Method and system for scour pit dynamic monitoring and analysis based on wave-current coupling

Abstract

The application relates to the technical field of discrete data processing of the surrounding terrain of a marine structure, and discloses a scour pit dynamic monitoring and analyzing method and system based on wave-current coupling, which comprises the following steps: according to the difference space gradient of the grid node shear stress and the starting shear stress threshold, dividing an initial terrain grid flow into a core sub-matrix and an auxiliary sub-matrix; configuring an asymmetric time step length cyclic iteration double sub-matrix to update an elevation vector; using a digital filter to calculate a space slope vector, when the space slope vector exceeds a critical rest angle threshold, using a diffusion weight matrix to evenly spread and rewrite an elevation change to a neighborhood node register address to output situation data; the application eliminates the calculation redundancy of the auxiliary sub-matrix through asymmetric step length scheduling, shortens the time sequence convergence time, limits the elevation mutation by using address rewriting, eliminates numerical divergence, and guarantees the time sequence stability.

Description

Technical Field

[0001] This invention relates to a dynamic monitoring and analysis method and system for scour pits based on wave-current coupling, belonging to the field of discrete data processing technology for the surrounding topography of marine structures. Background Technology

[0002] Currently, using a three-dimensional discrete mesh matrix to characterize the scour state of seabed foundations is a common technical approach. This method divides the seabed surrounding the structure into discrete nodes, calculates the dynamic response parameters of each node under the combined effects of waves and currents using hydrodynamic control equations, determines local elevation variations by calculating the shear stress of sediment particles, and the data processing system periodically overwrites the elevation values ​​of each node in memory, constructing a time-series topographic variation matrix and outputting variation trend data. This time-series iterative overwriting architecture based on discrete control equations provides data processing support under normal stable operating conditions and is a cornerstone of seabed scour state data analysis. However, under strongly nonlinear dynamic loads... In extreme sea conditions, the seabed data field carries highly variable and complex data streams. Wave-current interactions cause severe scouring around structures, resulting in abrupt changes in the elevation data of local topographic nodes. Existing global symmetric time-series scheduling mechanisms reveal data throughput bottlenecks and computing power imbalances when processing highly variable data streams. Because the global control equations force all nodes to adopt a uniform read / write refresh frequency, the system generates massive redundant calculations on nodes in stable regions far from the scouring core. Frequent data read / write throughput in the memory leads to low time-series convergence efficiency of the grid matrix. Most of the computing power is wasted in the idle overwriting of matrix nodes in stable regions, which cannot meet the timeliness requirements.

[0003] To alleviate the conflict between computational cost and convergence speed, conventional approaches typically involve reducing the global iteration step size or implementing mesh sparsification. However, reducing the step size inevitably leads to a surge in load, causing the system to interrupt computation due to bandwidth constraints. Mesh sparsification, on the other hand, loses high-frequency characteristics in abrupt change regions, resulting in nonlinear oscillations and gradient numerical overflow in the solver. Besides the numerical oscillations caused by spatial discrete meshes in dealing with data abrupt changes, from the perspective of the control mechanism of numerical simulation and prediction models, multi-field coupled prediction methods also suffer from imbalances in spatiotemporal scale adjustment and insufficient convergence stability. For example, Chinese invention patent CN112989681B discloses a system and method for predicting and analyzing seabed scouring and deposition in submarine cable routing areas, which uses a simultaneous hydrodynamic model, sediment transport model, and seabed deformation model for finite element solution. The method uses historical measured values ​​for post-correction prediction of submarine cable burial depth. This type of multi-field coupled prediction method uses a globally symmetrical unified time step for time-series extrapolation. It relies on the premise of gradual evolution of seabed topography. In the dynamic scenario of local severe scouring caused by strong nonlinear wave-current coupled dynamic load, the geomorphological evolution at the edge of the seabed scour pit exhibits extremely high frequency nonlinear discontinuous step characteristics. The spatial local elevation gradient can easily cross the critical collapse point of sediment collapse in an instant. The globally uniform extrapolation architecture and the post-macro correction mechanism lack the ability to actively intercept and reroute to correct sudden changes in local high-dimensional signals. The data differences within the sliding time window are amplified sharply, causing the cascading accumulation of local truncation errors in the solver. This leads to the overflow and divergence of values ​​in the system's internal temporary registers, making it impossible to guarantee the continuity of the data chain and the reliability of the solution for large-scale periodic situation prediction.

[0004] Therefore, how to design a data processing architecture that combines asymmetric topology splitting scheduling with adaptive convergence filtering to reduce the computational load of high-dimensional matrices in the stable region while suppressing data divergence in the abrupt region is the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A dynamic monitoring and analysis method for scour pits based on wave-current coupling, comprising the following steps: Step S1: Obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; Step S2: Call the numerical solver, divide the initial terrain grid flow into a core submatrix and an auxiliary submatrix according to the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, configure an asymmetric time step to iterate the core submatrix and the auxiliary submatrix, and update the node elevation vectors in the core submatrix. Step S3: Use the elevation gradient digital filter to read the elevation difference between the core node and the neighboring node in the core submatrix, and calculate the spatial slope vector. Step S4: When the magnitude of the spatial slope vector is greater than the threshold of the landslide critical repose angle that is synchronously updated under the time step, the elevation data allocation rule is triggered. Step S5: Using the diffusion weight matrix, the elevation change of the core node exceeding the critical repose angle threshold of the landslide is spread and rewritten to the register address of the neighboring node, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration. Step S6: After completing the preset cycle of iteration, output the scour pit development trend data, which includes the maximum scour depth, scour range, and submarine cable suspension span.

[0006] Preferably, step S1 includes the following sub-steps: step S11, calculating the extreme value distribution of hydrodynamic observation data to determine the design wave height and design flow velocity under the target return period; step S12, generating a wave-current load matrix containing multi-dimensional parameters by algebraically superimposing the design wave height and design flow velocity through a two-dimensional nonlinear spatial mapping relationship; step S13, dividing the geological grid matrix into multiple independent regions using topological splitting scheduling units to construct the initial topographic grid flow.

[0007] Preferably, step S5 includes the following sub-steps: Step S51, calling the elevation gradient digital filter to capture the elevation data overflow within the region of the core node; Step S52, reading the elevation change of the core node exceeding the critical repose angle threshold of the landslide, and constructing a three-dimensional elevation overflow vector; Step S53, multiplying the diffusion weight matrix with the three-dimensional elevation overflow vector to determine the apportionment weight coefficient of each neighboring node; Step S54, writing the elevation change into the register address of each neighboring node according to the apportionment weight coefficient, and reconstructing the terrain elevation data of the core sub-matrix.

[0008] Preferably, step S4 includes the following sub-steps: step S41, obtaining the seabed soil type, median grain size, and internal friction angle parameter of the target sea area; step S42, matching the corresponding static repose angle reference value according to the seabed soil type and median grain size; step S43, using the internal friction angle parameter to attenuate and correct the static repose angle reference value, generating a dynamic critical repose angle function that has a negative monotonic relationship with the real-time ocean current velocity, as the critical repose angle threshold for landslide.

[0009] Preferably, step S2 includes the following sub-steps: step S21, calculate the combined wave-current bottom shear stress at the bottom grid node and compare it with the critical shear stress for sediment initiation; step S22, when the combined wave-current bottom shear stress is greater than the critical shear stress for sediment initiation, calculate the local net sediment transport rate of the grid node; step S23, based on the sediment continuity equation, update the node elevation vector in the core submatrix using the divergence of the local net sediment transport rate.

[0010] Preferably, the method further includes the following steps: Step S7, storing the output scour pit development trend data into the submarine cable foundation status database in time series, and constructing the historical time series of maximum scour depth and the historical time series of submarine cable suspension span.

[0011] Preferably, step S7 includes the following sub-steps: Step S71, calculate the first time derivative of the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable to determine the dynamic evolution rate of the current scour pit depth and the expansion rate of the suspended span of the submarine cable.

[0012] Preferably, step S7 further includes the following sub-steps: step S72, determining whether the dynamic evolution rate is greater than a preset scour development rate limit threshold, or whether the span expansion rate is greater than a preset span extension rate limit threshold; step S73, issuing a warning signal for the risk of instability of the submarine cable foundation structure when the dynamic evolution rate is greater than the scour development rate limit threshold, or the span expansion rate is greater than the span extension rate limit threshold.

[0013] Preferably, the method further includes the following steps: Step S8, obtaining a preset physical threshold matrix for submarine cable operation safety; Step S9, inputting multiple sets of preset protective structure geometric parameter matrices into a numerical solver for cyclic updating, outputting secondary scour pit development trend data under the protective configuration, and extracting the temporal extreme value envelope matrix over the entire cycle; Step S10, comparing the boundary range of the temporal extreme value envelope matrix with the physical threshold matrix for submarine cable operation safety, and selecting the optimal protective design parameter matrix with the minimum maximum scour depth and the submarine cable suspension span not exceeding the standard.

[0014] A dynamic monitoring and analysis system for scour pits based on wave-current coupling is provided, which is used to implement a dynamic monitoring and analysis method for scour pits based on wave-current coupling. The system includes: The initial terrain grid flow construction module is used to obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; The matrix partitioning and elevation update module is used to call the numerical solver. Based on the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, the initial terrain grid flow is divided into a core sub-matrix and an auxiliary sub-matrix. The core sub-matrix and auxiliary sub-matrix are iterated in a loop with an asymmetric time step, and the node elevation vectors in the core sub-matrix are updated. The spatial slope vector calculation module is used to read the elevation difference between the core node and the neighboring node in the core submatrix using the elevation gradient digital filter, and to calculate the spatial slope vector. The allocation rule triggering module is used to trigger the elevation data allocation rule when the magnitude of the spatial slope vector is greater than the landslide critical repose angle threshold updated synchronously under the time step; the elevation data amortization and rewriting module is used to amortize and rewrite the elevation change of the core node that exceeds the landslide critical repose angle threshold to the register address of the neighboring node using the diffusion weight matrix, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration; The situation data output module is used to output the development situation data of the scour pit after completing the preset cycle iteration. The development situation data of the scour pit includes the maximum scour depth, scour range and the span of the submarine cable.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In the dynamic monitoring and analysis of scour pits with wave-current coupling, a shear stress response characteristic topology map is constructed using the bed shear stress calculation data of the grid nodes and the critical initiation shear stress threshold of sediment in the memory. Based on the topological edge weight distribution, the overall terrain grid is decoupled and split into a nonlinear abrupt core calculation submatrix and a boundary smooth gradual auxiliary calculation submatrix. By configuring an asymmetric temporal iteration step size scheduling strategy for the two submatrixes, a high-frequency subdivision step size value is configured for the core calculation submatrix, and a wide-frequency coarse step size value is configured for the auxiliary calculation submatrix. The heterogeneous splitting and differentiated scheduling mechanism eliminates the data processing redundancy in the smooth gradual region, shortens the temporal convergence time of the high-dimensional terrain matrix, and improves the data flow throughput efficiency.

[0016] 2. When updating high-frequency elevation data in the core submatrix, the digital filter reads the elevation difference between the core node and the neighboring nodes and calculates the local spatial slope feature vector. When the magnitude of this vector exceeds the threshold of the critical rest angle function for mud and sand landslide in the memory, the data rerouting correction rule is triggered. The elevation change exceeding the threshold is distributed and rewritten to the register address of the surrounding neighboring nodes using the spatial diffusion weight matrix. This address rewriting limits the filter overflow caused by the abrupt change in elevation data, causing the local slope feature vector to converge within the boundary, eliminating numerical divergence under nonlinear load calculation, and ensuring the temporal stability of the discrete terrain data stream under abrupt change.

[0017] 3. The data parsing module generates the initial terrain grid, the load condition generation module generates a load combination matrix containing multi-dimensional hydrodynamic parameters, the topology splitting scheduling unit and the filtering unit work together to run the response model and output situation data containing the suspended span and scour depth; the scheme optimization and evaluation module compares the situation data output by the second iteration calculation with the safety threshold in the memory, selects and outputs the optimal protection parameter matrix, and converts the terrain elevation variation state into a discrete digital matrix through a multi-module collaborative architecture, realizing a logical closed loop of multi-source feature perception and protection parameter optimization, bypassing the engineering dilemma of inaccurate prediction caused by nonlinear amplification effect in traditional analysis. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the overall steps of the wave-current coupling dynamic monitoring and analysis method for scour pits according to the present invention. Figure 2 This is the data determination and logic branch diagram of the wave-current coupled dynamic monitoring and analysis system for scour pits of the present invention.

[0019] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0021] A dynamic monitoring and analysis method for scour pits based on wave-current coupling includes the following steps: Step S1: Obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; Step S2: Call the numerical solver, divide the initial terrain grid flow into a core submatrix and an auxiliary submatrix according to the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, configure an asymmetric time step to iterate the core submatrix and the auxiliary submatrix, and update the node elevation vectors in the core submatrix. Step S3: Use the elevation gradient digital filter to read the elevation difference between the core node and the neighboring nodes in the core submatrix, and calculate the spatial slope vector. Step S4: When the magnitude of the spatial slope vector is greater than the threshold of the landslide critical repose angle that is synchronously updated under the time step, the elevation data allocation rule is triggered. Step S5: Using the diffusion weight matrix, the elevation change of the core node exceeding the critical repose angle threshold of the landslide is spread and rewritten to the register address of the neighboring node, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration. Step S6: After completing the preset cycle of iteration, output the scour pit development trend data, which includes the maximum scour depth, scour range, and submarine cable suspension span.

[0022] Preferably, step S1 includes the following sub-steps: step S11, calculating the extreme value distribution of hydrodynamic observation data to determine the design wave height and design flow velocity under the target return period; step S12, generating a wave-current load matrix containing multi-dimensional parameters by algebraically superimposing the design wave height and design flow velocity through a two-dimensional nonlinear spatial mapping relationship; step S13, dividing the geological grid matrix into multiple independent regions using topological splitting scheduling units to construct the initial topographic grid flow.

[0023] Preferably, step S5 includes the following sub-steps: Step S51, calling the elevation gradient digital filter to capture the elevation data overflow within the region of the core node; Step S52, reading the elevation change of the core node exceeding the critical repose angle threshold of the landslide, and constructing a three-dimensional elevation overflow vector; Step S53, multiplying the diffusion weight matrix with the three-dimensional elevation overflow vector to determine the apportionment weight coefficient of each neighboring node; Step S54, writing the elevation change into the register address of each neighboring node according to the apportionment weight coefficient, and reconstructing the terrain elevation data of the core sub-matrix.

[0024] Preferably, step S4 includes the following sub-steps: step S41, obtaining the seabed soil type, median grain size, and internal friction angle parameter of the target sea area; step S42, matching the corresponding static repose angle reference value according to the seabed soil type and median grain size; step S43, using the internal friction angle parameter to attenuate and correct the static repose angle reference value, generating a dynamic critical repose angle function that has a negative monotonic relationship with the real-time ocean current velocity, as the critical repose angle threshold for landslide.

[0025] Preferably, step S2 includes the following sub-steps: step S21, calculate the combined wave-current bottom shear stress at the bottom grid node and compare it with the critical shear stress for sediment initiation; step S22, when the combined wave-current bottom shear stress is greater than the critical shear stress for sediment initiation, calculate the local net sediment transport rate of the grid node; step S23, based on the sediment continuity equation, update the node elevation vector in the core submatrix using the divergence of the local net sediment transport rate.

[0026] Preferably, the method further includes the following steps: Step S7, storing the output scour pit development trend data into the submarine cable foundation status database in time series, and constructing the historical time series of maximum scour depth and the historical time series of submarine cable suspension span.

[0027] Preferably, step S7 includes the following sub-steps: Step S71, calculate the first time derivative of the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable to determine the dynamic evolution rate of the current scour pit depth and the expansion rate of the suspended span of the submarine cable.

[0028] Preferably, step S7 further includes the following sub-steps: step S72, determining whether the dynamic evolution rate is greater than a preset scour development rate limit threshold, or whether the span expansion rate is greater than a preset span extension rate limit threshold; step S73, issuing a warning signal for the risk of instability of the submarine cable foundation structure when the dynamic evolution rate is greater than the scour development rate limit threshold, or the span expansion rate is greater than the span extension rate limit threshold.

[0029] Preferably, the method further includes the following steps: Step S8, obtaining a preset physical threshold matrix for submarine cable operation safety; Step S9, inputting multiple sets of preset protective structure geometric parameter matrices into a numerical solver for cyclic updating, outputting secondary scour pit development trend data under the protective configuration, and extracting the temporal extreme value envelope matrix over the entire cycle; Step S10, comparing the boundary range of the temporal extreme value envelope matrix with the physical threshold matrix for submarine cable operation safety, and selecting the optimal protective design parameter matrix with the minimum maximum scour depth and the submarine cable suspension span not exceeding the standard.

[0030] A dynamic monitoring and analysis system for scour pits based on wave-current coupling is provided, which is used to implement a dynamic monitoring and analysis method for scour pits based on wave-current coupling. The system includes: The initial terrain grid flow construction module is used to obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; The matrix partitioning and elevation update module is used to call the numerical solver. Based on the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, the initial terrain grid flow is divided into a core sub-matrix and an auxiliary sub-matrix. The core sub-matrix and auxiliary sub-matrix are iterated in a loop with an asymmetric time step, and the node elevation vectors in the core sub-matrix are updated. The spatial slope vector calculation module is used to read the elevation difference between the core node and the neighboring node in the core submatrix using the elevation gradient digital filter, and to calculate the spatial slope vector. The allocation rule triggering module is used to trigger the elevation data allocation rule when the magnitude of the spatial slope vector is greater than the landslide critical repose angle threshold updated synchronously under the time step; the elevation data amortization and rewriting module is used to amortize and rewrite the elevation change of the core node that exceeds the landslide critical repose angle threshold to the register address of the neighboring node using the diffusion weight matrix, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration; The situation data output module is used to output the development situation data of the scour pit after completing the preset cycle iteration. The development situation data of the scour pit includes the maximum scour depth, scour range and the span of the submarine cable.

[0031] Example 1: This example provides a dynamic monitoring and analysis method for seabed scour pits based on wave-current coupling, to address the computational imbalance and numerical divergence problems in seabed scour pit data analysis under extreme sea conditions with intertwined strong nonlinear dynamic loads. Figure 1 As shown, the method specifically includes the following steps: Step S1: Obtain the geological grid matrix and wave-current load matrix to construct the initial topographic grid flow. As a preferred implementation, Step S1 can be implemented through the following sub-steps: Step S11: Calculate the extreme value distribution of hydrodynamic observation data to determine the design wave height and design current velocity under the target return period; Step S12: Generate a wave-current load matrix containing multidimensional parameters by algebraically superimposing the design wave height and design current velocity through a two-dimensional nonlinear mapping relationship. In this invention, the two-dimensional nonlinear mapping relationship is realized by introducing a nonlinear interaction model of the seabed boundary layer. The design wave height is transformed into a wave particle convection velocity vector at the seabed according to linear wave theory. Then, the wave velocity vector and the design current velocity vector are algebraically superimposed in two dimensions and multiplied by the nonlinear friction coupling coefficient of the seabed. This transforms the one-dimensional scalar design input into a multidimensional parameter matrix containing the force characteristics of two orthogonal directions in space, which is used to accurately characterize the bottom dynamic load characteristics under the combined action of waves and currents; Step S13: Divide the geological grid matrix into multiple independent regions using a topology splitting scheduling unit to construct the initial topographic grid flow.

[0032] Step S2: Call the numerical solver. Based on the spatial gradient of the difference between the shear stress at the grid nodes and the critical shear stress threshold, divide the initial terrain grid flow into a core sub-matrix and an auxiliary sub-matrix. Configure an asymmetric time step to iterate the core sub-matrix and the auxiliary sub-matrix, and update the node elevation vectors within the core sub-matrix. As a preferred implementation, step S2 can be implemented through the following sub-steps: Step S21: Calculate the combined wave-current bottom shear stress at the bottom grid nodes and compare it with the critical shear stress for sediment initiation. Specifically, the numerical solver extracts the bed shear stress calculation data of each grid node and calculates the difference between the bed shear stress calculation data of each grid node and the critical shear stress threshold for sediment initiation in memory. Based on the spatial gradient of this difference, construct a shear stress response characteristic topology map. This topology map uses the spatial three-dimensional coordinates of each grid node as topological vertices and the spatial gradient of shear stress between adjacent grid nodes as topological edge weights. Step S22: When the combined wave-current bottom shear stress is greater than the critical shear stress for sediment initiation, calculate the local net sediment transport rate of the grid nodes. The processor divides the overall terrain mesh based on the edge weight distribution of the topological map of shear stress response characteristics. Nodes with edge weights greater than or equal to a set gradient threshold, and whose gradient threshold working window is limited to 0.05 meters to 0.25 meters per meter, are completely cut and independently encapsulated to form the nonlinear catastrophe core computational submatrix. Residual vertices and edge smooth transition regions with edge weights less than the set gradient threshold are uniformly incorporated into the boundary steady transition auxiliary computational submatrix. In step S23, based on the sediment continuity equation, the node elevation vectors within the core submatrix are updated using the divergence of the local net sediment transport rate. During the iteration process, an asymmetric temporal iteration step size scheduling strategy is configured for the nonlinear catastrophe core computational submatrix and the boundary steady transition auxiliary computational submatrix. The core computational submatrix uses a high-frequency subdivision step size to update the elevation state vector, while the auxiliary computational submatrix uses a wide-frequency coarse step size for periodic refresh, ensuring that the ratio of the high-frequency subdivision step size to the wide-frequency coarse step size is less than or equal to 0.1. This eliminates data processing redundancy in the steady transition region and shortens the temporal convergence time of the high-dimensional terrain matrix.

[0033] Step S3: Use the elevation gradient digital filter to read the elevation difference between the core node and the neighboring node in the core submatrix, and calculate the spatial slope vector. In this step, the processor synchronously calls the elevation gradient digital filter to monitor whether the grid data diverges.

[0034] Step S4: When the magnitude of the spatial slope vector is greater than the synchronously updated critical angle of repose threshold for landslide under the time step, the elevation data allocation rule is triggered. As a preferred implementation method, step S4 can be implemented through the following sub-steps: Step S41: Obtain the seabed soil type, median grain size, and internal friction angle parameters of the target sea area; Step S42: Match the corresponding still water angle of repose benchmark value according to the seabed soil type and median grain size; Step S43: Use the internal friction angle parameter to attenuate and correct the still water angle of repose benchmark value, and generate a dynamic critical angle of repose function that has a negative monotonic relationship with the real-time ocean current velocity, as the critical angle of repose threshold for landslide. The threshold generated by this function in real time accurately characterizes the physical response process in which the shear strength of the bottom sediment and the stable angle of repose decrease monotonically with the increase of the flow velocity due to the drag force of the hydrodynamic force.

[0035] Step S5 involves using a diffusion weight matrix to evenly rewrite the elevation change of the core node exceeding the critical repose angle threshold into the register addresses of neighboring nodes. This stabilizes the modified spatial slope vector within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time-series iteration. As a preferred implementation, step S5 can be achieved through the following sub-steps: Step S51: Call the elevation gradient digital filter within the region of the core node to capture the elevation data overflow; Step S52: Read the elevation change of the core node exceeding the critical repose angle threshold and construct a three-dimensional elevation overflow vector; S53, multiply the diffusion weight matrix with the three-dimensional elevation overflow vector to determine the apportionment weight coefficient of each neighboring node; step S54, write the elevation change into the register address of each neighboring node according to the apportionment weight coefficient, and reconstruct the terrain elevation data of the core submatrix. This step performs register address rewriting and data apportionment at the computer hardware storage management level. In the numerical discrete solution system, an adaptive gravity collapse shear stress release chain is equivalently constructed to dissipate momentum and redistribute mass of the local overload elevation difference to the periphery, thus blocking the nonlinear numerical divergence caused by the infinite amplification of local gradients.

[0036] Step S6: After completing the preset cycle iteration, output the scour pit development trend data. The scour pit development trend data includes the maximum scour depth, scour range, and submarine cable suspension span. In order to achieve long-term full-cycle monitoring and protection design optimization of the submarine cable operation status, the method of this embodiment also includes the following extended steps: Step S7: Store the output data on the development trend of scour pits into the submarine cable foundation status database according to the time series, and construct the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable; Step S71: Calculate the first time derivative of the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable to determine the dynamic evolution rate of the current scour pit depth and the span expansion rate of the suspended submarine cable; Step S72: Determine whether the dynamic evolution rate is greater than the preset scour development rate limit threshold, or whether the span expansion rate is greater than the preset span extension rate limit threshold; Step S73: When the dynamic evolution rate is greater than the scour development rate limit threshold, or the span expansion rate is greater than the span extension rate limit threshold, issue a warning signal for the instability risk of the submarine cable foundation structure.

[0037] Step S8: Obtain the preset physical threshold matrix for submarine cable operation safety.

[0038] Step S9 involves inputting multiple sets of preset protective structure geometric parameter matrices into the numerical solver for cyclic updates, outputting secondary scour pit development trend data under the protective configuration, and extracting the temporal extreme value envelope matrix over the entire cycle. During this process, the geometric boundary of the protective structure is embedded into the solid wall boundary conditions of the bottom bed fluid dynamics control equation in the form of solid obstacles or porous media units, reconstructing the convection and diffusion terms of the local flow field control equation to directly capture the secondary scour variations after the protective structure changes its flow direction.

[0039] Step S10: Compare the boundary ranges of the time-series extreme value envelope matrix and the submarine cable operation safety physical threshold matrix to select the optimal protection design parameter matrix that minimizes the maximum scour depth and ensures that the submarine cable suspension span does not exceed the standard.

[0040] Example 2: This example provides a wave-current coupling-based dynamic monitoring and analysis system for scour pits to implement the method of Example 1. The system's electro-digital data processing architecture runs on a hardware computing platform with a dual-core, four-thread architecture and at least 16GB of dynamic temporary storage. The processor's clock speed is no less than 3.2GHz, and it possesses double-precision floating-point arithmetic capabilities to ensure the minimum bandwidth for reading and writing high-dimensional terrain matrices. Figure 2 As shown, the system includes the following core functional modules: Initial terrain grid flow construction module: used to obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow. It defines discrete terrain nodes with a spatial side length of 1.0 meter by reading the pre-stored initial grid topology matrix in the memory.

[0041] The matrix partitioning and elevation update module calls the numerical solver to partition the initial terrain grid flow into a core submatrix and an auxiliary submatrix based on the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold. It then configures an asymmetric time step to iterate the core submatrix and the auxiliary submatrix, updating the node elevation vectors within the core submatrix. This module uses a graph cut algorithm or a maximum connected component search strategy to traverse all edge weights of the topology graph, achieving accurate topology splitting and differentiated step size scheduling under heterogeneous loads.

[0042] Spatial slope vector calculation module: Used to read the elevation difference between the core node and the neighboring node in the core submatrix using the elevation gradient digital filter, and calculate the spatial slope vector to monitor whether the grid data diverges in real time.

[0043] The allocation rule triggering module and the elevation data amortization and rewriting module are used to trigger elevation data allocation rules when the magnitude of the spatial slope vector is greater than the landslide critical repose angle threshold updated synchronously under the time step. The elevation data amortization and rewriting module is used to amortize and rewrite the elevation changes of the core nodes that exceed the landslide critical repose angle threshold to the register addresses of neighboring nodes using the diffusion weight matrix. This keeps the modified spatial slope vector stable within the safe boundary, suppresses numerical divergence under nonlinear load calculation from the hardware storage management level, and ensures the convergence stability of the initial terrain grid flow time-series iteration.

[0044] Situation data output module: After completing the preset cycle iteration, it outputs the development situation data of the scour pit. The development situation data of the scour pit includes the maximum scour depth, scour range and the suspended span of the submarine cable. This module will also transfer the output data to the subsequent data analysis, early warning judgment and scheme optimization evaluation modules to realize the logical closed loop of multi-source feature perception and protection parameter optimization.

[0045] In addition, the system integrates a security fault tolerance and verification mechanism. When the system encounters a power outage or abnormal reset, it automatically recovers the state vector from the last successfully written non-volatile memory address and performs data self-checking through cyclic redundancy check code before each data transfer. This is to combat physical layer interference and ensure the continuity of the data chain for large-scale periodic situation prediction.

[0046] Example 3: In the data processing method of testing the synergistic effect of asymmetric topographic scheduling and adaptive convergence filtering in the electronic digital data processing system, the input data is generated by computational fluid dynamics simulation based on the two-dimensional shallow water wave equation and the sediment continuity equation. An initial topographic grid flow is constructed using a spatial discrete grid matrix of 10,000 x 10,000. Gaussian measurement noise with an amplitude fluctuation of 5% is superimposed on the input elevation data to simulate the discrete measurement random distortion caused by the disturbance of the water flow in the field environment. The ratio between the high-frequency subdivision step size and the wideband coarse step size used to divide the topographic grid matrix is ​​determined by the processor calculating the shear stress response characteristic topology of the grid nodes. The numerical range of the ratio controls the calculation of the nonlinear catastrophe core. The data refresh allocation frequency of the submatrix and boundary smooth gradient auxiliary calculation submatrix is ​​determined. When the variation intensity of the elevation gradient is in a high-frequency jump state, to ensure that the elevation gradient digital filter does not accumulate local truncation errors when handling strongly nonlinear wave-current coupled load conditions, the proportional relationship tends to be less than or equal to 0.10. In this state, the high-frequency subdivision step size is set to 0.1 seconds and the wideband coarse step size is set to 2.0 seconds, so that the calculated proportional relationship is 0.05. The experimental test includes an experimental group, a first control group, a second control group, and a third control group. The experimental group uses gradient data input including low-strain, medium-strain, and high-strain conditions, with corresponding step size proportional relationships set to 0.10, 0.0, and 0.0, respectively. The first control group had its mesh generation function for the shear stress response characteristic topology map disabled, and its step size ratio was maintained at a symmetrical refresh state of 1.0. The second control group had its data distribution path for the elevation gradient digital filter disabled. The third control group had its step size ratio set to 0.15. After the initial terrain mesh flow was input, the maximum magnitude of the spatial slope vector extracted by the medium-strain condition test group with a ratio of 0.05 stabilized at 24.3 degrees and did not exceed the critical repose angle threshold of 28.5 degrees. However, in the high-strain condition test group with a ratio of 0.01, when the time progressed to 12.4 minutes, the elevation gradient digital filter detected that the spatial slope vector magnitude of the local mesh node reached 28.8 degrees, exceeding the critical repose angle threshold of 28.5 degrees. The system executes elevation data allocation rules when the critical angle of repose for collapse is reached. The processor reads the elevation changes of the core nodes that exceed the critical angle of repose for collapse and constructs a three-dimensional elevation overflow vector. The diffusion weight matrix is ​​multiplied with the three-dimensional elevation overflow vector to determine the apportionment weight coefficient of each neighboring node. Based on the apportionment weight coefficient, the excess elevation changes are evenly rewritten to the register addresses of the surrounding neighboring nodes, and the spatial slope vector is controlled to return to the preset boundary. The timing convergence time of the first control group without the grid partitioning and refresh mechanism reached 68.4 minutes. The second control group with the data allocation path closed stopped numerical calculation at 3.5 minutes because the local gradient was too large and could not converge. The step size ratio was 0.The third control group (group 15) experienced 17.0 data overflow events in the temporary register inside the elevation gradient digital filter due to an excessively large time step, resulting in a spatial positional deviation of 0.45 meters in the maximum scour depth. In contrast, the experimental groups with time step ratios of 0.10, 0.05, and 0.01 had timing convergence times of 13.8 minutes, 11.2 minutes, and 8.1 minutes, respectively, with corresponding high-frequency noise suppression ratios of 89.2%, 93.4%, and 96.7%. Furthermore, the measurement deviations in the final output scour pit development data regarding the maximum scour depth, scour range, and cable suspension span were all within the 3.1% limit.

[0047] When the ratio of the high-frequency subdivision step size to the wideband coarse step size remains within the range of less than or equal to 0.10, and when the elevation gradient digital filter triggers the elevation data allocation rule to rewrite the elevation change amount of the register addresses of the surrounding neighboring nodes when the spatial slope vector magnitude is greater than the critical repose angle threshold of the landslide, the time-series update computational overhead of the boundary smooth gradient auxiliary computation submatrix is ​​reduced, and the local elevation change discrete trajectory of the nonlinear mutation core computation submatrix converges, thereby enabling the processor to output scour pit development trend data containing the maximum scour depth, scour range, and the suspended span of the submarine cable.

[0048] Example 4: During long-term wave-current interaction, the surge load of a specific period superimposes with the steady-state ocean current at the bottom in phase. Geological fissures are exposed at the seabed foundation edges surrounding marine structures, causing the digital data processing system to face a situation where local elevation gradients frequently cross critical collapse points. At this time, the grid nodes in the initial topographic grid flow exhibit highly dynamic variation due to abrupt changes in the intensity of scouring. Because the elevation data of the grid nodes in the abrupt change region undergoes discontinuous step changes within a short period, existing digital filters and solution loops often experience step-damped oscillations when processing discontinuous discrete-time signals under this condition due to the drastic differences in data items within the sliding time window. This results in discretization derivation paths for the elevation state vector being in multiple... After each iteration, the data deviates from the physical boundary, or the spatial gradient of the elevation difference temporarily stored in the register is too large, directly triggering numerical truncation in the bus register of the electronic digital data processing system, causing system overflow errors and abnormal computational interruptions. This local grid data divergence caused by load disturbances at the physical boundary limits the solution stability of the electronic digital data processing system for long-term situation prediction under specific operating conditions. To solve the digital divergence bottleneck under the above operating conditions, the method and system claimed in this invention provide a closed operating procedure that coordinates parameter calibration procedures and adaptive gradient amortization logic. The processor executes the initial state definition program, and the processor's performance specifications are set to have double-precision floating-point arithmetic capabilities and a basic clock frequency of not less than [missing information]. The processor reads the pre-stored initial mesh topology matrix from memory. This initial mesh topology matrix applies to the three-dimensional discrete terrain mesh of the target sea area, and the spatial edge length and edge characteristics of the mesh nodes are set as follows: It is used to provide boundary detection for scour patterns, and its operating environment is limited to running on a dual-core, quad-thread hardware computing platform, possessing at least... The invention utilizes a dynamic temporary storage memory space to ensure the baseline bandwidth for reading and writing high-dimensional terrain matrices. To structurally quantify the spatiotemporal convergence behavior of elevation gradient digital filters when processing abrupt discrete signals, this invention constructs an objectively measurable objection strength index. To quantify the non-monotonic variation behavior of multi-source data input streams during spatial topology reconstruction, this invention borrows the impulse response principle from discrete signal processing, treating grid nodes in the terrain grid stream that undergo unidirectional step changes in elevation as discrete objection impulse signals. Specifically, the objection strength index is defined by calculating the second-order difference value of the cumulative spatial slope vector change of the discrete objection impulse signal within a preset time step unit. During parameter calibration, the processor allocates multiple sets of initial terrain grid streams with different shear stress spatial gradients to the elevation gradient digital filter, examining the impact of different gradient thresholds on the grid data convergence path. In the process of setting the gradient threshold, when the spatial gradient of the difference between the experimental input bed shear stress calculation data and the critical starting shear stress threshold of sediment is at a low intensity level, the system operates in a safe state. However, when the spatial gradient of the difference increases to the point of generating highly dynamic scour abrupt changes, if the gradient threshold is set too low, the system will frequently misclassify stable nodes as core nodes, leading to an increase in the number of topological vertices in the nonlinear abrupt change core calculation submatrix, causing the processor to be overloaded by redundant overhead in register addressing. If the gradient threshold is set too high, some terrain nodes that have experienced severe scour will be missed and included in the boundary stable gradual change auxiliary calculation submatrix. This will cause the abrupt change node to perform periodic refresh using a wide-band coarse step size. Since the time step size is not refined in time, the elevation value of the node will have accumulated to a collapse span that exceeds physical common sense before the next time-series refresh arrives, and the temporary storage register inside the elevation gradient digital filter will frequently diverge in value.

[0049] Test data shows that when the set gradient threshold is greater than or equal to per meter At this point, the spatial convergence stability of the elevation gradient digital filter saturates, and when the set gradient threshold is less than per meter... At that time, the memory bus bandwidth utilization exceeded This leads to time-series updates experiencing latency due to waiting for hardware responses. Therefore, by calculating performance criteria under specific operating conditions, the parameter range for setting the gradient threshold is limited to per meter. Up to per meter Within the working window; from the perspective of engineering physics and mesh convergence boundary, when the set gradient threshold is below 0.05 meters per meter, its value is close to the random fluctuations in background shear stress caused by Gaussian measurement noise and water flow disturbance. This will cause the system to frequently misjudge and fragment normal stable region meshes into the core calculation matrix, resulting in a sudden increase in bus addressing load and unnecessary computational power consumption. When the set gradient threshold is above 0.25 meters per meter, it exceeds the critical upper limit for large-scale shear deformation of the bottom sediment. This will cause mesh nodes that have already undergone rapid scouring changes to be incorrectly ignored and remain in the auxiliary submatrix, making it impossible to refine the time step in time. This will lead to cross-cycle numerical overflow of the elevation gradient under a coarser time step. In the current offline calibration operation, the processor will... The input filter for the abrupt terrain mesh, with a differential spatial gradient of 0.35 meters per meter for the bed shear stress, is set to a gradient threshold of 0.12 meters per meter. The overall 3D terrain mesh matrix is ​​decoupled into a nonlinear abrupt core computational submatrix containing 1250 topological vertices and a boundary smooth gradient auxiliary computational submatrix containing 8750 topological vertices. At this point, the processor executes the elevation data allocation rules, setting the ratio of the high-frequency subdivision step size to the wideband coarse step size to 0.02. The time convergence time of the nonlinear abrupt core computational submatrix is ​​shortened to 7.4 minutes, and the peak value of the mesh discrete pulse oscillation wave, characterized by the objectivity intensity index, decreases from 45.2 before allocation to 1.1. The maximum modulus of the local spatial slope vector is locked at 26.1 degrees, and no data truncation or overflow occurs in the elevation direction.

[0050] After completing iterative updates at a set time step based on the parameter calibration procedure, the system outputs data on the development of scour pits, including the maximum scour depth, scour range, and the span of the suspended submarine cable. This transforms the severe seabed scour state under nonlinear wave-current coupling conditions into a stable flow of a high-dimensional discrete digital matrix, achieving a logical closed loop of boundary feature capture and low computational power allocation. This mechanism enables the electronic digital data processing system to fully implement the abstract mesh reconstruction logic in the processor's register addressing and step-size state vector allocation when facing scenarios with drastic changes in seabed topography. It also closes off local numerical oscillations by utilizing a parameter calibration procedure based on hard physical boundaries. The path utilizes the processor to perform corresponding fault tolerance and calibration. When the system encounters a power outage or abnormal reset, it automatically recovers the state vector from the last successfully written non-volatile memory address. Before each data transfer, it performs a data self-check using a cyclic redundancy check code to combat physical layer interference. Without adding external measurement components, it limits the risk of digital filter overflow caused by elevation changes to within a preset safety boundary. This not only ensures the continuity of the data chain during the dynamic reproduction of high-velocity scour pits but also eliminates the numerical overflow phenomenon caused by data mutations, thereby improving data processing timeliness and grid convergence reliability.

[0051] Example 5: When the system faces the heterogeneous hydrodynamic conditions of a newly deployed sea area, initial alignment misalignment occurs between the three-dimensional seabed discrete grid flow constructed in different batches and the historical wave-current dynamics model of the specific sea area. Due to the lack of on-site calibration for the specific seabed sediment gradation, dynamic impedance variation, and sensor reference drift, the elevation gradient digital filter is prone to generating pseudo-shear stress spatial gradients in the early stages of iteration, causing erroneous splitting of the nonlinear mutation core calculation sub-matrix. This can induce data block rerouting address overwriting conflicts in the processor, and even lead to excessive initial gain of discrete time-series signals, resulting in false shear oscillations in grid elevation. This limits the initial stability of the multi-source feature sensing mechanism in the new deployment environment and cannot provide a reliable basic simulation baseline for subsequent large-scale situation prediction. In order to eliminate the uncertainty of the initial calculation state caused by this, an on-site deployment pre-calibration procedure needs to be executed before data processing. The independent initialization calibration module configured in the hardware computing platform schedules the processor to read the discrete hydrodynamic instantaneous measured signals from the on-site sensors after de-identification processing and compares them with the critical initiation shear stress data of the seabed sediment generated by offline simulation.

[0052] The initialization calibration module acquires the reference residual voltage signal of the sensor under non-manifested scouring conditions, establishes a signal zeroing baseline, and clears the local temporary register of the processor. Then, it reads the changes in raw hydrodynamic physical parameters input from the seabed pressure sensor and current meter over 30 consecutive sampling periods. Under the initial stable state where the background noise variance is less than the preset signal jitter limit, the system initiates the gradient threshold in-situ calibration process. The processor extracts the median particle size and porosity of sediment recorded in the geological survey sample box data of the newly deployed sea area, using these as pre-inputs to determine the initial physical constant group. The numerical solver calculates the static shear stress distribution of the seabed under the corresponding normal flow velocity, and multiplies the calculated spatial shear stress variance envelope by... A stability correction coefficient of 1.25 is written into the processor's read-only register as a gradient threshold specific to this deployment environment. This allows the initial evolution trajectory to be locked through textual adaptive quantization deduction of the physical properties of this sea area. After this step is completed, the processor allocates a pulse calibration operator to inject an artificial step disturbance with an amplitude of 0.02 meters into the elevation state vector and monitors the response damping attenuation process of the elevation gradient digital filter within three high-frequency subdivision step sizes. If the second derivative of the output spatial slope vector shows a stable and monotonically decreasing trend, the initial state definition is determined to be complete. The calibration system locks the current register state as the sole starting point for global data iteration and outputs the calibrated initial high-dimensional discrete digital matrix for situation monitoring.

[0053] Example 6: In the engineering operation of long-term scour monitoring of the seabed foundation of an offshore booster station, the monitoring system needs to establish an initial situation analysis benchmark under the nonlinear load state of coupled ocean current velocity and wave morphology. Because the original geological grid matrix and wave load matrix collected on site have deviations due to sensor hardware clock drift and installation attitude angle, if they are directly used in the calculation without calibration, the calculated scour pit depth will deviate from the physical reality. Therefore, before the system is officially put into operation, the offline parameter calibration procedure is started. The processor reads and compares the temporal overlap area of ​​the geological survey sample box data and hydrodynamic observation data stored in the memory. The flow field vector is translated and calibrated using the real-time flow velocity reading of the benchmark flow point. The critical starting shear stress function threshold of the bed sediment after calibration is used as the endogenous judgment benchmark of the subsequent numerical solver. After the benchmark is completed, the processor stores the current topographic grid flow state vector into the non-volatile cache unit as the initial evolution reference for subsequent data updates.

[0054] During the monitoring period, the processor initiates adaptive monitoring logic to discretize and track the evolution of the terrain matrix. It calculates the scour pit depth change rate index to provide feedback on the current scour stage. When the processor detects that the depth change rate index is in the first development stage, it schedules the numerical solver to configure a larger computation time step to reduce the iterative computation load per unit period. As the scour pit development intensifies, the processor identifies the threshold of the depth change rate index entering the violent development stage and triggers a refined monitoring operator to reduce the computation time step to a millisecond-level subdivision step to enhance the transient capture capability of local scour depth abrupt change nodes. The processor compares the elevation difference of grid nodes in adjacent periods with the preset critical rest angle function threshold for mud and sand landslide in real time. Once it is identified that the shear stress vector magnitude generated by the local area grid elevation change rate exceeds the safety criterion, the processor executes the elevation rerouting allocation rule and allocates the excess amount to the surrounding register addresses according to the spatial diffusion weight matrix to ensure the numerical convergence of local spatial gradient data.

[0055] The calibration procedure and adaptive monitoring logic constitute the time-series convergence core of the monitoring and analysis system. By dynamically matching the calculation accuracy at different stages of scour development, the system can maintain the steady-state iterative update of the discrete terrain matrix under the scenario of sudden changes in ocean current load. This not only ensures the numerical accuracy of the scour pit morphology reproduction, but also avoids memory bus overload and time delay caused by global subdivision calculation. The system outputs situational data including the maximum scour depth, scour range and the span of the suspended submarine cable, providing quantitative conclusions with engineering timeliness for the optimization of protection schemes.

[0056] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dynamic monitoring and analysis method for scour pits based on wave-current coupling, characterized in that, Includes the following steps: Step S1: Obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; Step S2: Call the numerical solver, divide the initial terrain grid flow into a core submatrix and an auxiliary submatrix according to the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, configure an asymmetric time step to iterate the core submatrix and the auxiliary submatrix, and update the node elevation vectors in the core submatrix. Step S3: Use the elevation gradient digital filter to read the elevation difference between the core node and the neighboring nodes in the core submatrix, and calculate the spatial slope vector. Step S4: When the magnitude of the spatial slope vector is greater than the threshold of the landslide critical repose angle that is synchronously updated under the time step, the elevation data allocation rule is triggered. Step S5: Using the diffusion weight matrix, the elevation change of the core node exceeding the critical repose angle threshold of the landslide is spread and rewritten to the register address of the neighboring node, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration. Step S6: After completing the preset cycle of iteration, output the scour pit development trend data, which includes the maximum scour depth, scour range, and submarine cable suspension span.

2. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, Step S1 includes the following sub-steps: Step S11, calculate the extreme value distribution of hydrodynamic observation data to determine the design wave height and design flow velocity under the target return period; Step S12, generate a wave-current load matrix containing multi-dimensional parameters by algebraically superimposing the design wave height and design flow velocity through a two-dimensional nonlinear spatial mapping relationship; Step S13, divide the geological grid matrix into multiple independent regions using topological splitting scheduling units to construct the initial topographic grid flow.

3. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, Step S5 includes the following sub-steps: Step S51, call the elevation gradient digital filter in the region of the core node to capture the elevation data overflow; Step S52, read the elevation change of the core node that exceeds the critical repose angle threshold of the landslide, and construct a three-dimensional elevation overflow vector; Step S53, multiply the diffusion weight matrix with the three-dimensional elevation overflow vector to determine the apportionment weight coefficient of each neighboring node; Step S54, write the elevation change into the register address of each neighboring node according to the apportionment weight coefficient, and reconstruct the terrain elevation data of the core sub-matrix.

4. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, Step S4 includes the following sub-steps: Step S41, obtain the seabed soil type, median grain size and internal friction angle parameters of the target sea area; Step S42: Match the corresponding static angle of repose reference value according to the seabed soil type and median particle size; Step S43: Correct the static angle of repose reference value by using the internal friction angle parameter attenuation, and generate a dynamic critical angle of repose function that has a negative monotonic relationship with the real-time ocean current velocity, as the critical angle of repose threshold for landslide.

5. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, Step S2 includes the following sub-steps: Step S21, calculate the combined wave-current bottom shear stress at the bottom grid nodes and compare it with the critical shear stress for sediment initiation; Step S22, when the combined wave-current bottom shear stress is greater than the critical shear stress for sediment initiation, calculate the local net sediment transport rate of the grid nodes; Step S23, based on the sediment continuity equation, update the node elevation vector in the core submatrix using the divergence of the local net sediment transport rate.

6. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, The method also includes the following steps: Step S7, storing the output data on the development trend of scour pits into the submarine cable foundation status database in time series, and constructing the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable.

7. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 6, characterized in that, Step S7 includes the following sub-steps: Step S71, calculate the first time derivative of the historical time series of the maximum scour depth and the historical time series of the suspended span of the submarine cable to determine the dynamic evolution rate of the current scour pit depth and the expansion rate of the suspended span of the submarine cable.

8. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 7, characterized in that, Step S7 also includes the following sub-steps: Step S72, determining whether the dynamic evolution rate is greater than the preset scour development rate limit threshold, or whether the span expansion rate is greater than the preset span extension rate limit threshold; Step S73, issuing a warning signal for the risk of instability of the submarine cable foundation structure when the dynamic evolution rate is greater than the scour development rate limit threshold, or the span expansion rate is greater than the span extension rate limit threshold.

9. The method for dynamic monitoring and analysis of scour pits based on wave-current coupling according to claim 1, characterized in that, The method also includes the following steps: Step S8, obtaining the preset physical threshold matrix for submarine cable operation safety; Step S9, inputting multiple sets of preset protective structure geometric parameter matrices into the numerical solver for cyclic updating, outputting secondary scour pit development trend data under the protective configuration, and extracting the temporal extreme value envelope matrix over the entire cycle; Step S10, comparing the boundary range of the temporal extreme value envelope matrix with the physical threshold matrix for submarine cable operation safety, and selecting the optimal protective design parameter matrix with the minimum maximum scour depth and the submarine cable suspension span not exceeding the standard.

10. A dynamic monitoring and analysis system for scour pits based on wave-current coupling, used to implement the dynamic monitoring and analysis method for scour pits based on wave-current coupling as described in any one of claims 1 to 9, characterized in that, The system includes: The initial terrain grid flow construction module is used to obtain the geological grid matrix and wave-current load matrix, and construct the initial terrain grid flow; The matrix partitioning and elevation update module is used to call the numerical solver. Based on the differential spatial gradient between the shear stress of the grid nodes and the starting shear stress threshold, the initial terrain grid flow is divided into a core sub-matrix and an auxiliary sub-matrix. The core sub-matrix and auxiliary sub-matrix are iterated in a loop with an asymmetric time step, and the node elevation vectors in the core sub-matrix are updated. The spatial slope vector calculation module is used to read the elevation difference between the core node and the neighboring node in the core submatrix using the elevation gradient digital filter, and to calculate the spatial slope vector. The allocation rule triggering module is used to trigger the elevation data allocation rule when the magnitude of the spatial slope vector is greater than the landslide critical repose angle threshold updated synchronously under the time step; the elevation data amortization and rewriting module is used to amortize and rewrite the elevation change of the core node that exceeds the landslide critical repose angle threshold to the register address of the neighboring node using the diffusion weight matrix, so that the modified spatial slope vector is stabilized within the safe boundary, ensuring the convergence stability of the initial terrain grid flow time sequence iteration; The situation data output module is used to output the development situation data of the scour pit after completing the preset cycle iteration. The development situation data of the scour pit includes the maximum scour depth, scour range and the span of the submarine cable.

Images