Dynamic water flow analysis method and big data analysis system based on digital twinning

Abstract

The invention provides a digital twinning-based water flow dynamic analysis method and a big data analysis system, and the method comprises the steps: obtaining a historical water flow data set of a target water area, collecting a real-time water flow monitoring data set, and constructing a digital twinning water flow model corresponding to the target water area; performing multi-modal feature fusion processing on the historical water flow data set and the real-time water flow monitoring data set to generate a dynamic water flow feature set, loading the dynamic water flow feature set into a digital twin water flow model based on a spatial distribution topological structure of the digital twin water flow model to generate a water flow dynamic change rule set, and performing dynamic water flow change on the dynamic water flow change rule set. And generating a water flow optimization strategy set according to the water flow dynamic change rule set, and feeding back the water flow optimization strategy set to the water flow regulation and control system to execute dynamic regulation and control operation. By adopting the technical scheme of the invention, the space adaptability and the dynamic response capability of the water flow optimization strategy decision can be obviously improved.

Description

Technical Field

[0001] The present invention relates to the field of digital twins, and in particular to a water flow dynamics analysis method and a big data analysis system based on digital twins. Background Art

[0002] With the widespread application of digital twin technology in the field of water conservancy projects, water flow dynamic analysis technology has gradually become a key means of water area management. Existing technologies usually adopt isolated data processing modes, or extract static laws from historical hydrological data to build prediction models, or make local parameter adjustments based on real-time monitoring data, resulting in the construction of digital twins with spatiotemporal fragmentation. In traditional methods, spatial topology modeling often relies on fixed geographic information system parameters to solidify grid properties, while dynamic rule derivation uses preset empirical formulas for one-way calculations, making it difficult for water flow simulation results to simultaneously reflect the interactive effects of geological environment evolution and real-time water flow status. Such technical defects result in insufficient matching between the generated optimization strategies and the dynamic characteristics of real water areas, which easily leads to regulatory lag effects when responding to sudden water flow changes. Summary of the Invention

[0003] The present invention provides a water flow dynamic analysis method and a big data analysis system based on digital twins.

[0004] In a first aspect, an embodiment of the present invention provides a water flow dynamic analysis method based on digital twins, comprising: Acquire a historical water flow data set of the target water area and collect a real-time water flow monitoring data set, wherein the historical water flow data set includes a plurality of historical water flow state parameters, and the real-time water flow monitoring data set includes a plurality of real-time water flow monitoring parameters; Constructing a digital twin water flow model corresponding to the target water area, wherein the digital twin water flow model includes a spatial distribution topological structure of the target water area and water flow dynamic association rules; Performing multimodal feature fusion processing on the historical water flow data set and the real-time water flow monitoring data set to generate a dynamic water flow feature set; Based on the spatial distribution topology of the digital twin water flow model, the dynamic water flow feature set is loaded into the digital twin water flow model to generate a set of dynamic water flow change rules; A water flow optimization strategy set is generated according to the water flow dynamic change law set.

[0005] In a second aspect, an embodiment of the present invention provides a big data analysis system, including: a memory storing a computer program; A processor is used to load the computer program to implement the water flow dynamic analysis method based on digital twin as described above.

[0006] The digital twin-based water flow dynamics analysis method provided by this invention integrates the long-term evolution patterns of historical water flow data sets with the immediate state characteristics of real-time water flow monitoring data sets to construct a digital twin model with a spatially distributed topological structure and dynamic water flow association rules, achieving deep coupling between the physical water area and the virtual model. During the processing process, the dynamic water flow feature set generated by multimodal feature fusion is loaded into the model topology, forming a dynamic law derivation mechanism based on spatiotemporal correlation. This allows the generated water flow optimization strategy set to reflect the constraints of the water area's spatial structure and capture the temporal characteristics of water flow dynamic evolution. By organically integrating dynamic water flow association rules with the spatial topological network, this method enables the optimization strategy generation process to simultaneously consider the interaction between geological permeability characteristics, environmental coverage factors, and real-time water flow parameters, significantly improving the spatial adaptability and dynamic responsiveness of strategy decisions. Furthermore, the real-time update characteristics of the digital twin model are utilized to establish a two-way interactive channel between the virtual and real systems, ensuring that dynamic control operations can both act on physical control devices and reversely drive the iterative optimization of model parameters, forming an analysis and control system with adaptive evolution capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0008] Figure 1 This is a flow chart of a water flow dynamic analysis method based on digital twins provided in an embodiment of the present invention.

[0009] Figure 2 This is a schematic diagram of the composition of a big data analysis system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0011] See also Figure 1 , Figure 1A flowchart of a water flow dynamics analysis method based on digital twins is provided in an embodiment of the present invention. The water flow dynamics analysis method based on digital twins can be executed by a big data analysis system. The water flow dynamics analysis method based on digital twins may include the following steps: Step S100: Acquire historical water flow data sets of the target water area and collect real-time water flow monitoring data sets, wherein the historical water flow data sets include a plurality of historical water flow state parameters, and the real-time water flow monitoring data sets include a plurality of real-time water flow monitoring parameters.

[0012] The historical flow data set is the data accumulated from monitoring the flow conditions in the target water area over a period of time. This data covers multiple historical flow state parameters, such as the average flow velocity, mainstream direction, and water level changes at different depths. These parameters reflect the flow characteristics and patterns in the target water area over the historical time period. The real-time flow monitoring data set is the data obtained from real-time monitoring of the flow in the target water area at the current moment. It contains multiple real-time flow monitoring parameters, such as the instantaneous flow velocity, real-time flow direction, and real-time water level height. These parameters can promptly reflect the current flow status of the target water area.

[0013] When acquiring historical flow data, as permitted by laws and regulations, data can be extracted from the databases of relevant hydrological monitoring departments, water conservancy management agencies, or scientific research institutions. These institutions record flow data in the target water area over a long period of time and store it according to standards and formats. Real-time flow monitoring data is collected using various flow monitoring devices. For example, flow meters are used to measure water velocity. These flow meters operate based on electromagnetic induction or ultrasonic principles. Placed at different locations within the target water area, they can provide real-time flow velocity information. Water level gauges are used to monitor water levels. These can be pressure gauges or ultrasonic gauges, which convert water level information into electrical or digital signals for recording. For flow direction monitoring, flow meters can be used. These flow meters sense the direction of the water flow and convert it into specific angles or directions. These monitoring devices are installed at key locations within the target water area, such as the main stream of the river or at the confluence of tributaries. Real-time flow monitoring parameters are collected at intervals (e.g., every minute, every hour) and the data is transmitted to a data processing center for storage and analysis.

[0014] Step S200: Construct a digital twin water flow model corresponding to the target water area, where the digital twin water flow model includes the spatial distribution topology structure of the target water area and the dynamic association rules of the water flow.

[0015] The digital twin water flow model is a digital mapping of the target water body's flow system. By simulating the physical properties and flow patterns of the target water body, it can reflect the actual flow state of the target water body in real time. The spatial distribution topology describes the spatial relationships and connections between various locations in the target water body, such as flow channels between different regions and the connections between tributaries and the main stream. It presents the geographical layout and flow paths of the target water body in a graphical or digitized form. The dynamic association rules of water flow describe the laws governing the movement and changes of water flow in the target water body, including rules for the transmission of flow velocity, the balance of water level, and the coordination of flow direction. These rules reflect the behavioral characteristics and interrelationships of water flow under different conditions.

[0016] As an implementation method, in step S200, constructing a digital twin water flow model corresponding to the target water area may specifically include the following steps S210 to S250: Step S210: calling a preset three-dimensional grid processing tool to perform grid division processing on the geographic space of the target water area, generating a set of spatial grids of the target water area, and assigning an initial water flow attribute parameter set to each spatial grid.

[0017] The preset three-dimensional gridding tool is pre-developed or selected software or algorithm for gridding the geographic space of the target water area. This tool can divide the target water area into multiple three-dimensional spatial grids based on its geographic boundaries and topographic information, with each grid representing an area within the target water area. The spatial grid set is a collection of all spatial grids obtained from these divisions, which represents the geographic space of the target water area in a discrete manner. The initial water flow attribute parameter set is a set of water flow-related parameters pre-set for each spatial grid, including initial flow velocity, initial water level, initial flow direction, etc. These parameters are used to describe the initial state of the water flow within the spatial grid.

[0018] When using the pre-defined 3D gridding tool, you first need to obtain the geospatial data for the target water area, such as terrain elevation data and boundary coordinates, using a geographic information system (GIS). This data is then input into the 3D gridding tool, which then divides the target water area into grids based on pre-defined gridding rules (such as grid size, shape, and density). For example, using the open-source gridding software Tecplot, set the grid size to 10 meters x 10 meters x 5 meters, divide the target water area into grids, and generate a spatial grid set.

[0019] When assigning an initial set of flow attribute parameters to each spatial grid, reference can be made to both historical flow data and real-time flow monitoring data. For example, for a particular spatial grid, initial flow velocity, initial water level, and initial flow direction parameters can be assigned based on the average flow velocity, average water level, and mainstream direction at that location in the historical flow data. Alternatively, based on real-time flow monitoring data, initial flow attribute parameters similar to those in the monitoring data can be assigned to spatial grids closer to the monitoring point.

[0020] Step S220: Acquire a hydrogeological data set of the target water area, where the hydrogeological data set includes a permeability parameter set, a geological structure parameter set, and a surface cover parameter set.

[0021] The hydrogeological data set describes the subsurface and surface geological conditions and hydrological characteristics of the target watershed. The permeability parameter set reflects the permeability of the target watershed's strata to water flow, including information such as the permeability values ​​of different strata and their spatial distribution. The higher the permeability, the easier it is for water to flow through the strata. The geological structure parameter set encompasses the geological structural characteristics of the target watershed, such as the location, strike, and size of faults, the inclination angle of strata, and the presence of folds. These geological structures influence the path and velocity of water flow. The surface cover parameter set describes the surface cover of the target watershed, including vegetation type, coverage, and topographic relief. Surface cover affects water convergence, infiltration, and flow. There are various ways to obtain the hydrogeological data set for the target watershed. Permeability parameter sets can be obtained through field drilling tests and laboratory permeability testing. Core samples are collected by drilling at different locations in the target watershed. These core samples are then tested in the laboratory using a permeability tester to determine the permeability values ​​of different strata. Then, based on the location of the boreholes and the test results, a spatial distribution map of the permeability is drawn to form a set of permeability parameters. The set of geological structure parameters can be obtained through geological surveys and geophysical exploration methods. Geological surveyors conduct field surveys of the target waters, observing outcrops of strata, signs of faults, and other information, and record relevant information about the geological structure. Geophysical exploration methods such as seismic exploration and electrical exploration can detect deep information about underground geological structures. For example, a seismic exploration instrument is used to lay out survey lines in the target waters to collect seismic wave data. By processing and analyzing the seismic wave data, a three-dimensional image of the geological structure and related parameters are obtained.

[0022] Land cover parameter sets can be obtained through a combination of remote sensing technology and field surveys. Using satellite remote sensing images or aerial photography, image recognition and classification algorithms can be used to identify the vegetation type and coverage of the surface.

[0023] Step S230: performing dynamic association rule mining processing on the spatial grid set based on the hydrogeological data set to generate water flow dynamic association rules.

[0024] Dynamic association rule mining is the process of discovering the inherent patterns and relationships within hydrogeological data sets and spatial grids that underlie the movement and changes of water flows. By analyzing the relationships between permeability, geological structure, and surface cover parameters in the hydrogeological data set and the flow states at each grid node in the spatial grid set, rules describing the movement and changes of water flows under different conditions are mined. Dynamic flow association rules are a collection of these mined rules and provide a basis for subsequent simulation and analysis of flow dynamics in target waters.

[0025] As an implementation manner, step S230 may specifically include the following steps S231 to S237: Step S231: extracting a permeability parameter set, a geological structure parameter set, and a surface cover parameter set from the hydrogeological data set.

[0026] Within a hydrogeological data set, permeability parameter sets, geological structure parameter sets, and land cover parameter sets each describe the geological and hydrological characteristics of the target watershed from different perspectives. These parameter sets are extracted to enable subsequent targeted analysis and processing. In practice, parameter set extraction can be accomplished using data extraction tools or programming languages, depending on the data storage format and structure. For example, if the hydrogeological data set is stored in a relational database, SQL queries can be used to extract the permeability parameter sets, geological structure parameter sets, and land cover parameter sets from the database.

[0027] Step S232: Map the permeability parameter set to the corresponding grid nodes in the spatial grid set to generate a permeability distribution feature set for each grid node.

[0028] Mapping the permeability parameter set to the corresponding grid nodes in the spatial grid set is to associate the permeability parameters with the spatial grid and determine the permeability value corresponding to each grid node. The permeability distribution feature set is a collection of the permeability values ​​and spatial distribution information of each grid node, reflecting the permeability conditions at different locations in the target water area.

[0029] During mapping, a match can be made based on the spatial location of the grid node and the spatial coordinate information in the permeability parameter set. For example, for each grid node, its coordinates in geographic space are found. Then, the permeability value closest to the coordinates is searched in the permeability parameter set and assigned to the grid node. Spatial indexing algorithms (such as quadtrees and KD trees) can be used to improve matching efficiency.

[0030] Step S233: determining a geological fault zone distribution feature set between different grid nodes in the spatial grid set according to the geological structure parameter set, and generating a geological structure influencing factor set in combination with the permeability distribution feature set.

[0031] Geological fault zones are areas of rock fracture and displacement in the Earth's crust. The geological fault zone distribution feature set describes the location, strike, and scale of these fault zones between different grid nodes in a spatial grid set. The geological structure impact factor set is a set of factors derived by comprehensively considering the distribution characteristics of geological fault zones and permeability distribution characteristics. It reflects the degree to which geological structures influence water flow.

[0032] When determining a set of geological fault zone distribution characteristics based on a set of geological structural parameters, spatial analysis can be performed by combining relevant geological structural information (such as fault location and strike) with a set of spatial grids. For example, using geographic information system (GIS) software, the fault information in the set of geological structural parameters can be overlaid with the spatial grid set to determine whether each grid node is located on a geological fault zone and the distribution of fault zones across different grid nodes. When combining the set of permeability distribution characteristics to generate a set of geological structural impact factors, a weighted calculation approach can be used. Grid nodes located on geological fault zones can be assigned a higher impact factor because fault zones alter the path and velocity of water flow. Furthermore, considering the influence of permeability, areas with higher permeability have a relatively lower impact factor, as water flows more easily through them. For example, let the permeability of a grid node be P, and whether it is located on a fault zone be F (F = 1 if located on a fault zone, F = 0 otherwise). The geological structural impact factor I can be calculated using the following formula: I = F × k1 + (1-F) × k2 / P, where k1 and k2 are preset weighting coefficients that can be adjusted based on actual conditions. By calculating the geological structure influencing factor for each grid node, a geological structure influencing factor set is generated.

[0033] Step S234: extracting the vegetation cover type parameters and the terrain relief parameters from the surface cover parameter set, and generating a surface cover resistance coefficient set according to the vegetation cover type parameters and the terrain relief parameters.

[0034] The vegetation cover type parameter describes the type and distribution of surface vegetation in the target water area. Different vegetation cover types exert varying resistance to water flow. The terrain relief parameter reflects the elevation variation of the target water area's surface, which affects the velocity and direction of water flow. The surface cover resistance coefficient set is a set of coefficients calculated based on the vegetation cover type parameter and the terrain relief parameter, reflecting the resistance of the surface cover to water flow.

[0035] When extracting vegetation cover type parameters and terrain relief parameters from the surface cover parameter set, the extraction can be performed based on the storage structure and format of the data.

[0036] When generating a set of surface cover resistance coefficients based on vegetation cover type parameters and terrain relief parameters, a resistance coefficient calculation model can be established. For vegetation cover types, a basic resistance coefficient can be assigned to each vegetation type based on the characteristics of different vegetation (such as vegetation density, height, etc.). For terrain relief parameters, a terrain resistance coefficient can be calculated based on the slope and roughness of the terrain. These two coefficients are then combined to obtain the surface cover resistance coefficient. For example, let the basic resistance coefficient corresponding to the vegetation cover type be C v , the terrain resistance coefficient is C t , then the surface cover resistance coefficient C s It can be calculated by the following formula: C s =C v ×k3+C t ×k4; where k3 and k4 are preset weight coefficients that can be adjusted according to actual conditions. By calculating the surface cover resistance coefficient for each grid node, a surface cover resistance coefficient set is generated.

[0037] Step S235: Input the geological structure influencing factor set and the surface cover resistance coefficient set into the preset water flow dynamic impact analysis model, perform dynamic association analysis on the water flow transmission paths between adjacent grid nodes in the spatial grid set, and generate an initial water flow dynamic association rule set.

[0038] The pre-built flow dynamics impact analysis model is a pre-built model for analyzing the impact of geological structure and land cover on flow dynamics. This model can be a numerical model based on physical principles or a predictive model based on machine learning. The set of geological structure impact factors and the set of land cover resistance coefficients reflect the degree of influence of geological structure and land cover on water flow, respectively. These inputs are used to analyze the flow paths between adjacent grid nodes in a spatial grid set.

[0039] Dynamic correlation analysis is the process of analyzing the likelihood, direction, and speed of water flow between adjacent grid nodes. By analyzing geological structural factors and the surface resistance coefficient, it can be determined whether water flow between adjacent grid nodes is hindered and how the direction and speed of water flow change.

[0040] For example, the pre-set model for analyzing the dynamic impact of water flow is a neural network model based on machine learning. Its input layer contains two features: geological structure influence factors and surface cover resistance coefficients. The hidden layer contains multiple neurons, and the output layer outputs information such as the probability, direction, and speed of water flow transmission between adjacent grid nodes. The set of geological structure influence factors and the set of surface cover resistance coefficients are fed into the model as input data. The model uses forward propagation calculations to obtain information about water flow transmission between adjacent grid nodes and, based on this information, generates an initial set of water flow dynamic association rules.

[0041] Step S236: performing parameter adaptability verification processing on the initial water flow dynamic association rule set based on the initial water flow attribute parameter set of the spatial grid set, and determining the flow velocity transfer rule, water level balance rule and flow direction coordination rule in the water flow dynamic association rule.

[0042] Parameter adaptability verification checks whether the rules in the initial flow dynamics association rule set are compatible with the initial flow attribute parameters of the spatial grid set. By comparing the parameters in the rules with the initial flow attribute parameters, the rules are adjusted and optimized to ensure that they accurately describe the flow dynamics of the target water area.

[0043] Flow velocity transfer rules describe the flow velocity transfer relationship between adjacent grid nodes, such as the velocity attenuation pattern and transfer coefficient. Water level balance rules specify the equilibrium conditions for water levels between adjacent grid nodes. When the water level difference exceeds a threshold, water will flow from the node with the higher water level to the node with the lower water level. Flow direction coordination rules ensure that the flow direction is consistent throughout the target water area, avoiding unreasonable flow direction conflicts.

[0044] When performing parameter adaptability verification, numerical simulation and comparative analysis methods can be used. First, based on the initial water flow dynamics association rule set and the initial water flow attribute parameters of the spatial grid set, the water flow dynamics of the target water area are numerically simulated to obtain the simulated water flow state. Then, the simulated water flow state is compared with the actual initial water flow attribute parameters to analyze whether the parameters in the rule are reasonable. If it is found that there is a large deviation between the parameters in the rule and the actual situation, the rule is adjusted. For example, if the flow velocity transfer coefficient obtained by simulation causes the flow velocity to decay too quickly during the transfer process, which is inconsistent with the actual situation, the flow velocity transfer coefficient is appropriately adjusted. Through multiple simulations and adjustments, the flow velocity transfer rule, water level balance rule, and flow direction coordination rule in the water flow dynamics association rule are determined.

[0045] Step S237: performing rule topology integration processing on the flow velocity transmission rule, the water level balance rule, the flow direction coordination rule and the topological connection relationship of the spatial grid set to generate a water flow dynamic association rule.

[0046] Rule topology integration involves combining velocity transfer rules, water level balance rules, and flow direction coordination rules with the topological connectivity of a spatial grid set. This topological connectivity describes the connections and adjacency between grid nodes. By integrating rules with these topological connectivity relationships, we can determine the spatial scope and effect of the rules.

[0047] Graph theory can be used to perform rule topology integration. A spatial grid set is represented as a graph, with each grid node as a vertex and the connections between adjacent grid nodes as edges. Flow velocity transfer rules, water level balance rules, and flow direction coordination rules are then applied to the edges and vertices of the graph to determine the rule constraints for each edge and vertex. For example, for edges between adjacent grid nodes, the flow velocity transfer direction and transfer coefficient are determined based on the flow velocity transfer rules; for each vertex, the water level and flow direction constraints are determined based on the water level balance rules and flow direction coordination rules. By performing rule topology integration on the entire graph, dynamic association rules for water flows are generated.

[0048] Step S240: topologically integrate the spatial grid set and the water flow dynamic association rules to generate a spatial distribution topology structure of the digital twin water flow model, and configure a dynamic update interface for the spatial distribution topology structure.

[0049] Topological integration combines the spatial structure of a spatial grid set with the dynamic association rules of water flow. This integration creates a spatially distributed topological structure that reflects the dynamics of water flow in the target water area. The spatially distributed topological structure of the digital twin water flow model describes the spatial relationships between locations and the flow patterns within the target water area. The dynamic update interface is a feature of the spatially distributed topological structure that allows for real-time updates of the data and rules within it to adapt to changes in the actual flow dynamics within the target water area.

[0050] As an implementation manner, step S240 may specifically include the following steps S241 to S249: Step S241: Based on the topological connection relationship of each grid node in the spatial grid set, determine a set of water flow transfer paths between adjacent grid nodes.

[0051] The topological connectivity of each grid node in a spatial grid set describes the connection and adjacency between that node and its neighboring nodes. Based on these topological connectivity relationships, the flow paths between adjacent grid nodes can be determined. The flow path set is the set of all flow paths between adjacent grid nodes, reflecting the possible flow paths of water in the target water area.

[0052] A graph search algorithm can be used to determine the set of water flow paths. The spatial grid set is represented as a graph, with each grid node as a vertex and the connections between adjacent grid nodes as edges. A graph search algorithm (such as breadth-first search or depth-first search) is then used to traverse the graph and find paths between all adjacent vertices. These paths are the water flow paths between adjacent grid nodes. For example, using a breadth-first search algorithm, starting from a grid node, the search expands layer by layer to its adjacent nodes, recording the paths traversed during the search until all adjacent nodes are traversed. The water flow paths between this node and its other adjacent nodes are then determined. By searching all grid nodes, a set of water flow paths between adjacent grid nodes is generated.

[0053] Step S242: Mapping the flow velocity transfer rule in the water flow dynamic association rule to each transfer path in the water flow transfer path set to generate a transfer path flow velocity constraint condition set.

[0054] Flow rate transfer rules describe the flow rate transfer relationship between adjacent grid nodes. These rules are mapped to each flow path in the flow path set, defining the flow rate transfer constraints for each path. The flow path flow rate constraint set is a collection of flow rate constraints for each path, specifying the flow rate variation rules for each path.

[0055] When mapping, matching can be performed based on the parameters in the velocity transfer rule and the characteristics of the water flow transfer path. For example, the velocity transfer rule may specify the velocity transfer coefficient and attenuation law. For each water flow transfer path, the velocity transfer coefficient and attenuation coefficient of the path can be determined based on the properties of the grid nodes at its starting point and end point (such as permeability, geological structure influence factors, etc.). Then, based on these coefficients, the velocity constraint conditions are generated for the path. Assume that the velocity transfer rule is v out =v in ×k×e -αL , where v in is the input flow rate, v out is the output flow rate, k is the transfer coefficient, α is the attenuation coefficient, and L is the path length. For each flow transfer path, the values ​​of k and α are determined based on the properties of the grid nodes at its start and end points. Substituting the input flow rate for that path into the formula yields the output flow rate constraint for that path. By mapping all flow transfer paths, a set of flow rate constraints for each transfer path is generated.

[0056] Step S243: performing water level association matching processing on the initial water flow attribute parameter set of each grid node in the spatial grid set according to the water level balance rule to generate a grid node water level balance constraint condition set.

[0057] The water level balance rule specifies the conditions for maintaining water levels between adjacent grid nodes. Based on this rule, the initial flow attribute parameter set for each grid node in the spatial grid set is subjected to water level correlation matching. This checks whether the initial water level of each grid node satisfies the water level balance rule and generates corresponding water level constraints based on the rule. The grid node water level balance constraint set is a set of water level constraints for each grid node, ensuring that the water levels of each grid node in the target water area remain balanced within a preset range.

[0058] When performing water level association matching, the initial water levels of adjacent grid nodes can be compared. If the water level difference between adjacent grid nodes exceeds the threshold specified by the water level balance rule, the water level needs to be adjusted. For example, if the water level balance rule stipulates that the water level difference between adjacent grid nodes shall not exceed Δh, for two adjacent grid nodes i and j, their initial water levels are h i and h j , if |h i -h j | >Δh, the water level needs to be adjusted based on the flow direction (from high-water-level nodes to low-water-level nodes) so that the adjusted water level difference satisfies the water level balance rule. A set of grid node water level balance constraints is generated by performing water level correlation matching on all adjacent grid nodes.

[0059] Step S244: extracting the flow direction consistency parameter set in the flow direction coordination rule, and combining it with the transfer path flow velocity constraint condition set to determine the flow direction coordination constraint condition set between adjacent grid nodes.

[0060] The flow consistency parameter set in the flow coordination rule describes the consistency conditions that the water flow direction must meet. For example, the water flow direction must align with the mainstream direction to avoid countercurrent. The flow coordination constraint set between adjacent grid nodes is determined in conjunction with the transfer path velocity constraint set. This ensures that the flow direction between adjacent grid nodes complies with the flow coordination rule while taking into account the velocity constraint.

[0061] When extracting the flow consistency parameter set, relevant parameters such as the mainstream direction, the allowable range of flow deviation, etc. can be parsed from the flow coordination rules. Then, the flow coordination constraints between adjacent grid nodes are determined in combination with the transfer path flow velocity constraint set. For example, for two adjacent grid nodes i and j, the flow velocity direction from node i to node j is determined according to the transfer path flow velocity constraint set, and then the flow velocity direction is checked to see if it is within the allowable range of flow deviation specified by the flow consistency parameter set. If it is not within the range, the flow direction needs to be adjusted so that the adjusted flow direction satisfies the flow coordination rules. By analyzing all adjacent grid nodes, a flow coordination constraint set between adjacent grid nodes is generated.

[0062] Step S245: Loading the transfer path flow rate constraint set, the grid node water level balance constraint set, and the flow direction coordination constraint set into the topological connection relationship of the spatial grid set to generate a rule-constrained topological network of the grid nodes.

[0063] Loading the set of flow path velocity constraints, grid node water level balance constraints, and flow direction coordination constraints into the topological connectivity of the spatial grid set applies these constraints to the topological structure of the spatial grid set, adding corresponding rule constraints to the connections between each grid node and its adjacent nodes. The rule-constrained topological network of grid nodes is a topological network with rule constraints, describing the spatial relationships and rule constraints between each grid node in the target water area.

[0064] When loading constraints, each constraint can be represented as an attribute of an edge or vertex in the topological network. For example, for a set of flow rate constraints for a transfer path, the flow rate constraint for each transfer path is treated as an attribute of the edge corresponding to that path; for a set of water level balance constraints for grid nodes, the water level constraint for each grid node is treated as an attribute of that node; and for a set of flow direction coordination constraints, the flow direction constraints between adjacent grid nodes are treated as attributes of the edges. By loading these constraints into the topological connectivity of a set of spatial grids, a regular constraint topological network of grid nodes is generated.

[0065] Step S246: dynamically adjusting the initial topological connection weights of the spatial grid set according to the constraint condition set corresponding to each topological connection edge in the rule-constrained topological network to generate an optimized topological connection weight set.

[0066] The constraint set corresponding to each topological connection edge in a rule-constrained topological network specifies constraints such as flow velocity, water level, and flow direction. Based on these constraints, the initial topological connection weights of the spatial grid set are dynamically adjusted. This is done based on the stringency and impact of the constraints, allowing the topological connection weights to more accurately reflect the difficulty and likelihood of water flow along that edge. The optimized topological connection weight set, obtained after these adjustments, better reflects the dynamic characteristics of water flow in the target water area.

[0067] When performing dynamic adjustments, a weighted calculation method can be used. For each topological edge, a weight coefficient is assigned to each constraint in its corresponding constraint set (such as flow rate, water level, and flow direction). The adjustment weight for the edge is then calculated based on these weight coefficients and the strictness of the constraints.

[0068] Step S247: performing topological relationship reconstruction processing on the initial topological structure of the spatial grid set based on the optimized topological connection weight set to generate an initial spatial distribution topological structure.

[0069] Topological reconstruction involves adjusting and reconstructing the initial topological structure of a spatial grid set based on an optimized set of topological connection weights. By adjusting the weights of topological edges, the connection strengths and relationships between grid nodes are altered, thereby reconstructing a topological structure that better reflects the dynamic characteristics of the target watershed. The resulting initial spatial distribution topology is the resulting topological structure, which more accurately reflects the spatial relationships between locations and the patterns of water flow within the target watershed.

[0070] When reconstructing topological relationships, graph algorithms (such as the Dijkstra algorithm or the Floyd-Warshall algorithm) can be used to adjust the initial topological structure of the spatial grid set. For example, the Dijkstra algorithm can be used to calculate the shortest path from each grid node to every other node based on an optimized set of topological connection weights. The topological connections are then reconstructed based on these shortest paths. This calculation and reconstruction of all grid nodes generates the initial spatial distribution topology.

[0071] Step S248: Call the dynamic update interface in the water flow dynamic association rule to perform rule consistency verification on the initial spatial distribution topology structure. If it is detected that there are topological connection edges in the initial spatial distribution topology structure that are inconsistent with the water flow dynamic association rule, the topological connection edges are synchronized and corrected with the rule parameters to generate a corrected spatial distribution topology structure.

[0072] The dynamic update interface, a feature of the dynamic flow association rules, allows for rule consistency verification of the spatial distribution topology. This verification process examines whether the topological edges in the initial spatial distribution topology conform to the dynamic flow association rules. If any inconsistent topological edges are detected, the rule parameters for those edges must be synchronized and corrected to ensure consistency between the topology and the rules.

[0073] When performing rule consistency verification, the attributes of the topological connection edges in the initial spatial distribution topology (such as flow velocity, water level, flow direction, etc.) can be compared with the rules in the water flow dynamic association rules. If it is found that the attributes of a topological connection edge do not meet the rule requirements, the dynamic update interface is called to correct the rule parameters of the edge. For example, if the rule stipulates that the flow velocity transfer coefficient of a topological connection edge should be within a certain range, and the actual flow velocity transfer coefficient of the edge exceeds this range, the flow velocity transfer coefficient of the edge is adjusted through the dynamic update interface to make it meet the rule requirements. By verifying and correcting all topological connection edges, a corrected spatial distribution topology is generated.

[0074] Step S249: Bind the corrected spatial distribution topology structure to the dynamic update interface to generate a spatial distribution topology structure of the digital twin water flow model with real-time update capabilities.

[0075] Interface binding is the process of associating and binding the modified spatial distribution topology with the dynamic update interface. This binding allows the modified spatial distribution topology to update its data and rules in real time through the dynamic update interface, thus achieving real-time update capabilities for the digital twin water flow model.

[0076] During interface binding, a function pointer or reference to the dynamically updated interface can be associated with the relevant data structure of the revised spatially distributed topology. For example, in a programming language, the dynamically updated interface function can be registered as a callback function in the revised spatially distributed topology data structure. When the spatially distributed topology needs to be updated, the callback function is called to implement the update. Through interface binding, a spatially distributed topology structure of a digital twin water flow model with real-time update capabilities is generated.

[0077] Step S250: Load the initial water flow attribute parameter set to the grid node corresponding to the spatial distribution topology structure through the dynamic update interface to complete the construction of the digital twin water flow model.

[0078] The dynamic update interface is used to load the initial set of flow attribute parameters into the corresponding grid nodes in the spatially distributed topology. This involves writing the initial flow attribute parameters (such as initial velocity, initial water level, and initial flow direction) of each grid node into the properties of the corresponding grid node in the spatially distributed topology. Once this step is completed, the digital twin water flow model incorporates the spatially distributed topology of the target water area, the dynamic association rules for water flow, and the initial flow attribute parameters, thus completing the construction of the digital twin water flow model.

[0079] When loading the initial water flow attribute parameter set, you can traverse each grid node in the spatial distribution topology structure, find the corresponding parameters from the initial water flow attribute parameter set according to the node's identifier or index, and write the parameters to the node's attributes through the dynamic update interface. For example, when implementing it using a programming language, you can loop through the node list of the spatial distribution topology structure, and for each node, find the corresponding parameters from the initial water flow attribute parameter set according to its unique identifier, and then call the write function of the dynamic update interface to write the parameters to the node's attributes. In this way, the initial water flow attribute parameter set is loaded into the grid node corresponding to the spatial distribution topology structure, completing the construction of the digital twin water flow model.

[0080] Step S300: performing multimodal feature fusion processing on the historical water flow data set and the real-time water flow monitoring data set to generate a dynamic water flow feature set.

[0081] Multimodal feature fusion is the process of comprehensively analyzing and fusing different types of features from historical water flow data sets and real-time water flow monitoring data sets. The historical water flow data set contains the characteristics and patterns of water flow in the target water area over a period of time, while the real-time water flow monitoring data set reflects the current water flow status in the target water area. Through multimodal feature fusion, the features from these two data sets can be complemented and integrated to extract more comprehensive and accurate dynamic water flow characteristics. The dynamic water flow feature set is the feature set obtained after the fusion process, which can more effectively describe the dynamic changes in water flow in the target water area.

[0082] As an implementation manner, step S300 may specifically include the following steps S310 to S350: Step S310: extracting a static water flow feature set and a dynamic change trajectory set from the historical water flow data set. The static water flow feature set includes a flow velocity mean parameter, a flow direction stability parameter, and a water level fluctuation threshold parameter.

[0083] The static flow feature set is a set of features from the historical flow data set that reflects the average state and stability of the water flow in the target water area over a period of time. The mean flow velocity parameter is the average flow velocity of the water flow in the target water area over the historical period, reflecting the overall flow rate. The flow direction stability parameter describes the stability of the water flow direction over the historical period, such as the range and frequency of flow direction changes. The water level fluctuation threshold parameter is the maximum allowable range of water level fluctuations over the historical period, reflecting water level stability.

[0084] A dynamic change trajectory is a collection of historical flow data that records the state changes of water flow at different time points. It can be obtained by analyzing the time series of historical flow data, for example, recording the changes in flow velocity, flow direction, and water level at each time point to form a dynamic change trajectory.

[0085] Data processing and analysis methods can be used to extract static flow feature sets and dynamic change trajectory sets. For static flow feature sets, the mean velocity parameter is calculated by calculating the average flow velocity in historical flow data. The flow direction stability parameter is obtained by statistically analyzing flow direction changes, such as calculating the standard deviation or frequency of flow direction changes. The water level fluctuation range is analyzed to determine the water level fluctuation threshold parameter. For dynamic change trajectory sets, the historical flow data can be arranged in chronological order, and the flow velocity, flow direction, and water level data at each time point can be extracted to form a dynamic change trajectory.

[0086] Step S320: extracting real-time flow velocity parameters, real-time flow direction parameters and real-time water level parameters from the real-time water flow monitoring data set, and generating a real-time water flow fluctuation trajectory data set.

[0087] Real-time flow velocity, flow direction, and water level parameters are key parameters in real-time water flow monitoring data sets that reflect the current flow status of the target water area. The real-time flow velocity parameter indicates the current speed of the water flow in the target water area, the real-time flow direction parameter indicates the current direction of the water flow, and the real-time water level parameter indicates the current water level in the target water area.

[0088] The real-time water flow fluctuation trajectory dataset consists of continuously collected real-time flow velocity parameters, real-time flow direction parameters, and real-time water level parameters. It records the real-time fluctuation of water flow in the target water area over a period of time. To extract the real-time flow velocity parameters, real-time flow direction parameters, and real-time water level parameters, data can be obtained from real-time water flow monitoring equipment.

[0089] Step S330: performing spatiotemporal alignment processing on the static water flow feature set and the dynamic change trajectory set to generate a historical water flow spatiotemporal feature vector set.

[0090] Spatiotemporal alignment is the process of matching and aligning the static flow feature set and the dynamic change trajectory set in time and space. Because the static flow feature set and the dynamic change trajectory set may have different time scales and spatial ranges, spatiotemporal alignment is necessary to unify them into the same spatiotemporal coordinate system for subsequent analysis and fusion.

[0091] The historical water flow spatiotemporal feature vector set is a feature vector set obtained after spatiotemporal alignment processing, which contains the water flow characteristic information of the target water area at different time and space positions.

[0092] Temporal and spatial interpolation methods can be used for spatiotemporal alignment. In the temporal dimension, the static flow feature set is temporally interpolated based on the time series of the dynamically changing trajectory set, ensuring that the temporal resolution of the static flow feature set is consistent with that of the dynamically changing trajectory set. In the spatial dimension, the static flow feature set and the dynamically changing trajectory set are spatially interpolated based on the division of the spatial grid set, ensuring that their spatial resolution and range are consistent.

[0093] Step S340: Perform multi-dimensional mapping analysis on the real-time water flow fluctuation trajectory dataset and the historical water flow spatiotemporal feature vector set to determine the real-time water flow abnormal fluctuation interval set.

[0094] Multidimensional mapping analysis involves comparing and analyzing real-time flow fluctuation trajectory datasets with historical flow spatiotemporal feature vectors across multiple dimensions. By analyzing the differences between real-time flow fluctuation trajectories and historical spatiotemporal features, we can identify areas of abnormal flow fluctuations.

[0095] The real-time abnormal flow fluctuation interval set is a collection of time periods and spatial regions where abnormal fluctuations in real-time water flow occur. Abnormal fluctuations may manifest as sudden changes in flow velocity, direction, or water level that exceed the normal fluctuation range of historical water flow.

[0096] Statistical analysis and machine learning methods can be used when performing multi-dimensional mapping analysis. For example, the similarity or difference between the real-time water flow fluctuation trajectory dataset and the historical water flow spatiotemporal feature vector set in each dimension can be calculated, and a threshold can be set. When the difference exceeds the threshold, it is considered that abnormal fluctuations exist in that time period or spatial region. Machine learning algorithms (such as clustering algorithms and anomaly detection algorithms) can also be used to model and analyze the real-time water flow fluctuation trajectory dataset and the historical water flow spatiotemporal feature vector set to identify the intervals of abnormal fluctuations. For example, the real-time water flow fluctuation trajectory dataset and the historical water flow spatiotemporal feature vector set can be clustered using the K-means clustering algorithm, and points that are far away from the majority of data points can be identified as abnormal fluctuation points, thereby determining the set of real-time water flow abnormal fluctuation intervals.

[0097] Step S350: Based on the real-time water flow abnormal fluctuation interval set, the real-time flow velocity parameters, real-time flow direction parameters and real-time water level parameters are dynamically enhanced to generate a fused water flow dynamic feature set, and the fused water flow dynamic feature set is used as the dynamic water flow feature set.

[0098] Dynamic feature enhancement involves adjusting and enhancing real-time velocity, direction, and water level parameters based on a set of real-time abnormal flow fluctuation intervals. Within these intervals, the corresponding parameters are analyzed and processed in greater detail to highlight the characteristics of these fluctuations, thereby enhancing the expressiveness of dynamic flow characteristics.

[0099] The fused water flow dynamic feature set is a feature set obtained after dynamic feature enhancement processing. It combines the normal characteristics and abnormal fluctuation characteristics of real-time water flow and can more comprehensively describe the water flow dynamics in the target water area.

[0100] During dynamic feature enhancement, additional feature extraction and analysis can be performed on the real-time flow velocity, flow direction, and water level parameters within the abnormal fluctuation interval. For example, features such as the rate of change of flow velocity, sudden change angle of flow direction, and peak water level within the abnormal fluctuation interval can be calculated and added to the original real-time flow velocity, flow direction, and water level parameters to form a fused set of dynamic flow features.

[0101] Step S400: Based on the spatial distribution topology of the digital twin water flow model, a dynamic water flow feature set is loaded into the digital twin water flow model to generate a set of dynamic change rules of water flow.

[0102] Based on the spatially distributed topology of the digital twin water flow model, the dynamic water flow feature set is loaded into the digital twin water flow model. This involves associating and matching each feature in the dynamic water flow feature set with the corresponding grid node in the spatially distributed topology of the digital twin water flow model, thus transferring the feature information to the model. By loading the dynamic water flow feature set, the digital twin water flow model can simulate and analyze the dynamic changes in water flow in the target water area, thereby generating a set of dynamic flow change patterns.

[0103] As an implementation manner, step S400 may specifically include the following steps S410 to S450: Step S410: determining a set of water flow state constraints for each grid node according to the spatial distribution topology, wherein the set of water flow state constraints includes a flow velocity transmission rule, a water level balance rule, and a flow direction coordination rule between adjacent grid nodes.

[0104] The spatial distribution topology describes the spatial relationships and connections between grid nodes in the digital twin water flow model. This structure determines the set of flow state constraints for each grid node. The flow velocity transfer rule specifies the flow velocity transfer relationship between adjacent grid nodes. The water level balance rule ensures that the water levels between adjacent grid nodes remain balanced. The flow direction coordination rule ensures that the flow direction is consistent throughout the model.

[0105] To determine the set of flow constraints for each grid node, we traverse each grid node in the spatially distributed topology and, based on the connectivity between its adjacent nodes and the dynamic association rules of the flow, determine the flow velocity transfer rules, water level balance rules, and flow direction coordination rules for that node. For example, for a given grid node, constraints such as the flow velocity transfer coefficient, water level difference threshold, and allowable flow direction deviation range are determined based on the attributes of the topological edges connecting it to adjacent nodes. This analysis of all grid nodes generates a set of flow constraints for each grid node.

[0106] Step S420: performing matching and association processing on each dynamic feature in the dynamic water flow feature set and the corresponding grid node to generate a gridded dynamic feature distribution map.

[0107] Each dynamic feature in the dynamic flow feature set is matched and associated with its corresponding grid node. This involves associating the dynamic feature with the corresponding grid node in the spatial distribution topology of the digital twin water flow model based on its spatial location information. The gridded dynamic feature distribution map, consisting of the dynamic features corresponding to each grid node, visually displays the spatial distribution of the dynamic flow features in the target water area.

[0108] During the matching and association process, the grid node to which each dynamic feature belongs can be determined based on the spatial coordinates of the dynamic feature and the spatial position of the grid node. For example, a spatial indexing algorithm (such as a quadtree or KD tree) can be used to quickly find the grid node closest to the dynamic feature's coordinates and associate the dynamic feature with that grid node. By performing matching and association on all dynamic features, a gridded dynamic feature distribution map is generated.

[0109] Step S430: performing dynamic rule verification processing on the gridded dynamic feature distribution map based on the water flow state constraint condition set to determine a set of abnormal feature distribution areas.

[0110] Dynamic rule verification is the process of checking whether the dynamic characteristics of each grid node in the gridded dynamic feature distribution map conform to the rules in the set of flow state constraints. If the dynamic characteristics of a grid node violate the rules, such as the flow velocity exceeding the allowable range of the flow velocity transfer rule, the water level not meeting the water level balance rule, or the flow direction not meeting the flow coordination rule, the grid node is considered to have an abnormal characteristic. The abnormal characteristic distribution area set is the set of areas consisting of all grid nodes with abnormal characteristics. It reflects the areas in the target water area where abnormal water flow may occur.

[0111] During dynamic rule verification, for each grid node in the gridded dynamic feature distribution map, its dynamic features (such as flow velocity, water level, and flow direction) are compared with the set of flow state constraints for that node. For example, if the flow velocity transfer rule stipulates that the flow velocity attenuation coefficient between adjacent grid nodes should be within a set range, and the flow velocity change at a particular grid node does not comply with this rule, then that node is considered to have an abnormal feature. By traversing all grid nodes to identify nodes with abnormal features, the set of abnormal feature distribution areas is determined. A loop statement can be written in a programming language to compare and judge each grid node's features and constraints one by one.

[0112] Step S440: performing feature reconstruction processing on the gridded dynamic feature distribution map according to the abnormal feature distribution area set to generate an optimized dynamic feature distribution map.

[0113] Feature reconstruction is the process of correcting and optimizing the features of areas with abnormal feature distribution within a gridded dynamic feature distribution map. Based on the set of abnormal feature distribution areas, the causes of the abnormal features are analyzed. Combined with the dynamic association rules of water flow and the constraints of water flow states, the abnormal features are adjusted and reconstructed to make the entire map more consistent with the actual water flow dynamics of the target water area. During feature reconstruction, the dynamic features of the grid nodes within the abnormal feature distribution area can be corrected using methods such as interpolation and smoothing. For example, if the flow velocity characteristics within a certain area exhibit abnormal fluctuations, the moving average method can be used to smooth the flow velocity data within that area to eliminate the impact of the abnormal fluctuations. Alternatively, the features of the abnormal area can be estimated and corrected using an interpolation algorithm based on the features of adjacent normal areas. By reconstructing the features of the abnormal feature distribution area, an optimized dynamic feature distribution map is generated that more accurately reflects the water flow dynamics of the target water area.

[0114] Step S450: Dynamic law mining is performed on the optimized dynamic feature distribution map through the water flow dynamic association rules of the digital twin water flow model to generate a set of water flow dynamic change laws.

[0115] Dynamic pattern mining utilizes the dynamic association rules of the digital twin water flow model to conduct in-depth analysis and mining of the optimized dynamic feature distribution map to identify the dynamic patterns of water flow in the target water area. These dynamic association rules include velocity transmission rules, water level balance rules, and flow direction coordination rules. These rules can be used to analyze the changing trends and interrelationships of water flow over time and space.

[0116] When mining dynamic patterns, methods such as time series analysis and machine learning can be employed. For example, time series analysis can be used to analyze the flow velocity, water level, and flow direction data for each grid node in the optimized dynamic feature distribution map to identify periodic patterns and trends. Machine learning algorithms (such as decision trees and neural networks) can also be used to model the entire map and explore the correlations and change patterns between different features. By mining dynamic patterns on the optimized dynamic feature distribution map, a set of dynamic flow change patterns is generated. This set encompasses the temporal and spatial variations of the flow dynamics in the target water area, providing a basis for the subsequent formulation of water flow optimization strategies.

[0117] Step S500: Generate a water flow optimization strategy set according to the water flow dynamic change law set, and feed the water flow optimization strategy set back to the water flow control system to perform dynamic control operations.

[0118] Generating a set of water flow optimization strategies based on a set of dynamic flow patterns involves analyzing the changing trends and patterns of flow velocity, water level, and flow direction within that set of dynamic flow patterns. Combining these with pre-set flow control objectives and constraints, a series of strategies are developed to optimize the flow dynamics in the target water area. The flow optimization strategy set is a collection of these strategies, encompassing control solutions tailored to different situations, such as adjusting flow velocity, balancing water levels, and coordinating flow direction. This set of flow optimization strategies is fed back to the water flow control system, which can dynamically control the flow in the target water area based on these strategies to achieve the goal of optimizing water flow dynamics.

[0119] As an implementation manner, step S500 may specifically include the following steps S510 to S550: Step S510: extracting flow velocity variation pattern parameters, water level evolution trend parameters and flow direction deviation probability parameters from the set of water flow dynamic variation rules.

[0120] The flow velocity variation pattern parameter describes the temporal and spatial variation of the water velocity in the target water area, such as periodic and trend variations in flow velocity. The water level evolution trend parameter reflects the temporal variation of the water level, such as the rise or fall of the water level and the amplitude of water level fluctuation. The flow direction deviation probability parameter indicates the probability of the water flow direction deviating from the normal direction, reflecting the stability of the flow direction.

[0121] When extracting these parameters, data analysis and processing can be performed on the dynamic patterns of water flow. For velocity variation pattern parameters, time series analysis methods (such as Fourier transform and wavelet analysis) can be used to analyze velocity data to identify periodic and trend-based variation characteristics. For water level evolution trend parameters, the water level evolution trend can be determined by fitting a time series curve to the water level data and calculating parameters such as the slope and intercept of the curve. For flow direction deviation probability parameters, the number and proportion of deviations from the normal direction in the flow direction data can be counted to calculate the probability of flow direction deviation.

[0122] Step S520: Based on the preset water flow control constraints, a strategy matching analysis is performed on the flow velocity change pattern parameters, the water level evolution trend parameters and the flow direction deviation probability parameters to generate an initial water flow optimization solution set.

[0123] Preset flow control constraints are a set of restrictions set in advance based on the actual conditions of the target water area and control requirements, such as the permissible range of flow velocity, water level safety thresholds, and permissible deviations in flow direction. Strategy matching analysis compares and analyzes flow velocity variation pattern parameters, water level evolution trend parameters, and flow direction deviation probability parameters with the preset flow control constraints to identify a control strategy that meets these constraints.

[0124] During the strategy matching analysis, a corresponding control strategy is developed for each parameter based on its value and pre-set constraints. For example, if the flow velocity variation pattern parameter indicates that the flow velocity in a certain area exceeds the allowable range, a strategy is developed to reduce or increase the flow velocity in that area. If the water level evolution trend parameter indicates that the water level is about to exceed the safety threshold, a strategy is developed to drain or store water. By performing strategy matching analysis on all parameters, an initial set of water flow optimization solutions is generated, which includes control solutions for different parameters and situations.

[0125] Step S530: Call the dynamic update interface of the digital twin water flow model to perform simulation verification processing on the initial water flow optimization scheme set to determine the simulation control effect parameters corresponding to each initial water flow optimization scheme.

[0126] Calling the dynamic update interface of the digital twin water flow model to simulate and verify the initial set of water flow optimization solutions involves loading the initial water flow optimization solutions into the digital twin water flow model, executing these solutions through model simulation, and observing and analyzing the model's output to evaluate the control effect of each solution. Simulation control effect parameters reflect the control effect of each initial water flow optimization solution after execution in the model, such as the adjustment range of flow velocity, changes in water level, and the degree of improvement in flow direction.

[0127] As an implementation manner, step S530 may specifically include the following steps S531 to S535: Step S531: Load the initial water flow optimization solution set into the grid node of the digital twin water flow model through the dynamic update interface.

[0128] The dynamic update interface, provided by the digital twin water flow model, is used to update model data and rules. Through this interface, the control parameters (such as flow rate adjustment values, water level adjustment values, and flow direction adjustment angles) corresponding to each solution in the initial set of water flow optimization solutions can be loaded into the corresponding grid nodes of the digital twin water flow model. During loading, the control parameters are accurately assigned to the corresponding grid nodes based on the control area and control objectives specified in the solution.

[0129] Step S532: Based on the spatial distribution topology of the digital twin water flow model, simulate and execute the control operations corresponding to each initial water flow optimization scheme, and collect the virtual water flow parameter set after the simulation execution.

[0130] Based on the spatially distributed topology of the digital twin water flow model, the control operations corresponding to each initial water flow optimization solution are simulated. This involves simulating the changes in water flow after implementing these control operations within the digital twin water flow model, based on each solution's control parameters and the dynamic association rules for water flow. This simulation allows prediction of the effectiveness of each solution in actual application. The virtual water flow parameter set is the set of water flow parameters for each grid node in the model, including flow velocity, water level, flow direction, and other parameters, obtained after the simulation.

[0131] During simulation execution, the digital twin water flow model updates the flow state at each grid node based on the spatial distribution topology and dynamic flow association rules. For example, if a solution adjusts the flow velocity at a grid node, the model calculates the impact of that node on adjacent nodes based on the flow velocity transfer rules and updates the flow velocities of these adjacent nodes. During the simulation, flow parameters for each grid node are collected in real time to form a virtual flow parameter set. Numerical simulation algorithms (such as the finite difference method and the finite element method) can be used to implement model simulation execution and parameter collection.

[0132] Step S533: performing comparative analysis on the virtual water flow parameter set and the preset expected control target parameter set to determine the target deviation parameter of each initial water flow optimization scheme.

[0133] The preset expected control target parameter set is the control target that each initial water flow optimization scheme is expected to achieve, including target values ​​for flow rate, water level, and flow direction. Comparing and analyzing the virtual water flow parameter set with the expected control target parameter set involves comparing the differences between the virtual water flow parameters obtained after simulation execution and the expected control target parameters to evaluate the performance of each scheme. The target deviation parameter reflects the degree of deviation between each initial water flow optimization scheme and the expected control target. A greater deviation indicates a less than ideal scheme performance.

[0134] During the comparative analysis, for each initial flow optimization scenario, the difference between each parameter in the virtual flow parameter set and the corresponding parameter in the expected control target parameter set is calculated. For example, the difference between the flow velocity after a simulation of a scenario and the expected flow velocity, or the difference between the water level and the expected water level, is calculated. Based on these differences, a target deviation parameter is then calculated. Methods such as Euclidean distance and mean square error can be used to calculate the deviation.

[0135] Step S534: Adaptively correct the initial water flow optimization scheme according to the target deviation parameter to generate a set of corrected water flow optimization schemes.

[0136] Adaptively modifying the initial flow optimization schemes based on the target deviation parameter analyzes the existing problems in each scheme based on its target deviation parameter and adjusts and modifies the scheme's control parameters to improve its control effectiveness. The modified flow optimization scheme set is the resulting set of schemes, containing optimization schemes that better meet the desired control objectives.

[0137] When performing adaptive corrections, analyze the causes of deviations for solutions with large target deviations. For example, if a solution's flow rate adjustment is unsatisfactory, it may be due to an improperly set flow rate adjustment value. Adjust the flow rate adjustment value based on the deviation. An iterative optimization approach can be used to gradually adjust the solution's control parameters until the target deviation parameters meet the set requirements. By performing adaptive corrections on all initial flow optimization solutions, a set of corrected flow optimization solutions is generated.

[0138] Step S535: reload the modified water flow optimization scheme set into the digital twin water flow model for secondary simulation verification until the target deviation parameter is less than the preset deviation threshold, and use the verified modified water flow optimization scheme as the initial water flow optimization scheme set.

[0139] Reloading the revised flow optimization solution set into the digital twin water flow model for secondary simulation verification involves reloading the revised solutions into the digital twin water flow model, simulating their execution, and evaluating their control effects. The preset deviation threshold is the maximum allowable value of the pre-set target deviation parameter. When the target deviation parameter is less than this threshold, the solution's control effect is considered to meet the requirements.

[0140] During the secondary simulation verification process, steps S531-S533 are repeated, loading the revised flow optimization solution set into the model, simulating the solution execution, collecting the virtual flow parameter set, and calculating the target deviation parameter. If the target deviation parameter is greater than the preset deviation threshold, the solution is revised and the simulation verification process is repeated again until the target deviation parameter is less than the deviation threshold. The verified revised flow optimization solution is used as the initial flow optimization solution set for subsequent priority sorting.

[0141] Step S540: Prioritize the initial water flow optimization solution set according to the simulation control effect parameters to generate an optimization solution priority queue.

[0142] Prioritize the initial set of water flow optimization solutions based on simulated control effect parameters. This involves ranking the solutions in the initial set based on each solution's simulated control effect parameters (such as flow rate adjustment, water level improvement, and flow stability improvement), and determining the priority of each solution. The optimization solution priority queue ranks solutions from high to low priority, with higher-priority solutions being executed first.

[0143] When performing the priority sorting process, a weight coefficient can be assigned to each simulation control effect parameter, and then the comprehensive score of each scheme can be calculated based on these weight coefficients.

[0144] Step S550: Determine the top N optimization schemes with the highest priority in the optimization scheme priority queue as a water flow optimization strategy set, where N is a positive integer dynamically adjusted based on the execution capability parameter of the water flow control system.

[0145] The top N optimization solutions in the optimization solution priority queue are identified as the water flow optimization strategy set. This selects the top N solutions from the optimization solution priority queue as the final water flow optimization strategy. N is a positive integer that is dynamically adjusted based on the execution capability parameters of the water flow control system. The execution capability parameters of the water flow control system include the system's processing power, the number of devices, and their performance. Based on the execution capability of the water flow control system, the number of solutions N that the system can execute simultaneously is determined.

[0146] When determining the value of N, an execution capability model can be established based on the water flow control system's historical execution data and performance evaluation results. This model can dynamically calculate the value of N based on factors such as the system's processing capacity and device load. For example, if the water flow control system has strong processing capacity and a light device load, the value of N can be appropriately increased; otherwise, it can be decreased. By dynamically adjusting the value of N, the water flow optimization strategy set can be effectively executed by the water flow control system.

[0147] As an implementation manner, the method provided in the embodiment of the present invention may further include a process of dynamically updating the digital twin water flow model, which may specifically include the following steps S600 to S1000: Step S600: Periodically collect updated hydrological monitoring data sets of the target water area, where the updated hydrological monitoring data sets include newly added geological structure parameters, surface cover change parameters, and water flow environment interference parameters.

[0148] The periodic collection of updated hydrological monitoring data sets for target waters involves monitoring and collecting data on the hydrological conditions of the target waters at time intervals (e.g., daily, weekly, etc.). The updated hydrological monitoring data sets are data sets collected to reflect the latest hydrological conditions of the target waters. These include newly added geological structure parameters that describe the latest changes in the geological structure of the target waters, such as stratum displacement and the emergence of new faults. Land cover change parameters reflect changes in the land cover of the target waters, such as vegetation growth or destruction and changes in land use. Water flow environmental interference parameters represent external environmental factors that may affect water flow, such as changes in meteorological conditions and interference from human activities.

[0149] Step S700: Load the updated hydrological monitoring data set to the grid node corresponding to the spatial distribution topology structure through the dynamic update interface of the digital twin water flow model.

[0150] The updated hydrological monitoring data set is loaded into the corresponding grid nodes of the spatially distributed topology through the dynamic update interface of the digital twin water flow model. This interface is used to associate and match each parameter in the updated hydrological monitoring data set with the corresponding grid node in the spatially distributed topology, and the parameter information is written into the attributes of the grid node. In this way, the digital twin water flow model can update its status based on the latest hydrological monitoring data, more accurately reflecting the actual conditions of the target water area.

[0151] When loading data, the corresponding grid node is first determined based on the spatial location information of each parameter in the updated hydrological monitoring data set. For example, if a newly added geological structure parameter describes the stratum displacement of a certain area in the target water area, the corresponding grid node is found. Then, the value of this parameter is written to the geological structure attribute of the corresponding grid node through the write function of the dynamic update interface. For surface cover change parameters and water flow environment interference parameters, they are similarly loaded into the corresponding attributes of the corresponding grid node based on their spatial location and parameter type.

[0152] Step S800: performing incremental learning processing on the water flow dynamic association rules based on the newly added geological structure parameters and surface cover change parameters to generate an updated water flow dynamic association rule set.

[0153] Incremental learning of the flow dynamics association rules based on newly added geological structure parameters and land cover change parameters adjusts and optimizes the existing flow dynamics association rules based on the newly added geological structure and land cover change information, making them more adaptable to the new conditions in the target water area. Incremental learning fine-tunes the model by continuously introducing new data and information without requiring full model retraining. The updated flow dynamics association rule set, derived from incremental learning, includes flow velocity transfer rules, water level balance rules, and flow direction coordination rules that better reflect the current conditions in the target water area.

[0154] As an implementation manner, step S800 may specifically include the following steps S810 to S850: Step S810: extracting the stratum permeability change value in the newly added geological structure parameter and the vegetation coverage change value in the surface cover change parameter.

[0155] The newly added geological structure parameters include changes in formation permeability. Extracting the formation permeability change value extracts the specific value reflecting the formation permeability change from the newly added geological structure parameters. The formation permeability change value describes the degree of change in the formation's ability to penetrate water flow, which affects the flow rate and path of water within the formation. The surface cover change parameter includes information on changes in vegetation cover. Extracting the vegetation cover change value extracts the vegetation cover change value from the surface cover change parameter. Changes in vegetation cover affect the surface's ability to intercept and infiltrate water flow, thereby affecting water flow dynamics.

[0156] When extracting changes in formation permeability and vegetation cover, process the data according to its storage format and structure. If the newly added geological structure parameters and land cover change parameters are stored in a database, you can use SQL queries to extract the required values. For example, search the database for the field recording changes in formation permeability and extract its value as the formation permeability change value. For the land cover change parameter, search the database for the field recording changes in vegetation cover and extract the vegetation cover change value. If the data is stored in a file format, you can use data processing tools (such as the Python pandas library) to read the file and extract the required values.

[0157] Step S820: performing difference calculation processing on the formation permeability change value and the initial permeability parameter of the corresponding grid node in the digital twin water flow model to generate a formation permeability difference distribution map.

[0158] The permeability change is calculated by subtracting the initial permeability parameters of the corresponding grid nodes in the digital twin water flow model. For each grid node, the difference between the permeability change and the initial permeability parameters is calculated. The permeability difference distribution map, composed of the permeability differences for each grid node, visually illustrates the permeability changes at different locations within the target water area.

[0159] When performing the difference calculation process, each grid node in the spatial distribution topology structure of the digital twin water flow model is traversed, and the formation permeability change value of the node is subtracted from the initial permeability parameter to obtain the formation permeability difference of the node.

[0160] Step S830: Perform similarity matching processing on the vegetation coverage change value and the initial coverage parameter of the corresponding grid node in the digital twin water flow model to generate a vegetation coverage difference distribution map.

[0161] The vegetation coverage change values ​​are matched against the initial coverage parameters of the corresponding grid nodes in the digital twin water flow model. This process compares the difference between the vegetation coverage change value and the initial coverage parameters at each grid node. The vegetation coverage difference distribution map, composed of the vegetation coverage difference at each grid node, shows the changes in vegetation coverage at different locations in the target watershed.

[0162] When performing similarity matching, you can use difference or ratio calculation methods. For example, you can calculate the difference between the vegetation coverage change value and the initial coverage parameter for each grid node to obtain the vegetation coverage difference value. You can also calculate the ratio of the vegetation coverage change value to the initial coverage parameter to obtain the vegetation coverage change ratio. These difference values ​​are organized and stored, and a vegetation coverage difference distribution map can be drawn using a drawing tool.

[0163] Step S840: Dynamically adjust the velocity transfer coefficient, water level balance threshold, and flow direction coordination factor in the water flow dynamic association rule according to the stratum permeability difference distribution map and the vegetation cover difference distribution map.

[0164] Based on the distribution maps of stratum permeability differences and vegetation cover differences, the velocity transfer coefficient, water level balance threshold, and flow direction coordination factor in the water flow dynamic association rules are dynamically adjusted. This is done based on changes in stratum permeability and vegetation cover, adjusting key parameters in the water flow dynamic association rules to adapt to the new conditions in the target water area. The velocity transfer coefficient describes the flow velocity relationship between adjacent grid nodes, the water level balance threshold specifies the water level balance condition between adjacent grid nodes, and the flow direction coordination factor ensures that the flow direction is coordinated and consistent throughout the model.

[0165] When performing dynamic adjustment processing, for each grid node, the velocity transfer coefficient, water level balance threshold and flow direction coordination factor of the node are adjusted according to the difference in its stratum permeability and vegetation coverage. For example, if the stratum permeability of a grid node increases, it means that water flows more easily through the area, and the velocity transfer coefficient of the node can be appropriately increased; if the vegetation coverage decreases, the surface's ability to intercept water flow is weakened, and the water level balance threshold and flow direction coordination factor may need to be adjusted. A parameter adjustment model can be established to calculate the adjustment value of each parameter based on the size of the stratum permeability difference and vegetation coverage difference. For example, let the stratum permeability difference be ΔP diff , the vegetation cover difference is ΔC diff , the adjustment value of the flow velocity transfer coefficient is Δk, the adjustment value of the water level balance threshold is Δh, and the adjustment value of the flow direction coordination factor is Δf. The adjustment value can be calculated by the following formula: ; ; ; in The default weight coefficient can be adjusted according to the actual situation. By adjusting the parameters of all grid nodes, the dynamic association rules of water flow can be dynamically adjusted.

[0166] Step S850: The dynamically adjusted flow velocity transfer coefficient, water level balance threshold, and flow direction coordination factor are used as an updated water flow dynamic association rule set.

[0167] The dynamically adjusted velocity transfer coefficient, water level balance threshold, and flow direction coordination factor are used as the updated flow dynamics association rule set. This is achieved by organizing and storing the velocity transfer coefficient, water level balance threshold, and flow direction coordination factor obtained after dynamic adjustment to form a new rule set. The updated flow dynamics association rule set includes rules that better reflect the current reality of the target water area and can more accurately describe the water flow dynamics in the target water area.

[0168] When forming the updated water flow dynamic association rule set, the dynamically adjusted flow velocity transfer coefficient, water level balance threshold and flow direction coordination factor of each grid node are associated with the node's identifier or index and stored in a data structure (such as a dictionary, list, etc.).

[0169] Step S900: Dynamically adjust the weights of the grid nodes of the spatial distribution topology structure according to the water flow environment interference parameters to generate an optimized spatial distribution topology structure.

[0170] Dynamic weight adjustment of the grid nodes in the spatial distribution topology structure based on water flow environmental interference parameters involves adjusting the weight of each grid node in the spatial distribution topology based on water flow environmental interference parameters (such as changes in meteorological conditions and interference from human activities). The weight of a grid node reflects its importance and influence on water flow dynamics. By dynamically adjusting the weights, we can more accurately simulate water flow under different environmental conditions. The optimized spatial distribution topology, resulting from this dynamic weight adjustment, better reflects the dynamic characteristics of water flow in the target water area under the current water flow environmental interference.

[0171] When performing dynamic weight adjustment, analyze the impact of flow environment interference parameters on each grid node. For example, if a region is affected by heavy rainfall, the flow rate in that region will increase, and the weight of the grid nodes in that region can be appropriately increased. If human activities (such as the construction of a water conservancy project) interfere with a region, the path and speed of the water flow may be altered, and the weight of the grid nodes in that region may need to be adjusted accordingly. A weight adjustment model can be established to calculate the weight adjustment value for each grid node based on the type and intensity of the flow environment interference parameters.

[0172] Step S1000: Integrate the updated water flow dynamic association rule set with the optimized spatial distribution topology structure to generate an updated digital twin water flow model.

[0173] Integrating the updated set of dynamic water flow association rules with the optimized spatial distribution topology involves applying the rules from the updated set to the optimized spatial distribution topology, enabling the two to work together to form a complete, updated digital twin water flow model. This integration involves associating the parameters in the rules with the grid nodes in the topology to ensure that the rules are correctly executed within the topology.

[0174] During the integration process, for each grid node in the optimized spatial distribution topology, the corresponding velocity transfer coefficient, water level balance threshold, and flow direction coordination factor from the updated set of water flow dynamic association rules are written into the node's attributes. Simultaneously, the connection relationships and constraints between adjacent grid nodes in the topology are adjusted according to the requirements of the rules.

[0175] As an implementation manner, the method provided in the embodiment of the present invention may further include a process of adaptively controlling abnormal water flow, which may specifically include the following steps S1100 to S1900: Step S1100: Acquire in real time the virtual water flow parameter set of each grid node in the digital twin water flow model and the corresponding monitoring parameter set in the real-time water flow monitoring data set.

[0176] Real-time acquisition of the virtual flow parameter set for each grid node in the digital twin water flow model and the corresponding monitoring parameter set from the real-time water flow monitoring data set involves extracting the virtual flow parameters (such as virtual flow velocity, virtual water level, and virtual flow direction) for each grid node from the digital twin water flow model at each time point, while simultaneously obtaining the actual monitoring parameters (such as real-time flow velocity, real-time water level, and real-time flow direction) for the corresponding grid node from the real-time water flow monitoring equipment. The virtual flow parameter set is the set of water flow parameters for each grid node simulated by the digital twin water flow model, while the corresponding monitoring parameter set is the set of water flow parameters for each grid node collected by the actual monitoring equipment.

[0177] When acquiring a set of virtual flow parameters, the digital twin water flow model's interface functions can be used to query the virtual flow parameters for each grid node. For example, a programming language can be used to call the model's query function, passing in the grid node's identifier or index, to retrieve parameters such as the node's virtual flow velocity, virtual water level, and virtual flow direction. When acquiring the corresponding monitoring parameter set, real-time flow monitoring equipment (such as flow meters, water level gauges, and flow direction meters) transmits the collected data to the data processing center, which then uses the data processing center's interface functions to retrieve the monitoring parameters for the corresponding grid node.

[0178] Step S1200: Perform dynamic deviation verification processing on the virtual water flow parameter set and the monitoring parameter set to generate an initial parameter deviation set.

[0179] Dynamic deviation verification of the virtual flow parameter set and the monitoring parameter set involves comparing the virtual parameters of each grid node in the virtual flow parameter set with the actual monitoring parameters of that node in the corresponding monitoring parameter set to detect any parameter deviations. Dynamic deviation verification considers parameter variations in time and space, analyzing parameters over continuous time periods to identify abnormal deviations. The initial parameter deviation set consists of all deviated parameters, their deviation magnitudes, and timestamps. It reflects the discrepancies between the simulation results of the digital twin flow model and the actual monitoring data.

[0180] As an implementation manner, step S1200 may specifically include the following steps S1210 to S1280: Step S1210: configuring an independent verification window for the virtual flow velocity, virtual water level and virtual flow direction of each grid node, respectively. The independent verification window includes a dynamically adjusted time window length and parameter fluctuation tolerance.

[0181] Independent calibration windows are configured for each grid node's virtual velocity, virtual water level, and virtual flow direction. This allows each parameter (virtual velocity, virtual water level, and virtual flow direction) to have an independent calibration range and time range. These independent calibration windows include a dynamically adjusted time window length and parameter fluctuation tolerance. The time window length represents the time period used for calibration, while the parameter fluctuation tolerance specifies the maximum allowable parameter fluctuation within that time period. These dynamically adjusted time window lengths and parameter fluctuation tolerances can be adjusted based on actual conditions and historical data to accommodate varying water flow conditions and monitoring accuracy requirements.

[0182] When configuring an independent calibration window, you can determine the initial time window length and parameter fluctuation tolerance based on historical data statistics of each parameter's fluctuation range and frequency. For example, if historical data indicates that the flow rate parameter fluctuates frequently but with small amplitudes, you can set a shorter time window length and a smaller parameter fluctuation tolerance. During actual operation, the time window length and parameter fluctuation tolerance can be dynamically adjusted based on real-time monitoring data and model simulation results.

[0183] Step S1220: Collect the continuous change sequence of the virtual water flow parameters and the continuous change sequence of the monitoring parameters in the independent verification window.

[0184] Collecting continuous change sequences of virtual flow parameters and monitoring parameters within independent verification windows involves collecting continuous values ​​of virtual flow parameters (virtual flow velocity, virtual water level, and virtual flow direction) and corresponding monitoring parameters (real-time flow velocity, real-time water level, and real-time flow direction) at intervals within the time range specified by the independent verification window at each grid node, forming continuous change sequences. These continuous change sequences record the temporal changes of the parameters and are used for subsequent deviation verification analysis.

[0185] Step S1230: performing multi-stage alignment processing on the continuous change sequence of the virtual water flow parameters and the continuous change sequence of the monitoring parameters to generate a preliminary verification trajectory set.

[0186] Multi-stage alignment processing is performed on the continuous change sequence of virtual water flow parameters and the continuous change sequence of monitoring parameters. The two continuous change sequences are matched and aligned in time and value to ensure that they are comparable at the same time point. Multi-stage alignment processing can adopt time alignment and value alignment methods. Time alignment is to match the timestamps of the two sequences so that they have corresponding values ​​at the same time point; value alignment is to normalize or standardize the values ​​in the sequences so that they have the same scale and range. The preliminary verification trajectory set is the trajectory set obtained after the multi-stage alignment processing. It contains the comparative information of the virtual water flow parameters and the monitoring parameters at the same time point.

[0187] During multi-stage alignment, temporal alignment is performed first. Interpolation algorithms can be used to interpolate the temporal values ​​of two continuously changing sequences so that they have corresponding values ​​at the same time point. Next, numerical alignment is performed to normalize the values ​​in the sequences and map their range to the [0, 1] interval. This multi-stage alignment generates a preliminary set of verified trajectories, which can be represented as a graph or data structure for subsequent anomaly trajectory detection.

[0188] Step S1240: If there is an abnormal trajectory segment in the preliminary verification trajectory set whose trajectory offset exceeds the preset offset tolerance, a secondary verification mechanism is initiated; wherein the secondary verification mechanism includes: dynamically expanding the time window length of the independent verification window based on the time range of the abnormal trajectory segment, and reducing the parameter fluctuation tolerance.

[0189] If the initial verification trajectory set contains an abnormal trajectory segment whose trajectory offset exceeds the preset offset tolerance, it indicates that there may be a large deviation between the virtual water flow parameters and the monitored parameters, and the secondary verification mechanism needs to be activated for further verification. The preset offset tolerance is the maximum allowable value of trajectory offset. When the trajectory offset exceeds this value, it is considered an abnormal trajectory segment.

[0190] The secondary verification mechanism dynamically expands the independent verification window length based on the time range of the abnormal trajectory segment and reduces the tolerance for parameter fluctuations. Dynamically expanding the time window length allows for more data to be collected for a more comprehensive analysis of abnormal situations; reducing the tolerance for parameter fluctuations improves verification accuracy and more accurately identifies the presence of abnormal deviations.

[0191] When starting the secondary verification mechanism, first determine the time range of the abnormal trajectory segment Then the time window length of the independent verification window is extended to a larger range that includes the abnormal trajectory segment, for example, the time window length is extended to ,in and is the length of time for the extension. At the same time, reduce the parameter fluctuation tolerance, for example, change the original parameter fluctuation tolerance from Reduce to In this way, a secondary verification mechanism is activated to conduct a more in-depth analysis of abnormal situations.

[0192] Step S1250: recollecting the extended change sequence of the virtual water flow parameters and the extended change sequence of the monitoring parameters within the extended independent verification window.

[0193] Recollecting the extended change sequence of virtual flow parameters and monitoring parameters within the expanded independent verification window involves recollecting continuous values ​​of the virtual flow parameters and monitoring parameters at intervals within the time range specified by the dynamically expanded independent verification window to form an extended change sequence. The extended change sequence contains more data information for secondary verification analysis.

[0194] When re-collecting the extended change sequence, set the data collection time point according to the time range and time interval of the extended independent verification window. For example, if the time range of the extended independent verification window is , the time interval is , then at time point Collect the values ​​of virtual flow parameters and monitoring parameters. A data acquisition system and timer can be used to implement timed collection. The collected values ​​of virtual flow parameters and monitoring parameters are stored in two lists or arrays, forming an extended change sequence.

[0195] Step S1260: performing spatiotemporal segment alignment processing on the extended change sequence to generate a set of refined verification trajectories.

[0196] Performing spatiotemporal segmented alignment on the expanded change sequence allows for more precise temporal and spatial alignment of the expanded virtual flow parameter change sequence and the monitored parameter change sequence. This segmented alignment considers parameter variations over time and spatial locations, segmenting the sequence and performing separate alignment and analysis on each segment. The refined verification trajectory set, resulting from this segmented spatiotemporal alignment, contains more detailed comparative information on the virtual flow parameters and monitored parameters.

[0197] During the spatiotemporal segment alignment process, the extended variation sequence is first segmented based on the parameter variation characteristics and time range. For example, the sequence can be divided into segments based on the parameter's mutation points or periodic changes. Each segment is then subjected to temporal and numerical alignment, similar to the multi-stage alignment process in step S1230. This spatiotemporal segment alignment process generates a refined set of verification trajectories, which can be represented as a graph or data structure for subsequent abnormal deviation determination.

[0198] Step S1270: Calculate the trajectory similarity parameter of each trajectory segment in the refined verification trajectory set. If the trajectory similarity parameter is lower than a preset similarity tolerance threshold, it is determined that there is an abnormal deviation between the virtual water flow parameter and the monitoring parameter.

[0199] Calculating the trajectory similarity parameter for each segment in the refined verification trajectory set involves calculating the similarity between the virtual flow parameter trajectory and the monitored parameter trajectory. The trajectory similarity parameter reflects the degree of similarity between the two trajectories. A higher similarity indicates greater consistency between the virtual flow parameters and the monitored parameters; a lower similarity indicates a possible deviation between the two.

[0200] When calculating the trajectory similarity parameters, various methods can be used, such as Euclidean distance, cosine similarity, etc. The preset similarity tolerance threshold is the minimum allowable value of the trajectory similarity parameter. When the trajectory similarity parameter is lower than the threshold, it is determined that there is an abnormal deviation between the virtual water flow parameters and the monitoring parameters.

[0201] Step S1280: Integrate the parameter type, deviation amplitude, and timestamp corresponding to the abnormal deviation into an initial parameter deviation set.

[0202] The parameter type, deviation magnitude, and timestamp corresponding to the abnormal deviation are integrated into the initial parameter deviation set. This is to organize and store the relevant information of the parameters determined to have abnormal deviations into a set. Parameter types include flow rate, water level, flow direction, etc. The deviation magnitude is the difference between the virtual water flow parameter and the monitored parameter. The timestamp records the time when the abnormal deviation occurred.

[0203] When integrating the initial parameter deviation set, for each parameter that is determined to have abnormal deviation, its parameter type, deviation magnitude and timestamp are stored in the set as a tuple.

[0204] Step S1300: If there is an abnormal deviation parameter exceeding a preset reference range in the initial parameter deviation set, the corresponding abnormal grid node identifier and abnormal time period parameter set are extracted.

[0205] If the initial parameter deviation set contains abnormal deviation parameters that exceed the preset baseline range, it indicates that these abnormal deviations may have a significant impact on the water flow dynamics in the target water area and require further analysis and treatment. The preset baseline range is the pre-set allowable range for abnormal deviation parameters. When an abnormal deviation parameter exceeds this range, the parameter is considered severely abnormal.

[0206] Extracting the corresponding abnormal grid node identifiers and abnormal time period parameter sets involves identifying abnormal deviation parameters that exceed the preset baseline range from the initial parameter deviation set and recording the corresponding grid node identifiers and the time period parameters where the abnormality occurred. The abnormal grid node identifiers are used to locate the specific location of the abnormality, and the abnormal time period parameter set records the time range of the abnormality, providing a basis for subsequent dynamic backtracking analysis.

[0207] To extract the anomalous grid node identifiers and anomalous time period parameter sets, the initial parameter deviation set is traversed. For each anomalous deviation parameter, its deviation magnitude is checked to see if it exceeds a preset baseline range. If so, the corresponding grid node identifier and the timestamp of the anomalous occurrence are recorded. All eligible grid node identifiers and timestamps are collated to form a set of anomalous grid node identifiers and anomalous time period parameter sets.

[0208] Step S1400: Based on the abnormal period parameter set, a dynamic backtracking analysis is performed on the spatial distribution topology structure of the digital twin water flow model to restore the dynamic change trajectory of the water flow corresponding to the abnormal deviation parameters.

[0209] Dynamic backtracking analysis of the spatially distributed topology of the digital twin water flow model, based on the parameter set for the abnormal period, leverages the characteristics of the digital twin water flow model and the time range recorded in the parameter set for the abnormal period to reversely infer and analyze the water flow dynamics within that time period. The digital twin water flow model records the flow state of each grid node at different times and the association rules between them. Through backtracking analysis, the flow changes can be tracked backward along the time axis, thereby reconstructing the dynamic changes in the water flow corresponding to the abnormal deviation parameters.

[0210] When performing dynamic backtracking analysis and processing, the starting and ending points of the backtracking time are first determined based on the parameter set of the abnormal period. Then, starting from the moment the abnormality occurs, based on the dynamic association rules of water flow, the water flow status of each grid node at each previous moment is reversely deduced. For example, according to the flow velocity transmission rule, if the flow velocity of a grid node at the current moment is known, the influence of the adjacent grid nodes on its flow velocity at the previous moment can be reversely calculated, and then the flow velocity of the adjacent grid nodes can be deduced. For water level and flow direction, reverse deduction is also performed according to the corresponding rules. In the derivation process, it is necessary to take into account the influence of factors such as geological structure and surface cover on water flow. These factors are reflected in the dynamic association rules of water flow in the digital twin water flow model. Through continuous reverse deduction, the dynamic change trajectory of water flow corresponding to the abnormal deviation parameters is gradually restored. This trajectory records the changes in water flow during the abnormal period in detail with time and space as the dimensions.

[0211] Step S1500: performing rule matching analysis on the water flow dynamic change trajectory and the flow velocity transfer rule in the water flow dynamic association rule to determine the rule deviation parameter.

[0212] The matching degree analysis between the dynamic flow trajectory and the velocity transfer rules in the dynamic flow association rules is performed. This involves comparing and analyzing the velocity changes in the restored dynamic flow trajectory with the velocity transfer rules. The velocity transfer rules specify the velocity transfer relationship between adjacent grid nodes, such as velocity attenuation patterns and transfer coefficients. By comparing the actual velocity changes in the dynamic flow trajectory with the velocity changes specified by the rules, the degree of match between the two is evaluated.

[0213] When performing rule matching analysis, for each time point and grid node in the dynamic change trajectory of water flow, the actual flow velocity change of the node is compared with the flow velocity change expected by the flow velocity transmission rule. The difference between the two can be measured by calculating differences, ratios, etc. For example, the difference between the actual flow velocity change and the flow velocity change expected by the rule is calculated, and then the differences for all time points and grid nodes are statistically analyzed to obtain a comprehensive difference index. The ratio of the actual flow velocity change to the flow velocity change expected by the rule can also be calculated. The closer the ratio is to 1, the higher the degree of matching. The rule deviation parameter is a quantitative indicator that measures the degree of matching between the dynamic change trajectory of water flow and the flow velocity transmission rule. The larger the deviation parameter, the greater the difference between the actual water flow dynamic change and the rule provisions. The rule deviation parameter can be determined based on the difference index obtained by statistical analysis, for example, the average value or standard deviation of all differences is used as the rule deviation parameter.

[0214] Step S1600: If the rule deviation parameter exceeds a preset rule deviation threshold, the optimization strategy set associated with the abnormal grid node identifier in the water flow optimization strategy set is called.

[0215] If the rule deviation parameter exceeds the preset rule deviation threshold, it indicates that there is a significant deviation between the actual water flow dynamics and the flow velocity transmission rules. This deviation may affect the dynamic balance of the water flow in the target water area, and appropriate measures need to be taken to adjust it. The preset rule deviation threshold is the maximum allowable value of the rule deviation parameter. When the rule deviation parameter exceeds this threshold, it is considered that a serious abnormality has occurred in the water flow dynamics.

[0216] Calling the optimization strategy set associated with the abnormal grid node identifier in the water flow optimization strategy set is to find the optimization strategy corresponding to the abnormal grid node identifier from the previously generated water flow optimization strategy set. The water flow optimization strategy set is a series of optimization strategies formulated for different situations in the target water area, and each strategy is associated with a set grid node or area. Through the abnormal grid node identifier, the optimization strategy set related to it can be quickly located. For example, in the water flow optimization strategy set, the applicable grid node identifier is recorded for each optimization strategy. When an abnormality occurs, the corresponding optimization strategy is filtered out according to the abnormal grid node identifier. These optimization strategies may include measures such as adjusting flow rate, balancing water level, coordinating flow direction, etc., aimed at correcting the abnormality of water flow dynamics and restoring it to normal.

[0217] Step S1700: Generate a real-time control instruction set based on the optimization strategy set, and send the real-time control instruction set to the water flow control system.

[0218] Generating a set of real-time control instructions based on an optimization strategy set involves refining and transforming each strategy within the set, converting the control objectives and requirements within the strategy into specific, executable real-time control instructions. While an optimization strategy set typically describes the required control measures in a relatively abstract manner, such as "increase the flow velocity in a certain area" or "lower the water level in a certain area," a set of real-time control instructions includes specific control parameters and steps, such as "increase the flow velocity at a certain grid node to a value of v" or "open a certain drainage valve to lower the water level."

[0219] When generating a set of real-time control instructions, it's important to consider the actual conditions and capabilities of the water flow control system. For example, considering the control methods and parameter ranges of the devices in the water flow control system, the control objectives in the optimization strategy should be converted into instructions that can be executed by the devices. For a strategy to increase flow rate, it's necessary to determine the device (such as a water pump) and its operating parameters (such as power and speed) to achieve the increase. After converting each optimization strategy into corresponding real-time control instructions, these instructions are organized and summarized to form a set of real-time control instructions.

[0220] Sending a set of real-time control instructions to the water flow control system involves passing them through a data transmission interface. The water flow control system is responsible for receiving and executing these instructions, controlling the flow in the target water area in real time. This can be accomplished using wired or wireless communication, such as Ethernet, Wi-Fi, or GPRS, to the water flow control system's control center.

[0221] Step S1800: The water flow control system performs multi-level response control operations on the physical control equipment corresponding to the abnormal grid node according to the real-time control instruction set, and at the same time transmits the control feedback parameters back to the dynamic update interface of the digital twin water flow model.

[0222] The water flow control system performs multi-level response control operations on the physical control devices corresponding to the abnormal grid nodes based on the real-time control instruction set. After receiving the real-time control instruction set, the water flow control system operates the physical control devices corresponding to the abnormal grid nodes according to the instructions. Physical control devices include pumps, valves, sluice gates, etc., which are used to regulate the flow rate, water level, and flow direction of the water. Multi-level response control operations execute control instructions in stages and layers based on the complexity and urgency of the control. For example, in more urgent situations, rapid response measures may be taken first, such as immediately opening a drain valve. In situations requiring fine-tuning, the pump power or valve opening may be gradually adjusted.

[0223] During control operations, the water flow control system monitors the operating status of the physical control equipment and actual changes in water flow in real time, collecting control feedback parameters. These include equipment operating parameters (such as pump power and valve opening) and actual water flow parameters (such as flow rate, water level, and flow direction). These parameters reflect the actual effectiveness of the control operations.

[0224] The dynamic update interface, which transmits control feedback parameters back to the digital twin water flow model, transmits these control feedback parameters to the digital twin water flow model via a data transmission channel. This dynamic update interface, provided by the digital twin water flow model for updating model data, loads control feedback parameters into the model, enabling the digital twin water flow model to reflect the actual dynamic changes in the target water area in real time.

[0225] Step S1900: Load the control feedback parameters to the corresponding grid nodes of the spatially distributed topological structure through the dynamic update interface, trigger the real-time topological weight correction processing of the digital twin water flow model, and generate an updated virtual water flow parameter set to verify the effectiveness of the control.

[0226] Loading the control feedback parameters to the corresponding grid nodes in the spatially distributed topology structure through a dynamic update interface accurately writes the control feedback parameters transmitted back by the water flow control system into the properties of the corresponding grid nodes in the spatially distributed topology of the digital twin water flow model. The spatially distributed topology describes the spatial relationships and connections between the grid nodes in the digital twin water flow model. Each grid node has corresponding water flow properties (such as flow velocity, water level, and flow direction). Through the dynamic update interface, the actual flow velocity, water level, and flow direction parameters in the control feedback parameters are updated to the corresponding grid node properties, allowing the model to accurately reflect the latest water flow dynamics in the target water area.

[0227] To trigger the real-time topology weight correction process in the digital twin water flow model, after loading the control feedback parameters into the model, the topological weights of each grid node in the spatially distributed topological structure are adjusted based on the new flow state. Topological weights reflect the importance and influence of grid nodes in water flow dynamics, and as water flow states change, topological weights need to be adjusted accordingly. To verify the effectiveness of the control, an updated set of virtual water flow parameters is generated. After the real-time topology weight correction process is completed, the virtual water flow parameters of each grid node in the digital twin water flow model are recalculated based on the updated topological structure and water flow dynamic association rules to form an updated set of virtual water flow parameters. The updated set of virtual water flow parameters is compared and analyzed with the real-time water flow monitoring data. If the deviation between the two is reduced, the control operation is effective; if the deviation remains large, further adjustment of the control strategy may be necessary. In this way, the effectiveness of the water flow control operation is verified and the water flow dynamics in the target water area are restored to normal.

[0228] It is understandable that the various algorithms involved in the above-mentioned introductions of the embodiments of the present invention, such as spatial index algorithms, difference algorithms, etc., can be learned from the relevant content in the prior art. In order to save space, they will not be expanded too much in the embodiments of this application. In addition, when implementing the scheme of this application, those skilled in the art can supplement the details according to the common knowledge in this field. For example, according to the common knowledge in this field, normalization can be used to eliminate dimensional conflicts before feature fusion, interpolation can be used to eliminate dimensional differences, and thresholds can be reasonably set based on historical data, experience or business scenario requirements. The model can be trained based on a general model training method, and the number of layers in the model structure can be set based on actual needs, the activation function can be selected, etc. This application will no longer provide redundant introductions to the overly detailed implementation process.

[0229] See also Figure 2 , Figure 2This is a schematic diagram of the structure of a big data analysis system provided in an embodiment of the present invention. The big data analysis system, for example, is a backend server system and includes at least a processor 101, a communication interface 102, and a memory 103. The processor 101, communication interface 102, and memory 103 may be connected via a bus or other means. The processor 101 (also known as the Central Processing Unit (CPU)) is the computing and control core of the big data analysis system, capable of parsing various instructions within the big data analysis system and processing various data within the big data analysis system. The communication interface 102 may optionally include a standard wired interface or a wireless interface (such as Wi-Fi or a mobile communication interface), and may be used to send and receive data under the control of the processor 101. The communication interface 102 may also be used for data transmission and interaction within the big data analysis system. The memory 103 is a storage device within the big data analysis system, used to store programs and data. It is understood that the memory 103 herein may include both the built-in memory of the big data analysis system and the extended memory supported by the big data analysis system. The memory 103 provides a storage space, which stores the operating system of the big data analysis system, but the present invention does not limit this.

[0230] In one embodiment, the processor 101 executes the water flow dynamic analysis method based on digital twin provided in the above embodiment of the present invention by running the computer program in the memory 103.

Claims

1. A water flow dynamic analysis method based on digital twin, characterized by: include: Acquire a historical water flow data set of the target water area and collect a real-time water flow monitoring data set, wherein the historical water flow data set includes a plurality of historical water flow state parameters, and the real-time water flow monitoring data set includes a plurality of real-time water flow monitoring parameters; Constructing a digital twin water flow model corresponding to the target water area, wherein the digital twin water flow model includes a spatial distribution topological structure of the target water area and water flow dynamic association rules; Performing multimodal feature fusion processing on the historical water flow data set and the real-time water flow monitoring data set to generate a dynamic water flow feature set; Based on the spatial distribution topology of the digital twin water flow model, the dynamic water flow feature set is loaded into the digital twin water flow model to generate a set of dynamic water flow change rules; A water flow optimization strategy set is generated according to the water flow dynamic change law set.

2. The method according to claim 1, wherein The constructing of a digital twin water flow model corresponding to the target water area includes: Calling a preset three-dimensional grid processing tool to perform grid division processing on the geographic space of the target water area, generating a set of spatial grids of the target water area, and assigning an initial water flow attribute parameter set to each spatial grid; Acquire a hydrogeological data set of the target water area, wherein the hydrogeological data set includes a permeability parameter set, a geological structure parameter set, and a surface cover parameter set; Performing dynamic association rule mining on the spatial grid set based on the hydrogeological data set to generate the water flow dynamic association rules; Performing topological integration processing on the spatial grid set and the water flow dynamic association rule to generate a spatial distribution topological structure of the digital twin water flow model, and configuring a dynamic update interface for the spatial distribution topological structure; The initial water flow attribute parameter set is loaded into the grid node corresponding to the spatial distribution topology structure through the dynamic update interface to complete the construction of the digital twin water flow model.

3. The method according to claim 1, wherein The performing multimodal feature fusion processing on the historical water flow data set and the real-time water flow monitoring data set to generate a dynamic water flow feature set includes: Extracting a static water flow feature set and a dynamic change trajectory set from the historical water flow data set, wherein the static water flow feature set includes a flow velocity mean parameter, a flow direction stability parameter, and a water level fluctuation threshold parameter; Extracting real-time flow velocity parameters, real-time flow direction parameters, and real-time water level parameters from the real-time water flow monitoring data set, and generating a real-time water flow fluctuation trajectory data set; Performing spatiotemporal alignment processing on the static water flow feature set and the dynamic change trajectory set to generate a historical water flow spatiotemporal feature vector set; Perform multi-dimensional mapping analysis on the real-time water flow fluctuation trajectory dataset and the historical water flow spatiotemporal feature vector set to determine a real-time water flow abnormal fluctuation interval set; Based on the real-time water flow abnormal fluctuation interval set, the real-time flow velocity parameters, real-time flow direction parameters and real-time water level parameters are dynamically enhanced to generate a fused water flow dynamic feature set, and the fused water flow dynamic feature set is used as the dynamic water flow feature set.

4. The method according to claim 3, wherein The method of loading the dynamic water flow feature set into the digital twin water flow model based on the spatial distribution topology of the digital twin water flow model to generate a set of dynamic water flow change rules includes: Determine a set of water flow state constraints for each grid node according to the spatial distribution topology, wherein the set of water flow state constraints includes a flow velocity transmission rule, a water level balance rule, and a flow direction coordination rule between adjacent grid nodes; Matching and associating each dynamic feature in the dynamic water flow feature set with a corresponding grid node to generate a gridded dynamic feature distribution map; Performing dynamic rule verification processing on the gridded dynamic feature distribution map based on the water flow state constraint condition set to determine a set of abnormal feature distribution areas; Performing feature reconstruction processing on the gridded dynamic feature distribution map according to the abnormal feature distribution area set to generate an optimized dynamic feature distribution map; The optimized dynamic feature distribution map is subjected to dynamic law mining processing through the water flow dynamic association rules of the digital twin water flow model to generate the water flow dynamic change law set.

5. The method according to claim 4, wherein Generating a set of water flow optimization strategies according to the set of water flow dynamic change rules includes: Extracting flow velocity change pattern parameters, water level evolution trend parameters and flow direction deviation probability parameters from the set of water flow dynamic change rules; Based on the preset water flow control constraints, the flow velocity change pattern parameters, water level evolution trend parameters and flow direction deviation probability parameters are subjected to strategy matching analysis and processing to generate an initial water flow optimization solution set; Calling the dynamic update interface of the digital twin water flow model to perform simulation verification processing on the initial water flow optimization scheme set, and determining the simulation control effect parameters corresponding to each initial water flow optimization scheme; Prioritizing the initial water flow optimization scheme set according to the simulation control effect parameters to generate an optimization scheme priority queue; The top N optimization schemes with the highest priority in the optimization scheme priority queue are determined as the water flow optimization strategy set, wherein N is a positive integer dynamically adjusted based on the execution capability parameter of the water flow control system.

6. The method according to claim 5, wherein The calling of the dynamic update interface of the digital twin water flow model to perform simulation verification processing on the initial water flow optimization solution set includes: Loading the initial water flow optimization solution set to the grid nodes of the digital twin water flow model through the dynamic update interface; Based on the spatial distribution topology of the digital twin water flow model, simulate and execute the control operations corresponding to each initial water flow optimization scheme, and collect the virtual water flow parameter set after the simulation execution; Comparing and analyzing the virtual water flow parameter set with a preset expected control target parameter set to determine a target deviation parameter for each initial water flow optimization scheme; Adaptively correcting the initial water flow optimization scheme according to the target deviation parameter to generate a set of corrected water flow optimization schemes; The modified water flow optimization scheme set is reloaded into the digital twin water flow model for secondary simulation verification until the target deviation parameter is less than the preset deviation threshold, and the verified modified water flow optimization scheme is used as the initial water flow optimization scheme set.

7. The method according to claim 1, wherein The method also includes a process of adaptively controlling water flow anomalies, including: Acquire in real time a virtual water flow parameter set of each grid node in the digital twin water flow model and a corresponding monitoring parameter set in the real-time water flow monitoring data set; Performing dynamic deviation verification processing on the virtual water flow parameter set and the monitoring parameter set to generate an initial parameter deviation set; If there is an abnormal deviation parameter exceeding a preset reference range in the initial parameter deviation set, extracting the corresponding abnormal grid node identifier and abnormal time period parameter set; Based on the abnormal period parameter set, a dynamic backtracking analysis is performed on the spatial distribution topology structure of the digital twin water flow model to restore the dynamic change trajectory of the water flow corresponding to the abnormal deviation parameter; Performing rule matching analysis on the water flow dynamic change trajectory and the flow velocity transfer rule in the water flow dynamic association rule to determine the rule deviation parameter; If the rule deviation parameter exceeds a preset rule deviation threshold, calling the optimization strategy set associated with the abnormal grid node identifier in the water flow optimization strategy set; Generate a real-time control instruction set based on the optimization strategy set, and send the real-time control instruction set to the water flow control system; The water flow control system performs multi-level response control operations on the physical control equipment corresponding to the abnormal grid node according to the real-time control instruction set, and at the same time transmits the control feedback parameters back to the dynamic update interface of the digital twin water flow model; The control feedback parameters are loaded into the corresponding grid nodes of the spatially distributed topological structure through the dynamic update interface, triggering the real-time topological weight correction processing of the digital twin water flow model, and generating an updated virtual water flow parameter set to verify the effectiveness of the control.

8. The method according to claim 7, wherein The performing dynamic deviation verification processing on the virtual water flow parameter set and the monitoring parameter set to generate an initial parameter deviation set includes: An independent verification window is configured for each grid node's virtual flow velocity, virtual water level, and virtual flow direction. The independent verification window includes a dynamically adjusted time window length and parameter fluctuation tolerance. Collecting a continuous change sequence of the virtual water flow parameter and a continuous change sequence of the monitoring parameter within the independent verification window; Performing multi-stage alignment processing on the continuous change sequence of the virtual water flow parameter and the continuous change sequence of the monitoring parameter to generate a preliminary verification trajectory set; If there is an abnormal trajectory segment in the preliminary verification trajectory set whose trajectory deviation exceeds the preset deviation tolerance, a secondary verification mechanism is initiated; The secondary verification mechanism includes: dynamically extending the time window length of the independent verification window based on the time range of the abnormal trajectory segment, and reducing the parameter fluctuation tolerance; Recollecting the extended change sequence of the virtual water flow parameter and the extended change sequence of the monitoring parameter within the extended independent verification window; Performing spatiotemporal segment alignment processing on the extended change sequence to generate a set of refined verification trajectories; Calculating a trajectory similarity parameter for each trajectory segment in the refined verification trajectory set, and if the trajectory similarity parameter is lower than a preset similarity tolerance threshold, determining that there is an abnormal deviation between the virtual water flow parameter and the monitoring parameter; The parameter type, deviation amplitude and timestamp corresponding to the abnormal deviation are integrated into the initial parameter deviation set.

9. The method according to claim 2, wherein The performing dynamic association rule mining processing on the spatial grid set based on the hydrogeological data set to generate the water flow dynamic association rules includes: Extracting a permeability parameter set, a geological structure parameter set, and a surface cover parameter set from the hydrogeological data set; Mapping the permeability parameter set to corresponding grid nodes in the spatial grid set to generate a permeability distribution feature set for each grid node; Determining a geological fault zone distribution feature set between different grid nodes in the spatial grid set according to the geological structure parameter set, and generating a geological structure influencing factor set in combination with the permeability distribution feature set; Extracting vegetation cover type parameters and terrain relief parameters from the surface cover parameter set, and generating a surface cover resistance coefficient set according to the vegetation cover type parameters and terrain relief parameters; Inputting the geological structure influencing factor set and the surface cover resistance coefficient set into a preset water flow dynamic impact analysis model, performing dynamic association analysis on the water flow transmission paths between adjacent grid nodes in the spatial grid set, and generating an initial water flow dynamic association rule set; Performing parameter adaptability verification processing on the initial water flow dynamic association rule set based on the initial water flow attribute parameter set of the spatial grid set, and determining the flow velocity transmission rule, water level balance rule and flow direction coordination rule in the water flow dynamic association rule; The flow velocity transmission rule, the water level balance rule and the flow direction coordination rule are subjected to rule topology integration processing with the topological connection relationship of the spatial grid set to generate the water flow dynamic association rule.

10. A big data analysis system, characterized in that: include: a memory storing a computer program; A processor, configured to load the computer program to implement the water flow dynamic analysis method based on digital twins as described in any one of claims 1 to 9.

Images