Microplastic transportation path prediction method and device based on complex hydrodynamic environment, equipment and medium

Abstract

The invention discloses a microplastic transportation path prediction method, device and equipment based on a complex hydrodynamic environment and a medium, and relates to the technical field of environment monitoring and water pollution governance, and the method comprises the steps: integrating a watershed digital elevation model, a river channel vector diagram, microplastic attributes and real-time hydrological data; a dynamic adjacency matrix is used for capturing water flow direction changes, multi-modal feature fusion is combined for modeling micro-plastic physical attributes and topographic resistance, the problem of monitoring data scarcity is solved through a cross-basin transfer learning strategy, a micro-plastic transfer path is accurately predicted, high-precision simulation of micro-plastic transfer in a complex hydrodynamic environment is achieved, and high-precision simulation of micro-plastic transfer in a complex hydrodynamic environment is achieved. The pollution prevention and control efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of environmental monitoring and water pollution control, and in particular to a method, device, equipment and medium for predicting the transport path of microplastics based on a complex hydrodynamic environment. Background Art

[0002] Microplastics (plastic particles with a particle size less than 5 mm) as an emerging environmental pollutant have been widely detected in river, lake and marine ecosystems. Due to their persistence, bioaccumulation and ecological toxicity, microplastics pose a potential threat to aquatic ecosystems and human health. Their sources are diverse, including cosmetics, textile fibers, degradation of agricultural plastic films, etc., and the transport path is affected by various factors such as hydrodynamic force, topography and particle characteristics.

[0003] Currently, the research on predicting the transport path of microplastics based on a complex hydrodynamic environment mainly focuses on three types of technologies: hydrodynamic models based on physical mechanisms, statistical learning models based on data-driven, and hybrid models that combine physics and data. The hydrodynamic model uses the Navier-Stokes equation and the particle sedimentation formula to simulate the water flow movement, and constructs a river channel topological network by combining GIS data; data-driven models such as LSTM, random forest, etc. extract features from historical data through machine learning methods to improve the calculation efficiency; the hybrid model attempts to combine the advantages of the above two methods, for example, using the hydrodynamic equation as a neural network constraint condition to improve the prediction accuracy and adaptability.

[0004] Despite the progress, the existing technologies still have limitations in many aspects. First, although the hydrodynamic model can accurately simulate the water flow movement, due to relying on the complex Navier-Stokes equation, it requires high-performance computing resources in a three-dimensional simulation scenario, which is difficult to meet the real-time response requirements of sudden pollution events. At the same time, this model assumes that microplastics are homogeneous spherical particles and does not fully consider the influence of different materials and shapes. Second, the data-driven model performs well in processing time series or local spatial features, but has limited ability in expressing the complex topological relationship between the upstream and downstream of the basin and between tributaries and main streams, and its "black box" characteristic limits its application in pollution control decision-making. In addition, although the hybrid model attempts to combine physical modeling and data-driven methods, the coupling mechanism between physics and data is relatively rough, the training process is unstable, and there is a lack of a dynamic weight adjustment mechanism, making it difficult to balance the computational weights between real-time hydrological data and long-term physical laws. Therefore, there is an urgent need for an efficient and accurate path prediction method to realize the prediction of the transport path of microplastics based on a complex hydrodynamic environment. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device, equipment, and medium for predicting the transport path of microplastics based on a complex hydrodynamic environment, aiming to solve the technical problem of how to improve the prediction accuracy of microplastic paths in a complex hydrodynamic environment.

[0006] To achieve the above objective, this application proposes a method for predicting the transport path of microplastics based on a complex hydrodynamic environment, including: Obtain watershed digital elevation model data, river channel vector maps, microplastic properties, and hydrological data; Divide the watershed digital elevation model data and the river channel vector maps to obtain a set of nodes, where the nodes include elevation, slope aspect, river channel bifurcation points, and watershed outlets; Based on the hydrological data and the watershed digital elevation model data, calculate to generate a dynamic adjacency matrix; Fuse the microplastic properties and the watershed terrain features to obtain fused terrain features. The watershed terrain features are obtained based on the watershed digital elevation model data and the river channel vector maps, and the watershed terrain features include slope data, roughness coefficient, and river channel curvature; Based on the dynamic adjacency matrix and the fused terrain features, calculate to obtain the target microplastic concentration; Determine the concentration change direction and the concentration change intensity according to the target microplastic concentration; Based on the concentration change direction and the concentration change intensity, determine the transport nodes and the transport distance; When the transport node is the watershed outlet in the set of nodes, combine the transport distances to obtain the target transport path prediction result.

[0007] In one embodiment, the step of calculating based on the hydrological data and the watershed digital elevation model data to generate a dynamic adjacency matrix includes: Calculate through the hydrological data to obtain the flow direction similarity; Calculate according to the flow direction similarity to obtain the flow direction consistency between two adjacent nodes. The specific formula is: Where and are the flow velocities of nodes and node respectively, and is the flow direction similarity; Obtain the terrain resistance according to the watershed digital elevation model data. The specific formula is: Where is the node and The elevation difference between is the node and the horizontal spatial distance between is the resistance factor; According to the flow direction consistency and the terrain resistance, a dynamic adjacency matrix is obtained. The specific formula is: where is the flow direction consistency, is the terrain resistance, represents the indicator function, and only adjacent nodes in the downstream direction are allowed to participate in the calculation.

[0008] In one embodiment, the step of fusing the microplastic attributes and the watershed terrain features to obtain the fused terrain features, where the watershed terrain features are obtained based on the watershed digital elevation model data and the river channel vector map, and the watershed terrain features include slope data, roughness coefficient, and river channel curvature, includes: Encoding the watershed terrain features to obtain an initial feature vector, which is specifically expressed as: where, is the slope data, is the roughness coefficient, is the river channel curvature; According to the initial feature vector, through double-layer GCN processing, a first feature vector and a second feature vector are obtained. The specific formula is: where, and respectively represent the weights of the first layer and the second layer of the GCN, is the node the first-order neighbor set of; Encoding the microplastic attributes to obtain an attribute encoding feature. The specific formula is: represents the microplastic density, is the shape factor, represents the function of converting the material category into a numerical value; Based on the attribute encoding feature and the second feature vector, a fused terrain vector is obtained. The specific formula is: Among them, are the mapping matrices of terrain features, key vectors, and microplastic properties respectively, is the dimension of the key vector, is the attention weight, represents the normalization function, is the fused feature vector; Based on the fused terrain vector, the fused terrain feature is obtained.

[0009] In one embodiment, the step of calculating the target microplastic concentration based on the dynamic adjacency matrix and the fused terrain feature includes: Based on the dynamic adjacency matrix and the fused terrain feature, update and calculate through the graph propagation algorithm to obtain the microplastic concentration. The specific formula is: Among them, is the node at time prediction of the microplastic concentration, is the learnable weight matrix; According to the basin digital elevation model data and the river channel vector map, calculate to obtain the water volume corresponding to the node; According to the hydrological data, obtain the flow velocity from the node to the corresponding adjacent node; Initialize the update weight; According to the water volume and the flow velocity, combine with the microplastic concentration to obtain the concentration loss value through the mass conservation equation. The formula is: Among them, is the water volume corresponding to the node , is the node to flow velocity, represents the flow velocity from the node to flow velocity; Update the concentration loss value until the update weight reaches the preset value to obtain the target loss value; Combine the target loss value and the microplastic concentration for calculation to obtain the target microplastic concentration.

[0010] In one embodiment, the step of determining the concentration change direction and concentration change intensity according to the target microplastic concentration includes: Determine the predicted concentration distribution according to the target microplastic concentration; Calculate the concentration gradient of the node according to the predicted concentration distribution to determine the concentration change direction and the concentration change intensity. The formula is: Wherein, represents the predicted concentration distribution, represents the node position vector of, and are the abscissa direction and the ordinate direction of the spatial coordinate axis.

[0011] In one embodiment, before the step of determining the transport node and the transport path based on the concentration change direction and the concentration change intensity, it includes: Obtain the target watershed data, and the target watershed data includes the microplastic concentration and the topographic attribute data of the target watershed; Select a preset proportion of the target watershed data and use the Kriging spatial interpolation algorithm to generate pseudo-labels for the unlabeled watershed. The specific formula is: Wherein, represents the Kriging spatial interpolation algorithm, represents the microplastic concentration of the target watershed, represents the topographic attribute data of the target watershed; Evaluate the pseudo-labels of the unlabeled watershed to obtain the corresponding confidence level; When the confidence level exceeds the preset confidence level, obtain the unlabeled watershed data corresponding to the confidence level; Adopt an optimization strategy for the target watershed data and the unlabeled watershed data to obtain a loss value. The specific formula is: Wherein, represents the maximum mean difference loss value, represents the mean square error loss value, represents the trade-off coefficient, represents the unlabeled watershed data; Update the loss value until the trade-off coefficient reaches the preset value, and execute the step of determining the concentration change direction and the concentration change intensity according to the target microplastic concentration.

[0012] In one embodiment, the step of determining the transport node and the transport path based on the concentration change direction and the concentration change intensity includes: Determine a set of downstream adjacent nodes consistent with the change direction according to the concentration change direction; Select a node corresponding to a preset concentration change intensity among the concentration change intensities from the set of downstream adjacent nodes as the transport node; Based on the basin digital elevation model data and the river channel vector map, combine the current node and the transport node to obtain the transport distance.

[0013] In addition, to achieve the above object, the present application also proposes a microplastic transport path prediction device based on a complex hydrodynamic environment. The microplastic transport path prediction device based on a complex hydrodynamic environment includes: An acquisition module, configured to acquire basin digital elevation model data, river channel vector maps, microplastic properties, and hydrological data; A division module, configured to divide the basin digital elevation model data and the river channel vector map to obtain a set of nodes, where the nodes include elevation, slope direction, river channel bifurcation points, and basin outlets; A calculation module, configured to calculate based on the hydrological data and the basin digital elevation model data to generate a dynamic adjacency matrix; A fusion module, configured to fuse the microplastic properties and the basin terrain features to obtain fused terrain features, where the basin terrain features are obtained based on the basin digital elevation model data and the river channel vector map, and the basin terrain features include slope data, roughness coefficient, and river channel curvature; The calculation module is further configured to calculate based on the dynamic adjacency matrix and the fused terrain features to obtain the target microplastic concentration; The calculation module is further configured to determine the concentration change direction and the concentration change intensity according to the target microplastic concentration; The calculation module is further configured to determine the transport node and the transport distance based on the concentration change direction and the concentration change intensity; A result module, configured to combine the transport distances when the transport node is the basin outlet in the set of nodes to obtain a target transport path prediction result.

[0014] In addition, to achieve the above object, the present application also proposes a medium, where the medium is a computer-readable medium, and a computer program is stored on the medium. When the computer program is executed by a processor, the steps of the microplastic transport path prediction method based on a complex hydrodynamic environment as described above are implemented.

[0015] In addition, to achieve the above object, the present application also provides a computer program product, where the computer program product includes a computer program. When the computer program is executed by a processor, the steps of the microplastic transport path prediction method based on a complex hydrodynamic environment as described above are implemented.

[0016] This application obtains basin digital elevation model data, river channel vector maps, microplastic properties, and hydrological data, divides the basin digital elevation model data and river channel vector maps to obtain a node set, calculates based on the hydrological data and basin digital elevation model data to generate a dynamic adjacency matrix, fuses the microplastic properties and basin terrain features to obtain fused terrain features, then calculates based on the dynamic adjacency matrix and fused terrain features to obtain the target microplastic concentration, determines the concentration change direction and concentration change intensity according to the target microplastic concentration, determines the transport nodes and transport distances based on the concentration change direction and concentration change intensity, and finally, when the transport node is the basin outlet in the node set, combines the transport distances to obtain the target transport path prediction result. By integrating the basin digital elevation model, river channel vector maps, microplastic properties, and real-time hydrological data, using the dynamic adjacency matrix and fused terrain features, the microplastic transport path is accurately predicted, realizing high-precision simulation of microplastic migration in a complex hydrodynamic environment and improving the pollution prevention and control efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for description in the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 FIG. is a schematic flowchart of the first embodiment of the microplastic transport path prediction method based on a complex hydrodynamic environment according to the present application; Figure 2 FIG. is a schematic flowchart of the second embodiment of the microplastic transport path prediction method based on a complex hydrodynamic environment according to the present application; Figure 3 FIG. is a schematic flowchart of the third embodiment of the microplastic transport path prediction method based on a complex hydrodynamic environment according to the present application; Figure 4 FIG. is a schematic module structure diagram of the microplastic transport path prediction device based on a complex hydrodynamic environment in the first embodiment of the microplastic transport path prediction method based on a complex hydrodynamic environment according to the present application; Figure 5 FIG. is a schematic device structure diagram of the hardware operating environment involved in the microplastic transport path prediction method based on a complex hydrodynamic environment in the embodiments of the present application.

[0019] The implementation, functional features, and advantages of the present application will be further described with reference to the embodiments and the drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0021] To better understand the technical solution of the present application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0022] Globally, microplastics (plastic particles with a particle size less than 5 mm) as a new type of environmental pollutant have been widely present in river, lake and marine ecosystems. Due to their persistence, bioaccumulation and potential ecological toxicity, microplastics can not only adsorb harmful substances such as persistent organic pollutants and heavy metals, but may also be transmitted to humans through the food chain, posing a serious threat to aquatic ecosystems and human health. However, current research on the transport law of microplastics faces many challenges, especially there are still technical bottlenecks in source tracing, prediction accuracy and treatment response. The traditional method based on fixed-point sampling is difficult to achieve dynamic tracking of the source and migration path of microplastics in the whole basin; existing prediction models mainly rely on hydrodynamic empirical formulas and use static hydrological parameters for simulation, which are difficult to accurately reflect the dynamic effects of rainfall runoff mutation, water conservancy facility regulation, etc. on the water flow direction and velocity, resulting in limited prediction accuracy. In addition, in the absence of accurate transport path prediction technology, pollution interception measures are often lagging, especially in sudden pollution events such as rainstorm scouring, and high-precision pollutant diffusion prediction cannot be provided in time, affecting the efficiency of pollution prevention and control.

[0023] Therefore, in order to overcome the above problems, the present application proposes a method for predicting the transport path of microplastics based on a complex hydrodynamic environment, which is efficient and accurate. The main solution of the embodiment of the present application is: obtaining basin digital elevation model data, river channel vector maps, microplastic attributes and hydrological data, dividing the basin digital elevation model data and river channel vector maps to obtain a node set, calculating based on the hydrological data and basin digital elevation model data to generate a dynamic adjacency matrix, fusing the microplastic attributes and basin terrain features to obtain a fused terrain feature, then calculating based on the dynamic adjacency matrix and the fused terrain feature to obtain the target microplastic concentration, determining the concentration change direction and concentration change intensity according to the target microplastic concentration, determining the transport nodes and transport distances based on the concentration change direction and concentration change intensity, and finally when the transport node is the basin outlet in the node set, combining the transport distances to obtain the target transport path prediction result.

[0024] Based on the above, the embodiment of the present application also provides a method for predicting the transport path of microplastics based on a complex hydrodynamic environment, referring to Figure 1 , Figure 1 is the flow chart of the first embodiment of the method for predicting the transport path of microplastics based on a complex hydrodynamic environment of the present application.

[0025] In this embodiment, the method for predicting the transport path of microplastics based on a complex hydrodynamic environment includes steps S10 to S80: Step S10, obtain the basin digital elevation model data, river channel vector map, microplastic properties, and hydrological data.

[0026] It should be noted that the basin digital elevation model (DEM) data provides detailed information about the terrain, including altitude and slope, etc. These information are crucial for understanding the water flow direction and velocity. By parsing the high-resolution (resolution ≤ 10m) DEM data, the basin boundary can be accurately divided, and key nodes such as elevation points, slope aspects, river channel bifurcation points, and basin outlets can be identified, laying a foundation for subsequent construction of the dynamic adjacency matrix. The river channel vector map provides the network structure information of the river system, which helps to understand the water flow path and river channel connection. Combining with the natural river channel connection relationship, an initial edge set can be established to further refine the water flow pattern within the basin. In addition, the microplastic properties include physical parameters such as particle density, shape factor, material classification (such as PE / PET), etc. They directly affect the migration behavior of microplastics in water bodies, such as sedimentation rate and suspension time. Accurately encoding these properties and integrating them with the terrain features can more realistically simulate the migration trajectory of microplastics under complex hydrodynamic conditions. Hydrological data, such as flow velocity, rainfall, tidal changes, etc., are used to update the water flow state within the basin in real time, reflecting the impact of hydrological dynamic changes on the microplastic transport path. Using these data to generate a dynamic adjacency matrix can capture the change of water flow direction over time, thus realizing the accurate simulation of the microplastic migration process.

[0027] Step S20, divide the basin digital elevation model data and the river channel vector map to obtain a node set.

[0028] It should be noted that by loading the high-resolution DEM data (resolution ≤ 10 meters), we can extract the terrain information within the basin, including altitude, slope, and slope aspect, etc. These geographical attributes are of great significance for understanding the water flow direction, flow velocity, and potential deposition areas. To further refine the research area, the Voronoi algorithm can be used to process the DEM data to generate a node set , and each node corresponds to a key node within the basin. This process helps to accurately capture the changes in hydrological conditions within the basin and provides a basis for subsequent modeling. At the same time, the river channel vector map provides the network structure information of the river system, including the starting point, ending point, branching points of the river, and their connection relationships.

[0029] Furthermore, based on this information, an initial edge set can be established to represent the natural river channel connection between each node. The specific formula is: where It is a binary attribute of the static adjacency matrix. Specifically, if there is a direct waterway connection between two adjacent nodes, the corresponding edge is marked as 1; conversely, if there is no direct connection, it is marked as 0. This method constructs a static adjacency matrix to preliminarily describe the water flow pattern within the basin. The initial edge set is obtained by combining multiple static adjacency matrices.

[0030] Step S30: Calculate based on hydrological data and basin digital elevation model data to generate a dynamic adjacency matrix.

[0031] It should be noted that the dynamic adjacency matrix accurately reflects the dynamic changes in water flow direction and velocity by fusing real-time hydrological data and DEM data. Two main factors need to be considered in constructing the dynamic adjacency matrix: flow direction similarity calculation and terrain resistance. The flow direction similarity calculation depends on the real-time obtained velocity data to evaluate the consistency of the flow direction between nodes. Specifically, the cosine similarity is used to measure the consistency of the flow velocity direction between two nodes, and a flow direction consistency degree is set. If the angle is less than the set value, it is considered that there is a co-directional transmission between the two nodes; otherwise, the weight is forced to be adjusted to zero. In addition, considering the influence of terrain on water flow, terrain resistance needs to be introduced. The terrain resistance is generated by combining the elevation difference and the horizontal distance with the resistance factor to ensure that water flow transmission is only allowed in the downstream direction.

[0032] Specifically, calculate through hydrological data to obtain the flow direction similarity. Calculate the flow direction similarity between adjacent nodes through vector analysis methods. The specific formula is: where and are the flow velocities of nodes and node respectively, represents the included angle of the flow velocity direction between the two nodes. The closer the flow direction similarity is to 1, the more consistent the flow direction is. Calculate according to the flow direction similarity to obtain the flow direction consistency degree between two adjacent nodes. The specific formula is: where, and are the flow velocities of nodes and node respectively, is the flow direction similarity; Obtain the terrain resistance according to the basin digital elevation model data. The specific formula is: where, is the elevation difference between nodes and respectively, is the node and the horizontal spatial distance between is the resistance factor.

[0033] According to the flow direction consistency and terrain resistance, a dynamic adjacency matrix is obtained. The sigmoid function is used to map the adjacency weights to the range of (0, 1) for facilitating the weight normalization process in subsequent GNN training. The specific formula is: where is the flow direction consistency, is the terrain resistance, represents the indicator function, which is used to screen the qualified node pairs and only allows the adjacent nodes in the downstream direction to participate in the calculation.

[0034] Step S40: Integrate the microplastic properties and the watershed terrain features to obtain the integrated terrain features.

[0035] It should be noted that the above-mentioned watershed terrain features are obtained based on the watershed digital elevation model data and the river channel vector map. The watershed terrain features include slope data, roughness coefficient, and river channel curvature. The slope data reflects the ability of water flow to accelerate, the roughness coefficient quantifies the frictional resistance of the ground surface to water flow, and the river channel curvature describes the meandering degree of the river channel, which affects the complexity of water flow and the possibility of turbulence formation. The microplastic properties include key parameters such as particle density, shape factor, and material classification (such as PE / PET), etc. These factors directly affect the sedimentation rate and suspension time of microplastics in water bodies. For example, microplastic particles with different densities will sink or float at different speeds; while the shape factor determines their resistance and stability in water flow, and fibrous particles are more likely to be carried by water flow and remain suspended compared to spherical particles.

[0036] Specifically, the watershed terrain features (such as slope, roughness coefficient, and river channel curvature) are encoded into an initial feature vector and processed through a two-layer GCN to capture the spatial dependence relationship between adjacent nodes. At the same time, the microplastic properties are also mapped into numerical forms, and the sedimentation velocity is calculated according to Stokes' formula and added to the model as a physical prior. Next, using the multi-head cross-attention mechanism, the dynamic interaction and adaptive integration between the two are realized to obtain the integrated terrain vector, and then the integrated terrain features are obtained according to the vector transformation.

[0037] Step S50: Calculate based on the dynamic adjacency matrix and the integrated terrain features to obtain the target microplastic concentration.

[0038] It should be noted that the dynamic adjacency matrix provides a basic framework for simulating the migration of microplastics by capturing the changes in water flow direction, flow velocity within the basin, and the influence of topographic resistance in real time. This matrix not only reflects the natural river channel connections but also can respond to changes in hydrological conditions caused by factors such as rainfall variations and sluice and dam regulation. Combining the fused topographic features with the dynamic adjacency matrix is used to achieve accurate prediction of the microplastic concentration distribution. The fused topographic features include the basin topographic features and microplastic properties, and these factors jointly affect the migration behavior of microplastics in water bodies. Using the Graph Neural Network (GNN) model, calculations are performed by combining the dynamic adjacency matrix and the fused topographic features to obtain the target microplastic concentration. The graph propagation calculation updates the predicted value of the microplastic concentration on each node through a message-passing mechanism based on the dynamic adjacency matrix and node features. Each step of message passing takes into account the information from adjacent nodes and its own physical properties, enabling the model to dynamically adapt to changes in hydrological conditions. In addition, to ensure that the prediction results conform to the law of mass conservation, a physical loss calculation step is introduced to force the model prediction to satisfy the discrete conservation equation, prevent unreasonable physical contradictions, and add the mass conservation constraint as a regularization term to the loss function for supervising and optimizing the physical laws of concentration changes.

[0039] Specifically, based on the dynamic adjacency matrix and the fused topographic features, calculations are updated through the graph propagation algorithm to obtain the microplastic concentration. The specific formula is as follows: where, is the predicted microplastic concentration of node at time , and is the learnable weight matrix. Calculations are performed based on the basin digital elevation model data and the river channel vector map to obtain the water volume corresponding to the node. The flow velocity from the node to the corresponding adjacent node is obtained according to the hydrological data, the update weights are initialized, and the concentration loss value is obtained through the mass conservation equation by combining the water volume and the flow velocity with the microplastic concentration. The formula is as follows: where, is the water volume corresponding to node , is the flow velocity from node to , and represents the flow from node to For the flow velocity, the mass conservation equation takes into account the total amount of microplastics flowing from one node to another. The concentration loss value is updated until the update weight reaches a preset value to obtain the target loss value. This process ensures that the model can dynamically adapt to changes in hydrological conditions and provide accurate microplastic concentration predictions. By combining the target loss value and the microplastic concentration, the target microplastic concentration is calculated.

[0040] Step S60: Determine the concentration change direction and concentration change intensity based on the target microplastic concentration.

[0041] It should be noted that after obtaining the target microplastic concentration of each node, the concentration change direction can be determined by analyzing the concentration difference between adjacent nodes. Specifically, if the microplastic concentration of a node is significantly higher than that of its neighboring nodes, it can be inferred that the microplastics are diffusing from the high-concentration area to the low-concentration area, that is, the concentration change direction is from the high-concentration area to the low-concentration area.

[0042] Specifically, based on the target microplastic concentration, the predicted concentration distribution is determined, and then the concentration gradient of the node is calculated according to the predicted concentration distribution to determine the concentration change direction and concentration change intensity. To quantify the intensity of this concentration change, we can calculate the concentration gradient between adjacent nodes. The concentration gradient not only reflects the diffusion trend of microplastics in the water body but also indirectly reveals the influence of factors such as water flow velocity and topographic resistance on the migration of microplastics. The specific formula is: where represents the predicted concentration distribution, represents the position vector of node , and are the abscissa and ordinate directions of the spatial coordinate axes. The larger the absolute value of the concentration gradient, the more obvious the migration trend of microplastics in this direction and the faster the migration speed.

[0043] Step S70: Determine the transport node and transport distance based on the concentration change direction and concentration change intensity.

[0044] It should be noted that after determining the direction and intensity of the concentration change, the transport nodes can be further selected. In the dynamic adjacency matrix of the basin, each node has multiple possible downstream nodes. By comparing the concentration gradient directions and intensities of these downstream nodes, the node with the most consistent concentration gradient direction and the largest intensity with the current node is selected as the transport target node. This process not only considers the natural trend of microplastic migration but also combines the topological structure of the basin to ensure that the migration path conforms to the hydrological conditions. At the same time, the calculation of the transport distance is also crucial. Based on the DEM data and the river channel vector map, the actual transport distance from the current node to the target node can be accurately calculated. This distance not only reflects the spatial range of microplastic migration but also, in combination with the intensity of the concentration change, can further deduce the migration rate and time of microplastics.

[0045] Specifically, according to the direction of the concentration change, a set of downstream adjacent nodes consistent with the change direction is determined, and then the node corresponding to the preset concentration change intensity among the concentration change intensities in the set of downstream adjacent nodes is selected as the transport node. Specifically, the node with the largest absolute value of the concentration gradient is selected as the transport node of the current node. Combining the current node and the transport node based on the basin digital elevation model data and the river channel vector map, the transport distance is obtained. Specifically, based on the initial edge set, that is, the natural river channel connection situation between each node, to determine the transport distance. When there is a direct river channel connection between node and node , the transport distance is the length of the center line of the river channel. When there is no direct river channel connection between node and node , the shortest path length between node and node is calculated through the basin digital elevation model data, so as to determine the transport distance of microplastics from node to node .

[0046] In this way, not only can the migration path of microplastics be traced, but also its diffusion speed and range in the basin can be predicted, providing important spatio-temporal information for pollution control.

[0047] Step S80, when the transport node is the basin outlet in the node set, the transport distances are combined to obtain the predicted result of the target transport path.

[0048] It should be noted that starting from the initial release position of microplastics, through the analysis of the concentration change direction and intensity at each node, the migration path is gradually traced until it reaches the basin outlet. During this process, the transport distances between each pair of nodes have been calculated and recorded. The accumulation of these transport distances forms the complete transport path of microplastics from the source to the basin outlet. By integrating these paths, the migration trajectory of microplastics within the entire basin can be obtained, including important information such as the key nodes it passes through, the path length, and the migration time.

[0049] In this embodiment, by obtaining the basin digital elevation model data, river channel vector map, microplastic attributes, and hydrological data, the basin digital elevation model data and the river channel vector map are partitioned to obtain a node set. Based on the hydrological data and the basin digital elevation model data, a dynamic adjacency matrix is generated. The microplastic attributes and the basin topographic features are fused to obtain the fused topographic features. Then, based on the dynamic adjacency matrix and the fused topographic features, calculations are performed to obtain the target microplastic concentration. According to the target microplastic concentration, the concentration change direction and the concentration change intensity are determined. Based on the concentration change direction and the concentration change intensity, the transport nodes and the transport distances are determined. Finally, when the transport node is the basin outlet in the node set, the transport distances are combined to obtain the predicted result of the target transport path. By integrating the basin digital elevation model, river channel vector map, microplastic attributes, and real-time hydrological data, and using the dynamic adjacency matrix and the fused topographic features, the microplastic transport path is accurately predicted, realizing a high-precision simulation of microplastic migration in a complex hydrodynamic environment and improving the pollution prevention and control efficiency.

[0050] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar content as in the above-mentioned first embodiment can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 2 , step S40 of the microplastic transport path prediction method based on a complex hydrodynamic environment further includes steps S201 to S205: Step S201, encoding the basin topographic features to obtain an initial feature vector.

[0051] It should be noted that the slope information in the high-resolution DEM data is extracted, which is the degree of inclination of the ground surface relative to the horizontal plane. The slope directly affects the water flow velocity and direction. Steep slopes usually accelerate the water flow, while gentle areas may cause sedimentation. Next, the roughness coefficient is calculated, which is an index quantifying the surface roughness and reflects the frictional force between the water flow and the riverbed or riverbank. A higher roughness coefficient means greater resistance, which will slow down the water flow and affect the migration speed of microplastics. Finally, the river channel curvature is evaluated by measuring the curvature of the river channel centerline to determine the meandering degree of the river. Highly curved river channels will lead to complex flow state changes and increase the residence time of microplastics in local areas.

[0052] Further, after obtaining the above features, they are integrated to obtain an initial feature vector, which is specifically expressed as: Among them, is the slope data, is the roughness coefficient, is the channel curvature.

[0053] Step S202: Process the initial feature vector through a two-layer GCN to obtain a first feature vector and a second feature vector.

[0054] It should be noted that by using a two-layer graph convolutional network (GCN) to process these initial feature vectors, the complex relationships between adjacent nodes can be captured, thus better simulating the actual water flow situation. GCN is a deep learning method specifically used for graph-structured data, which can effectively capture the spatial correlation between nodes.

[0055] Specifically, in the process of the first-layer GCN processing, the input is the initial feature vector of each node, which is linearly transformed through the first-layer weights, and then a non-linear activation function (such as ReLU) is added. The formula is as follows: Next, in the second-layer GCN processing, the second-layer weights are used to further refine the features. The formula is: Among them, and respectively represent the first-layer weights and the second-layer weights of the GCN. is the set of first-order neighbors of node . Through this two-layer GCN processing method, the understanding of the hydrological conditions in the basin can be improved, especially for those areas with complex terrain and variable water flow patterns. For example, steep slopes may accelerate the water flow and carry more microplastic particles, while areas with a high roughness coefficient may slow down the water flow speed and increase the residence time of microplastics.

[0056] Step S203: Encode the microplastic attributes to obtain attribute-encoded features.

[0057] It should be noted that the properties of microplastics, including density, shape factor, and material classification (such as PE / PET), all affect their migration behavior in water bodies. The density of microplastics determines their floating or sinking state in water: particles with a density less than that of water tend to float, while those with a density greater than water will sink. Secondly, the shape factor reflects the morphology of the particles. Fibrous or sheet-shaped microplastics are more likely to be carried by the water flow and remain suspended compared to spherical particles. To convert these physical characteristics into a form that can be processed by the model, we use a specific function to convert the material category into a numerical form. The specific formula is: represents the density of microplastics, is the shape factor, represents the function that converts the material category into a numerical value. In addition, based on Stokes' sedimentation formula, the theoretical sedimentation velocity of each microplastic particle is calculated and added to the model as a physical prior. The specific formula is: where, represents the sedimentation velocity of the microplastic particle, represents the density of the microplastic particle, represents the density of water, represents the acceleration due to gravity, represents the migration of the microplastic particle, represents the dynamic viscosity of water. It can be seen from the above formula that the greater the density difference between the microplastic particle and water, the faster the sedimentation velocity. The larger the radius of the microplastic particle, the faster the sedimentation velocity, and the greater the viscosity of water, the slower the sedimentation velocity. This coding method not only quantifies the basic physical properties of microplastics but also allows the model to adjust its prediction results based on these properties, thus more accurately simulating the migration trajectory of microplastics under complex hydrodynamic conditions.

[0058] Step S204: Obtain a fused terrain vector based on the attribute coding features and the second feature vector.

[0059] It should be noted that the attribute coding features and the second feature vector, these two types of key information, are combined through the multi-head cross-attention mechanism. In this process, the geographical features serve as the query (Query), while the microplastic attributes serve as the key-value (Key-Value). Through a specific mapping matrix , the terrain features, key vectors, and microplastic attributes are projected into the same space, and the attention weights are calculated. The purpose of this is to dynamically adjust the weight distribution between different features to capture the most relevant parts. Finally, through processing with a normalization function such as Layer Normalization, a fused feature vector is obtained. The specific formula is: Among them, are the mapping matrices of topographic features, key vectors, and microplastic properties respectively, is the dimension of the key vector, is the attention weight, represents the normalization function, is the fused feature vector. This fusion method not only integrates the information of terrain and microplastic properties but also enhances the expression ability and adaptability of the model.

[0060] Step S205: Obtain the fused topographic features based on the fused terrain vectors.

[0061] It should be noted that the fused vectors not only contain the information of natural geographical elements such as terrain slope, roughness coefficient, and river channel curvature but also consider the physical properties such as microplastic particle density, shape factor, and material classification.

[0062] Next, use these fused terrain vectors to construct the final fused topographic features. This process usually involves further processing of the fused vectors, such as enhancing the expression ability of the model through non-linear transformation (such as ReLU activation function) or adopting normalization techniques to ensure the scale consistency between different features. In addition, additional physical constraints or prior knowledge (such as the law of conservation of mass) can be introduced to improve the reliability and interpretability of model prediction. The fused topographic features can more accurately simulate the migration behavior of microplastics under complex hydrodynamic conditions. For example, for areas with special topographic features (such as steep slopes or highly meandering rivers), the changes in water flow velocity and direction can be better captured, thus providing a more accurate prediction of microplastic concentration distribution.

[0063] In this embodiment, the topographic features of the watershed are encoded, the enhanced spatial correlation feature vectors are obtained by using GCN processing, and combined with the encoding of microplastic physical properties, the fused terrain vectors are generated through the multi-head cross-attention mechanism. Finally, the fused topographic features are refined based on these fused terrain vectors, improving the simulation accuracy of microplastic migration behavior under complex hydrological conditions and providing strong support for precise pollution prevention and control.

[0064] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar content as in the above-mentioned embodiment one can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 3 , before step S70 of predicting the microplastic transport path based on the complex hydrodynamic environment, there are also steps S301 to S306: Step S301: Obtain the target watershed data.

[0065] It should be noted that the above target watershed data includes the microplastic concentration and topographic attribute data of the target watershed. Regarding the acquisition of microplastic concentration data, this usually needs to be completed through a combination of field sampling and laboratory analysis. In the selected target watershed, several monitoring points are set according to factors such as water body flow characteristics, population density, and industrial distribution, and water samples or sediment samples are collected regularly. After these samples are processed, advanced analysis techniques (such as microscopic infrared spectroscopy, pyrolysis gas chromatography-mass spectrometry, etc.) are used to measure the microplastic concentration. In addition, in the case of sparse monitoring data, a cross-basin transfer learning strategy combined with the Kriging spatial interpolation algorithm can also be used to generate pseudo-labels based on existing data to supplement the microplastic concentration information in unlabeled areas.

[0066] Secondly, the topographic attribute data includes key geographical information such as elevation, slope, roughness coefficient, and channel curvature. These data mainly come from digital elevation models (DEMs). To ensure accuracy, it is recommended to use DEM data with a resolution of no more than 10 meters, and relevant topographic parameters are extracted through GIS software. At the same time, river vector maps also need to be collected to identify river bifurcation points, the starting point, and the ending point of the watershed.

[0067] Step S302: Select a preset proportion of the target watershed data and use the Kriging spatial interpolation algorithm to generate pseudo-labels for unlabeled watersheds.

[0068] It should be noted that Kriging interpolation is a geostatistical method that not only considers the distance between sample points but also uses semi-variogram analysis to quantify spatial autocorrelation, thereby providing an optimal estimate and error variance for each prediction location.

[0069] Specifically, select 10% of the representative monitoring points in the target watershed as the initial data source. The selection of these monitoring points should be based on the principle of uniform distribution of watershed characteristics to ensure coverage of different topographic and hydrological conditions, so as to improve the accuracy of the interpolation results. First, calculate the semi-variogram model between the microplastic concentration of known monitoring points and geographical attributes (such as slope, elevation, etc.), and then interpolate the unknown area according to this model to generate pseudo-labels. In this way, data gaps can be effectively filled, and the microplastic concentration of the entire watershed can be obtained. The specific formula is: Among them, represents the Kriging spatial interpolation algorithm, represents the microplastic concentration of the target watershed, represents the topographic attribute data of the target watershed.

[0070] Step S303: Evaluate the pseudo-labels of unlabeled watersheds to obtain the corresponding confidence levels.

[0071] It should be noted that the Monte Carlo simulation method is adopted. By randomly sampling the data of known monitoring points multiple times and repeating the interpolation process, we can observe the distribution of prediction results under different sampling conditions, and then calculate the confidence interval and probability distribution of the pseudo-labels for each unlabeled basin.

[0072] Step S304, when the confidence level exceeds the preset confidence level, obtain the unlabeled basin data corresponding to the confidence level.

[0073] It should be noted that after generating the pseudo-labels, a confidence level threshold needs to be set. In this embodiment, it is set to 0.85 with reference to existing research. Only the regions with prediction probabilities exceeding this threshold are retained to ensure the reliability of the interpolation results. For each predicted region, if the predicted value of the microplastic concentration falls within the set high-confidence interval, the data of this region is marked as a reliable pseudo-label.

[0074] In addition, the unlabeled basin data corresponding to these high confidence levels not only includes the estimated values of microplastic concentrations, but also covers the associated confidence scores, geographical attributes (such as slope, elevation, etc.), and other factors that may affect the migration of microplastics. Using these detailed information, the dynamic change trend of microplastics in the target basin can be simulated more accurately. Next, incorporating these data into the domain adaptation fine-tuning stage helps to adjust the model parameters to better adapt to the specific conditions of the target basin. A step-by-step refinement method is used to integrate these high-confidence pseudo-labels.

[0075] Step S305, adopt an optimization strategy for the target basin data and the unlabeled basin data to obtain a loss value.

[0076] It should be noted that the domain difference alignment loss is calculated by methods such as the Maximum Mean Discrepancy (MMD). The aim is to narrow the feature distribution gap between the target basin (i.e., the data-rich training set) and the unlabeled basin, obtain the loss value, and then adjust the parameters to reduce the loss value so as to enhance the cross-domain generalization ability of the model. The specific formula is: where, represents the maximum mean discrepancy loss value, represents the mean square error loss value, represents the trade-off coefficient, takes 0.1 at the beginning of training and linearly increases to 0.9 round by round to achieve a smooth transition from MSE to MMD, represents the unlabeled basin data.

[0077] Step S306, update the loss value until the trade-off coefficient reaches the preset value, and execute the steps of determining the concentration change direction and concentration change intensity according to the target microplastic concentration.

[0078] It should be noted that a relatively small trade - off coefficient is set in the initial stage (e.g., = 0.1), and this coefficient is gradually increased as the number of training rounds increases until it reaches a preset maximum value (such as = 1.0). By doing so, the model can rely more on data - driven methods to converge quickly in the initial stage, and gradually pay more attention to physical constraints in the later stage, thereby ensuring that the final result not only conforms to the observed data but also follows the laws of nature. When the trade - off coefficient reaches the preset value, the step of determining the direction and intensity of concentration change according to the target micro - plastic concentration is executed. This step uses the optimized model to predict the distribution of micro - plastic concentration at each node in the basin at future times, and calculates the direction (i.e., flow direction) and intensity (i.e., flow velocity and diffusion rate) of the concentration change.

[0079] In this embodiment, by obtaining the data of the target basin, generating pseudo - labels using Kriging spatial interpolation, screening high - quality pseudo - labels through confidence, and finally adopting an optimization strategy to calculate the loss value and dynamically adjust the trade - off coefficient until the model is stable, the data coverage and prediction accuracy in few - sample regions are improved, and the reliability of the results is ensured by combining physical constraints.

[0080] Based on the first embodiment of the present application, the present application also provides a micro - plastic transport path prediction device based on a complex hydrodynamic environment. Please refer to Figure 4 , and the device includes: An acquisition module 10 for acquiring basin digital elevation model data, river channel vector maps, micro - plastic attributes, and hydrological data.

[0081] A partitioning module 20 for partitioning the basin digital elevation model data and the river channel vector maps to obtain a node set, where the nodes include elevation, slope direction, river channel bifurcation points, and basin outlets.

[0082] A calculation module 30 for calculating based on the hydrological data and the basin digital elevation model data to generate a dynamic adjacency matrix.

[0083] A fusion module 40 for fusing the micro - plastic attributes and the basin terrain features to obtain fused terrain features, where the basin terrain features are obtained based on the basin digital elevation model data and the river channel vector maps, and the basin terrain features include slope data, roughness coefficient, and river channel curvature.

[0084] The calculation module 30 is further configured to calculate based on the dynamic adjacency matrix and the fused terrain features to obtain the target micro - plastic concentration.

[0085] The calculation module 30 is further configured to determine the concentration change direction and the concentration change intensity according to the target micro - plastic concentration.

[0086] The calculation module 30 is further configured to determine a transport node and a transport path based on the concentration change direction and the concentration change intensity.

[0087] The result module 50 is configured to combine the transport paths when the transport node is the basin outlet in the node set to obtain a predicted result of the target transport path.

[0088] The microplastic transport path prediction device based on a complex hydrodynamic environment provided by the present application adopts the microplastic transport path prediction method in the above embodiment, and can solve the technical problem of how to improve the prediction accuracy of the microplastic path in a complex hydrodynamic environment. Compared with the prior art, the beneficial effects of the microplastic transport path prediction device based on a complex hydrodynamic environment provided by the present application are the same as those of the microplastic transport path prediction method based on a complex hydrodynamic environment provided by the above embodiment, and other technical features in the microplastic transport path prediction device based on a complex hydrodynamic environment are the same as the features disclosed in the method of the above embodiment, which will not be elaborated herein.

[0089] In one embodiment, the calculation module 30 is further configured to calculate through hydrological data to obtain a flow direction similarity; calculate according to the flow direction similarity to obtain a flow direction consistency between two adjacent nodes; obtain a terrain resistance according to the basin digital elevation model data; and obtain a dynamic adjacency matrix according to the flow direction consistency and the terrain resistance.

[0090] In one embodiment, the fusion module 40 is further configured to encode the basin terrain features to obtain an initial feature vector; process the initial feature vector through a double-layer GCN to obtain a first feature vector and a second feature vector; encode the microplastic attributes to obtain an attribute encoding feature; obtain a fused terrain vector based on the attribute encoding feature and the second feature vector; and obtain a fused terrain feature based on the fused terrain vector.

[0091] In one embodiment, the calculation module 30 is further configured to update and calculate through a graph propagation algorithm based on the dynamic adjacency matrix and the fused terrain feature to obtain a microplastic concentration; calculate according to the basin digital elevation model data and the river channel vector map to obtain the water volume corresponding to the node; obtain the flow velocity from the node to the corresponding adjacent node according to the hydrological data; initialize an update weight; obtain a concentration loss value according to the water volume and the flow velocity in combination with the microplastic concentration through a mass conservation equation; update the concentration loss value until the update weight reaches a preset value to obtain a target loss value; and calculate in combination with the target loss value and the microplastic concentration to obtain a target microplastic concentration.

[0092] In one embodiment, the calculation module 30 is further configured to determine a predicted concentration distribution according to the target microplastic concentration; and calculate the concentration gradient of the node according to the predicted concentration distribution to determine the concentration change direction and the concentration change intensity.

[0093] In one embodiment, the calculation module 30 is further configured to obtain target basin data, where the target basin data includes the microplastic concentration and topographic attribute data of the target basin; select a preset proportion of the target basin data and use the Kriging spatial interpolation algorithm to generate pseudo-labels for unlabeled basins; evaluate the pseudo-labels of the unlabeled basins to obtain corresponding confidence levels; when the confidence level exceeds the preset confidence level, obtain the unlabeled basin data corresponding to the confidence level; adopt an optimization strategy for the target basin data and the unlabeled basin data to obtain a loss value; update the loss value until the trade-off coefficient reaches a preset value, and perform the step of determining the concentration change direction and concentration change intensity according to the target microplastic concentration.

[0094] In one embodiment, the calculation module 30 is further configured to determine a set of downstream adjacent nodes consistent with the change direction according to the concentration change direction; select a node corresponding to a preset concentration change intensity in the concentration change intensity from the set of downstream adjacent nodes as the transport node; based on the basin digital elevation model data and the river channel vector map, combine the current node and the transport node to obtain the transport distance.

[0095] The present application provides a microplastic transport path prediction device based on a complex hydrodynamic environment. The microplastic transport path prediction device based on a complex hydrodynamic environment includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the microplastic transport path prediction method based on a complex hydrodynamic environment in the first embodiment above.

[0096] Refer to the following Figure 5 , which shows a schematic structural diagram of a microplastic transport path prediction device suitable for implementing the embodiments of the present application. The microplastic transport path prediction device based on a complex hydrodynamic environment in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The microplastic transport path prediction device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0097] As Figure 5As shown, the microplastic transport path prediction device based on a complex hydrodynamic environment may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the microplastic transport path prediction device based on a complex hydrodynamic environment are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the microplastic transport path prediction device based on a complex hydrodynamic environment to communicate with other devices wirelessly or wiredly to exchange data. Although various microplastic transport path prediction devices based on a complex hydrodynamic environment are shown in the figure, it should be understood that it is not required to implement or have all the shown ones. More or fewer can be implemented or had alternatively.

[0098] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network via the communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0099] The microplastic transport path prediction device based on a complex hydrodynamic environment provided by this application adopts the microplastic transport path prediction method in the above-mentioned embodiment, and can solve the technical problem of how to improve the prediction accuracy of the microplastic path in a complex hydrodynamic environment. Compared with the prior art, the beneficial effects of the microplastic transport path prediction device based on a complex hydrodynamic environment provided by this application are the same as those of the microplastic transport path prediction method based on a complex hydrodynamic environment provided by the above-mentioned embodiment, and other technical features in the microplastic transport path prediction device based on a complex hydrodynamic environment are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0100] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0101] As mentioned above, only the specific implementation manners of this application are described, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0102] This application provides a computer-readable medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the microplastic transport path prediction method based on a complex hydrodynamic environment in the above-mentioned embodiment.

[0103] The computer-readable medium provided by this application can, for example, be a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor devices, or any combination of the above. More specific examples of computer-readable media can include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable medium can be any tangible medium that contains or stores a program that can be executed by an instruction, or used by, or in combination with, a device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0104] The above computer-readable medium can be included in a microplastic transport path prediction device based on a complex hydrodynamic environment; or it can exist separately and not be assembled into a microplastic transport path prediction device based on a complex hydrodynamic environment.

[0105] The above computer-readable medium carries one or more programs. When the above one or more programs are executed by a microplastic transport path prediction device based on a complex hydrodynamic environment, the microplastic transport path prediction device based on a complex hydrodynamic environment can be programmed in one or more programming languages or combinations thereof to write computer program code for performing the operations of this application. The above programming languages include object-oriented programming languages - such as Java, Smalltalk, C++; and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0106] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based means for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0107] The modules described in the embodiments of the present application can be implemented in software or in hardware. In some cases, the name of the module does not constitute a limitation on the unit itself.

[0108] The readable medium provided by the present application is a computer-readable medium, and the computer-readable medium stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned microplastic transport path prediction method based on a complex hydrodynamic environment, which can solve the technical problem of how to improve the accuracy of microplastic path prediction in a complex hydrodynamic environment. Compared with the prior art, the beneficial effects of the computer-readable medium provided by the present application are the same as those of the microplastic transport path prediction method based on a complex hydrodynamic environment provided by the above embodiments, and will not be elaborated here.

[0109] The present application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of the above-mentioned microplastic transport path prediction method based on a complex hydrodynamic environment.

[0110] The computer program product provided by the present application can solve the technical problem of how to improve the accuracy of microplastic path prediction in a complex hydrodynamic environment. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as those of the microplastic transport path prediction method based on a complex hydrodynamic environment provided by the above embodiments, and will not be elaborated here.

[0111] The above are only some embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structural transformation made under the technical concept of the present application by using the content of the specification and drawings of the present application, or any direct / indirect application in other related technical fields, is included in the patent protection scope of the present application.

Claims

1. A method for predicting the transport path of microplastics based on a complex hydrodynamic environment, characterized in that, The method includes: Obtaining watershed digital elevation model data, river channel vector maps, microplastic attributes, and hydrological data; Dividing the watershed digital elevation model data and the river channel vector maps to obtain a node set, where the nodes include elevation, slope direction, river channel bifurcation points, and watershed outlets; Calculating based on the hydrological data and the watershed digital elevation model data to generate a dynamic adjacency matrix; Fusing the microplastic attributes and the watershed topographic features to obtain fused topographic features, where the watershed topographic features are obtained based on the watershed digital elevation model data and the river channel vector maps, and the watershed topographic features include slope data, roughness coefficient, and river channel curvature; Calculating based on the dynamic adjacency matrix and the fused topographic features to obtain the target microplastic concentration; Determining the concentration change direction and concentration change intensity according to the target microplastic concentration; Determining the transport nodes and transport distances based on the concentration change direction and the concentration change intensity; When the transport node is the watershed outlet in the node set, combining the transport distances to obtain the target transport path prediction result.

2. The method according to claim 1, wherein The step of calculating based on the hydrological data and the watershed digital elevation model data to generate a dynamic adjacency matrix includes: Calculating through the hydrological data to obtain the flow direction similarity; Calculating according to the flow direction similarity to obtain the flow direction consistency between two adjacent nodes, and the specific formula is: Among them, and are the flow velocities of nodes and node respectively, and is the flow direction similarity; Obtaining the topographic resistance according to the watershed digital elevation model data, and the specific formula is: Among them, is the elevation difference between nodes and ; is the horizontal spatial distance between nodes and ; is the resistance factor. Obtaining the dynamic adjacency matrix according to the flow direction consistency and the topographic resistance, and the specific formula is: wherein is the flow direction consistency degree, is the terrain resistance, represents an indicator function, and only adjacent nodes in the downstream direction are allowed to participate in the calculation.

3. The method according to claim 1, characterized in that The step of fusing the microplastic attributes and the watershed topographic features to obtain the fused topographic features, where the watershed topographic features are obtained based on the watershed digital elevation model data and the river channel vector maps, and the watershed topographic features include slope data, roughness coefficient, and river channel curvature includes: Encoding the watershed topographic features to obtain an initial feature vector, which is specifically represented as: Among them, is the slope data, is the roughness coefficient, is the river channel curvature; Processing through a two-layer GCN according to the initial feature vector to obtain a first feature vector and a second feature vector, and the specific formula is: Among them, and represent the weights of the first layer and the second layer of the GCN respectively, is the set of first-order neighbors of node ; Encoding the microplastic attributes to obtain an attribute-encoded feature, and the specific formula is: represents the density of microplastics, is the shape factor, represents a function that converts the material category to a numerical value; Based on the attribute-encoded feature and the second feature vector, obtaining a fused topographic vector, and the specific formula is: Among them, are the mapping matrices of terrain features, key vectors, and microplastic properties respectively, is the dimension of the key vector, is the attention weight, represents the normalization function, is the fused feature vector; Obtaining the fused topographic features based on the fused topographic vector.

4. The method according to claim 1, wherein The step of calculating based on the dynamic adjacency matrix and the fused topographic features to obtain the target microplastic concentration includes: Updating and calculating through a graph propagation algorithm based on the dynamic adjacency matrix and the fused topographic features to obtain the microplastic concentration, and the specific formula is: Among them, is the node in the time prediction of microplastic concentration, is the learnable weight matrix; Calculating according to the watershed digital elevation model data and the river channel vector maps to obtain the water volume corresponding to the nodes; Obtaining the flow velocity from the nodes to the corresponding adjacent nodes according to the hydrological data; Initializing the update weight; Obtaining the concentration loss value through the mass conservation equation according to the water volume and the flow velocity in combination with the microplastic concentration, and the formula is: Among them, is the water volume corresponding to the node . is the flow velocity from the node to . represents the flow velocity from the node to . Updating the concentration loss value until the update weight reaches a preset value to obtain the target loss value; Calculate based on the target loss value and the microplastic concentration to obtain the target microplastic concentration.

5. The method according to claim 1, characterized in that, The step of determining the concentration change direction and concentration change intensity according to the target microplastic concentration includes: Determine the predicted concentration distribution according to the target microplastic concentration; Calculate the concentration gradient of the nodes according to the predicted concentration distribution to determine the concentration change direction and concentration change intensity. The formula is: Among them, represents the predicted concentration distribution, represents the node position vector, and are the abscissa direction and ordinate direction of the spatial coordinate axes.

6. The method according to claim 1, wherein Before the step of determining the transport node and transport distance based on the concentration change direction and the concentration change intensity, it includes: Obtain the target watershed data, which includes the microplastic concentration and topographic attribute data of the target watershed; Select a preset proportion of the target watershed data and use the Kriging spatial interpolation algorithm to generate pseudo-labels for the unlabeled watershed. The specific formula is: Among them, represents the Kriging spatial interpolation algorithm, represents the microplastic concentration in the target basin, represents the topographic attribute data of the target basin; Evaluate the pseudo-labels of the unlabeled watershed to obtain the corresponding confidence level; When the confidence level exceeds the preset confidence level, obtain the unlabeled watershed data corresponding to the confidence level; Adopt an optimization strategy for the target watershed data and the unlabeled watershed data to obtain a loss value. The specific formula is: Among them, represents the maximum mean difference loss value, represents the mean squared error loss value, represents the trade-off coefficient, represents the unlabeled watershed data; Update the loss value until the trade-off coefficient reaches the preset value, and execute the step of determining the concentration change direction and concentration change intensity according to the target microplastic concentration.

7. The method according to claim 1, characterized in that The step of determining the transport node and transport distance based on the concentration change direction and the concentration change intensity includes: Determine the set of downstream adjacent nodes consistent with the change direction according to the concentration change direction; Select the nodes corresponding to the preset concentration change intensity in the concentration change intensity from the set of downstream adjacent nodes as the transport nodes; Based on the watershed digital elevation model data and the river channel vector map, combine the current node and the transport node to obtain the transport distance.

8. A microplastic transport path prediction device based on a complex hydrodynamic environment, characterized in that, The device includes: An acquisition module for acquiring watershed digital elevation model data, river channel vector maps, microplastic attributes, and hydrological data; A division module for dividing the watershed digital elevation model data and the river channel vector map to obtain a set of nodes, where the nodes include elevation, slope direction, river channel bifurcation points, and watershed outlets; A calculation module for calculating based on the hydrological data and the watershed digital elevation model data to generate a dynamic adjacency matrix; A fusion module for fusing the microplastic attributes and watershed topographic features to obtain fused topographic features. The watershed topographic features are obtained based on the watershed digital elevation model data and the river channel vector map, and the watershed topographic features include slope data, roughness coefficient, and river channel curvature; The calculation module is also used to calculate based on the dynamic adjacency matrix and the fused topographic features to obtain the target microplastic concentration; The calculation module is also used to determine the concentration change direction and concentration change intensity according to the target microplastic concentration; The calculation module is also used to determine the transport node and transport distance based on the concentration change direction and the concentration change intensity; A result module for combining the transport distances when the transport node is the watershed outlet in the set of nodes to obtain a target transport path prediction result.

9. A microplastic transport path prediction device based on a complex hydrodynamic environment, characterized in that, The device includes: a memory, a processor, and a microplastic transport path prediction program stored on the memory and running on the processor. The microplastic transport path prediction program based on a complex hydrodynamic environment is configured to implement the steps of the microplastic transport path prediction method based on a complex hydrodynamic environment according to any one of claims 1-7.

10. A medium, characterized in that, The medium stores a microplastic transport path prediction program based on a complex hydrodynamic environment. When the microplastic transport path prediction program based on a complex hydrodynamic environment is executed by a processor, it implements the steps of the microplastic transport path prediction method based on a complex hydrodynamic environment according to any one of claims 1-7.

Images