A method, apparatus, equipment and storage medium for assessing the integrity of hydrogen transportation pipelines.

Abstract

This application discloses a method, apparatus, equipment, and storage medium for assessing the integrity of hydrogen transportation pipelines, relating to the field of oil and gas pipeline safety technology. The method establishes a pipeline simulation model incorporating pre-existing cracks. Based on this model, it simulates the diffusion and accumulation of hydrogen atoms within the pipe wall to calculate the stress field distribution and effective stress intensity factor range at the crack tip under different operating conditions, thus obtaining the fatigue crack propagation rate. Then, using hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio as input features, and the fatigue crack propagation rate as the model label, a machine learning regression model is trained. The real-time operating conditions of the pipeline to be assessed are then input into the trained machine learning regression model to predict the fatigue crack propagation rate. The fracture ratio and load ratio of the pipeline under the current crack state are calculated and mapped onto a failure assessment graph, enabling real-time and accurate assessment of the current integrity status of the hydrogen transportation pipeline.

Description

Technical Field

[0001] This application relates to the field of oil and gas pipeline safety technology, and in particular to a method, apparatus, equipment and storage medium for assessing the integrity of hydrogen pipelines. Background Technology

[0002] Utilizing natural gas pipelines for hydrogen-blended transportation (HCNG) has become an important approach to achieving large-scale, low-cost hydrogen energy transportation. However, the introduction of hydrogen presents unprecedented challenges to pipeline safety. Current research on the hydrogen embrittlement properties of high-strength pipeline steels such as X80, X52, X60, X65, X70, X90, and X100, as well as conventional natural gas pipeline steels such as L360, mainly relies on slow strain rate tensile tests (SSRT) or autoclave fatigue tests. These tests have extremely high equipment requirements, pose safety hazards such as hydrogen leaks and explosions, and are time-consuming and costly. They also cannot exhaustively cover various pressure fluctuations and hydrogen blending ratios that may occur in actual pipeline operation, resulting in significant limitations in physical testing. Furthermore, the diffusion of hydrogen atoms within the steel is driven by stress gradients, while accumulated hydrogen, in turn, reduces the material's fracture toughness. Existing engineering assessment methods often decouple these two processes, making it difficult to accurately describe the true mechanical behavior of the "high stress-high hydrogen concentration" region at the crack tip. This lack of multi-field coupling quantitative assessment results in low assessment efficiency. Summary of the Invention

[0003] This application provides a method, apparatus, equipment, and storage medium for assessing the integrity of hydrogen pipelines, addressing issues such as high testing risk, complex coupling mechanisms, and low assessment efficiency in existing technologies. To solve this problem, this application utilizes the high-throughput characteristics of finite element analysis software to replace physical testing, generating massive amounts of hydrogen-force coupling characteristic data formed by stress field distribution and effective stress intensity factor range values. Then, it uses the nonlinear fitting capability of machine learning regression models to establish a prediction model. Finally, it combines the commonly used Failure Assessment Diagram (FAD) in engineering to accurately determine the integrity of the pipeline to be assessed.

[0004] In a first aspect, embodiments of this application provide a method for assessing the integrity of a hydrogen transportation pipeline, including: A pipeline simulation model containing pre-existing cracks was constructed using finite element analysis software, and the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model were set. The diffusion and accumulation of hydrogen atoms in the pipe wall are simulated in the pipeline simulation model, and the stress field distribution and effective stress intensity factor range value at the crack tip are calculated under different working conditions. The local hydrogen concentration at the crack tip is determined based on the stress field distribution. The fatigue crack propagation rate is calculated by modifying the Paris formula based on the local hydrogen concentration and the effective stress intensity factor range. The fatigue crack propagation rate is then used as the model label. The hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio are preprocessed to obtain the model input features; A training dataset is constructed based on the model input features and the model labels, and the machine learning regression model is trained using the training dataset. A target loss function is constructed by using the fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model, thereby generating the trained machine learning regression model. The real-time operating conditions of the pipeline to be evaluated are obtained and input into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline to be evaluated. The fracture ratio of the pipeline under evaluation in the current crack state is determined based on the predicted fatigue crack propagation rate, and the load ratio of the pipeline under evaluation in the current crack state is calculated. The fracture ratio and load ratio of the pipeline under evaluation in the current crack state are mapped to the failure evaluation diagram to obtain the integrity evaluation result of the pipeline under evaluation.

[0005] In conjunction with the first aspect, in one possible implementation, the simulation of the diffusion and accumulation of hydrogen atoms within the pipe wall in the pipeline simulation model, and the calculation of the stress field distribution and effective stress intensity factor range at the crack tip under different operating conditions, include: The stress field distribution at the crack tip of the pipeline simulation model under different working conditions is extracted, and the hydrostatic stress gradient is calculated based on the extracted stress field distribution. The hydrogen atom distribution concentration update subroutine is invoked to update the hydrogen atom distribution concentration through the hydrostatic stress gradient, thereby simulating the diffusion and aggregation process of hydrogen atoms within the pipe wall. The post-processing script is invoked to calculate the stress field distribution and obtain the effective stress intensity factor range value.

[0006] In conjunction with the first aspect, in one possible implementation, the modified Paris formula is:

[0007] In the formula, a represents the crack depth, and N represents the number of fatigue load cycles. denoted by , C(H) represents the fatigue crack propagation rate, C(H) represents the fatigue crack propagation coefficient related to the local hydrogen concentration at the crack tip, m(H) represents the fatigue crack propagation exponent related to the local hydrogen concentration at the crack tip, and ΔK represents the effective stress intensity factor range value at the crack tip.

[0008] In conjunction with the first aspect, in one possible implementation, the preprocessing of the hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio to obtain model input features includes: Based on the linear elastic fracture mechanics theory, the pressure amplitude inside the pipeline and the crack depth ratio are calculated to obtain the benchmark effective stress intensity factor; The partial pressure of hydrogen is processed by taking the square root to obtain the square root feature, and the square root feature is mapped to the equivalent concentration benchmark value. The reference effective stress intensity factor, the equivalent concentration reference value, and the crack depth ratio are subjected to Z-Score normalization to obtain the model input features.

[0009] In conjunction with the first aspect, in one possible implementation, the target loss function is:

[0010] In the formula, The total loss value calculated by the objective loss function is represented; MSE represents the mean square error between the predicted fatigue crack propagation rate and the model label. This represents the first weighting coefficient of the pre-defined fracture mechanics prior penalty term. The second weighting coefficient represents the pre-defined fracture mechanics prior penalty term, and all are constants greater than 0; P represents the predicted fatigue crack propagation rate; ΔP represents the pipe internal pressure amplitude in the model input features; H2 This represents the partial pressure of hydrogen in the model input features; This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the magnitude of the internal pressure in the pipe. This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the hydrogen partial pressure.

[0011] In conjunction with the first aspect, in one possible implementation, determining the fracture ratio of the pipeline under evaluation in the current crack state based on the predicted fatigue crack propagation rate and calculating the load ratio of the pipeline under evaluation in the current crack state includes: The current crack depth of the pipeline to be evaluated is updated based on the fatigue load cycle step and the predicted fatigue crack propagation rate. Based on the updated current crack depth and the current operating internal pressure of the pipeline to be evaluated, determine the stress intensity factor and reference stress of the pipeline to be evaluated in the current crack state. The stress intensity factor is compared with the material fracture toughness of the pipeline under the current hydrogen partial pressure to obtain the fracture ratio under the current crack state; The load ratio under the current crack state is obtained by comparing the reference stress with the yield strength of the material of the pipeline to be evaluated.

[0012] In conjunction with the first aspect, in one possible implementation, mapping the fracture ratio and load ratio of the pipeline under evaluation in the current crack state to a failure evaluation map to obtain the integrity evaluation result of the pipeline under evaluation includes: The fracture ratio and load ratio of the pipeline under the current crack state are mapped onto a failure assessment diagram constructed with the fracture ratio as the vertical axis and the load ratio as the horizontal axis to obtain the failure assessment point. When the failure assessment point is located inside the safety boundary curve of the failure assessment diagram, the pipeline to be assessed is determined to be in a safe state, and the corresponding integrity assessment result is intact. If the failure assessment point is not inside the safety boundary curve, then calculate the shortest geometric distance from the failure assessment point to the safety boundary curve; When the shortest geometric distance of the failure assessment point is less than the preset safety margin threshold and the fracture ratio of the failure assessment point is greater than the preset limit value, the pipeline to be assessed is determined to be in a failure state dominated by hydrogen-induced embrittlement fracture, and the corresponding integrity assessment result is hydrogen-induced embrittlement fracture. When the failure assessment point is outside the safety boundary curve, and the shortest geometric distance of the failure assessment point is not less than the preset safety margin threshold, or the fracture ratio of the failure assessment point is not greater than the preset limit value, the pipeline to be assessed is determined to be in a failure state dominated by plastic collapse, and the corresponding integrity assessment result is plastic collapse.

[0013] Secondly, embodiments of this application provide a hydrogen pipeline integrity assessment device, comprising: The simulation model construction module is used to construct a pipeline simulation model containing pre-existing cracks using finite element analysis software, and to set the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model. The input feature acquisition module is used to simulate the diffusion and accumulation process of hydrogen atoms in the pipe wall in the pipeline simulation model, and to calculate the stress field distribution and effective stress intensity factor range value at the crack tip under different working conditions. The hydrogen atom simulation module is used to determine the local hydrogen concentration at the crack tip based on the stress field distribution. Based on the local hydrogen concentration and the effective stress intensity factor range, the fatigue crack propagation rate is calculated by modifying the Paris formula, and the fatigue crack propagation rate is used as the model label. The model label determination module is used to preprocess the hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio to obtain model input features; The model training module is used to construct a training dataset based on the model input features and the model labels, and to train the machine learning regression model using the training dataset. A target loss function is constructed by using a fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model and generate the trained machine learning regression model. The model recognition module is used to obtain the real-time operating conditions of the pipeline to be evaluated and input the real-time operating conditions into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline to be evaluated. The pipeline parameter calculation module is used to determine the fracture ratio of the pipeline under the current crack state based on the predicted fatigue crack propagation rate, and to calculate the load ratio of the pipeline under the current crack state. The pipeline integrity assessment module is used to map the fracture ratio and load ratio of the pipeline under the current crack state to the failure assessment diagram to obtain the integrity assessment result of the pipeline under the assessment.

[0014] Thirdly, embodiments of this application provide an apparatus comprising: a processor; a memory for storing processor-executable instructions; wherein, when the processor executes the executable instructions, it implements the method as described in the first aspect or any possible implementation of the first aspect.

[0015] Fourthly, embodiments of this application provide a non-volatile computer-readable storage medium, the non-volatile computer-readable storage medium including storage for storing a computer program or instructions that, when executed, cause the method described in the first aspect or any possible implementation of the first aspect to be implemented.

[0016] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: This application embodiment embeds the hydrogen diffusion equation into finite element analysis software to establish a pipeline simulation model containing pre-existing cracks, and divides the crack tip with a high-precision mesh. This model can simulate the diffusion and accumulation process of hydrogen atoms within the pipe wall, calculate the stress field distribution and effective stress intensity factor range at the crack tip under different operating conditions. Then, using the modified Paris formula for the hydrogen embrittlement influence factor, the stress field distribution and effective stress intensity factor range are calculated to obtain the fatigue crack propagation rate. Next, using hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio as input features, and crack propagation rate as the model label, a training dataset is constructed and a machine learning regression model is trained. After training, this model can output prediction results in milliseconds without needing to run complex finite element calculations again. Finally, the real-time operating conditions of the pipeline to be evaluated are input into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline under evaluation, and calculate the fracture ratio and load ratio of the pipeline under evaluation in the current crack state, mapping them to a failure assessment diagram to accurately assess the integrity of the pipeline under evaluation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application or the prior art 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.

[0018] Figure 1 A flowchart of a hydrogen pipeline integrity assessment method provided in this application embodiment; Figure 2 This is a schematic diagram of a pipeline simulation model provided in an embodiment of this application; Figure 3 A schematic diagram of a failure assessment diagram provided in an embodiment of this application; Figure 4 This is a schematic diagram of a hydrogen pipeline integrity assessment device provided in an embodiment of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] The following description of some technologies involved in the embodiments of this application is provided to aid understanding and should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and brevity, some descriptions of well-known functions and structures are omitted in the following description.

[0021] Figure 1 This is a flowchart illustrating a hydrogen pipeline integrity assessment method provided in this application embodiment. It is merely one execution order shown in this embodiment and does not represent the only execution order for a hydrogen pipeline integrity assessment method. Where the final result can be achieved, [further details are needed]. Figure 1 The steps shown can be performed in parallel or in reverse order.

[0022] A simulation model of the pipeline containing pre-existing cracks was established using finite element analysis software. For example... Figure 2 The diagram shown is a schematic of the pipeline simulation model established in this embodiment.

[0023] Among them, pre-fabricated cracks refer to pipeline cracks that are pre-set for establishing a pipeline simulation model.

[0024] Specifically, a pipeline simulation model was established using ABAQUS (a finite element analysis software). To simulate real working conditions, the crack depth ratio, pipeline internal pressure amplitude, and hydrogen partial pressure (or hydrogen concentration boundary conditions on the pipe wall) were set as variable input parameters.

[0025] The crack depth ratio (a / t) is set with four gradients (e.g., 0.1, 0.3, 0.5, and 0.7) to represent different stages from crack initiation to deep propagation. Indicates the crack depth. This indicates the pipe wall thickness.

[0026] Pipeline internal pressure amplitude (ΔP): simulates the fluctuation of the pressure transported in the pipeline. It is exemplarily set to three working conditions: 4MPa, 6MPa, and 8MPa, with a stress ratio of 0.1.

[0027] Hydrogen partial pressure: Hydrogen atom diffusion is simulated using a heat and mass transfer analogy method. Boundary conditions for hydrogen concentration are set on the inner wall of the pipe, such as setting the corresponding hydrogen partial pressure P. H2 The pressure was set to 0 MPa (pure natural gas), 0.1 MPa (low proportion of hydrogen blending), and 1.0 MPa (pure hydrogen or high proportion of hydrogen blending), respectively.

[0028] After obtaining the pipeline simulation model, the diffusion and accumulation process of hydrogen atoms in the pipeline wall is simulated in the pipeline simulation model, and the stress field distribution and effective stress intensity factor range value at the crack tip are calculated under different working conditions.

[0029] Specifically, a Python script was written to control the finite element solver for batch calculations. During the calculation process, the stress field distribution at the crack tip of the pipeline simulation model under different operating conditions was first extracted. Based on the extracted stress field distribution, the hydrostatic stress gradient was calculated. Then, a hydrogen atom distribution concentration update subroutine was called to update the hydrogen atom distribution concentration according to the hydrostatic stress gradient, thus simulating the diffusion and accumulation process of hydrogen atoms within the pipe wall. After the calculation was completed, a post-processing script was called to read the simulation-generated history output file (.odb file) and automatically extract the maximum effective stress intensity factor K at the crack front node during a complete internal pressure fluctuation cycle. max and the minimum effective stress intensity factor K min And according to K max and K min Calculate the effective stress intensity factor range ΔK. Simultaneously, extract the average hydrogen concentration C within the crack tip ductile zone. H .

[0030] Furthermore, the specific implementation process of updating the hydrogen atom distribution concentration based on the hydrostatic stress gradient includes: by calling a user-defined subroutine (i.e., the hydrogen atom distribution concentration update subroutine, such as UMATHT in ABAQUS), taking the node hydrostatic stress and initial hydrogen concentration under the current increment step as input, the user-defined subroutine performs additional driving force correction on Fick's second law based on the hydrostatic stress gradient, and finally outputs the updated node / cell hydrogen atom distribution concentration.

[0031] Furthermore, after obtaining the updated hydrogen atom distribution concentration, the updated hydrogen atom distribution concentration and the effective stress intensity factor range value are time-aligned using a post-processing script, and the output includes the stress field distribution and the effective stress intensity factor range value.

[0032] After obtaining the stress field distribution and effective stress intensity factor range of the crack tip under different working conditions in the pipeline simulation model, the local hydrogen concentration at the crack tip is first determined based on the stress field distribution. Then, the local hydrogen concentration and effective stress intensity factor range are calculated using the modified Paris formula for the hydrogen embrittlement influence factor to obtain the fatigue crack propagation rate, which is then used as the model label.

[0033] Specifically, based on the classic Paris formula A hydrogen embrittlement correction factor is introduced, and the fatigue crack propagation coefficient parameter C, the fatigue crack propagation index m, and the hydrogen concentration C are established. H Functional relationship:

[0034]

[0035] In the formula, This represents the fatigue crack propagation coefficient related to the local hydrogen concentration at the crack tip. This represents the basic fatigue crack propagation coefficient under hydrogen-free conditions (or initial normal conditions). This represents the hydrogen embrittlement correction fitting parameters related to the material's fatigue crack propagation coefficient. This indicates the local hydrogen concentration within the plastic zone at the crack tip. The fatigue crack propagation index represents the relationship between the local hydrogen concentration at the crack tip and the crack growth rate. This represents the basic fatigue crack propagation index under hydrogen-free conditions (or initial normal conditions). This represents the hydrogen embrittlement correction fitting parameters related to the material's fatigue crack propagation index.

[0036] Based on the above functional relationship, the modified Paris formula is:

[0037] In the formula, a represents the crack depth, and N represents the number of fatigue load cycles. denoted by , C(H) represents the fatigue crack propagation rate, C(H) represents the fatigue crack propagation coefficient related to the local hydrogen concentration at the crack tip, m(H) represents the fatigue crack propagation exponent related to the local hydrogen concentration at the crack tip, and ΔK represents the effective stress intensity factor range value at the crack tip.

[0038] By modifying the Paris formula as described above, the mechanical data of the hundreds of sets of "local hydrogen concentration-effective stress intensity factor range values" (i.e., effective stress intensity factor range values) are converted into fatigue crack propagation rates (da / dN).

[0039] Furthermore, after establishing the pipeline simulation model, the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model are set, and the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio are preprocessed to obtain the model input features.

[0040] Specifically, the preprocessing process in this embodiment includes: calculating the pressure amplitude and crack depth ratio in the pipeline based on the linear elastic fracture mechanics theory to obtain the benchmark effective stress intensity factor; taking the square root of the hydrogen partial pressure to obtain the square root feature, and mapping the square root feature to the equivalent concentration benchmark value; and performing Z-score normalization on the benchmark effective stress intensity factor, the equivalent concentration benchmark value, and the crack depth ratio to obtain the model input features.

[0041] Furthermore, the formula for calculating the benchmark effective stress intensity factor is as follows:

[0042] In the formula, Indicates the reference effective stress intensity factor; This represents a dimensionless shape correction factor, which is related to the pipe geometry and the shape of the pre-existing crack. It can be found in standard manuals (such as API 579 (American Petroleum Institute standard) and BS 7910 (British Standards Institution standard)). Indicates the pressure amplitude inside the pipeline. Indicates the outer diameter of the pipe. Indicates the pipe wall thickness. Indicates the depth of the crack.

[0043] Furthermore, the process of mapping the square root feature to the equivalent concentration benchmark value specifically includes: Considering that the dissociation of hydrogen molecules on the steel pipe surface and the diffusion and dissolution within the crystal lattice obey Siefvert's Law, that is, the steady-state hydrogen concentration inside the material is proportional to the square root of the external hydrogen partial pressure, this embodiment sets the hydrogen partial pressure P... H2 The square root is extracted to obtain the square root feature, which is then mapped to the equivalent concentration benchmark C, characterizing the driving force of hydrogen-induced degradation inside the pipe wall. proxy The mapping formula is: C proxy

[0044] In the formula, S represents the hydrogen solubility constant of the pipe material, and P H2 This indicates the partial pressure of hydrogen under the current operating conditions.

[0045] This nonlinear mapping eliminates the long-tail effect of data caused by surface coverage saturation under high-pressure hydrogen conditions.

[0046] Furthermore, the Z-Score standardization process specifically includes: The benchmark effective stress intensity factor ΔK base Equivalent concentration benchmark value C proxy A new fusion feature vector is formed by combining auxiliary features such as the crack depth ratio without changing dimensions. The Z-Score standardization method is used to uniformly map it to a dimensionless space with a mean of 0 and a variance of 1, which is then used as the model input feature of the machine learning regression model.

[0047] By explicitly injecting fracture mechanics formulas and hydrogen diffusion laws as prior knowledge into the model input features, the blind search of machine learning regression models during multi-physics field (stress field and concentration field) coupling training process is avoided. With only a small amount of finite element simulation data (small sample), the convergence speed of machine learning regression models and the generalization prediction accuracy for unknown working conditions are greatly improved.

[0048] Through the above data processing, a training dataset is finally formed that includes model input features (crack depth ratio, pipe internal pressure amplitude, hydrogen partial pressure) and model labels (fatigue crack propagation rate (da / dN)).

[0049] After obtaining the training dataset, the machine learning regression model is trained using the training dataset. A target loss function is constructed by using the fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model, thereby generating the trained machine learning regression model.

[0050] The machine learning regression models used in this embodiment include, but are not limited to, neural network models based on the backpropagation algorithm, support vector regression models, and gradient boosting tree models.

[0051] Specifically, the machine learning regression model in this embodiment includes a backpropagation (BP) neural network with 3 hidden layers.

[0052] Training set: Select 80% of the data in the training dataset.

[0053] Test set: Select the remaining 20% ​​of the data in the training dataset.

[0054] Training process: The model input features are fed into the machine learning regression model, and the mean squared error (MSE) between the predicted fatigue crack propagation rate and the model label is minimized using the gradient descent method. To better suit the application scenario of this embodiment, in addition to calculating the mean squared error (MSE) between the predicted fatigue crack propagation rate and the model label during the model training phase, a penalty term based on prior knowledge of fracture mechanics is added to the original loss function of the machine learning regression model to obtain the target loss function.

[0055] The target loss function is as follows:

[0056] In the formula, The total loss value calculated by the objective loss function is represented; MSE represents the mean squared error between the predicted fatigue crack propagation rate and the model label. This represents the first weighting coefficient of the pre-defined fracture mechanics prior penalty term. The second weighting coefficient represents the pre-defined fracture mechanics prior penalty term, and all are constants greater than 0; P represents the predicted fatigue crack propagation rate; ΔP represents the pipe internal pressure amplitude in the model input features; H2 This represents the partial pressure of hydrogen in the model input features; This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the magnitude of the internal pressure in the pipe. This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the hydrogen partial pressure.

[0057] A penalty term is triggered when the partial derivative of the predicted fatigue crack propagation rate is less than zero (i.e., violating the physical law of monotonically increasing crack rates). This penalty term ensures that the machine learning regression model can still guarantee that the output results conform to actual physical laws even when training data for certain extreme conditions is lacking.

[0058] After training the machine learning regression model, the real-time operating conditions of the pipeline to be evaluated are obtained and input into the trained machine learning regression model for identification, to predict the fatigue crack propagation rate of the pipeline to be evaluated, and to determine the fracture ratio of the pipeline to be evaluated under the current crack state based on the predicted fatigue crack propagation rate. At the same time, the load ratio of the pipeline to be evaluated under the current crack state is calculated.

[0059] Specifically, based on the fatigue load cycle step size and the predicted fatigue crack propagation rate, the current crack depth of the pipeline to be evaluated is updated; based on the updated current crack depth and the current operating internal pressure of the pipeline to be evaluated, the stress intensity factor and reference stress of the pipeline to be evaluated in the current crack state are determined; the stress intensity factor is compared with the material fracture toughness of the pipeline to be evaluated under the current hydrogen partial pressure to obtain the fracture ratio in the current crack state; the reference stress is compared with the material yield strength of the pipeline to be evaluated to obtain the load ratio in the current crack state.

[0060] Among them, the type I stress intensity factor K I The calculation formula is:

[0061] In the formula, This represents a dimensionless shape correction factor, which is related to the pipe geometry and the shape of the pre-existing crack. It can be found in standard manuals (such as API 579 (American Petroleum Institute standard) and BS 7910 (British Standards Institution standard)). This indicates the current operating internal pressure of the pipeline to be evaluated. Indicates the outer diameter of the pipe. Indicates the pipe wall thickness. Indicates the first The current crack depth.

[0062] Reference stress The calculation formula is:

[0063] In the formula, This represents the correction factor (Folias factor) related to the crack length and pipe curvature of the pipe being evaluated. Among them,

[0064] In the formula, Indicates the relationship with the first Current crack depth The corresponding real-time total axial length of the crack (i.e., the current crack depth) is based on the initial crack length obtained by the internal detection device, and is adjusted according to the preset crack length-to-depth ratio (L / a) with the current crack depth. It is obtained through iteration and synchronous updates.

[0065] Fracture ratio The calculation formula is:

[0066] In the formula, Indicates the stress intensity factor. This indicates the fracture toughness of the material in the pipeline to be evaluated under the current hydrogen partial pressure.

[0067] Load ratio The calculation formula is:

[0068] In the formula, Indicates the reference stress. This indicates the yield strength of the material of the pipe to be evaluated.

[0069] In this embodiment Based on a preset load cycle step size By using the explicit Euler integral formula Dynamically update the current crack depth of the pipeline to be evaluated .

[0070] like Figure 3 The diagram shown is a schematic of the failure assessment diagram provided in this embodiment. The failure assessment diagram uses the fracture ratio as the vertical axis and the load ratio as the horizontal axis.

[0071] Wherein, the vertical axis (fracture ratio Kr): Kr=K I / K mat (H), where, K represents the fracture ratio. I K represents the stress intensity factor (in this embodiment, the current crack state refers to the state where the crack tip is open under the current internal pressure amplitude of the pipeline), and Kmat(H) represents the fracture toughness of the material of the pipeline to be evaluated. The key here is K. mat (H) (fracture toughness of the material) is not a constant, but a value that decreases dynamically based on experimental data or theoretical models under the current hydrogen partial pressure.

[0072] Abscissa (load ratio Lr): Lr = σ ref / σ y It characterizes the degree of plastic instability.

[0073] Specifically, after calculating the fracture ratio and load ratio of the pipeline under the current crack state, these ratios are mapped onto a failure assessment diagram to obtain failure assessment points. Then, the integrity assessment result of the pipeline is determined based on the positional relationship between the failure assessment points and the safety boundary curve of the failure assessment diagram.

[0074] When the failure assessment point is located inside the safety boundary curve of the failure assessment diagram, the pipeline is determined to be in a safe state, and the corresponding integrity assessment result is intact.

[0075] If the failure assessment point is not inside the safety boundary curve, then determine whether the failure assessment point is outside the safety boundary curve. Calculate the shortest geometric distance from the failure assessment point to the safety boundary curve. If the shortest geometric distance is less than the preset safety margin threshold (e.g., 5% of the coordinate system range) and the ordinate (fracture ratio) of the failure assessment point is greater than the preset limit value, then the integrity assessment result corresponding to the failure state dominated by hydrogen-induced embrittlement fracture of the pipeline under assessment is determined to be hydrogen-induced embrittlement fracture.

[0076] When the failure assessment point is outside the safety boundary curve and the shortest geometric distance of the failure assessment point is not less than the preset safety margin threshold, or the fracture ratio of the failure assessment point is not greater than the preset limit value, the pipeline to be assessed is determined to be in a failure state dominated by plastic collapse, and the corresponding integrity assessment result is plastic collapse.

[0077] Furthermore, after obtaining the integrity assessment results, a hydrogen pipeline integrity assessment method also includes determining whether the hydrogen pipeline is in a safe zone (i.e., whether it is in a safe state) based on the integrity assessment results; when the integrity assessment result is intact, it indicates that the hydrogen pipeline is in a safe zone, and a safe operation indicator is output; when the integrity assessment result is hydrogen-induced brittle fracture or plastic collapse, it indicates that the hydrogen pipeline is not in a safe zone, and an alarm trigger indicator or a maintenance suggestion indicator is output, so that staff can monitor the status of the hydrogen pipeline in real time and remind staff to inspect or repair the damaged pipeline when damage occurs.

[0078] In this embodiment, digital simulation completely avoids the safety risks and expensive consumables of high-pressure hydrogen physical testing. It also considers the strong coupling effect of hydrogen diffusion and stress field, which is closer to the actual failure mechanism of high-strength pipeline steel of X80, X52, X60, X65, X70, X90, X100 and L360 steel than traditional decoupling methods. Finally, the trained machine learning model is integrated into the pipeline SCADA (data acquisition and monitoring control system) system to realize real-time online assessment of pipeline integrity.

[0079] While this application provides method operation steps as shown in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in this embodiment is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the method can be executed sequentially according to this embodiment or the accompanying drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment).

[0080] like Figure 4 As shown in the illustration, this application also provides a hydrogen pipeline integrity assessment device. The device includes: The simulation model building module is used to construct a pipeline simulation model containing pre-existing cracks using finite element analysis software, and to set the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model.

[0081] The input feature acquisition module is used to simulate the diffusion and accumulation process of hydrogen atoms in the pipe wall in the pipeline simulation model, and to calculate the stress field distribution and effective stress intensity factor range value at the crack tip under different working conditions.

[0082] The hydrogen atom simulation module is used to determine the local hydrogen concentration at the crack tip based on the stress field distribution. Based on the local hydrogen concentration and the effective stress intensity factor range, the fatigue crack propagation rate is calculated by modifying the Paris formula, and the fatigue crack propagation rate is used as the model label.

[0083] The model label determination module is used to preprocess the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio to obtain the model input features.

[0084] The model training module is used to construct a training dataset based on the model input features and model labels, and to train the machine learning regression model using the training dataset. It constructs a target loss function through a fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model, thereby generating the trained machine learning regression model.

[0085] The model recognition module is used to obtain the real-time operating conditions of the pipeline to be evaluated and input the real-time operating conditions into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline to be evaluated.

[0086] The pipeline parameter calculation module is used to determine the fracture ratio of the pipeline under the current crack state based on the predicted fatigue crack propagation rate, and to calculate the load ratio of the pipeline under the current crack state.

[0087] The pipeline integrity assessment module is used to map the fracture ratio and load ratio of the pipeline under the current crack state to the failure assessment diagram to obtain the integrity assessment results of the pipeline under the assessment.

[0088] Some modules in the apparatus described in this application can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0089] The apparatus or module described in the above embodiments can be implemented by a computer chip or physical entity, or by a product with a certain function. For ease of description, the above apparatus is described by dividing it into various modules according to their functions. When implementing the embodiments of this application, the functions of each module can be implemented in one or more software and / or hardware. Of course, a module that implements a certain function can also be implemented by combining multiple sub-modules or sub-units.

[0090] The methods, apparatus, or modules described in this application can be implemented in a computer-readable program code manner. The controller can be implemented in any suitable manner, such as a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of a memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code manner, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included within it for implementing various functions can also be considered as structures within the hardware component. Alternatively, the device used to implement various functions can be viewed as either a software module that implements the method or a structure within a hardware component.

[0091] This application also provides an apparatus, the apparatus comprising: a processor; a memory for storing processor-executable instructions; wherein, when the processor executes the executable instructions, it implements the method described in this application.

[0092] This application also provides a non-volatile computer-readable storage medium storing a computer program or instructions thereon, which, when executed, enables the method described in this application embodiment to be implemented.

[0093] Furthermore, in the various embodiments of the present invention, each functional module can be integrated into a processing module, or each module can exist independently, or two or more modules can be integrated into a single module.

[0094] The aforementioned storage media include, but are not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions.

[0095] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary hardware. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, or it can be embodied in the process of data migration. The computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0096] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. All or part of this application can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0097] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit this application. Although this application 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 or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of this application.

Claims

1. A method for assessing the integrity of a hydrogen transportation pipeline, characterized in that, include: A pipeline simulation model containing pre-existing cracks was constructed using finite element analysis software, and the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model were set. The diffusion and accumulation of hydrogen atoms in the pipe wall are simulated in the pipeline simulation model, and the stress field distribution and effective stress intensity factor range value at the crack tip are calculated under different working conditions. The local hydrogen concentration at the crack tip is determined based on the stress field distribution. The fatigue crack propagation rate is calculated by modifying the Paris formula based on the local hydrogen concentration and the effective stress intensity factor range. The fatigue crack propagation rate is then used as the model label. The hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio are preprocessed to obtain the model input features; A training dataset is constructed based on the model input features and the model labels, and the machine learning regression model is trained using the training dataset. A target loss function is constructed by using the fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model, thereby generating the trained machine learning regression model. The real-time operating conditions of the pipeline to be evaluated are obtained and input into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline to be evaluated. The fracture ratio of the pipeline under evaluation in the current crack state is determined based on the predicted fatigue crack propagation rate, and the load ratio of the pipeline under evaluation in the current crack state is calculated. The fracture ratio and load ratio of the pipeline under evaluation in the current crack state are mapped to the failure evaluation diagram to obtain the integrity evaluation result of the pipeline under evaluation.

2. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The process of hydrogen atom diffusion and accumulation within the pipe wall is simulated in the pipeline simulation model, and the stress field distribution and effective stress intensity factor range values ​​at the crack tip are calculated under different operating conditions, including: The stress field distribution at the crack tip of the pipeline simulation model under different working conditions is extracted, and the hydrostatic stress gradient is calculated based on the extracted stress field distribution. The hydrogen atom distribution concentration update subroutine is invoked to update the hydrogen atom distribution concentration through the hydrostatic stress gradient, thereby simulating the diffusion and aggregation process of hydrogen atoms within the pipe wall. The post-processing script is invoked to calculate the stress field distribution and obtain the effective stress intensity factor range value.

3. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The modified Paris formula is: In the formula, a represents the crack depth, and N represents the number of fatigue load cycles. denoted by , C(H) represents the fatigue crack propagation rate, C(H) represents the fatigue crack propagation coefficient related to the local hydrogen concentration at the crack tip, m(H) represents the fatigue crack propagation exponent related to the local hydrogen concentration at the crack tip, and ΔK represents the effective stress intensity factor range value at the crack tip.

4. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The preprocessing of the hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio yields model input features, including: Based on the linear elastic fracture mechanics theory, the pressure amplitude inside the pipeline and the crack depth ratio are calculated to obtain the benchmark effective stress intensity factor; The partial pressure of hydrogen is processed by taking the square root to obtain the square root feature, and the square root feature is mapped to the equivalent concentration benchmark value. The reference effective stress intensity factor, the equivalent concentration reference value, and the crack depth ratio are subjected to Z-Score normalization to obtain the model input features.

5. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The target loss function is: In the formula, The total loss value calculated by the objective loss function is represented; MSE represents the mean square error between the predicted fatigue crack propagation rate and the model label. This represents the first weighting coefficient of the pre-defined fracture mechanics prior penalty term. The second weighting coefficient represents the pre-defined fracture mechanics prior penalty term, and all are constants greater than 0; P represents the predicted fatigue crack propagation rate; ΔP represents the pipe internal pressure amplitude in the model input features; H2 This represents the partial pressure of hydrogen in the model input features; This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the magnitude of the internal pressure in the pipe. This represents the partial derivative of the predicted fatigue crack propagation rate with respect to the hydrogen partial pressure.

6. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The determination of the fracture ratio of the pipeline under evaluation in the current crack state based on the predicted fatigue crack propagation rate, and the calculation of the load ratio of the pipeline under evaluation in the current crack state, include: The current crack depth of the pipeline to be evaluated is updated based on the fatigue load cycle step and the predicted fatigue crack propagation rate. Based on the updated current crack depth and the current operating internal pressure of the pipeline to be evaluated, determine the stress intensity factor and reference stress of the pipeline to be evaluated in the current crack state. The stress intensity factor is compared with the material fracture toughness of the pipeline under the current hydrogen partial pressure to obtain the fracture ratio under the current crack state; The load ratio under the current crack state is obtained by comparing the reference stress with the yield strength of the material of the pipeline to be evaluated.

7. The method for assessing the integrity of a hydrogen pipeline according to claim 1, characterized in that, The process of mapping the fracture ratio and load ratio of the pipeline under the current crack state to a failure assessment diagram to obtain the integrity assessment result of the pipeline under the assessment includes: The fracture ratio and load ratio of the pipeline under the current crack state are mapped onto a failure assessment diagram constructed with the fracture ratio as the vertical axis and the load ratio as the horizontal axis to obtain the failure assessment point. When the failure assessment point is located inside the safety boundary curve of the failure assessment diagram, the pipeline to be assessed is determined to be in a safe state, and the corresponding integrity assessment result is intact. If the failure assessment point is not inside the safety boundary curve, then calculate the shortest geometric distance from the failure assessment point to the safety boundary curve; When the shortest geometric distance of the failure assessment point is less than the preset safety margin threshold and the fracture ratio of the failure assessment point is greater than the preset limit value, the pipeline to be assessed is determined to be in a failure state dominated by hydrogen-induced embrittlement fracture, and the corresponding integrity assessment result is hydrogen-induced embrittlement fracture. When the failure assessment point is outside the safety boundary curve, and the shortest geometric distance of the failure assessment point is not less than the preset safety margin threshold, or the fracture ratio of the failure assessment point is not greater than the preset limit value, the pipeline to be assessed is determined to be in a failure state dominated by plastic collapse, and the corresponding integrity assessment result is plastic collapse.

8. A device for assessing the integrity of a hydrogen pipeline, characterized in that, include: The simulation model construction module is used to construct a pipeline simulation model containing pre-existing cracks using finite element analysis software, and to set the hydrogen partial pressure, pipeline internal pressure amplitude, and crack depth ratio of the pipeline simulation model. The input feature acquisition module is used to simulate the diffusion and accumulation process of hydrogen atoms in the pipe wall in the pipeline simulation model, and to calculate the stress field distribution and effective stress intensity factor range value at the crack tip under different working conditions. The hydrogen atom simulation module is used to determine the local hydrogen concentration at the crack tip based on the stress field distribution. Based on the local hydrogen concentration and the effective stress intensity factor range, the fatigue crack propagation rate is calculated by modifying the Paris formula, and the fatigue crack propagation rate is used as the model label. The model label determination module is used to preprocess the hydrogen partial pressure, the pipeline internal pressure amplitude, and the crack depth ratio to obtain model input features; The model training module is used to construct a training dataset based on the model input features and the model labels, and to train the machine learning regression model using the training dataset. A target loss function is constructed by using a fracture mechanics prior penalty term and the original loss function of the model to adjust the weights and biases in the machine learning regression model and generate the trained machine learning regression model. The model recognition module is used to obtain the real-time operating conditions of the pipeline to be evaluated and input the real-time operating conditions into the trained machine learning regression model to predict the fatigue crack propagation rate of the pipeline to be evaluated. The pipeline parameter calculation module is used to determine the fracture ratio of the pipeline under the current crack state based on the predicted fatigue crack propagation rate, and to calculate the load ratio of the pipeline under the current crack state. The pipeline integrity assessment module is used to map the fracture ratio and load ratio of the pipeline under the current crack state to the failure assessment diagram to obtain the integrity assessment result of the pipeline under the assessment.

9. An apparatus for performing a method for assessing the integrity of a hydrogen pipeline, characterized in that, include: processor; Memory used to store processor-executable instructions; When the processor executes the executable instructions, it implements the method as described in any one of claims 1 to 7.

10. A non-volatile computer-readable storage medium, characterized in that, Includes storage of computer programs or instructions that, when executed, cause the method as described in any one of claims 1 to 7 to be implemented.

Images