore, stress monitoring and safety assessment of pipelines affected by surface displacement is an important requirement in the field of oil and gas storage and transportation.

[0003] Current pipeline stress monitoring methods in the industry can only collect strain data from specific cross-sections of the pipeline, and cannot achieve continuous monitoring of the entire pipeline. Due to limitations such as monitoring costs, construction difficulties, and pipeline laying environment, the layout of on-site monitoring points is obviously sparse, and only sporadic and discrete monitoring data can be obtained. With these limited discrete data, it is impossible to accurately identify the actual deformation mode of the pipeline under the action of geological disasters such as landslides and settlement, and it is impossible to deduce the stress distribution state of the entire pipeline. Ultimately, it is difficult for technicians to make a comprehensive and accurate evaluation of the actual safety status of the entire pipeline, and it is impossible to identify the stress-exceeding sections and safety risk points in a timely manner, forming a technical shortcoming in pipeline safety monitoring and evaluation.

[0004] Based on the shortcomings of the existing technologies, the industry urgently needs a stress analysis method for buried pipelines that can overcome the limitations of sparse monitoring points, accurately identify pipeline deformation patterns, and realize stress field inversion and continuous safety status evaluation of the entire pipeline. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method for stress inversion and correction of buried pipelines based on pattern matching, which solves the problem that the prior art cannot achieve continuous monitoring of the entire pipeline.

[0006] To solve the above problems, the technical solution of the present invention is as follows: A method for stress inversion and correction of buried pipelines based on pattern matching, comprising:

[0007] S1. Discretize the long-distance pipeline into spatial beam elements. Based on the principle of virtual work, the internal pressure and temperature loads are equivalent to nodal loads. The reference internal force distribution considering only the effects of temperature and pressure is obtained by solving the stiffness equation.

[0008] S2. By using strain sensors installed on the pipeline monitoring section and inertial measurement units (IMUs) mounted inside the pipeline, strain data and geometric deformation data of the monitoring section are collected simultaneously.

[0009] S3. Based on the plane section assumption, the deformation equation of the monitored section is fitted by strain data. The total internal force of the monitored section is obtained by integration. The reference internal force caused by temperature and pressure is eliminated, and the internal force component caused by surface displacement is obtained. The plane of bending moment action is checked by the curvature direction detected by IMU.

[0010] S4. Construct the IMU curvature waveform as a feature vector, divide the terrain feature area by K-means clustering, establish a deformation pattern database, filter candidate patterns by confidence conditions, determine the matching weight by exponential weighting, select the optimal target deformation pattern, and output RBF interpolation parameters.

[0011] S5. Using the internal force components of the monitored sections as samples, an interpolation model is constructed using RBF interpolation parameters. The weight coefficients are solved by the least squares method, and the surface displacement internal force distribution of the unmonitored sections along the entire line is derived.

[0012] S6. The internal force distribution obtained by deduction is superimposed on the reference internal force distribution to calculate the stress distribution along the entire pipeline. The stress distribution along the entire pipeline is compared with the allowable stress of the pipeline material to identify and locate high-risk pipe sections with excessive stress.

[0013] Furthermore, in S1, the method of equating internal pressure and temperature loads to nodal loads based on the principle of virtual work includes:

[0014] Equivalent nodal loads caused by pipeline internal pressure ,

[0015] in, The pressure inside the pipeline. The diameter of the pipe's mid-section. For pipe wall thickness, For the pipe material, Poisson's ratio This represents the cross-sectional area of ​​the pipe.

[0016] Equivalent nodal loads caused by temperature changes ,

[0017] in, The coefficient of linear expansion of the pipe is... The elastic modulus of the pipe. For operating temperature, This is the initial installation temperature.

[0018] Furthermore, in S1, solving the stiffness equation includes: constructing the overall stiffness matrix of the entire pipe segment. Solve the finite element equilibrium equations: The displacement vectors of each node are obtained. Then, the reference axial force of each node is calculated by using the element stiffness matrix. and reference bending moment .

[0019] Furthermore, in S3, the process of fitting the deformation equation of the monitored section based on the plane section assumption using strain data and obtaining the total internal force of the monitored section through integration includes:

[0020] The least squares method is used to fit the plane section deformation equation: ;

[0021] Then, the total axial force at the cross section is calculated by cross section integration. ,

[0022] Total bending moment ,

[0023] in, Coordinates on the cross section Axial strain at the location. For axial average strain , For curvature, The elastic modulus of the pipe. The cross-sectional area of ​​the pipe. Let be the moment of inertia of the cross section.

[0024] Furthermore, in S3, the step of eliminating the reference internal forces caused by temperature and pressure to obtain the internal force components caused by surface displacement includes:

[0025] Components of axial force change caused by surface displacement: ,

[0026] Components of bending moment variation caused by surface displacement: .

[0027] Furthermore, in S5, the interpolation model is constructed based on the RBF interpolation parameters determined by matching, and the radial basis function is used to construct the internal force function along the entire line, the expression of which is: ,

[0028] in, For the axial coordinate of the pipeline, For RBF center, As weight, For radial basis kernel functions, The number of centers.

[0029] Furthermore, in S5, the radial basis kernel function is a Gaussian function, expressed as: ,

[0030] in, The width parameter is used; the objective function expression for minimizing the sum of squared errors is: 2 , n To monitor the number of cross sections.

[0031] Furthermore, in S6, the step of superimposing the derived internal force distribution onto the reference internal force distribution includes:

[0032] Corrected axial force: ,

[0033] Corrected bending moment: .

[0034] Furthermore, in S6, the calculation of the stress distribution along the entire pipeline includes:

[0035] Based on the corrected internal forces, the maximum equivalent stress at each section is calculated using the von Mises stress formula, expressed as follows:

[0036] ,

[0037] in, For the maximum equivalent stress, For axial stress, For circumferential stress, Radial stress, Let the moment of inertia of the cross section be... This refers to the inner diameter of the pipe.

[0038] Furthermore, in S6, when comparing the overall stress distribution with the allowable stress of the pipe material, the maximum equivalent stress is used. Minimum yield strength of pipe material The ratio is used to classify risk levels and identify pipe sections with excessive stress:

[0039] Low risk

[0040] Medium risk.

[0041] This is considered a high-risk activity.

[0042] This is a high-risk pipe section with excessive stress.

[0043] Compared with existing technologies, this invention has the following advantages: by establishing a finite element reference internal force model that only considers the effects of temperature and pressure, the internal pressure and temperature loads are equivalent to nodal loads based on the principle of virtual work and the reference internal force distribution is solved. Then, the internal force components caused by temperature and pressure are accurately removed from the total internal force of the monitored section, and the internal force change components caused by surface displacement alone are obtained. This achieves physical decoupling of internal forces under different loads, lays the foundation for accurate analysis of the additional stress of surface displacement, and greatly improves the accuracy of the inversion of the internal force components of surface displacement.

[0044] Synchronously acquiring strain data from strain sensors and geometric deformation data from IMUs, the system not only fits the deformation equations using strain data and integrates to solve for the total internal force, but also uses the curvature direction detected by the IMU to verify the plane of action of the bending moment. This corrects potential misjudgments of the bending moment vector direction during the strain data fitting process, ensuring that the internal force inversion results are consistent with the actual physical deformation laws of the pipeline. Through the complementarity and verification of multi-source data, the reliability of the internal force inversion of the monitoring section is significantly improved, and the impact of data errors on the calculation results is effectively reduced.

[0045] Based on IMU curvature waveforms, feature vectors are constructed and terrain feature regions are divided using K-means clustering. A deformation pattern database containing mean matrix, covariance matrix, and confidence coefficients is specifically built. By combining Mahalanobis distance and exponential weighting function, accurate pattern matching is achieved. Compared with traditional simple distance matching methods, this method considers the correlation between curvature features and introduces confidence constraints, resulting in stronger matching accuracy and robustness. Simultaneously, the actual deformation pattern of the pipeline is determined, and the associated RBF interpolation parameters are called. Using these as shape constraints, an interpolation model is constructed. Combined with the internal force component sample points of the monitoring section, the objective function of minimizing the sum of squared errors is established using the least squares method to solve for the weight coefficients and deduce the distribution of surface displacement and internal forces along the entire line. At the same time, the interpolation parameters are dynamically adjusted according to the deformation pattern, such as densifying the RBF interpolation center and reducing the Gaussian kernel width parameter in high curvature stress concentration sections, to accurately capture local deformation features. This ensures that the interpolation process strictly follows the actual geometric deformation characteristics of the pipeline, achieving accurate deduction from discrete monitoring points to continuous internal force distribution along the entire line, and significantly improving the engineering reliability and accuracy of internal force calculations for unmonitored sections.

[0046] By superimposing the derived surface displacement internal force distribution along the entire pipeline with the temperature and pressure baseline internal force distribution point by point, the corrected internal force along the entire pipeline considering the influence of surface displacement is obtained. Based on the corrected internal force, the maximum equivalent stress along the entire pipeline is calculated using the von Mises stress formula. The risk level is divided according to the stress ratio, which realizes the continuous reconstruction of the stress field of the entire pipeline and the accurate location of high-risk pipeline sections with stress exceeding the limit. This breaks through the limitation of sparse monitoring points and realizes the continuous and quantitative evaluation of the safety status of the entire pipeline. It can accurately locate pipeline sections with stress exceeding the limit caused by geological disasters such as landslides and subsidence. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating the overall process of this invention.

[0048] Figure 2 This is a schematic diagram of the three-dimensional spatial trajectory of a pipeline, as exemplified by this invention.

[0049] Figure 3 This is a schematic diagram of the strain monitoring point installation of the present invention;

[0050] Figure 4 This is a schematic diagram of the strain gauge layout of the present invention;

[0051] Figure 5 This is the construction step of the radial basis function interpolation model of the present invention;

[0052] Figure 6 This is a diagram showing the simulation results of the continuous distribution of surface displacement internal forces at typical nodes in an embodiment of the present invention.

[0053] Figure 7 This is a diagram showing the security assessment and risk threshold distribution of typical nodes in an embodiment of the present invention. Detailed Implementation

[0054] This application discloses a method for stress inversion and correction of buried pipelines based on pattern matching. Addressing the technical challenges of sparse monitoring points, inaccurate deformation pattern identification, and low overall stress field inversion accuracy in long-distance oil and gas pipelines affected by geological disasters, the method achieves accurate inversion and internal force correction from discrete monitoring points to the continuous stress field along the entire pipeline through physical model decoupling, multi-source data fusion, and pattern matching constraint interpolation. The method steps are as follows:

[0055] S1. Discretize the long-distance pipeline into spatial beam elements. Based on the principle of virtual work, the internal pressure and temperature loads are equivalent to nodal loads. The reference internal force distribution considering only the effects of temperature and pressure is obtained by solving the stiffness equation.

[0056] S2. By using strain sensors installed on the pipeline monitoring section and inertial measurement units (IMUs) mounted inside the pipeline, strain data and geometric deformation data of the monitoring section are collected simultaneously.

[0057] S3. Based on the plane section assumption, the deformation equation of the monitored section is fitted by strain data. The total internal force of the monitored section is obtained by integration. The reference internal force caused by temperature and pressure is eliminated, and the internal force component caused by surface displacement is obtained. The plane of bending moment action is checked by the curvature direction detected by IMU.

[0058] S4. Based on the curvature waveform obtained by IMU, construct curvature feature vectors. Divide the pipeline into several terrain feature areas by K-means clustering, and construct a database of typical deformation patterns for each feature area, including mean matrix, covariance matrix and confidence coefficient. Calculate the Mahalanobis distance between the measured curvature feature vector and each typical deformation pattern. Combine with pre-set confidence conditions to screen candidate patterns. Determine the pattern matching weight through an exponential weighting function, select the optimal matching target deformation pattern, and determine the radial basis function (RBF) interpolation parameters.

[0059] S5. Using the internal force components of the monitored section as samples, an interpolation model is constructed based on the RBF interpolation parameters determined by matching. The RBF weight coefficients are solved by establishing the objective function of minimizing the sum of squared errors through the least squares method, and the internal force distribution caused by surface displacement of the unmonitored sections of the entire pipeline is derived.

[0060] S6. The internal force distribution obtained by deduction is superimposed on the reference internal force distribution to calculate the stress distribution along the entire pipeline. The stress distribution along the entire pipeline is compared with the allowable stress of the pipeline material to identify and locate high-risk pipe sections with excessive stress.

[0061] This embodiment uses a long-distance buried oil and gas pipeline of X70 steel grade and an outer diameter of 1016 mm as the research object to verify the feasibility and accuracy of the method. Its three-dimensional spatial trajectory is as follows: Figure 2 As shown, the methodology is as follows: First, a baseline internal force model of the pipeline is constructed by stripping away temperature and pressure loads. Then, the internal force components of the monitored cross-section surface displacement are inverted using multi-source measured data. Next, the internal force is extrapolated along the entire pipeline using mode matching as a constraint. Finally, the actual stress field along the entire pipeline is obtained through superposition and correction, and risk identification is completed. Each step is progressive and interconnected. The previous step provides the data foundation and technical support for the next step, and the next step iterates and applies the results of the previous step, forming a complete technical closed loop from model construction to single-point inversion, full-line extrapolation, stress correction, and risk identification. The detailed implementation of each step is as follows:

[0062] S1. Discretize the long-distance pipeline into spatial beam elements. Based on the principle of virtual work, the internal pressure and temperature loads are equivalent to nodal loads. By solving the stiffness equation, the baseline internal force distribution considering only the effects of temperature and pressure is obtained. This step is the basic modeling stage of the entire stress inversion and correction method. The core purpose is to isolate the influence of the two conventional loads, temperature and internal pressure, during pipeline operation and construct a baseline internal force distribution model that is only affected by temperature and pressure. This provides an accurate reference for the subsequent separation of additional internal forces caused by surface displacement. This solves the problem of coupling temperature and pressure with surface displacement internal forces in traditional methods, making it impossible to analyze additional stresses separately. The details are as follows:

[0063] Pipeline discretization and spatial model construction: Based on the actual geometric parameters (outer diameter) of the long-distance pipeline to be monitored. , inner diameter Wall thickness Mid-surface diameter Moment of inertia of cross section Cross-sectional area ) and material properties (elastic modulus) Poisson's ratio Coefficient of thermal expansion The minimum yield strength is used to construct a three-dimensional spatial model of the pipeline system. Two-node spatial straight beam elements are used to discretize the pipeline into several elements, with each node having 6 degrees of freedom: linear displacement along the x, y, and z axes and angular displacement around the x, y, and z axes. This fully characterizes the tensile, bending, and torsional deformation characteristics of the pipeline. In this embodiment, a 100m pipeline is discretized into 10 elements and 11 nodes. Node 0 is set as a fixed end, constraining all degrees of freedom, while node 10 is set as a sliding end, constraining only the degree of freedom perpendicular to the pipeline axis. This simulates the actual constraint conditions of buried pipelines. The basic information and material properties of the selected pipeline are shown in Table 1 below.

[0064] Table 1 Basic Information and Material Properties of Pipelines

[0065]

[0066] Thermo-pressure loads are equivalent to nodal loads: Based on the principle of virtual work, the circumferential stress caused by the internal pressure of the pipe and the axial thermal stress caused by temperature changes are transformed into equivalent load vectors that can be directly applied to the nodes of the beam element. The specific calculation formula strictly follows the settings of the initial draft of the invention.

[0067] Equivalent nodal loads caused by internal pipeline pressure: ,

[0068] in, The pressure inside the pipeline. The diameter of the pipe's mid-section. For pipe wall thickness, For the pipe material, Poisson's ratio Let be the cross-sectional area of ​​the pipe. In this embodiment, we substitute . , , , , Calculated It is distributed to all nodes in the form of a uniformly distributed axial force.

[0069] Equivalent nodal loads caused by temperature changes: ,

[0070] in, The coefficient of linear expansion of the pipe is... The elastic modulus of the pipe. For operating temperature, As the initial installation temperature, this embodiment substitutes... , , Calculated Similarly, it is applied to all nodes in the form of a uniformly distributed axial force.

[0071] Solving the stiffness equation yields the baseline internal force distribution: Based on the stiffness matrix formula for a two-node spatial straight beam element, the overall stiffness matrix of the entire pipe segment is constructed. Establish the finite element equilibrium equations: The displacement vectors of each node are obtained by solving. Then, the reference axial force of each node is calculated by using the element stiffness matrix. and reference bending moment This results in a baseline internal force distribution for the entire pipeline considering only the effects of temperature and pressure. The calculation results of the baseline internal forces at typical nodes in this embodiment are shown in Table 2 below:

[0072] Table 2 Calculation results of reference internal forces at typical nodes

[0073]

[0074] S2. Strain and geometric deformation data of the monitoring section are simultaneously acquired using strain sensors deployed at the pipeline monitoring section and an inertial measurement unit (IMU) mounted inside the pipeline. This step is the actual data acquisition stage, which is the data source for subsequent internal force inversion and mode matching. The core purpose is to obtain multi-source synchronous actual measurement data of the monitoring section, solving the problems of traditional methods that rely only on single strain data, have insufficient information dimensions, and are easily affected by measurement errors. At the same time, it provides strain and geometric deformation data support for internal force inversion in S3 and mode matching in S4, respectively, as detailed below:

[0075] Monitoring device deployment: Monitoring sections are set up at key locations such as sections prone to geological disasters and sections with potential stress concentration in the pipeline. Multiple strain sensors are evenly distributed around the circumference of each monitoring section. In this embodiment, four strain sensors are evenly distributed around the circumference of each section. Figure 4 As shown, the sensors are positioned at 0°, 90°, 180°, and 270° respectively, with a measurement accuracy of ±1με. They are used to collect axial strain data of the cross section. A high-precision inertial measurement unit (IMU) is installed inside the pipe and operates synchronously with the internal detector, with a sampling frequency of not less than 100Hz.

[0076] Multi-source data synchronous acquisition: Within a preset time period, strain sensor data (axial strain values ​​at each measuring point) of the monitoring section are acquired synchronously. The strain data and geometric deformation data (the rate of change of the pipe axis inclination, curvature information, and continuous curvature waveform of the pipe centerline) are used to ensure temporal synchronization and spatial matching of the two types of data. This means that strain data and geometric deformation data at the same monitoring time and the same pipe section correspond one-to-one. In this embodiment, four monitoring sections are set at x=20m, 40m, 60m, and 80m, corresponding to nodes 2, 4, 6, and 8, respectively. Figure 3 As shown, strain and IMU data were acquired synchronously, where x=20m, 40m, and 60m are inversion sample points, and x=80m is an independent verification point.

[0077] S3. Based on the plane section assumption, the deformation equation of the monitored section is fitted using strain data. The total internal force of the monitored section is obtained through integration. The reference internal force caused by temperature and pressure is eliminated, and the internal force component caused by surface displacement is obtained. The curvature direction detected by IMU is used to verify the plane of action of bending moment. This step is the single-point internal force inversion stage of the monitored section. It is the core bridge connecting the measured data and the additional stress of surface displacement. Based on the reference internal force distribution of S1 and the multi-source measured data of S2, the internal force component caused by surface displacement alone of the monitored section is accurately inverted through the process of fitting-integration-elimination-verification. The logic between the steps is as follows: first, the total internal force is calculated from the strain data; then, the reference internal force of temperature and pressure in S1 is eliminated to obtain the internal force component of surface displacement; finally, the results are verified and corrected using IMU data, as detailed below:

[0078] Fitting the deformation equation and integrating to obtain the total internal force: Based on the plane section assumption, the least squares method is used to fit the strain data collected by S2 into the deformation equation of the monitoring section. ,

[0079] in, Coordinates on the cross section Axial strain at the location. For axial average strain , Let be the curvature of the cross section in the y and z directions;

[0080] Then, the total axial force and total bending moment of the section are calculated by section integration:

[0081] Total axial force: ;

[0082] Total bending moment: ,in, Let be the moment of inertia of the pipe cross section.

[0083] In this embodiment, x=40m, and the monitoring section is fitted to obtain... , , Calculated , .

[0084] The surface displacement internal force components are obtained by subtracting the reference axial force corresponding to the same monitoring section in S1 from the total internal force calculated above. and reference bending moment The internal force change component caused solely by surface displacement is obtained, and the calculation formula is as follows:

[0085] Axial force variation components: ;

[0086] Components of bending moment variation: .

[0087] In this embodiment, x=40m, and the internal force data of the monitored section are calculated based on the reference internal force data in Table 2. , This achieves precise separation of the internal force components of surface displacement from the internal forces of temperature and pressure.

[0088] IMU Curvature Direction Verification of Bending Moment Action Plane: Using the direction of change in pipe axis curvature and the rate of change in inclination detected by the IMU in S2, the actual action plane of the bending moment caused by surface displacement is determined. The bending moment vector direction calculated from the strain data is verified to correct misjudgments of the bending moment direction caused by strain sensor measurement and fitting errors, ensuring... The sign of the curve is consistent with the actual physical deformation law of the pipeline. In this embodiment, the IMU detected that the rate of change of the inclination angle of the x=40m section about the y-axis is 1.2° / m and about the z-axis is 0.9° / m. The verification confirmed that the plane of action of the bending moment vector of this section is the xy plane. The positive and negative signs are consistent with the actual deformation.

[0089] S4. Based on the curvature waveform acquired by the IMU, a curvature feature vector is constructed. K-means clustering is used to divide the pipeline along the pipeline into several terrain feature regions, and a database of typical deformation patterns, including mean matrix, covariance matrix, and confidence coefficients, is constructed for each feature region. The Mahalanobis distance between the measured curvature feature vector and each typical deformation pattern is calculated. Candidate patterns are screened based on pre-set confidence conditions. The pattern matching weight is determined using an exponential weighting function, and the optimal matching target deformation pattern is selected to determine the radial basis function (RBF) interpolation parameters. This step, the deformation pattern matching stage, is a crucial constraint step in realizing the transition from single-point inversion to full-line extrapolation. Using the IMU curvature waveform data from S2 as input and the pre-constructed deformation pattern database as a reference, it provides shape constraints and parameter basis for the RBF interpolation model construction in S5. This solves the problems of traditional interpolation methods lacking physical constraints and the exponential results being disconnected from the actual pipeline deformation. Specifically:

[0090] Deformation pattern database pre-construction: The database stores typical deformation patterns corresponding to different types of surface displacement (landslides, subsidence, fault displacement, etc.). Each deformation pattern includes at least a pattern label, a curvature waveform template, and an associated set of RBF interpolation parameters (kernel function width parameter). Interpolation center location distribution parameters, number of centers In this process, the RBF interpolation parameters are pre-optimized based on the mechanical and geometric characteristics of different deformation modes, and then extracted based on the curvature waveform detected by the IMU. y Directional curvature , z Directional curvature Wavelength of curvature change L c Three core features to construct a curvature feature vector K =[ L c ] T The K-means clustering algorithm was used to perform cluster analysis on the curvature feature vectors of the entire pipeline, dividing the pipeline into several terrain feature regions with similar deformation characteristics. Each major category of surface displacement was further subdivided, and for each subdivided terrain feature region, a mean matrix was constructed based on historical IMU detection data. M i Covariance matrix Σ i Confidence coefficient C i The typical deformation mode sub-library stores the optimized RBF interpolation parameters into the corresponding sub-library.

[0091] Mahalanobis distance calculation and candidate pattern selection: The continuous curvature waveform of the pipe centerline acquired by the IMU in S2 is transformed into a measured curvature feature vector. K j Calculate its Mahalanobis distance with each typical deformation mode in the deformation mode database. The calculation formula is as follows: Select the option that satisfies the Mahalanobis distance. And confidence coefficient C i Patterns with a value >0.85 are used as candidate matching patterns to effectively filter out invalid patterns and improve the reliability of matching results.

[0092] Exponentially weighted matching: The matching weight of each candidate pattern is calculated using an exponentially weighted function, as shown in the formula: The candidate pattern with the highest weight is selected as the target deformation pattern for optimal matching.

[0093] Retrieve associated RBF interpolation parameters: Based on the target deformation pattern obtained through matching, directly retrieve the associated radial basis function (RBF) interpolation parameters from the database, including the Gaussian kernel width parameter. Interpolation Center Location distribution, number of centers This provides direct parameter basis for the construction of the S5 interpolation model. In this embodiment, the high curvature deformation mode caused by local settlement is identified by matching at x=30~50m, and the RBF parameters corresponding to this mode are retrieved: high curvature section Encrypted interpolation center, low curvature section Sparse interpolation center.

[0094] S5. Using the internal force components of the monitored sections as samples, construct an interpolation model based on the matched and determined RBF interpolation parameters. Solve for the RBF weighting coefficients by establishing an objective function that minimizes the sum of squared errors using the least squares method. This allows for the deduction of the internal force distribution caused by surface displacement at unmonitored sections along the entire pipeline. The steps are as follows: Figure 5 As shown, this step is the stage of extrapolating the internal forces of surface displacement along the entire line. It is the core implementation link to realize the transition from discrete monitoring points to the continuous entire line, as detailed below:

[0095] Constructing the RBF interpolation model: The internal force function along the entire pipe is constructed using radial basis functions.

[0096] ,

[0097] in, For the axial coordinate of the pipeline, The RBF interpolation center determined by S4, These are the weighting coefficients. For radial basis kernel functions, The number of interpolation centers determined for S4;

[0098] The kernel function chosen is the Gaussian function, and its expression is: ,

[0099] in, The Gaussian kernel width parameter is retrieved from S4. This represents the radial distance between the interpolation point and the interpolation center.

[0100] Input sample points and solve for weighting coefficients: Invert the surface displacement internal force components of the monitoring section obtained from S3 ( , () as known sample points for the interpolation model, using the pipe axial coordinates As input variables, and surface displacement internal force components as training target values, establish an objective function that minimizes the sum of squared errors: The RBF weighting coefficients are obtained by solving the objective function using the least squares method. This ensures that the calculation results of the interpolation model at the sample points are highly consistent with the measured inversion results.

[0101] Deducing the distribution of surface displacement and internal forces along the entire pipeline: Calculate the axial coordinates of all sections along the entire pipeline. Substituting the weighted RBF interpolation model, the axial force variation components caused by ground displacement at any location along the entire pipeline are calculated. and bending moment variation components This allows for the continuous deduction of the internal force components of surface displacement at unmonitored sections. The deduced surface displacement internal force components for typical nodes in this embodiment are shown in Table 3 below. Figure 6 As shown:

[0102] Table 3. Results of Surface Displacement Internal Force Components at Typical Nodes

[0103]

[0104] S6. Superimpose the derived internal force distribution onto the baseline internal force distribution to calculate the overall stress distribution of the pipeline. Compare the overall stress distribution with the allowable stress of the pipeline material to identify and locate high-risk pipe sections with excessive stress. This step is the overall internal force correction and risk identification stage, and is the final implementation stage. Based on the baseline internal force distribution in S1 and the overall surface displacement internal force distribution in S5, the actual stress field of the entire pipeline is obtained through the process of superposition correction, stress calculation, and risk classification, and high-risk pipe sections are identified. The details are as follows:

[0105] Superimposed to obtain the corrected internal forces along the entire pipeline: The distribution of surface displacement internal forces along the entire pipeline derived from S5 ( , ), compared with the reference internal force distribution obtained by S1 considering only the effects of temperature and pressure ( , By superimposing the values ​​point by point, the corrected internal forces along the entire pipeline, considering the combined effects of surface displacement and temperature and pressure, are obtained. The calculation formula is as follows:

[0106] Corrected axial force: ;

[0107] Corrected bending moment: .

[0108] Calculate the maximum equivalent stress along the entire pipeline: Based on the modified internal forces mentioned above, the maximum equivalent stress at each section along the entire pipeline is calculated using the von Mises stress formula. The calculation formula is as follows:

[0109] ,

[0110] in, For axial stress, For circumferential stress, Radial stress, The outer diameter of the pipe. Given the inner diameter of the pipe, the actual stress of each section is accurately calculated by comprehensively considering the coupling effects of axial force, bending moment, and internal pressure.

[0111] Risk level classification and high-risk section identification: The calculated maximum equivalent stress across the entire line The minimum yield strength of the pipe material In comparison, using stress ratio To classify risk levels according to standards, high-risk pipe sections with excessive stress are identified and located. The risk classification standards are as follows:

[0112] Low risk: ;

[0113] Medium risk: ;

[0114] Higher risk: ;

[0115] High risk: (Stress exceeded).

[0116] The security assessment results of typical nodes in this embodiment are shown in Table 4 below. Figure 7 As shown, the stress ratio ranges from 0.41 to 0.59, which are all low to medium risk, with no high-risk pipe sections exceeding the stress limit.

[0117] Table 4. Security Assessment Results of Typical Nodes

[0118]

[0119] Method accuracy verification: To verify the accuracy of the method of this invention, a method of inversion at 3 monitoring points and independent verification at 1 monitoring point was adopted. The stress result of the verification point (x=80m) obtained by the inversion of this invention was compared with the measured stress result of the verification point (not involved in the correction process): the axial strain of the cross section at x=80m was measured by the strain sensor to be 5.49×10⁻ 4The measured axial stress at this cross-section was calculated to be 115.29 MPa based on Hooke's Law. Combined with the circumferential stress, the equivalent stress at this verification point was calculated to be 201.7 MPa according to the von Mises stress formula. The relative error between this and the inverted value (201.8 MPa) in this invention is only 0.05%, and the error is controlled within the allowable range of engineering (±2%), which is far lower than the allowable error in engineering. This proves that this method can accurately realize the inversion and correction of the continuous stress field from discrete monitoring points to the entire pipeline. At the same time, this method can complete the reconstruction of the stress field and risk identification of the entire pipeline using only the limited measured data of conventional monitoring equipment, which greatly reduces the cost and difficulty of on-site monitoring and implementation. It is applicable to the safety monitoring and assessment of various long-distance buried oil and gas pipelines affected by geological disasters.

[0120] The above specific 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 examples, 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 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 stress inversion and correction of buried pipelines based on pattern matching, characterized in that, include: S1. Discretize the long-distance pipeline into spatial beam elements. Based on the principle of virtual work, the internal pressure and temperature loads are equivalent to nodal loads. The reference internal force distribution considering only the effects of temperature and pressure is obtained by solving the stiffness equation. S2. By using strain sensors installed on the pipeline monitoring section and inertial measurement units (IMUs) mounted inside the pipeline, strain data and geometric deformation data of the monitoring section are collected simultaneously. S3. Based on the plane section assumption, the deformation equation of the monitored section is fitted by strain data. The total internal force of the monitored section is obtained by integration. The reference internal force caused by temperature and pressure is eliminated, and the internal force component caused by surface displacement is obtained. The plane of bending moment action is checked by the curvature direction detected by IMU. S4. Construct the IMU curvature waveform as a feature vector, divide the terrain feature area by K-means clustering, establish a deformation pattern database, filter candidate patterns by confidence conditions, determine the matching weight by exponential weighting, select the optimal target deformation pattern, and output RBF interpolation parameters. S5. Using the internal force components of the monitored sections as samples, an interpolation model is constructed using RBF interpolation parameters. The weight coefficients are solved by the least squares method, and the surface displacement internal force distribution of the unmonitored sections along the entire line is derived. S6. The internal force distribution obtained by deduction is superimposed on the reference internal force distribution to calculate the stress distribution along the entire pipeline. The stress distribution along the entire pipeline is compared with the allowable stress of the pipeline material to identify and locate high-risk pipe sections with excessive stress.

2. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 1, characterized in that: In S1, the method of equating internal pressure and temperature loads to nodal loads based on the principle of virtual work includes: Equivalent nodal loads caused by pipeline internal pressure , in, The pressure inside the pipeline. The diameter of the pipe's mid-section. For pipe wall thickness, For the pipe material, Poisson's ratio This represents the cross-sectional area of ​​the pipe. Equivalent nodal loads caused by temperature changes , in, The coefficient of linear expansion of the pipe material. The elastic modulus of the pipe. For operating temperature, This is the initial installation temperature.

3. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 2, characterized in that, In S1, solving the stiffness equation includes: constructing the global stiffness matrix for the entire pipe segment. Solve the finite element equilibrium equations: The displacement vectors of each node are obtained. Then, the reference axial force of each node is calculated by using the element stiffness matrix. and reference bending moment .

4. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 3, characterized in that: In S3, the deformation equation of the monitored section is fitted from strain data based on the plane section assumption, and the total internal force of the monitored section is obtained by integration, including: The least squares method is used to fit the plane section deformation equation: ; Then, the total axial force at the cross section is calculated by cross section integration. , Total bending moment , in, Coordinates on the cross section Axial strain at the location. For axial average strain , For curvature, The elastic modulus of the pipe. The cross-sectional area of ​​the pipe. Let be the moment of inertia of the cross section.

5. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 4, characterized in that: In S3, the step of eliminating the reference internal forces caused by temperature and pressure to obtain the internal force components caused by surface displacement includes: Components of axial force change caused by surface displacement: , Components of bending moment variation caused by surface displacement: .

6. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 5, characterized in that: In S5, the interpolation model is constructed using RBF interpolation parameters, and the internal force function along the entire line is constructed using radial basis functions. The expression is as follows: , in, For the axial coordinates of the pipeline, As the RBF center, As weight, For radial basis kernel functions, The number centered on.

7. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 6, characterized in that: In S5, the radial basis kernel function is a Gaussian function, and its expression is: , in, For width parameters; The objective function expression for minimizing the sum of squared errors of the interpolation model using the Gaussian function as the radial basis kernel function is: 2 ,in, n To monitor the number of cross sections.

8. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 7, characterized in that: In S6, the step of superimposing the derived internal force distribution onto the reference internal force distribution includes: Corrected axial force: , Corrected bending moment: .

9. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 8, characterized in that: In S6, the calculated stress distribution along the entire pipeline includes: Based on the corrected internal forces, the maximum equivalent stress at each section is calculated using the von Mises stress formula, expressed as follows: ， in, For the maximum equivalent stress, For axial stress, For circumferential stress, Radial stress, Let the moment of inertia of the cross section be... This refers to the inner diameter of the pipe.

10. The method for stress inversion and correction of buried pipelines based on pattern matching according to claim 9, characterized in that: In S6, when comparing the overall stress distribution with the allowable stress of the pipe material, the maximum equivalent stress is used. Minimum yield strength of pipe material The ratio is used to classify risk levels and identify pipe sections with excessive stress: Low risk Medium risk. This is considered a high-risk activity. This is a high-risk pipe section with excessive stress.

Images