Submarine pipeline scour prediction method based on multi-scale turbulence-particle interaction

Abstract

The invention discloses a submarine pipeline scour prediction method based on multi-scale turbulence-particle interaction. Comprise a pipeline geometric model, a fluid domain and boundary conditions, an initial flow field is generated, and scouring pit parameters and particle distribution are set. And then, simulating fluid-particle interaction, iteratively updating the particle position and the flow field until a preset number of times is reached, and outputting and visualizing local scouring prediction data. Further, constructing a state matrix for the plurality of pipelines, and performing initial state classification through eigenvalue decomposition and mean-shift clustering to obtain a first category; and performing secondary classification according to the distance between the pipelines and the similarity of the flow field to obtain a second category, and evaluating the prediction precision of the model in real time by comparing the deviation of the two categories. According to the method, real-time simulation analysis of multi-scale turbulence and particles in the scouring state of the submarine pipeline is achieved, the vibration frequency of the whole flow field domain is effectively analyzed, and reliable data support is provided for evaluating the safety state of the pipeline.

Description

Technical Field

[0001] This invention relates to the field of subsea pipeline data analysis, and more specifically, to a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction. Background Technology

[0002] Currently, the main problems of subsea pipeline scour are the expansion of scour pits beneath the pipeline and the increase in the pipeline's overhang length. When the overhang length exceeds the bearing capacity of the pipeline material, it can easily lead to pipeline rupture, leakage, and other accidents, causing not only huge economic losses but also catastrophic damage to the marine ecosystem. Furthermore, monitoring and early warning of pipeline vibration frequencies are crucial for assessing the pipeline's safety status; however, existing finite element method (FEM) software methods struggle to obtain flow field information around the pipeline and cannot accurately analyze the vibration frequencies across the entire flow field.

[0003] To address these issues, researchers have proposed various methods for predicting subsea pipeline scour, such as physical experiments and numerical simulations. However, these methods suffer from drawbacks such as high costs in terms of manpower and resources, and limited computational accuracy. Furthermore, traditional scour simulation methods are often based on single-phase simulations, which cannot effectively simulate the motion of particles and the flow of water within them during scour, thus exhibiting certain limitations. Therefore, there is an urgent need to develop a method for predicting localized scour of subsea pipelines that can comprehensively consider multiple environmental factors and improve prediction accuracy, thereby ensuring the safe operation of subsea pipelines. Summary of the Invention

[0004] This invention overcomes the shortcomings of existing technologies and proposes a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction.

[0005] The first aspect of this invention provides a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction, comprising:

[0006] Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions.

[0007] Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters;

[0008] Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4.

[0009] Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal;

[0010] Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category.

[0011] Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

[0012] In this solution, step 1 specifically includes:

[0013] The pipeline geometry model was created using Pro / E software, and the boundary conditions included pipeline length, inner diameter, outer diameter, and material parameters.

[0014] The boundary conditions of the fluid domain model include water flow velocity, water flow pressure, and environmental parameters around the pipe.

[0015] In this solution, step 2 specifically includes:

[0016] The initial flow field distribution around the pipe was generated by a three-dimensional scouring model, and the SST k-ω turbulence model was used for fluid dynamics calculations.

[0017] Set the initial scour pit geometry parameters, including pit depth and pit diameter;

[0018] Initialize the particle distribution parameters, which include particle density, particle diameter, and initial position.

[0019] In this solution, step 3 specifically includes:

[0020] A preset time step is set, and the Eulerian-Lagrange method is used to simulate the interaction between particles and fluid, calculate the flow field and particle motion state at the current time step, and analyze the fluid-particle interaction force. The particle position and flow field distribution are calculated and updated iteratively. In each time step, the particle trajectory is calculated based on the fluid-particle interaction force and the data is updated.

[0021] Determine if the preset number of iterations has been reached. If not, return to step 1; otherwise, proceed to step 4.

[0022] In this solution, step 4 specifically includes:

[0023] The longitudinal expansion rate of the scour pit is predicted by simulating local scour of the pipeline using a 3D scour model, and the vibration frequency of the flow field around the pipeline is calculated. The predicted output is displayed through the user terminal, and the spatial distribution of the simulated and predicted flow field lines of the subsea pipeline and the scour pit is visualized in 3D using Maya software.

[0024] In this solution, step 5 specifically includes:

[0025] Set multiple subsea pipelines and a preset cycle;

[0026] Within a preset period, multiple time steps are used as one-dimensional data, and model parameters are used as two-dimensional data to construct the state matrix of the pipeline.

[0027] The state matrix is ​​decomposed into eigenvectors. Euclidean distance is introduced to calculate the similarity between eigenvectors, and the similarity is mapped to the similarity between state matrices.

[0028] The mean-shift clustering algorithm is introduced to perform cluster analysis on the state matrix. Density clustering is performed by calculating the similarity of the state matrix, and multiple pipeline categories are generated.

[0029] Based on each pipeline category, the flow field distribution and particle distribution status of the corresponding subsea pipeline are analyzed, and category information is set. Based on multiple pipeline categories and category information, the first pipeline status category data is generated.

[0030] In this solution, step 6 specifically includes:

[0031] In the three-dimensional scour model, based on the similarity of the distance and flow field distribution of multiple subsea pipelines, a secondary state classification is performed on multiple subsea pipelines to generate a second pipeline state category.

[0032] Determine the classification deviation between the first pipeline state category and the first pipeline state category, analyze the difference between the number of categories and the pipelines in each similar category group, and evaluate the prediction accuracy of the three-dimensional scour model in real time based on the classification deviation;

[0033] The simulation process parameters are adjusted in real time based on the accuracy of the prediction.

[0034] A second aspect of the present invention also provides a subsea pipeline scour prediction system based on multi-scale turbulence-particle interaction. The system includes: a memory, a processor, and a data interface. The data interface is used to connect to a user terminal. The memory includes a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction. When the processor executes the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction, it performs the following steps:

[0035] Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions.

[0036] Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters;

[0037] Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4.

[0038] Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal;

[0039] Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category.

[0040] Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

[0041] A third aspect of the present invention also provides a computer-readable storage medium comprising a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction, wherein when the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction is executed by a processor, it implements the steps of the subsea pipeline scour prediction method based on multi-scale turbulence-particle interaction as described in any of the preceding claims.

[0042] This invention introduces a multi-scale turbulence-particle interaction analysis process, which can effectively simulate the motion state of particles and the motion state of water flow within particles during scouring, overcoming the limitations of traditional single-phase simulation and significantly improving the accuracy of pipeline condition prediction. In addition, this invention uses advanced flow field simulation technology, which can accurately obtain flow field information around the pipeline and effectively analyze the vibration frequency of the entire flow field, providing a reliable basis for assessing the safety status of the pipeline.

[0043] Furthermore, this invention achieves refined simulation of water flow field and scour topography through the construction of a multi-scale model, effectively solving the problems of insufficient accuracy in flow field simulation and insufficient refinement in scour topography simulation in existing technologies. Moreover, the constructed prediction model enables multi-dimensional parameter analysis of physical and dimensionless quantities, improving the accuracy and reliability of predictions and overcoming the shortcomings of low dimensionality in the analysis of physical and dimensionless quantities in existing technologies. This invention can also monitor and assess submarine pipeline scour disasters in real time, providing effective technical support for timely detection and handling of potential safety hazards. Attached Figure Description

[0044] Figure 1 A flowchart of a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to the present invention is shown.

[0045] Figure 2 A simplified flowchart of the pipeline prediction simulation of the present invention is shown;

[0046] Figure 3 A block diagram of a subsea pipeline scour prediction system based on multi-scale turbulence-particle interaction is shown. Detailed Implementation

[0047] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0049] Figure 1 The flowchart of a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to the present invention is shown.

[0050] like Figure 1 As shown, the first aspect of this invention provides a method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction, comprising:

[0051] Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions.

[0052] Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters;

[0053] Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4.

[0054] Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal;

[0055] Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category.

[0056] Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

[0057] It is worth mentioning that this invention performs multi-dimensional fusion analysis of flow field parameters, particle distribution, and prediction parameters to determine the state shift of multiple pipelines at different time steps. Cluster analysis is introduced to classify matrices containing multi-dimensional parameter information, defining different states of subsea pipelines and storing classification information to obtain the first pipeline state category. This category includes multiple groups of pipelines, each group exhibiting a correlation between smooth flow distribution and scour prediction parameters. Each group corresponds to a specific flow field and particle distribution characteristic. This invention uses a matrix to represent the flow field, particle distribution characteristics, and scour prediction characteristics of the pipelines, fusing matrix similarity for clustering. The mean-shift algorithm is also introduced, fusing multiple matrix data density forms for clustering, effectively classifying different pipeline states. Simultaneously, it optimizes the early warning status and monitoring requirements for pipelines in different state categories. Furthermore, it combines pipeline distance and flow field characteristic similarity to determine the effectiveness and accuracy of the model's predictions, correcting simulation parameters of the three-dimensional scour model, such as iteration count and model boundary conditions, thereby optimizing the applicability of the prediction model to real-time pipeline conditions.

[0058] Compared to traditional technologies that rely on single-dimensional numerical analysis and human experience for predictive analysis, this invention can provide an intelligent and automated simulation of the scouring process of subsea pipelines, and provide strong data support for early warning analysis.

[0059] Figure 2 A simplified flowchart of the pipeline prediction simulation of the present invention is shown;

[0060] like Figure 2 The diagram shows a simplified flowchart corresponding to steps 1 to 4.

[0061] According to an embodiment of the present invention, step 1 specifically includes:

[0062] The pipeline geometry model was created using Pro / E software, and the boundary conditions included pipeline length, inner diameter, outer diameter, and material parameters.

[0063] The boundary conditions of the fluid domain model include water flow velocity, water flow pressure, and environmental parameters around the pipe.

[0064] In a preferred embodiment, the pipeline geometry model is modeled using Pro / E software, with an inner diameter of 0.8 meters, an outer diameter of 0.9 meters, and a length of 150 meters. The fluid domain model is 150 meters long, 150 meters wide, and 50 meters high. Boundary conditions are set for the pipeline geometry model, and the pipeline material is high-strength alloy steel with an elastic modulus of 200 GPa and a Poisson's ratio of 0.3. The fluid is seawater with a density of 1020 kg / m³. 3 The viscosity is 1.3×10^-3 Pa·s. In the boundary conditions of the fluid domain model, the water flow velocity is 2.0 m / s, the water flow pressure is 0.25 MPa, and the environmental parameters around the pipe are water temperature 35℃ and water pressure 0.15 MPa.

[0065] According to an embodiment of the present invention, step 2 specifically comprises:

[0066] The initial flow field distribution around the pipe was generated by a three-dimensional scouring model, and the SST k-ω turbulence model was used for fluid dynamics calculations.

[0067] Set the initial scour pit geometry parameters, including pit depth and pit diameter;

[0068] Initialize the particle distribution parameters, which include particle density, particle diameter, and initial position.

[0069] In a preferred embodiment, an initial flow field distribution around the pipe is generated using the SST k-ω turbulence model for fluid dynamics calculations. The flow field grid has 2 million elements and a grid size of 0.3 m. The scour pit geometry parameters include a pit depth of 0.2 m and a pit diameter of 0.4 m. The initial particle distribution parameters specify sand and gravel particles with a density of 1700 kg / m³. 3 The particles have a diameter of 1.0 mm and are initially evenly distributed around the pipe.

[0070] According to an embodiment of the present invention, step 3 specifically comprises:

[0071] A preset time step is set, and the Eulerian-Lagrange method is used to simulate the interaction between particles and fluid, calculate the flow field and particle motion state at the current time step, and analyze the fluid-particle interaction force. The particle position and flow field distribution are calculated and updated iteratively. In each time step, the particle trajectory is calculated based on the fluid-particle interaction force and the data is updated.

[0072] Determine if the preset number of iterations has been reached. If not, return to step 1; otherwise, proceed to step 4.

[0073] In a preferred embodiment, the preset time step is set to 0.005 s, and the particle trajectory is calculated based on the fluid-particle interaction force, updating the particle position and flow field distribution. The preset time iteration count can be set to 5000 times.

[0074] According to an embodiment of the present invention, step 4 specifically comprises:

[0075] The longitudinal expansion rate of the scour pit is predicted by simulating local scour of the pipeline using a 3D scour model, and the vibration frequency of the flow field around the pipeline is calculated. The predicted output is displayed through the user terminal, and the spatial distribution of the simulated and predicted flow field lines of the subsea pipeline and the scour pit is visualized in 3D using Maya software.

[0076] In a preferred embodiment, the longitudinal expansion rate of the scour pit is output, with an average expansion rate of 0.08 m / s, and the flow field vibration frequency around the pipe is 0.8 Hz. A three-dimensional visualization is generated. Flow field lines and a spatial distribution map of the scour pit are generated using Maya software. The predicted output includes the longitudinal expansion rate and the flow field vibration frequency.

[0077] According to an embodiment of the present invention, step 5 specifically comprises:

[0078] Set multiple subsea pipelines and a preset cycle;

[0079] Within a preset period, multiple time steps are used as one-dimensional data, and model parameters are used as two-dimensional data to construct the state matrix of the pipeline.

[0080] The state matrix is ​​decomposed into eigenvectors. Euclidean distance is introduced to calculate the similarity between eigenvectors, and the similarity is mapped to the similarity between state matrices.

[0081] The mean-shift clustering algorithm is introduced to perform cluster analysis on the state matrix. Density clustering is performed by calculating the similarity of the state matrix, and multiple pipeline categories are generated.

[0082] Based on each pipeline category, the flow field distribution and particle distribution status of the corresponding subsea pipeline are analyzed, and category information is set. Based on multiple pipeline categories and category information, the first pipeline status category data is generated.

[0083] It should be noted that the preset period includes multiple time steps. The state matrix is ​​a feature matrix reflecting the current state and predicted state of the pipeline. The model parameters include flow field velocity, particle position, interaction force, and scour prediction data. The scour prediction data includes the average spread rate and flow field vibration frequency. This matrix is ​​a feature matrix that integrates real-time state and predicted state, used to characterize the scour state of the subsea pipeline under multi-dimensional parameters.

[0084] According to an embodiment of the present invention, step 6 specifically comprises:

[0085] In the three-dimensional scour model, based on the similarity of the distance and flow field distribution of multiple subsea pipelines, a secondary state classification is performed on multiple subsea pipelines to generate a second pipeline state category.

[0086] Determine the classification deviation between the first pipeline state category and the first pipeline state category, analyze the difference between the number of categories and the pipelines in each similar category group, and evaluate the prediction accuracy of the three-dimensional scour model in real time based on the classification deviation;

[0087] The simulation process parameters are adjusted in real time based on the accuracy of the prediction.

[0088] In this embodiment, the closer the distance and the more similar the flow field distribution, the higher the probability of grouping them together. This is analyzed based on the actual pipeline conditions. Alternatively, a fixed number of groups can be defined, for example, limiting the pipelines to 5 groups, and then dividing them into 5 groups based on distance and flow field distribution. In the classification deviation analysis, each pipeline can be numbered individually, and the classification differences can be analyzed. The greater the classification deviation, the lower the prediction accuracy. Due to the influence of multiple environmental factors affecting subsea pipelines, the analysis of actual pipeline models may have certain prediction deviations. This invention utilizes pipeline distance and flow field for secondary classification to evaluate prediction accuracy. It can perform real-time and rapid accuracy evaluation of the model prediction process and results. Compared to traditional techniques that rely on manual experience to review data and evaluate models, this invention's method can effectively improve the intelligence and automation of pipeline analysis.

[0089] According to an embodiment of the present invention, it further includes:

[0090] Obtain the average spread rate and flow field vibration frequency at N preset time steps;

[0091] Two characteristic sequences were constructed based on the average spread rate and the flow field vibration frequency;

[0092] Introduce the LSTM prediction model to capture the long-term and short-term change features of two feature sequences, set the prediction step length to predict the two feature sequences, and obtain the rate prediction sequence and the frequency prediction sequence;

[0093] Set a time window to move in the prediction sequence. In each move, perform a linear fit on multiple values within the current time window. If the fitting degrees of both prediction sequences are lower than the preset degree, mark the current time window as an intervention period;

[0094] By evaluating the fitting degrees of the rate prediction sequence and the frequency prediction sequence, screen out all intervention periods;

[0095] Set the monitoring plan and early warning plan for the submarine pipeline according to the intervention period.

[0096] In the specific application process of the embodiment, it should be noted that in each feature sequence, the length is N. The time window is used to gradually move and analyze the fluctuation time period in the prediction sequence, and is used as the fluctuation time period of the pipeline environment state, marked as the intervention period. For example, the time window is set to a length of N (N < M), and each time it moves one unit, then the time window can move (M - N) times in the prediction sequence. Each time it moves, perform a linear fit on multiple values within the window and obtain the fitting coefficient. Based on the fitting coefficient, judge the fitting degree. If it is lower than the preset coefficient, it is judged as a fluctuation time period. If both prediction sequences in a time window are judged to be lower than the preset fitting degree, set the intervention period.

[0097] The movement in the prediction sequence includes the window movement analysis of both the rate prediction sequence and the frequency prediction sequence. By setting the intervention period, it is possible to perform two-dimensional time series prediction (average expansion rate and flow field vibration frequency dimension) through the pipeline state. At the same time, through the oscillation fluctuation section of the prediction time series for period marking, and the fluctuation section is based on linear fitting, it is possible to quickly analyze the abnormal fluctuation section, and by dynamically adjusting the standard of the fitting degree, the present invention can be applied to pipeline predictions in different environments, improving the multi-scenario application of the system.

[0098] Effectively solve the lag in the early warning of submarine pipeline equipment, be able to predict the pipeline early warning state in advance and set solutions, and reduce the dependence on manual experience for early warning evaluation.

[0099] Figure 3 Shows a block diagram of a submarine pipeline scour prediction system based on multi-scale turbulence-particle interaction of the present invention.

[0100] A second aspect of this invention provides a subsea pipeline scour prediction system based on multi-scale turbulence-particle interaction. The system includes: an M3 memory, an M2 processor, and an M1 data interface. The data interface is used to connect to a user terminal. The memory stores simulation process parameters involved in the three-dimensional scour model, such as flow field state, distribution, particle position, fluid-particle interaction force, scour prediction data, and model parameters. The memory includes a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction. When executed by the processor, the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction performs the following steps:

[0101] Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions.

[0102] Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters;

[0103] Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4.

[0104] Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal;

[0105] Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category.

[0106] Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

[0107] It is worth mentioning that this invention performs multi-dimensional fusion analysis of flow field parameters, particle distribution, and prediction parameters to determine the state shift of multiple pipelines at different time steps. Cluster analysis is introduced to classify matrices containing multi-dimensional parameter information, defining different states of subsea pipelines and storing classification information to obtain the first pipeline state category. This category includes multiple groups of pipelines, each group exhibiting a correlation between smooth flow distribution and scour prediction parameters. Each group corresponds to a specific flow field and particle distribution characteristic. This invention uses a matrix to represent the flow field, particle distribution characteristics, and scour prediction characteristics of the pipelines, fusing matrix similarity for clustering. The mean-shift algorithm is also introduced, fusing multiple matrix data density forms for clustering, effectively classifying different pipeline states. Simultaneously, it optimizes the early warning status and monitoring requirements for pipelines in different state categories. Furthermore, it combines pipeline distance and flow field characteristic similarity to determine the effectiveness and accuracy of the model's predictions, correcting simulation parameters of the three-dimensional scour model, such as iteration count and model boundary conditions, thereby optimizing the applicability of the prediction model to real-time pipeline conditions.

[0108] Compared to traditional technologies that rely on single-dimensional numerical analysis and human experience for predictive analysis, this invention can provide an intelligent and automated simulation of the scouring process of subsea pipelines, and provide strong data support for early warning analysis.

[0109] According to an embodiment of the present invention, step 1 specifically includes:

[0110] The pipeline geometry model was created using Pro / E software, and the boundary conditions included pipeline length, inner diameter, outer diameter, and material parameters.

[0111] The boundary conditions of the fluid domain model include water flow velocity, water flow pressure, and environmental parameters around the pipe.

[0112] In a preferred embodiment, the pipeline geometry model is modeled using Pro / E software, with an inner diameter of 0.8 meters, an outer diameter of 0.9 meters, and a length of 150 meters. The fluid domain model is 150 meters long, 150 meters wide, and 50 meters high. Boundary conditions are set for the pipeline geometry model, and the pipeline material is high-strength alloy steel with an elastic modulus of 200 GPa and a Poisson's ratio of 0.3. The fluid is seawater with a density of 1020 kg / m³. 3 The viscosity is 1.3×10^-3 Pa·s. In the boundary conditions of the fluid domain model, the water flow velocity is 2.0 m / s, the water flow pressure is 0.25 MPa, and the environmental parameters around the pipe are water temperature 35℃ and water pressure 0.15 MPa.

[0113] According to an embodiment of the present invention, step 2 specifically comprises:

[0114] The initial flow field distribution around the pipe was generated by a three-dimensional scouring model, and the SST k-ω turbulence model was used for fluid dynamics calculations.

[0115] Set the initial scour pit geometry parameters, including pit depth and pit diameter;

[0116] Initialize the particle distribution parameters, which include particle density, particle diameter, and initial position.

[0117] In a preferred embodiment, an initial flow field distribution around the pipe is generated using the SST k-ω turbulence model for fluid dynamics calculations. The flow field grid has 2 million elements and a grid size of 0.3 m. The scour pit geometry parameters include a pit depth of 0.2 m and a pit diameter of 0.4 m. The initial particle distribution parameters specify sand and gravel particles with a density of 1700 kg / m³. 3 The particles have a diameter of 1.0 mm and are initially evenly distributed around the pipe.

[0118] According to an embodiment of the present invention, step 3 specifically comprises:

[0119] A preset time step is set, and the Eulerian-Lagrange method is used to simulate the interaction between particles and fluid, calculate the flow field and particle motion state at the current time step, and analyze the fluid-particle interaction force. The particle position and flow field distribution are calculated and updated iteratively. In each time step, the particle trajectory is calculated based on the fluid-particle interaction force and the data is updated.

[0120] Determine if the preset number of iterations has been reached. If not, return to step 1; otherwise, proceed to step 4.

[0121] In a preferred embodiment, the preset time step is set to 0.005 s, and the particle trajectory is calculated based on the fluid-particle interaction force, updating the particle position and flow field distribution. The preset time iteration count can be set to 5000 times.

[0122] According to an embodiment of the present invention, step 4 specifically comprises:

[0123] The longitudinal expansion rate of the scour pit is predicted by simulating local scour of the pipeline using a 3D scour model, and the vibration frequency of the flow field around the pipeline is calculated. The predicted output is displayed through the user terminal, and the spatial distribution of the simulated and predicted flow field lines of the subsea pipeline and the scour pit is visualized in 3D using Maya software.

[0124] In a preferred embodiment, the longitudinal expansion rate of the scour pit is output, with an average expansion rate of 0.08 m / s, and the flow field vibration frequency around the pipe is 0.8 Hz. A three-dimensional visualization is generated. Flow field lines and a spatial distribution map of the scour pit are generated using Maya software. The predicted output includes the longitudinal expansion rate and the flow field vibration frequency.

[0125] According to an embodiment of the present invention, step 5 specifically comprises:

[0126] Set multiple subsea pipelines and a preset cycle;

[0127] Within a preset period, multiple time steps are used as one-dimensional data, and model parameters are used as two-dimensional data to construct the state matrix of the pipeline.

[0128] The state matrix is ​​decomposed into eigenvectors. Euclidean distance is introduced to calculate the similarity between eigenvectors, and the similarity is mapped to the similarity between state matrices.

[0129] The mean-shift clustering algorithm is introduced to perform cluster analysis on the state matrix. Density clustering is performed by calculating the similarity of the state matrix, and multiple pipeline categories are generated.

[0130] Based on each pipeline category, the flow field distribution and particle distribution status of the corresponding subsea pipeline are analyzed, and category information is set. Based on multiple pipeline categories and category information, the first pipeline status category data is generated.

[0131] It should be noted that the preset period includes multiple time steps. The state matrix is ​​a feature matrix reflecting the current state and predicted state of the pipeline. The model parameters include flow field velocity, particle position, interaction force, and scour prediction data. The scour prediction data includes the average spread rate and flow field vibration frequency. This matrix is ​​a feature matrix that integrates real-time state and predicted state, used to characterize the scour state of the subsea pipeline under multi-dimensional parameters.

[0132] According to an embodiment of the present invention, step 6 specifically comprises:

[0133] In the three-dimensional scour model, based on the similarity of the distance and flow field distribution of multiple subsea pipelines, a secondary state classification is performed on multiple subsea pipelines to generate a second pipeline state category.

[0134] Determine the classification deviation between the first pipeline state category and the first pipeline state category, analyze the difference between the number of categories and the pipelines in each similar category group, and evaluate the prediction accuracy of the three-dimensional scour model in real time based on the classification deviation;

[0135] The simulation process parameters are adjusted in real time based on the accuracy of the prediction.

[0136] In this embodiment, the closer the distance and the more similar the flow field distribution, the higher the probability of grouping them together. This is analyzed based on the actual pipeline conditions. Alternatively, a fixed number of groups can be defined, for example, limiting the pipelines to 5 groups, and then dividing them into 5 groups based on distance and flow field distribution. In the classification deviation analysis, each pipeline can be numbered individually, and the classification differences can be analyzed. The greater the classification deviation, the lower the prediction accuracy. Due to the influence of multiple environmental factors affecting subsea pipelines, the analysis of actual pipeline models may have certain prediction deviations. This invention utilizes pipeline distance and flow field for secondary classification to evaluate prediction accuracy. It can perform real-time and rapid accuracy evaluation of the model prediction process and results. Compared to traditional techniques that rely on manual experience to review data and evaluate models, this invention's method can effectively improve the intelligence and automation of pipeline analysis.

[0137] A third aspect of the present invention also provides a computer-readable storage medium comprising a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction, wherein when the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction is executed by a processor, it implements the steps of the subsea pipeline scour prediction method based on multi-scale turbulence-particle interaction as described in any of the preceding claims.

[0138] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0139] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0140] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0141] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0143] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction, characterized in that, include: Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions. Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters; Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4. Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal; Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category. Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

2. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 1 specifically includes: The pipeline geometry model was created using Pro / E software, and the boundary conditions included pipeline length, inner diameter, outer diameter, and material parameters. The boundary conditions of the fluid domain model include water flow velocity, water flow pressure, and environmental parameters around the pipe.

3. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 2 specifically involves: The initial flow field distribution around the pipe was generated by a three-dimensional scouring model, and the SST k-ω turbulence model was used for fluid dynamics calculations. Set the initial scour pit geometry parameters, including pit depth and pit diameter; Initialize the particle distribution parameters, which include particle density, particle diameter, and initial position.

4. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 3 specifically involves: A preset time step is set, and the Eulerian-Lagrange method is used to simulate the interaction between particles and fluid, calculate the flow field and particle motion state at the current time step, and analyze the fluid-particle interaction force. The particle position and flow field distribution are calculated and updated iteratively. In each time step, the particle trajectory is calculated based on the fluid-particle interaction force and the data is updated. Determine if the preset number of iterations has been reached. If not, return to step 1; otherwise, proceed to step 4.

5. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 4 specifically involves: The longitudinal expansion rate of the scour pit is predicted by simulating local scour of the pipeline using a 3D scour model, and the vibration frequency of the flow field around the pipeline is calculated. The predicted output is displayed through the user terminal, and the spatial distribution of the simulated and predicted flow field lines of the subsea pipeline and the scour pit is visualized in 3D using Maya software.

6. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 5 specifically involves: Set multiple subsea pipelines and a preset cycle; Within a preset period, multiple time steps are used as one-dimensional data, and model parameters are used as two-dimensional data to construct the state matrix of the pipeline. The state matrix is ​​decomposed into eigenvectors. Euclidean distance is introduced to calculate the similarity between eigenvectors, and the similarity is mapped to the similarity between state matrices. The mean-shift clustering algorithm is introduced to perform cluster analysis on the state matrix. Density clustering is performed by calculating the similarity of the state matrix, and multiple pipeline categories are generated. Based on each pipeline category, the flow field distribution and particle distribution status of the corresponding subsea pipeline are analyzed, and category information is set. Based on multiple pipeline categories and category information, the first pipeline status category data is generated.

7. The method for predicting subsea pipeline scour based on multi-scale turbulence-particle interaction according to claim 1, characterized in that, Step 6 specifically involves: In the three-dimensional scour model, based on the similarity of the distance and flow field distribution of multiple subsea pipelines, a secondary state classification is performed on multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the first pipeline state category, analyze the difference between the number of categories and the pipelines in each similar category group, and evaluate the prediction accuracy of the three-dimensional scour model in real time based on the classification deviation; The simulation process parameters are adjusted in real time based on the accuracy of the prediction.

8. A subsea pipeline scour prediction system based on multi-scale turbulence-particle interaction, characterized in that, The system includes: a memory, a processor, and a data interface. The data interface is used to connect to a user terminal. The memory includes a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction. When the processor executes the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction, it performs the following steps: Step 1: Establish a three-dimensional scour model of the subsea pipeline using CFD software. The three-dimensional scour model includes the pipeline geometry model, the fluid domain model, and the boundary conditions. Step 2: Generate the initial flow field distribution using a 3D scour model, set the geometric parameters of the scour pits, and initialize the particle distribution parameters; Step 3: Simulate the motion state of the fluid and particles based on the fluid-particle interaction force, iteratively calculate and update the particle position and flow field distribution, and perform time iteration calculation. If the preset number of iterations is not reached, return to step 1; if the preset number of iterations is reached, proceed to step 4. Step 4: Output the local scour prediction data of the subsea pipeline and visualize the prediction through the user terminal; Step 5: Set up multiple subsea pipelines for analysis, construct the pipeline state matrix according to the preset time step and model parameters, evaluate the similarity of the matrix by decomposing the eigenvalues ​​and eigenvectors of the matrix, introduce mean-shift clustering to perform cluster analysis on the state matrix, classify the state of multiple subsea pipelines, and generate the first pipeline state category. Step 6: Based on the similarity of distance and flow field distribution among multiple subsea pipelines, perform secondary state classification on the multiple subsea pipelines to generate a second pipeline state category. Determine the classification deviation between the first pipeline state category and the second pipeline state category, and evaluate the prediction accuracy of the three-dimensional scour model in real time.

9. The subsea pipeline scour prediction system based on multi-scale turbulence-particle interaction according to claim 8, characterized in that, Step 1 specifically includes: The pipeline geometry model was created using Pro / E software, and the boundary conditions included pipeline length, inner diameter, outer diameter, and material parameters. The boundary conditions of the fluid domain model include water flow velocity, water flow pressure, and environmental parameters around the pipe.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction. When the subsea pipeline scour prediction program based on multi-scale turbulence-particle interaction is executed by a processor, it implements the steps of the subsea pipeline scour prediction method based on multi-scale turbulence-particle interaction as described in any one of claims 1 to 7.

Images