A coastal channel maintenance decision-making method and system based on a siltation mathematical model

Abstract

The application discloses a coastal channel maintenance decision-making method and system based on a siltation mathematical model. The method comprises the following steps: constructing a channel full-life-cycle database; coupling a tidal current, a wave and a sediment model by using an unstructured grid and an MCT coupler to establish a channel siltation mathematical model; correcting siltation parameters, optimizing the mathematical model and predicting future siltation by using an LSTM time series prediction model, and constructing a siltation risk atlas; generating early warning information according to the atlas, determining dredging engineering quantity and formulating a dredging scheme. The application realizes accurate prediction of channel siltation and quantitative decision-making of a maintenance scheme, and improves the scientificity and economy of channel maintenance.

Description

Technical Field

[0001] This invention relates to the field of waterway maintenance decision-making technology, and in particular to a coastal waterway maintenance decision-making method and system based on a siltation mathematical model. Background Technology

[0002] Channel siltation is a common problem in waterway operation, directly affecting navigation capacity and safety. Analysis and scientific maintenance of channel siltation are core tasks of waterway management. Currently, some waterways have low levels of informatization and intelligentization in their construction and maintenance, and many technical problems exist.

[0003] First, waterway maintenance lacks intelligent methods, still relying primarily on on-site surveys and measurements. This prevents automated prediction based on data and models, resulting in a lack of scientific basis for dredging priorities and frequencies, high maintenance costs, and potential navigational safety risks. Second, waterway data resources are scattered; multi-source, heterogeneous historical and measured data lack centralized management and standardized integration, failing to provide a comprehensive and accurate data foundation for high-precision siltation analysis. Third, existing siltation prediction models lack accuracy, relying heavily on single empirical formulas or simple mechanistic models. These models cannot accurately reflect the spatiotemporal characteristics of sediment deposition and fail to effectively integrate with data-driven technologies, making it difficult to achieve large-scale, long-term, and high-precision predictions. Furthermore, existing systems lack an integrated intelligent decision-making system. Early warning, engineering quantity calculation, and maintenance decisions rely heavily on manual processes, resulting in low efficiency and large errors, failing to form a closed loop of "prediction-early warning-decision-feedback." Finally, current early warning methods are mostly based on threshold judgments from single data sources, lacking the ability to couple the siltation evolution trend with navigational risks, leading to delayed early warning responses and a disconnect from subsequent maintenance decisions.

[0004] In summary, existing technologies are insufficient to meet the requirements of modern waterways for safety, efficiency, and lean management. There is an urgent need for a full-process intelligent waterway maintenance decision-making solution that integrates multi-source data, intelligent prediction, automatic early warning, and precise decision-making. Summary of the Invention

[0005] To address at least one of the aforementioned technical problems, this invention proposes a coastal waterway maintenance decision-making method and system based on a siltation mathematical model.

[0006] The first aspect of this invention provides a coastal waterway maintenance decision-making method based on a siltation mathematical model, comprising: Collect multi-source heterogeneous data of the target waterway area, clean and convert the multi-source heterogeneous data to obtain a standardized waterway dataset, and construct a waterway construction and maintenance full life cycle database from the standardized waterway dataset. The target channel area is divided into grids using an unstructured grid, and the standardized channel dataset is coupled in real time with the current tidal model, wave model and sediment model based on the MCT coupler to construct a mathematical model for channel siltation. The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of the waterway within a preset time period are obtained according to the waterway siltation mathematical model. The features of the waterway hydraulic parameters are input into the data-driven model to correct the siltation parameters and obtain the corrected siltation parameters. The channel siltation mathematical model is optimized based on the corrected siltation parameters. Based on the optimized channel siltation mathematical model, the siltation situation in the future preset time period is predicted, and a siltation risk map is constructed. Based on the siltation risk map, siltation early warning information is generated, and based on the siltation early warning information, the dredging volume at each location in the target waterway area is determined; The dredging plan for the target waterway area is determined based on the amount of dredging work.

[0007] In this solution, the process involves collecting multi-source heterogeneous data from the target waterway area, cleaning and converting the data to obtain a standardized waterway dataset, and then constructing a waterway construction and maintenance lifecycle database from this standardized waterway dataset. Specifically: Collect multi-source heterogeneous data for the target waterway area. The multi-source heterogeneous data includes real-time collected underwater topographic point cloud data, hydrological time series data, wave spectrum data, meteorological data, water quality data, as well as historical waterway survey datasets, maintenance and dredging project records, and waterway construction drawings. Data cleaning is performed on the multi-source heterogeneous data. The data cleaning includes removing abnormal redundant data based on a preset reasonableness threshold range, correcting data mutation error values ​​using interpolation algorithms, and supplementing missing data based on adjacent spatiotemporal data points. After data cleaning, the multi-source heterogeneous data is unified in terms of coordinate system and data format to obtain a standardized waterway dataset. The standardized waterway dataset is then stored according to the collection time sequence and spatial location to construct a waterway construction and maintenance full life cycle database.

[0008] In this scheme, an unstructured mesh is used to divide the target waterway area into meshes, and the standardized waterway dataset is coupled in real time with the tidal current model, wave model, and sediment model based on the MCT coupler to construct a mathematical model for waterway siltation. Specifically: Unstructured grids are used to divide the target waterway area into grids, and the grids are densified in the main waterway area and the siltation-sensitive area. Based on the finite volume method, a set of flow control equations including continuity equation, momentum equation and state equation is established for the grid. Coriolis force, wind stress and Manning drag coefficient terms are introduced into the momentum equation of the flow control equation set. The Euler-Lagrange method is used to discretize and solve the flow control equation set to form a tidal flow model. The third-generation wave spectrum model SWAN is introduced to establish a wave propagation model based on the wave action conservation equation. Wave breaking, bottom friction, white crown dissipation and wave-wave nonlinear interaction source terms are introduced into the wave propagation model, and the outer sea boundary is set as the incident wave spectrum boundary condition to form a wave model. A control equation for the transport of viscous sediment based on a three-dimensional convection-diffusion equation is established. The bottom shear stress term under the combined action of water flow and waves is introduced into the control equation for the transport of viscous sediment. Combined with the formulas for critical initiation shear stress, settling velocity and resuspension flux of sediment, the scouring, transport, flocculation and settling process of viscous sediment is described, thus forming a sediment model. A coupling framework is constructed based on the model coupling tool MCT to connect the tidal current model, wave model, and sediment model. In this coupling framework, the effective wave height, wave period, and wave propagation direction calculated by the wave model are transmitted to the tidal current model in real time to update the bottom friction. The flow velocity, flow direction, and water level calculated by the tidal current model are transmitted to the sediment model in real time as driving conditions for sediment transport. The shear stress at the bottom of the water flow calculated by the tidal current model and the wave model are vector superimposed with the shear stress at the bottom of the wave, and the total shear stress after superposition is transferred to the sediment model to drive the scouring and initiation calculation of the sediment model, and are integrated to form a mathematical model for channel siltation.

[0009] In this scheme, the LSTM time-series prediction model is selected as the data-driven model. Based on the channel siltation mathematical model, the channel hydraulic parameters within a preset time period are obtained. The characteristics of these hydraulic parameters are then input into the data-driven model to correct the siltation parameters, resulting in corrected siltation parameters. Specifically: The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of each grid node in the target waterway area calculated by the waterway siltation mathematical model within a preset historical time period are used as the input feature sequence of the LSTM model. The waterway hydraulic parameters include flow velocity, sediment concentration, water level and wave parameters. The siltation parameters of each grid node are measured within the same preset time period and used as training labels to supervise the training of the LSTM model. The siltation parameters include siltation thickness, siltation rate, and siltation volume. The nonlinear mapping relationship from channel hydraulic parameters to siltation parameters is identified. The predicted hydraulic parameters of the waterway, calculated by the waterway siltation mathematical model for a future preset time period, are input into the trained LSTM model, and the siltation parameter correction values ​​for the corresponding time period and location are output to obtain the corrected siltation parameters.

[0010] In this scheme, the optimization of the channel siltation mathematical model based on the corrected siltation parameters, the prediction of siltation conditions within a preset time period based on the optimized channel siltation mathematical model, and the construction of a siltation risk map are specifically as follows: Obtain the initial silting prediction parameters output by the channel silting mathematical model, and calculate the prediction deviation sequence between the initial silting prediction parameters and the corrected silting parameters for each grid node within a preset historical time period; Based on the predicted deviation sequence, the distribution pattern of the predicted deviation in the time and space dimensions is identified, and based on the distribution pattern, the direction and magnitude of the optimization adjustment of the model parameters related to sediment transport and settling processes in the channel siltation mathematical model are determined. Based on the optimized adjustment direction and magnitude, the model parameters in the channel siltation mathematical model are adjusted to obtain the optimized channel siltation mathematical model. Based on the optimized channel siltation mathematical model, by inputting boundary condition driving data within a future preset time period, the predicted values ​​of siltation thickness, siltation rate, and siltation volume of each grid node in the target channel area within the future preset time period are obtained. Based on the navigation safety standards, a threshold for the bottom elevation of the channel and a warning threshold for the siltation rate are set. The predicted siltation thickness of each grid node is compared with the threshold for the bottom elevation of the channel, and the predicted siltation rate is compared with the warning threshold for the siltation rate to determine the siltation risk of each grid node. Based on the aforementioned siltation risk, a siltation risk map is constructed, which includes the risk levels of different grid nodes and the spatial distribution of predicted siltation volume.

[0011] In this scheme, the step of generating siltation early warning information based on the siltation risk map and determining the dredging volume at each location in the target waterway area based on the siltation early warning information specifically involves: Extract the location coordinates, siltation risk level, predicted siltation thickness, and predicted siltation rate of each grid node in the siltation risk map to form a multi-dimensional node feature vector; Based on the DBSCAN clustering algorithm, the latitude and longitude coordinates of the grid node location are used as spatial attributes, the Euclidean distance of the grid node is used as a measure of spatial proximity, and the weighted combination of the siltation risk level, the predicted value of siltation thickness, and the predicted value of siltation rate is used as a multi-dimensional feature measure to calculate the Euclidean distance between nodes. Using nodes whose Euclidean distance is less than the preset neighborhood radius and whose number of nodes in the neighborhood of the core point is not less than the preset minimum number of points as core points, starting from the core point, the grid nodes that meet the preset density reachability conditions are clustered and expanded, and the grid nodes that are spatially continuous and have similar backfilling characteristics are aggregated into different independent clusters. Calculate the predicted average backfill thickness and the predicted average backfill rate of all grid nodes within each independent cluster. Determine the regional range of each independent cluster based on the location of all grid nodes within the independent cluster. Construct backfill early warning information for each independent cluster, including the regional range, predicted average backfill thickness, and rate. For each independent cluster, the dredging thickness of each node is calculated based on the current predicted channel bottom elevation, the initial channel design bottom elevation, and the navigation maintenance bottom elevation standard of each grid node within it. The dredging thickness is then multiplied by the area of ​​each independent cluster region to obtain the dredging volume of the grid node.

[0012] In this plan, the step of determining the dredging scheme for the target waterway area based on the dredging volume specifically includes: Based on the siltation risk level, predicted average siltation thickness, siltation rate and area range in the siltation early warning information of each independent cluster in the target waterway area, calculate the dredging urgency index of each independent cluster. Based on the navigation requirements of the waterway, data on design water depth, waterway width, slope gradient, navigation density and vessel tonnage are extracted from the waterway maintenance database to calculate the improvement index of navigation efficiency of each independent cluster of dredging. Data on the working status, construction capacity, scheduling distance, and available capacity of the sludge discharge area of ​​the construction vessels are obtained. Based on the dredging volume, dredging urgency index, and improvement index for navigation efficiency, a multi-objective optimization algorithm is adopted, with the goal of minimizing the total construction time, to determine the dredging operation plan, including the priority order of construction areas, the scheduling plan of construction vessels, the allocation of dredging volume, the construction window period, and the construction path.

[0013] A second aspect of the present invention also provides a coastal waterway maintenance decision-making system based on a siltation mathematical model. The system includes a memory and a processor. The memory includes a coastal waterway maintenance decision-making method program based on a siltation mathematical model. When the processor executes the coastal waterway maintenance decision-making method program based on the siltation mathematical model, it implements the steps of the coastal waterway maintenance decision-making method based on the siltation mathematical model as described in any of the preceding claims.

[0014] This invention discloses a coastal waterway maintenance decision-making method and system based on a siltation mathematical model. It includes: constructing a waterway lifecycle database; establishing a waterway siltation mathematical model by coupling tidal current, wave, and sediment models using an unstructured grid and an MCT coupler; correcting siltation parameters using an LSTM time-series prediction model, optimizing the mathematical model, predicting future siltation, and constructing a siltation risk map; generating early warning information based on the map, determining dredging work volume, and formulating dredging plans. This invention achieves accurate prediction of waterway siltation and quantitative decision-making on maintenance plans, improving the scientific and economic efficiency of waterway maintenance. Attached Figure Description

[0015] Figure 1 A flowchart of a coastal waterway maintenance decision-making method based on a siltation mathematical model is shown in this invention. Figure 2 The flowchart illustrating the mathematical model for constructing channel siltation according to the present invention is shown. Figure 3 The flowchart illustrating the correction of siltation parameters according to the present invention is shown; Figure 4 A block diagram of a coastal waterway maintenance decision system based on a siltation mathematical model is shown. Detailed Implementation

[0016] 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.

[0017] 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.

[0018] Figure 1 The flowchart of a coastal waterway maintenance decision-making method based on a siltation mathematical model is shown.

[0019] like Figure 1 As shown, the first aspect of the present invention provides a coastal waterway maintenance decision-making method based on a siltation mathematical model, comprising: Collect multi-source heterogeneous data of the target waterway area, clean and convert the multi-source heterogeneous data to obtain a standardized waterway dataset, and construct a waterway construction and maintenance full life cycle database from the standardized waterway dataset. The target channel area is divided into grids using an unstructured grid, and the standardized channel dataset is coupled in real time with the current tidal model, wave model and sediment model based on the MCT coupler to construct a mathematical model for channel siltation. The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of the waterway within a preset time period are obtained according to the waterway siltation mathematical model. The features of the waterway hydraulic parameters are input into the data-driven model to correct the siltation parameters and obtain the corrected siltation parameters. The channel siltation mathematical model is optimized based on the corrected siltation parameters. Based on the optimized channel siltation mathematical model, the siltation situation in the future preset time period is predicted, and a siltation risk map is constructed. Based on the siltation risk map, siltation early warning information is generated, and based on the siltation early warning information, the dredging volume at each location in the target waterway area is determined; The dredging plan for the target waterway area is determined based on the amount of dredging work.

[0020] It should be noted that by collecting multi-source heterogeneous data from the target waterway area and performing data cleaning and format conversion, abnormal data and noise interference can be effectively eliminated, achieving a unified expression of multiple data types in both time and space dimensions. This allows for the construction of a complete, continuous, and high-quality database covering the entire lifecycle of waterway construction and maintenance. Based on this, unstructured grids are used to finely divide the waterway area, and the tidal current, wave, and sediment models are dynamically coupled using an MCT coupler. This comprehensively reflects the interactions between multiple physical processes, significantly improving the simulation accuracy of the waterway siltation mathematical model for complex hydrodynamic and sediment transport processes. Furthermore, an LSTM time-series prediction model is introduced to further enhance the model. By learning and correcting the output hydraulic parameters, nonlinear evolution patterns in historical data can be uncovered, enabling dynamic correction of siltation parameters and compensating for prediction biases in traditional mechanistic models under uncertain conditions. Based on the correction results, the siltation mathematical model is optimized, and future siltation evolution is predicted, generating a siltation risk map with spatial distribution characteristics, making the risk levels of different areas of the waterway more intuitive. Generating siltation early warning information and calculating dredging volumes at each location helps to accurately identify key siltation areas and their treatment needs, achieving a refined assessment of dredging scale. Finally, by combining dredging volumes to formulate waterway maintenance plans, rational allocation of construction resources and work arrangements can be achieved.

[0021] According to an embodiment of the present invention, the process of collecting multi-source heterogeneous data of the target waterway area, cleaning and format conversion of the multi-source heterogeneous data to obtain a standardized waterway dataset, and constructing a waterway construction and maintenance full lifecycle database from the standardized waterway dataset, specifically involves: Collect multi-source heterogeneous data for the target waterway area. The multi-source heterogeneous data includes real-time collected underwater topographic point cloud data, hydrological time series data, wave spectrum data, meteorological data, water quality data, as well as historical waterway survey datasets, maintenance and dredging project records, and waterway construction drawings. Data cleaning is performed on the multi-source heterogeneous data. The data cleaning includes removing abnormal redundant data based on a preset reasonableness threshold range, correcting data mutation error values ​​using interpolation algorithms, and supplementing missing data based on adjacent spatiotemporal data points. After data cleaning, the multi-source heterogeneous data is unified in terms of coordinate system and data format to obtain a standardized waterway dataset. The standardized waterway dataset is then stored according to the collection time sequence and spatial location to construct a waterway construction and maintenance full life cycle database.

[0022] It should be noted that the interpolation algorithm includes inverse distance weighted interpolation and Kriging interpolation.

[0023] Figure 2 The flowchart of the mathematical model for constructing channel siltation according to the present invention is shown.

[0024] According to an embodiment of the present invention, the method of using an unstructured mesh to divide the target channel area into meshes, and then coupling the standardized channel dataset with the tidal current model, wave model, and sediment model in real time based on the MCT coupler to construct a channel siltation mathematical model, specifically: Unstructured grids are used to divide the target waterway area into grids, and the grids are densified in the main waterway area and the siltation-sensitive area. Based on the finite volume method, a set of flow control equations including continuity equation, momentum equation and state equation is established for the grid. Coriolis force, wind stress and Manning drag coefficient terms are introduced into the momentum equation of the flow control equation set. The Euler-Lagrange method is used to discretize and solve the flow control equation set to form a tidal flow model. The third-generation wave spectrum model SWAN is introduced to establish a wave propagation model based on the wave action conservation equation. Wave breaking, bottom friction, white crown dissipation and wave-wave nonlinear interaction source terms are introduced into the wave propagation model, and the outer sea boundary is set as the incident wave spectrum boundary condition to form a wave model. A control equation for the transport of viscous sediment based on a three-dimensional convection-diffusion equation is established. The bottom shear stress term under the combined action of water flow and waves is introduced into the control equation for the transport of viscous sediment. Combined with the formulas for critical initiation shear stress, settling velocity and resuspension flux of sediment, the scouring, transport, flocculation and settling process of viscous sediment is described, thus forming a sediment model. A coupling framework is constructed based on the model coupling tool MCT to connect the tidal current model, wave model, and sediment model. In this coupling framework, the effective wave height, wave period, and wave propagation direction calculated by the wave model are transmitted to the tidal current model in real time to update the bottom friction. The flow velocity, flow direction, and water level calculated by the tidal current model are transmitted to the sediment model in real time as driving conditions for sediment transport. The shear stress at the bottom of the water flow calculated by the tidal current model and the wave model are vector superimposed with the shear stress at the bottom of the wave, and the total shear stress after superposition is transferred to the sediment model to drive the scouring and initiation calculation of the sediment model, and are integrated to form a mathematical model for channel siltation.

[0025] It should be noted that the model systematically reflects the hydrodynamic processes of the waterway by using the continuity equation to characterize the water body's mass conservation relationship, the momentum equation to depict the motion and changes of the water body under external forces, and the state equation to describe the intrinsic relationship between the water body's physical properties. Introducing Coriolis force, wind stress, and Manning drag coefficient terms into the momentum equation allows for a comprehensive consideration of the effects of Earth's rotation, atmospheric forces, and seabed friction on water flow, making the model closer to the real marine environment. The Eulerian method is used to describe the variation characteristics of the flow field in a fixed space, and the Lagrange method is combined to track the trajectory of particles or water masses, thus contributing to the control of... Discrete solutions to the equations enable stable numerical calculations of complex hydrodynamic processes, thus forming a tidal current model. Furthermore, a wave model is constructed based on the principle of wave action conservation to characterize the energy evolution of waves during propagation and their interaction with the seabed and water body. Simultaneously, a sediment transport model is established based on the three-dimensional convection-diffusion equations, using the bottom shear stress generated by the combined action of water flow and waves as the driving mechanism to describe the initiation, transport, and sedimentation processes. Finally, a model coupling mechanism enables information interaction and feedback between tidal currents, waves, and sediment, allowing each physical process to influence and evolve collaboratively. The siltation-sensitive areas include areas located inside bends, diffusion sections, transition zones between widening and depth, near harbor entrances, and localized areas where water flow deceleration or backflow is significant.

[0026] Figure 3 A flowchart illustrating the correction of siltation parameters according to the present invention is shown.

[0027] According to an embodiment of the present invention, the step of selecting an LSTM time-series prediction model as a data-driven model, obtaining the channel hydraulic parameters within a preset time period based on the channel siltation mathematical model, and inputting the channel hydraulic parameter features into the data-driven model to correct the siltation parameters to obtain corrected siltation parameters is specifically as follows: The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of each grid node in the target waterway area calculated by the waterway siltation mathematical model within a preset historical time period are used as the input feature sequence of the LSTM model. The waterway hydraulic parameters include flow velocity, sediment concentration, water level and wave parameters. The siltation parameters of each grid node are measured within the same preset time period and used as training labels to supervise the training of the LSTM model. The siltation parameters include siltation thickness, siltation rate, and siltation volume. The nonlinear mapping relationship from channel hydraulic parameters to siltation parameters is identified. The predicted hydraulic parameters of the waterway, calculated by the waterway siltation mathematical model for a future preset time period, are input into the trained LSTM model, and the siltation parameter correction values ​​for the corresponding time period and location are output to obtain the corrected siltation parameters.

[0028] It should be noted that due to the extreme complexity of natural processes such as tides, waves, and sediment transport, and the influence of uncertainties in boundary conditions, model parameters (such as critical shear force and erosion coefficient), and model simplification assumptions, the simulation results of purely mechanistic channel siltation mathematical models over long periods or under specific conditions may exhibit systematic deviations from measured siltation data. In other words, the model's predicted outputs (such as siltation thickness and sedimentation volume) may not be accurate enough. Therefore, by introducing an LSTM time-series prediction model, a nonlinear mapping bridge is established from the "causes" (channel hydraulic parameters) predicted by the mechanistic model to the final "results" (measured siltation parameters). The LSTM model learns and captures this complex causal relationship using historical data, enabling intelligent real-time correction of the mechanistic model's output. This is equivalent to adding an "error compensator" to the physical model, dynamically correcting the systematic deviations of the mechanistic model, thereby significantly improving the accuracy of key predicted parameters such as final siltation thickness and rate.

[0029] According to an embodiment of the present invention, the optimization of the channel siltation mathematical model based on the corrected siltation parameters, the prediction of siltation conditions within a preset time period based on the optimized channel siltation mathematical model, and the construction of a siltation risk map are specifically as follows: Obtain the initial silting prediction parameters output by the channel silting mathematical model, and calculate the prediction deviation sequence between the initial silting prediction parameters and the corrected silting parameters for each grid node within a preset historical time period; Based on the predicted deviation sequence, the distribution pattern of the predicted deviation in the time and space dimensions is identified, and based on the distribution pattern, the direction and magnitude of the optimization adjustment of the model parameters related to sediment transport and settling processes in the channel siltation mathematical model are determined. Based on the optimized adjustment direction and magnitude, the model parameters in the channel siltation mathematical model are adjusted to obtain the optimized channel siltation mathematical model. Based on the optimized channel siltation mathematical model, by inputting boundary condition driving data within a future preset time period, the predicted values ​​of siltation thickness, siltation rate, and siltation volume of each grid node in the target channel area within the future preset time period are obtained. Based on the navigation safety standards, a threshold for the bottom elevation of the channel and a warning threshold for the siltation rate are set. The predicted siltation thickness of each grid node is compared with the threshold for the bottom elevation of the channel, and the predicted siltation rate is compared with the warning threshold for the siltation rate to determine the siltation risk of each grid node. Based on the aforementioned siltation risk, a siltation risk map is constructed, which includes the risk levels of different grid nodes and the spatial distribution of predicted siltation volume.

[0030] It should be noted that the aforementioned channel siltation mathematical model structurally couples multiple key physical processes such as tidal currents, waves, and sediment transport. It can describe the behavior of water flow, wave propagation, and sediment scouring, transport, and settling based on physical laws and conservation equations. After inputting boundary condition driving data for a preset future time period, the model can dynamically calculate hydraulic elements such as flow velocity, flow direction, sediment concentration, and wave parameters at various points on the channel during the future time period by numerically solving these governing equations. These hydraulic elements are the direct driving force for sediment movement. Based on this, the model calculates the sediment transport flux and settling rate, thereby predicting the future siltation thickness, rate, and total amount at various locations. The distribution pattern refers to the periodicity, trend, or abruptness of the model prediction deviation over time, as well as the systematic regional distribution pattern related to channel topography and hydrological conditions in spatial location. For example, the deviation may be consistently higher or lower in certain specific areas (such as the wharf front or channel bends) or specific time periods (such as during floods or after strong winds and waves). The model parameters include critical shear force, erosion coefficient, riverbed roughness (i.e., Manning coefficient), and sedimentation rate parameters; the boundary condition driving data include the flow process of the upstream open boundary, the tidal process of the downstream open boundary, the incident wave spectrum of the offshore boundary, and meteorological and hydrological data such as wind speed, wind direction, and sediment concentration.

[0031] According to an embodiment of the present invention, the step of generating siltation early warning information based on the siltation risk map, and determining the dredging volume at each location in the target waterway area based on the siltation early warning information, specifically includes: Extract the location coordinates, siltation risk level, predicted siltation thickness, and predicted siltation rate of each grid node in the siltation risk map to form a multi-dimensional node feature vector; Based on the DBSCAN clustering algorithm, the latitude and longitude coordinates of the grid node location are used as spatial attributes, the Euclidean distance of the grid node is used as a measure of spatial proximity, and the weighted combination of the siltation risk level, the predicted value of siltation thickness, and the predicted value of siltation rate is used as a multi-dimensional feature measure to calculate the Euclidean distance between nodes. Using nodes whose Euclidean distance is less than the preset neighborhood radius and whose number of nodes in the neighborhood of the core point is not less than the preset minimum number of points as core points, starting from the core point, the grid nodes that meet the preset density reachability conditions are clustered and expanded, and the grid nodes that are spatially continuous and have similar backfilling characteristics are aggregated into different independent clusters. Calculate the predicted average backfill thickness and the predicted average backfill rate of all grid nodes within each independent cluster. Determine the regional range of each independent cluster based on the location of all grid nodes within the independent cluster. Construct backfill early warning information for each independent cluster, including the regional range, predicted average backfill thickness, and rate. For each independent cluster, the dredging thickness of each node is calculated based on the current predicted channel bottom elevation, the initial channel design bottom elevation, and the navigation maintenance bottom elevation standard of each grid node within it. The dredging thickness is then multiplied by the area of ​​each independent cluster region to obtain the dredging volume of the grid node.

[0032] It should be noted that by introducing the DBSCAN clustering algorithm to intelligently analyze the siltation risk map, grid nodes with similar siltation risk characteristics (such as risk level, siltation thickness, and siltation rate) and spatially continuous distribution within the waterway area are identified and aggregated. This allows for the efficient and accurate division of several independent siltation areas with relatively consistent internal conditions from massive, scattered point-like prediction data. By calculating the average siltation parameters for each independent cluster to generate structured early warning information, and accurately calculating the dredging thickness and engineering volume of each node within each cluster according to waterway design standards, the macro-level risk warning is seamlessly transformed into specific and quantifiable engineering tasks.

[0033] According to an embodiment of the present invention, determining the dredging scheme for the target waterway area based on the dredging volume specifically involves: Based on the siltation risk level, predicted average siltation thickness, siltation rate and area range in the siltation early warning information of each independent cluster in the target waterway area, calculate the dredging urgency index of each independent cluster. Based on the navigation requirements of the waterway, data on design water depth, waterway width, slope gradient, navigation density and vessel tonnage are extracted from the waterway maintenance database to calculate the improvement index of navigation efficiency of each independent cluster of dredging. Data on the working status, construction capacity, scheduling distance, and available capacity of the sludge discharge area of ​​the construction vessels are obtained. Based on the dredging volume, dredging urgency index, and improvement index for navigation efficiency, a multi-objective optimization algorithm is adopted, with the goal of minimizing the total construction time, to determine the dredging operation plan, including the priority order of construction areas, the scheduling plan of construction vessels, the allocation of dredging volume, the construction window period, and the construction path.

[0034] It should be noted that, firstly, the comprehensive impact of each dredging area on safety risks and operational benefits was quantified, namely the dredging urgency index and the improvement index on navigation efficiency. Subsequently, multiple practical factors, including dredging workload, construction resources (vessel status, construction capacity, scheduling distance), and engineering constraints (sludge discharge area capacity), were integrated to construct an optimization model with minimizing total construction time as the core objective. By solving this model, a dredging operation plan that achieves an optimal balance in time, resources, and results can be automatically generated, specifically covering construction priority, vessel scheduling, workload allocation, construction sequence, and route planning, significantly improving the execution efficiency of maintenance operations.

[0035] Figure 4 A block diagram of a coastal waterway maintenance decision system based on a siltation mathematical model is shown.

[0036] A second aspect of the present invention also provides a coastal waterway maintenance decision-making system based on a siltation mathematical model. The system includes a memory 401, a processor 402, and a communication interface 403. The memory includes a coastal waterway maintenance decision-making method program based on a siltation mathematical model. The communication interface is used for data connection and communication between the memory and the processor. When the coastal waterway maintenance decision-making method program based on the siltation mathematical model is executed by the processor, it implements the steps of the coastal waterway maintenance decision-making method based on the siltation mathematical model as described in any of the above claims.

[0037] This invention discloses a coastal waterway maintenance decision-making method and system based on a siltation mathematical model. It includes: constructing a waterway lifecycle database; establishing a waterway siltation mathematical model by coupling tidal current, wave, and sediment models using an unstructured grid and an MCT coupler; correcting siltation parameters using an LSTM time-series prediction model, optimizing the mathematical model, predicting future siltation, and constructing a siltation risk map; generating early warning information based on the map, determining dredging work volume, and formulating dredging plans. This invention achieves accurate prediction of waterway siltation and quantitative decision-making on maintenance plans, improving the scientific and economic efficiency of waterway maintenance.

[0038] 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.

[0039] 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.

[0040] 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.

[0041] 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.

[0042] 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.

[0043] 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 variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A coastal waterway maintenance decision-making method based on a siltation mathematical model, characterized in that, Includes the following steps: Collect multi-source heterogeneous data of the target waterway area, clean and convert the multi-source heterogeneous data to obtain a standardized waterway dataset, and construct a waterway construction and maintenance full life cycle database from the standardized waterway dataset. The target channel area is divided into grids using an unstructured grid, and the standardized channel dataset is coupled in real time with the current tidal model, wave model and sediment model based on the MCT coupler to construct a mathematical model for channel siltation. The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of the waterway within a preset time period are obtained according to the waterway siltation mathematical model. The features of the waterway hydraulic parameters are input into the data-driven model to correct the siltation parameters and obtain the corrected siltation parameters. The channel siltation mathematical model is optimized based on the corrected siltation parameters. Based on the optimized channel siltation mathematical model, the siltation situation in the future preset time period is predicted, and a siltation risk map is constructed. Based on the siltation risk map, siltation early warning information is generated, and based on the siltation early warning information, the dredging volume at each location in the target waterway area is determined; The dredging plan for the target waterway area is determined based on the amount of dredging work.

2. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The process involves collecting multi-source heterogeneous data from the target waterway area, cleaning and format conversion of the data to obtain a standardized waterway dataset, and then constructing a waterway construction and maintenance lifecycle database from this standardized data dataset. Specifically: Collect multi-source heterogeneous data for the target waterway area. The multi-source heterogeneous data includes real-time collected underwater topographic point cloud data, hydrological time series data, wave spectrum data, meteorological data, water quality data, as well as historical waterway survey datasets, maintenance and dredging project records, and waterway construction drawings. Data cleaning is performed on the multi-source heterogeneous data. The data cleaning includes removing abnormal redundant data based on a preset reasonableness threshold range, correcting data mutation error values ​​using interpolation algorithms, and supplementing missing data based on adjacent spatiotemporal data points. After data cleaning, the multi-source heterogeneous data is unified in terms of coordinate system and data format to obtain a standardized waterway dataset. The standardized waterway dataset is then stored according to the collection time sequence and spatial location to construct a waterway construction and maintenance full life cycle database.

3. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The target channel area is divided into grids using an unstructured grid, and the standardized channel dataset is coupled in real time with the tidal current model, wave model, and sediment model based on the MCT coupler to construct a mathematical model for channel siltation. Specifically: Unstructured grids are used to divide the target waterway area into grids, and the grids are densified in the main waterway area and siltation-sensitive areas. Based on the finite volume method, a set of flow control equations including continuity equation, momentum equation and state equation is established for the grid. Coriolis force, wind stress and Manning drag coefficient terms are introduced into the momentum equation of the flow control equation set. The Euler-Lagrange method is used to discretize and solve the flow control equation set to form a tidal flow model. The third-generation wave spectrum model SWAN is introduced to establish a wave propagation model based on the wave action conservation equation. Wave breaking, bottom friction, white crown dissipation and wave-wave nonlinear interaction source terms are introduced into the wave propagation model, and the outer sea boundary is set as the incident wave spectrum boundary condition to form a wave model. A control equation for the transport of viscous sediment based on a three-dimensional convection-diffusion equation is established. The bottom shear stress term under the combined action of water flow and waves is introduced into the control equation for the transport of viscous sediment. Combined with the formulas for critical initiation shear stress, settling velocity and resuspension flux of sediment, the scouring, transport, flocculation and settling process of viscous sediment is described, thus forming a sediment model. A coupling framework is constructed based on the model coupling tool MCT to connect the tidal current model, wave model, and sediment model. In this coupling framework, the effective wave height, wave period, and wave propagation direction calculated by the wave model are transmitted to the tidal current model in real time to update the bottom friction. The flow velocity, flow direction, and water level calculated by the tidal current model are transmitted to the sediment model in real time as driving conditions for sediment transport. The shear stress at the bottom of the water flow calculated by the tidal current model and the wave model are vector superimposed with the shear stress at the bottom of the wave, and the total shear stress after superposition is transferred to the sediment model to drive the scouring and initiation calculation of the sediment model, and are integrated to form a mathematical model for channel siltation.

4. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The LSTM time-series prediction model is selected as the data-driven model. Based on the channel siltation mathematical model, the channel hydraulic parameters within a preset time period are obtained. The characteristics of these hydraulic parameters are then input into the data-driven model to correct the siltation parameters, resulting in corrected siltation parameters. Specifically: The LSTM time series prediction model is selected as the data-driven model. The hydraulic parameters of each grid node in the target waterway area calculated by the waterway siltation mathematical model within a preset historical time period are used as the input feature sequence of the LSTM model. The waterway hydraulic parameters include flow velocity, sediment concentration, water level and wave parameters. The siltation parameters of each grid node are measured within the same preset time period and used as training labels to supervise the training of the LSTM model. The siltation parameters include siltation thickness, siltation rate, and siltation volume. The nonlinear mapping relationship from channel hydraulic parameters to siltation parameters is identified. The predicted hydraulic parameters of the waterway, calculated by the waterway siltation mathematical model for a future preset time period, are input into the trained LSTM model, and the siltation parameter correction values ​​for the corresponding time period and location are output to obtain the corrected siltation parameters.

5. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The process involves optimizing the channel siltation mathematical model based on the corrected siltation parameters, predicting the siltation situation within a preset time period based on the optimized model, and constructing a siltation risk map. Specifically: Obtain the initial silting prediction parameters output by the channel silting mathematical model, and calculate the prediction deviation sequence between the initial silting prediction parameters and the corrected silting parameters for each grid node within a preset historical time period; Based on the predicted deviation sequence, the distribution pattern of the predicted deviation in the time and space dimensions is identified, and based on the distribution pattern, the direction and magnitude of the optimization adjustment of the model parameters related to sediment transport and settling processes in the channel siltation mathematical model are determined. Based on the optimization adjustment direction and magnitude, the model parameters in the channel siltation mathematical model are adjusted to obtain the optimized channel siltation mathematical model; Based on the optimized channel siltation mathematical model, by inputting boundary condition driving data within a future preset time period, the predicted values ​​of siltation thickness, siltation rate, and siltation volume of each grid node in the target channel area within the future preset time period are obtained. Based on the navigation safety standards, a threshold for the bottom elevation of the channel and a warning threshold for the siltation rate are set. The predicted siltation thickness of each grid node is compared with the threshold for the bottom elevation of the channel, and the predicted siltation rate is compared with the warning threshold for the siltation rate to determine the siltation risk of each grid node. Based on the aforementioned siltation risk, a siltation risk map is constructed, which includes the risk levels of different grid nodes and the spatial distribution of predicted siltation volume.

6. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The step of generating siltation early warning information based on the siltation risk map, and determining the dredging volume at each location in the target channel area based on the siltation early warning information, specifically involves: Extract the location coordinates, siltation risk level, predicted siltation thickness, and predicted siltation rate of each grid node in the siltation risk map to form a multi-dimensional node feature vector; Based on the DBSCAN clustering algorithm, the latitude and longitude coordinates of the grid node location are used as spatial attributes, the Euclidean distance of the grid node is used as a measure of spatial proximity, and the weighted combination of the siltation risk level, the predicted value of siltation thickness, and the predicted value of siltation rate is used as a multi-dimensional feature measure to calculate the Euclidean distance between nodes. Using nodes whose Euclidean distance is less than the preset neighborhood radius and whose number of nodes in the neighborhood of the core point is not less than the preset minimum number of points as core points, starting from the core point, the grid nodes that meet the preset density reachability conditions are clustered and expanded, and the grid nodes that are spatially continuous and have similar backfilling characteristics are aggregated into different independent clusters. Calculate the predicted average backfill thickness and the predicted average backfill rate of all grid nodes within each independent cluster. Determine the regional range of each independent cluster based on the location of all grid nodes within the independent cluster. Construct backfill early warning information for each independent cluster, including the regional range, predicted average backfill thickness, and rate. For each independent cluster, the dredging thickness of each node is calculated based on the current predicted channel bottom elevation, the initial channel design bottom elevation, and the navigation maintenance bottom elevation standard of each grid node within it. The dredging thickness is then multiplied by the area of ​​each independent cluster region to obtain the dredging volume of the grid node.

7. The coastal waterway maintenance decision-making method based on a siltation mathematical model according to claim 1, characterized in that, The dredging plan for the target waterway area, determined based on the dredging volume, is specifically as follows: Based on the siltation risk level, predicted average siltation thickness, siltation rate and area range in the siltation early warning information of each independent cluster in the target waterway area, calculate the dredging urgency index of each independent cluster. Based on the navigation requirements of the waterway, data on design water depth, waterway width, slope gradient, navigation density and vessel tonnage are extracted from the waterway maintenance database to calculate the improvement index of navigation efficiency of each independent cluster of dredging. Data on the working status, construction capacity, scheduling distance, and available capacity of the sludge discharge area of ​​the construction vessels are obtained. Based on the dredging volume, dredging urgency index, and improvement index for navigation efficiency, a multi-objective optimization algorithm is adopted, with the goal of minimizing the total construction time, to determine the dredging operation plan, including the priority order of construction areas, the scheduling plan of construction vessels, the allocation of dredging volume, the construction window period, and the construction path.

8. A coastal waterway maintenance decision-making system based on a siltation mathematical model, characterized in that, The coastal waterway maintenance decision system based on the siltation mathematical model includes a storage unit and a processor. The storage unit includes a coastal waterway maintenance decision method program based on the siltation mathematical model. When the processor executes the coastal waterway maintenance decision method program based on the siltation mathematical model, it implements the steps of the coastal waterway maintenance decision method based on the siltation mathematical model as described in any one of claims 1 to 7.

Images