Water network model construction method and system based on digital twinning

Abstract

The application discloses a water network model construction method and system based on digital twinning. The model construction method comprises the following steps: integrating data resources and obtaining water network and geographical environment information; determining attribute items and topological relations, building a concept model; abstracting information water network and constructing a water network model; binding a buoy to collect fish school and water flow data, and cleaning standardized data; calculating the correlation between fish school movement and water flow; optimizing water network model parameters, optimizing ecological niche, and analyzing and quantifying the relationship between fish school and environment. The water network model construction method can construct an accurate and comprehensive water network model by integrating existing data resources in the region, improve the accuracy of the model, and provide a solid foundation for subsequent analysis and optimization. By collecting fish school trajectory data and water flow data, the correlation between fish school movement and water flow can be analyzed in depth, which helps to reveal the internal rules of the aquatic ecosystem.

Description

Technical Field

[0001] This invention belongs to the field of water conservancy engineering and information technology, specifically relating to a method and system for constructing a water network model based on digital twins. Background Technology

[0002] In the fields of water conservancy engineering and information technology, the construction of water network models has always been an important research topic. Traditional methods for constructing water network models often rely on static data and simplified physical assumptions, making it difficult to accurately reflect the dynamic characteristics and complexity of water network systems. With the rapid development of information technology, especially the rise of digital twin technology, new ideas and methods have been provided for the construction of water network models. Digital twin technology is a simulation process that integrates multiple disciplines, multiple physical quantities, multiple scales, and multiple probabilities. It can create digital models in virtual space that correspond one-to-one with entities or systems in the real world. Through digital twin technology, real-time monitoring, accurate prediction, and efficient optimization of entities or systems can be achieved. However, many challenges remain when applying digital twin technology to the construction of water network models.

[0003] Water network systems are highly complex and dynamic. They comprise multiple elements such as channels, nodes, and water flows, which interact with each other through intricate topological relationships. Furthermore, water network systems are influenced by various factors, including geographical environment, climate conditions, and human activities; changes in these factors can alter the system's state. Therefore, accurately acquiring and integrating various information about water network systems to construct a digital model that truly reflects their characteristics is a pressing issue. Traditional water network modeling methods often focus only on the physical characteristics of water flow, neglecting ecological factors within the system. In reality, ecological factors in water network systems have a significant impact on their stability and sustainability. To address this, we propose a water network modeling method and system based on digital twins. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a water network model based on digital twins.

[0005] Another objective of this invention is to provide a water network model construction system based on digital twins.

[0006] This invention is achieved through the following technical solution:

[0007] The water network model construction method based on digital twins described in this invention includes the following steps:

[0008] Step 1: By integrating existing data resources in the target area, obtain information on physical water network elements and geographic environment;

[0009] The physical water network element information includes: channels, nodes, and water flow; the geographic environment information includes: geographic information, hydrology, geology, topography, and environment.

[0010] Step 2: Determine the attribute items of water network elements and the topological relationships between elements, establish the physical water network information structure, and construct a conceptual model of the water network based on the physical water network information structure;

[0011] Step 3: Abstract the physical water network into an information water network, and construct a data model of the water network elements and a virtual space environment for the physical water network;

[0012] Step 4: Attach detection buoys to the fish school to collect fish movement trajectory data. Use the water flow sensor on the buoy and fixed monitoring equipment to collect water flow data, and clean and standardize the fish movement trajectory and water flow data.

[0013] Step 5: Based on the fish movement trajectory and water flow data after cleaning and standardization, calculate the angle and proportional relationship between the fish movement and the water flow speed and direction, and analyze the correlation.

[0014] Step 6: Based on the irregular areas of the fish movement trajectory and the movement trajectory of the fish at the boundary of the water network, optimize the physical parameters of the water network model, optimize the ecological niche distribution based on the movement trajectory of the fish and the water flow, analyze the interaction between the fish and the environment, quantify the relationship and integrate it into the model.

[0015] The method for obtaining physical water network element information in step 1 of this invention is as follows:

[0016] Channel information acquisition involves identifying the channel's location and orientation using GIS and satellite remote sensing imagery, determining the channel's centerline and boundaries through on-site surveys, measuring the channel length along its centerline in GIS, and averaging the width after on-site measurements at multiple locations. For large rivers, the average width of different river sections is analyzed using remote sensing imagery, and the depth is estimated based on water level data and river cross-sectional shape. The initial roughness coefficient is determined based on the channel's material by consulting hydraulic engineering manuals.

[0017] Node information acquisition involves using GIS data and field surveys to determine the location information of nodes, analyzing the boundary condition attributes of nodes, determining the connection relationship between nodes and surrounding channels, and the inflow and outflow conditions of water at nodes.

[0018] Water flow information acquisition involves obtaining data on water flow rate, velocity, and water level through hydrological monitoring stations.

[0019] The method for obtaining geographic environment information in step 1 of this invention is as follows:

[0020] Geographic information acquisition: using GIS data to acquire the natural and human geographical elements of the area where the water network is located;

[0021] Hydrological information acquisition involves obtaining past hydrological parameters and historical hydrological data through hydrological monitoring stations.

[0022] Geological information acquisition involves reviewing geological survey reports and geological maps to understand the stratigraphic structure, rock types, and geological structure information of the area where the water network is located.

[0023] Topographic information acquisition: Using DEM data, topographic information of the area where the water network is located is obtained. The topographic information includes: elevation, slope, and aspect.

[0024] Environmental information is obtained by regularly collecting water samples from water quality monitoring points and analyzing them to obtain water quality parameters, and by obtaining ecological survey data to understand the ecological environment around the water network.

[0025] Step 2 of this invention, which involves establishing a physical water network information structure and constructing a conceptual model of the water network based on that structure, includes:

[0026] The water network element attributes include: channel attributes, node attributes, and water flow attributes; among which, channel attributes include: length, width, depth, roughness coefficient, and water volume-water level relationship curve; node attributes include: location, size, artificial node functional attributes, and natural node functional attributes; and water flow attributes include: flow rate, flow velocity, flow direction, seasonal variation, and flood frequency.

[0027] The topological relationships between the elements include: connectivity analysis, determination of upstream and downstream relationships, and construction of hierarchical relationships;

[0028] Connectivity analysis uses GIS to analyze the connectivity between channels and between channels and nodes. SWAT is used to determine the connectivity of the water network. Upstream and downstream relationships are determined based on topographic information and the natural flow direction of water. At the confluence of rivers, the upstream and downstream relationships are determined by analyzing the flow direction and flow distribution. Hierarchical relationship construction uses GIS to automatically divide watersheds based on topographic data and flow direction. Within each watershed, the hierarchical relationship of the water system is further determined based on the length, flow, and watershed area parameters of the rivers. The hierarchical relationship includes: main rivers and tributaries, first-order tributaries and second-order tributaries.

[0029] Based on the physical water network information structure, and using object-oriented modeling, the channels, nodes, and water flow elements in the physical water network are used as object classes to construct a conceptual model of the water network.

[0030] The method for abstracting the physical water network into an information water network and constructing a data model of water network elements in step 3 of this invention is as follows:

[0031] The data model for channel elements is constructed by using the channel centerline as a baseline and dividing the channel into several segments (the core division criteria are adapting to terrain changes, differences in channel cross-sectional characteristics, and data accuracy requirements; priority is given to splitting based on key nodes such as abrupt changes in river topography and changes in cross-sectional dimensions; the number of segments can be flexibly adjusted according to the actual water network scale and monitoring accuracy). Within each segment, it is assumed that depth and width change linearly. The parameters of each segment are determined based on existing data resources in the integrated target area. The roughness coefficient is dynamically adjusted according to the channel material and actual water flow conditions; the adjustment formula is as follows:

[0032] ;

[0033] In the formula, This represents the adjusted roughness coefficient. represents the initial roughness coefficient, and k represents the adjustment coefficient determined empirically. Indicates the actual flow rate. The design flow velocity is indicated; the determination of the adjustment coefficient k is based on the characteristics of the channel material, combined with historical data and actual measurement experience of similar water network projects. First, the initial range is set according to the material such as concrete, natural soil, and pebble coverage; then, the model simulation is compared with the actual water flow data, and the flow velocity deviation is fed back for iterative calibration to ensure that the adjusted roughness coefficient can accurately match the actual water flow state, taking into account both calculation efficiency and simulation accuracy.

[0034] The data model for node elements constructs a three-dimensional coordinate system with the node center as the origin. The connection relationship is determined by the coordinates of the connection points between nodes and channels. The flow distribution model for nodes is as follows:

[0035] ;

[0036] In the formula, This represents the traffic of the i-th branch channel. This represents the total inflow to the node. This represents the water head at the entrance of the i-th branch channel. Let m represent the head at the entrance of the j-th branch channel, and m represent the number of branch channels.

[0037] The data model for water flow elements uses time series data to record the changes in flow rate, velocity, and direction over time, and obtains continuous function expressions through interpolation and fitting of hydrological monitoring data.

[0038] The construction process of the physical water network virtual space environment described in step 3 of this invention is as follows:

[0039] Using Geographic Information System (GIS) and computer graphics technology, the topographic information of the area where the water network is located is transformed into a three-dimensional topographic model. The topographic information includes, but is not limited to, elevation, slope and aspect.

[0040] Terrain construction is carried out using digital elevation model (DEM) data, and topographic features are generated through interpolation and rendering of the DEM data.

[0041] The water network elements are drawn in a 3D terrain model according to their actual geographical location and shape. The channels are drawn based on their centerline and width and depth data to generate a 3D solid model. The nodes are constructed according to their position and size information to construct the corresponding geometric shape. Combined with environmental information, the virtual space environment is rendered to obtain the physical water network virtual space environment.

[0042] The method for cleaning and standardizing the fish movement trajectory and water flow data described in step 4 of this invention is as follows:

[0043] Based on previously collected data on fish movement trajectories, the average speed of each fish in different time periods and the magnitude of positional change at adjacent time points are calculated. A frequency distribution histogram of speed and positional change is plotted on the calculated speed and positional change data, and speed thresholds and positional change thresholds are set.

[0044] The real-time collected fish movement trajectory data is judged based on speed threshold and position change threshold. When the fish movement speed is greater than the speed threshold and the position change at adjacent time points is greater than the position change threshold, it is judged as abnormal data and removed.

[0045] The data is smoothed using a calculation formula, which is as follows:

[0046] ;

[0047] In the formula, The coordinates of the fish swarm at time t are represented by z, which represents the size of the moving average window. and Represents the coordinates of the fish group at the u-th time.

[0048] For water flow data, missing values ​​are handled for data such as flow rate, flow velocity, and water level. If data is missing at a certain moment, it is supplemented by linear interpolation based on data from the preceding and following moments.

[0049] The method for calculating the angle and proportional relationship between fish movement and water flow velocity and direction, as described in step 5 of this invention, is as follows:

[0050] Let the velocity vector of the school of fish at a certain moment be... The water flow velocity vector is The formula for calculating the angle between the fish movement and the water flow velocity is:

[0051] ;

[0052] In the formula, This represents the angle between the movement of the fish school and the speed of the water flow. Represents the dot product of two vectors. and The modulus represents the movement of the fish school and the speed of the water flow, respectively;

[0053] The formula for calculating the proportion of fish moving in the same direction as the water flow is:

[0054] ;

[0055] In the formula, This indicates the proportion where the direction of fish movement is the same as the direction of water flow. This indicates the number of samples in which the fish are moving in the same direction as the water flow. Indicates the total number of samples;

[0056] By calculating the angle and proportion relationships in different regions and time periods, the correlation between fish movement and water flow is analyzed. If the angle is small and the proportion is high, it indicates that the fish movement in that area is greatly affected by the water flow. Based on this, the areas in the water network where the water flow has a significant impact on fish behavior can be identified.

[0057] The method for optimizing the physical parameters of the water network model based on the irregular area of ​​the fish movement trajectory and the movement trajectory of the fish at the boundary of the water network in step 6 of this invention is as follows:

[0058] If the fish's movement trajectory frequently changes direction or they gather at the edge of the channel in a certain area, it indicates that the channel width is incorrect. The width is adjusted by the distribution density of the fish's movement trajectory, using the following formula:

[0059] ;

[0060] In the formula, This indicates the adjusted channel width. This indicates the channel width before adjustment. Indicates the adjustment factor. This indicates the distribution density of the fish's movement trajectory. The trajectory distribution density threshold is defined as follows: an initial threshold is set based on historical fish trajectory density statistics of similar water networks, combined with the water network type and fish habits; this threshold is then optimized through model iteration and calibration. Adjustment coefficients are also included. The determination of the width of the channel is based on the type of water network, such as natural rivers and artificial canals, the original design accuracy of the channel, and the activity characteristics of the target fish. First, the initial value is set with reference to the engineering experience of optimizing the width of similar water networks. The value is usually in the range of 0.1-0.5. Then, it is verified by small-scale field measurement data. Combined with the adaptability of the adjusted channel width to the movement trajectory of the fish, iterative calibration is performed to ensure that the width adjustment range fits the actual hydrological environment and avoids over-correction or under-correction.

[0061] Nodes include confluence points and divergence points. By observing the movement trajectories of fish at these nodes, the shape of the nodes and water flow distribution parameters can be optimized.

[0062] Let the water flow distribution parameters of the branches be respectively and , , Calculations are based on node topography, channel cross-section data, and historical flow meter engineering empirical formulas, among which... According to the proportions of the fish entering the two branches, respectively and ,in ;

[0063] The adjusted water flow distribution parameters are as follows:

[0064] ;

[0065] In the formula, This represents the node traffic adjustment coefficient. The adjusted water flow distribution parameters, The original water flow distribution parameters; node flow adjustment coefficients. The determination of the flow distribution is based on the node type (merging point / branching point), the hydraulic characteristics of the branch channel (cross-sectional size, roughness) and the experience of optimizing the flow of nodes in similar water networks. The initial value range is usually set to 0.2-0.6. Then, the model simulation and the fitting feedback of the actual branch selection ratio of the fish are used for iterative calibration to ensure that the adjusted flow distribution can match the actual water flow state reflected by the movement of the fish and balance the adjustment sensitivity and stability.

[0066] Adjust the water flow velocity field based on the relationship between the fish movement trajectory and the water flow velocity and direction.

[0067] The specific method for optimizing the ecological niche distribution based on the fish movement trajectory and water flow conditions, analyzing the interaction between the fish and the environment, quantifying the relationship, and incorporating it into the model in step 6 of this invention is as follows:

[0068] The ecological niche characteristics of a water network are defined as food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability.

[0069] The movement trajectory of the fish and the water flow are obtained by attaching detection buoys to the fish;

[0070] Ecological niche characteristics were normalized to obtain parameter values ​​for food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability;

[0071] Taking into account various niche parameters, the comprehensive niche value is determined, and the expression for the comprehensive niche value is as follows:

[0072] ;

[0073] In the formula, E represents the comprehensive niche value, and F, T, D, and V represent the parameter values ​​of food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively. These represent the weighting coefficients for food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively.

[0074] Adjust the niche distribution in the water network model based on the comprehensive niche value.

[0075] The water network model construction system based on digital twins of the present invention is used to implement the water network model construction method based on digital twins, including: a regional data resource integration module, a water network model construction module, an optimized data acquisition and processing module, and a model optimization module;

[0076] The regional data resource integration module is used to integrate existing data resources in the target area to provide basic data support for model building.

[0077] The water network model construction module is electrically connected to the regional data resource integration module. The water network model construction module is used to construct a water network model based on the integrated data resources.

[0078] The optimized data acquisition and processing module is electrically connected to the water network model construction module. The optimized data acquisition and processing module is used to collect fish movement trajectory data and water flow data, providing a dynamic data source for model optimization, and cleaning and standardizing the fish movement trajectory data and water flow data.

[0079] The model optimization module is electrically connected to the optimized data acquisition and processing module. The model optimization module is used to optimize the physical parameters and niche distribution data of the water network model based on the processed fish movement trajectory and water flow data.

[0080] Compared with the prior art, the present invention has the following beneficial effects:

[0081] The water network model construction method proposed in this invention integrates existing regional data resources to construct an accurate and comprehensive water network model, improving the model's accuracy and providing a solid foundation for subsequent analysis and optimization. By collecting data on fish movement trajectories and water flow conditions, it is possible to deeply analyze the correlation between fish movement and water flow, helping to reveal the inherent laws of the aquatic ecosystem and providing a scientific basis for the rational utilization and protection of water resources. Optimizing the physical parameters and niche distribution data of the water network model based on fish movement trajectories and water flow conditions enables the water network model to better adapt to changes in the actual environment, improving the model's practicality and reliability. Attached Figure Description

[0082] Figure 1 Flowchart of the method for constructing a water network model based on digital twins;

[0083] Figure 2 System module framework diagram for constructing a water network model based on digital twins. Detailed Implementation

[0084] The technical solution of the present invention will be further described in detail below through specific embodiments.

[0085] Example 1

[0086] Reference Figure 1 As shown, the water network model construction method based on digital twins is as follows:

[0087] Step 1: By integrating existing data resources in the target area, obtain physical water network element information and geographic environment information. The physical water network element information includes: channels, nodes and water flow. The geographic environment information includes: geographic information, hydrology, geology, topography and environment.

[0088] Step 2: Determine the attribute items of water network elements and the topological relationships between elements, establish the physical water network information structure, and construct a conceptual model of the water network based on the physical water network information structure;

[0089] Step 3: Abstract the physical water network into an information water network, and construct a data model of the water network elements and a virtual space environment for the physical water network;

[0090] Step 4: Attach detection buoys to the fish school to collect fish movement trajectory data. Use the water flow sensor on the buoy and fixed monitoring equipment to collect water flow data, and clean and standardize the fish movement trajectory and water flow data.

[0091] Step 5: Based on the fish movement trajectory and water flow data after cleaning and standardization, calculate the angle and proportional relationship between the fish movement and the water flow speed and direction, and analyze the correlation.

[0092] Step 6: Based on the irregular areas of the fish movement trajectory and the movement trajectory of the fish at the boundary of the water network, optimize the physical parameters of the water network model, optimize the ecological niche distribution based on the movement trajectory of the fish and the water flow, analyze the interaction between the fish and the environment, quantify the relationship and integrate it into the model.

[0093] The specific method is as follows:

[0094] Step 1: The method for obtaining physical water network element information is as follows: Using GIS and satellite remote sensing imagery, the location and direction of channels are identified. Further confirmation and refinement of the image data through on-site surveys are conducted to clarify the centerline and boundaries of the channels. Channel length is obtained by measuring along the centerline in GIS, and width is calculated by averaging measurements at multiple locations on-site. For large rivers, combining remote sensing imagery with analysis of the average width of different river sections provides a more comprehensive and accurate understanding of channel width characteristics. Depth is estimated based on water level data and river cross-sectional shape, while the initial roughness coefficient is determined according to the channel material by consulting a hydraulic engineering manual. Node information is obtained by determining their location using GIS data and on-site surveys, deeply analyzing boundary condition attributes, and clarifying their connection relationships with surrounding channels and water inflow and outflow conditions. Water flow information is obtained by collecting flow rate, velocity, and water level data through hydrological monitoring stations.

[0095] The methods for obtaining geographic environmental information are as follows: geographic information utilizes GIS data to obtain natural and human geographic elements of the area where the water network is located; hydrological information utilizes hydrological monitoring stations to obtain past hydrological parameters and historical hydrological data; geological information is obtained by reviewing geological survey reports and geological maps to understand stratigraphic structure, rock type, geological structure, and other information; topographic information is obtained by using DEM data to obtain topographic information such as elevation, slope, and aspect; environmental information is obtained by setting up water quality monitoring points to regularly collect water samples for analysis to obtain water quality parameters, and at the same time, ecological survey data is used to obtain the ecological environment status around the water network, thereby improving the understanding of the water network from an ecological perspective.

[0096] Step 2: Determine the attribute items of the water network elements and the topological relationships between elements, establish the physical water network information structure, and then construct a conceptual model of the water network. The attribute items of the water network elements include channel attributes, node attributes, and flow attributes. The length, width, depth, roughness coefficient, and water volume-water level relationship curve in the channel attributes are used to describe the hydraulic characteristics of the channel and the changes in flow. Node attributes include location, size, functional attributes of artificial nodes, and functional attributes of natural nodes. Flow attributes include flow rate, velocity, direction, seasonal variation, and flood frequency. The topological relationships between elements include connectivity analysis, determination of upstream and downstream relationships, and construction of hierarchical relationships.

[0097] Connectivity analysis utilizes GIS to analyze the connectivity relationships between channels and between channels and nodes. SWAT analysis is also used to determine the connectivity of the water network, which determines the flow path and range of water within the network. Upstream and downstream relationships are determined based on topographic information and the natural flow direction of water, further clarified at river confluences by analyzing flow direction and distribution. Hierarchical relationships are constructed using GIS to automatically divide watersheds based on topographic data and flow direction. Within each watershed, hierarchical relationships are determined based on parameters such as river length, flow rate, and watershed area, including major rivers and tributaries, first-order tributaries, and second-order tributaries. Based on the physical water network information structure, object-oriented modeling is used to construct a conceptual model of the water network, treating channels, nodes, and flow elements as object classes.

[0098] Step 3: Abstract the physical water network into an information water network, construct a data model of the water network elements and a virtual space environment of the physical water network, so that the actual water network can be transformed into a quantifiable and simulated research object.

[0099] The method for abstracting the physical water network into an information water network and constructing a data model of water network elements is as follows:

[0100] The data model for channel elements is constructed by using the channel centerline as a baseline and dividing the channel into several segments (the core division criteria are adapting to terrain changes, differences in channel cross-sectional characteristics, and data accuracy requirements; priority is given to splitting based on key nodes such as abrupt changes in river topography and changes in cross-sectional dimensions; the number of segments can be flexibly adjusted according to the actual water network scale and monitoring accuracy). Within each segment, it is assumed that depth and width change linearly. The parameters of each segment are determined based on existing data resources in the integrated target area. The roughness coefficient is dynamically adjusted according to the channel material and actual water flow conditions; the adjustment formula is as follows:

[0101] ;

[0102] In the formula, This represents the adjusted roughness coefficient. represents the initial roughness coefficient, and k represents the adjustment coefficient determined empirically. Indicates the actual flow rate. The design flow velocity is indicated. The adjustment coefficient k is determined based on the characteristics of the channel material, combined with historical data and actual measurement experience from similar water network projects. Initial ranges are first set according to materials such as concrete, natural soil, and pebble coverage. Then, the model simulation is compared with the actual water flow data, and the flow velocity deviation is fed back for iterative calibration to ensure that the adjusted roughness coefficient can accurately match the actual water flow state, balancing computational efficiency and simulation accuracy.

[0103] The data model for node elements constructs a three-dimensional coordinate system with the node center as the origin. The connection relationship is determined by the coordinates of the connection points between nodes and channels. The flow distribution model for nodes is as follows:

[0104] ;

[0105] In the formula, This represents the traffic of the i-th branch channel. This represents the total inflow to the node. This represents the water head at the entrance of the i-th branch channel. Let m represent the head at the entrance of the j-th branch channel, and m represent the number of branch channels.

[0106] The data model for water flow elements uses time series data to record the changes in flow rate, velocity, and direction over time, and obtains continuous function expressions through interpolation and fitting of hydrological monitoring data.

[0107] The process of constructing a virtual space environment for a physical water network is as follows:

[0108] Using Geographic Information System (GIS) and computer graphics technology, the topographic information of the area where the water network is located is transformed into a three-dimensional topographic model. Topographic information includes, but is not limited to, elevation, slope, and aspect. Digital Elevation Model (DEM) data is used for terrain construction, generating topographic features through interpolation and rendering of the DEM data. Water network elements are drawn in the three-dimensional topographic model according to their actual geographical location and shape. Channels are drawn based on their centerlines and width / depth data to generate three-dimensional solid models. Nodes are constructed with their corresponding geometric shapes based on their location and size information. Combined with environmental information, the virtual space environment is rendered to obtain the physical water network virtual space environment.

[0109] Step 4: Attach detection buoys to the fish school to collect data on their movement trajectories. Use water flow sensors on the buoys and fixed monitoring equipment to collect water flow data, and then clean and standardize this data as follows:

[0110] Based on previously collected data on fish movement trajectories, the average speed of each fish in different time periods and the magnitude of positional change at adjacent time points are calculated. A frequency distribution histogram of speed and positional change is plotted on the calculated speed and positional change data, and speed thresholds and positional change thresholds are set.

[0111] The real-time collected fish movement trajectory data is judged based on speed threshold and position change threshold. When the fish movement speed is greater than the speed threshold and the position change at adjacent time points is greater than the position change threshold, it is judged as abnormal data and removed.

[0112] The data is smoothed using a calculation formula, which is as follows:

[0113] ;

[0114] In the formula, The coordinates of the fish swarm at time t are represented by z, which represents the size of the moving average window. and Represents the coordinates of the fish group at the u-th time.

[0115] For water flow data, missing values ​​are handled for data such as flow rate, flow velocity, and water level. If data is missing at a certain moment, it is supplemented by linear interpolation based on data from the preceding and following moments.

[0116] Step 5: Based on the fish movement trajectory and water flow data after cleaning and standardization, calculate the angle and proportional relationship between fish movement and water flow velocity and direction, analyze their correlation, and reveal the mechanism of the influence of the aquatic network environment on organisms from the perspective of biological behavior. The method is as follows:

[0117] Let the velocity vector of the school of fish at a certain moment be... The water flow velocity vector is The formula for calculating the angle between the fish movement and the water flow velocity is:

[0118] ;

[0119] In the formula, This represents the angle between the movement of the fish school and the speed of the water flow. Represents the dot product of two vectors. and The modulus represents the movement of the fish school and the speed of the water flow, respectively;

[0120] The formula for calculating the proportion of fish moving in the same direction as the water flow is:

[0121] ;

[0122] In the formula, This indicates the proportion where the direction of fish movement is the same as the direction of water flow. This indicates the number of samples in which the fish are moving in the same direction as the water flow. Indicates the total number of samples;

[0123] By calculating the angle and proportion relationships in different regions and time periods, the correlation between fish movement and water flow is analyzed. If the angle is small and the proportion is high, it indicates that the fish movement in that area is greatly affected by the water flow. Based on this, the areas in the water network where the water flow has a significant impact on fish behavior can be identified.

[0124] Step 6: Optimize the physical parameters of the water network model based on the irregular areas of the fish movement trajectory and the movement trajectory at the boundary of the water network. The method is as follows:

[0125] If the fish's movement trajectory frequently changes direction or they gather at the edge of the channel in a certain area, it indicates that the channel width is incorrect. The width is adjusted by the distribution density of the fish's movement trajectory, using the following formula:

[0126] ;

[0127] In the formula, This indicates the adjusted channel width. This indicates the channel width before adjustment. Indicates the adjustment factor. This indicates the distribution density of the fish's movement trajectory. The trajectory distribution density threshold is defined as follows: an initial threshold is set based on historical fish trajectory density statistics of similar water networks, combined with the water network type and fish habits; this threshold is then optimized through model iteration and calibration. Adjustment coefficients are also included. The determination of the width is based on the water network type (natural rivers, artificial canals, etc.), the original design accuracy of the channel, and the activity characteristics of the target fish. First, the initial value is set with reference to the engineering experience of optimizing the width of similar water networks, usually ranging from 0.1 to 0.5. Then, the value is verified by small-scale field measurement data, and iterative calibration is performed by combining the adaptability of the adjusted channel width with the movement trajectory of the fish to ensure that the width adjustment range fits the actual hydrological environment and avoids over-correction or under-correction.

[0128] Nodes include confluence points and divergence points. By observing the movement trajectories of fish at these nodes, the shape of the nodes and water flow distribution parameters can be optimized.

[0129] Let the water flow distribution parameters of the branches be respectively and , , The initial values ​​are calculated based on node topography, channel cross-sectional data, and historical flow meter engineering empirical formulas, whereby... According to the proportions of the fish entering the two branches, respectively and ,in ;

[0130] The adjusted water flow distribution parameters are as follows:

[0131] ;

[0132] In the formula, This represents the node traffic adjustment coefficient. The adjusted water flow distribution parameters, The original water flow distribution parameters; node flow adjustment coefficients. The determination of the flow distribution is based on the node type (merging point / branching point), the hydraulic characteristics of the branch channel (cross-sectional size, roughness) and the experience of optimizing the flow of nodes in similar water networks. The initial value range is usually set to 0.2-0.6. Then, the model simulation and the fitting feedback of the actual branch selection ratio of the fish are used for iterative calibration to ensure that the adjusted flow distribution can match the actual water flow state reflected by the movement of the fish and balance the adjustment sensitivity and stability.

[0133] Adjust the water flow velocity field based on the relationship between the fish movement trajectory and the water flow velocity and direction.

[0134] Based on fish movement trajectories and water flow conditions, ecological niche distribution was optimized, and the interaction between fish and the environment was analyzed, quantified, and incorporated into the model. The method is as follows:

[0135] The ecological niche characteristics of a water network are defined as food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability.

[0136] The movement trajectory of the fish and the water flow are obtained by attaching detection buoys to the fish;

[0137] Ecological niche characteristics were normalized to obtain parameter values ​​for food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability;

[0138] Taking into account various niche parameters, the comprehensive niche value is determined, and the expression for the comprehensive niche value is as follows:

[0139] ;

[0140] In the formula, E represents the comprehensive niche value, and F, T, D, and V represent the parameter values ​​of food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively. These represent the weighting coefficients for food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively.

[0141] Adjust the niche distribution in the water network model based on the comprehensive niche value.

[0142] Reference Figure 2 As shown, a water network model construction system based on digital twins is used to implement the water network model construction method based on digital twins, including: a regional data resource integration module, a water network model construction module, an optimized data acquisition and processing module, and a model optimization module;

[0143] The regional data resource integration module is used to integrate existing data resources in the target area to provide basic data support for model building.

[0144] The water network model construction module is electrically connected to the regional data resource integration module. The water network model construction module is used to construct a water network model based on the integrated data resources.

[0145] The optimized data acquisition and processing module is electrically connected to the water network model construction module. The optimized data acquisition and processing module is used to collect fish movement trajectory data and water flow data, providing a dynamic data source for model optimization, and cleaning and standardizing the fish movement trajectory data and water flow data.

[0146] The model optimization module is electrically connected to the optimized data acquisition and processing module. The model optimization module is used to optimize the physical parameters and niche distribution data of the water network model based on the processed fish movement trajectory and water flow data.

[0147] Based on the above steps, the specific implementation method is as follows:

[0148] This study selected a tributary of the Yangtze River in its middle and lower reaches as the research object. It integrated the provincial water resources department's GIS database, Landsat 8 satellite remote sensing images, historical data from hydrological monitoring stations over the past 10 years, and regional geological survey reports. Using GIS, the river channel (passage) direction was identified. Field surveys of 30 cross-sections were conducted to determine the centerline and boundaries, resulting in an average width of 10m. Combined with water level data and the trapezoidal cross-section shape, the average depth was estimated at 3.5m. The initial roughness coefficient of the concrete channel was set at 0.014. Using GIS and field surveys, 12 node locations were identified, clarifying confluence / diversion attributes and water inflow / outflow conditions. The average flow rate of 25 cubic meters per second was obtained from hydrological stations. Data such as average flow velocity of 1.2 m / s were collected simultaneously, along with DEM topographic data (30 m resolution), COD and ammonia nitrogen parameters from 15 water quality monitoring points, and the surrounding vegetation cover.

[0149] The connectivity between the river channel and nodes was analyzed using ArcGIS, and the connectivity of the water network was verified using a SWAT model. Based on topographic elevation and flow direction, the upstream and downstream relationships and hierarchical relationships of three first-order tributaries and five second-order tributaries were determined. Object-oriented modeling was used to encapsulate the river channel, nodes, and flow into object classes to construct a conceptual model. The river channel was divided into 20 segments of 1km length, assuming linear changes in depth and width within each segment. Based on an actual flow velocity of 1.2m / s and a design flow velocity of 1.0m / s, the roughness coefficient was adjusted to 0.015 using a formula. A three-dimensional coordinate system was established with the node center as the origin, and the flow distribution ratio was calculated based on the head of each branch. A continuous function of the flow parameters was fitted using time series analysis. Three-dimensional terrain with elevations of 20-150m was generated using ArcGIS and Unity3D, and the three-dimensional entities of the river channel and the geometry of the nodes were drawn. After rendering, a virtual space environment was formed.

[0150] Fifty carp and crucian carp were selected and attached to small integrated water quality and flow velocity buoys (sampling frequency 10 seconds / time). Simultaneously, water flow data were collected from 10 fixed monitoring stations within the river channel. Based on previous data, a velocity threshold of 0.8 m / s and a position change threshold of 0.5 m were set. Three sets of abnormal trajectory data were removed, and the coordinates were smoothed using a moving average window of z=5. Linear interpolation was used to supplement two missing flow velocity data points. Calculations showed that the average angle between fish movement and water flow in the shallow waters of the middle reaches of the basin was 30°, with 85% of the movement following the same direction, indicating that this area is significantly affected by water flow.

[0151] To address the frequent turning of fish at the downstream bend, based on a trajectory distribution density of 0.8 (threshold 0.5), the original 10m wide channel was optimized to 10.09m by adjusting coefficient k1=0.3. At a certain diversion node, the initial flow allocation coefficient q1=0.6, based on the fish entry ratio p1=0.7, was adjusted to q1=0.64 after k3=0.4, simultaneously correcting the flow velocity field. In niche optimization, the weights for food resources, water temperature, water depth, and flow velocity were set to 0.3, 0.25, 0.25, and 0.2 respectively. Substituting these into normalized parameter values, the comprehensive value was calculated to be 0.715, adjusting the niche level of this area to "suitable." This ultimately resulted in a digital twin water network model that closely reflects the actual hydrological and ecological characteristics of the watershed.

[0152] Although the present invention has been described in detail above with general descriptions, specific embodiments, and experiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A method for constructing a water network model based on digital twinning, characterized in that, The method comprises the following steps: Step 1: obtaining physical water network element information and geographical environment information by integrating existing data resources in the target area; The physical water network element information comprises channels, nodes and water flow; and the geographical environment information comprises geographical information, hydrology, geology, topography and environment; Step 2: determining the attribute items of the physical water network element information and the topological relationship between the element information, establishing a physical water network information structure, and constructing a conceptual model of the water network based on the physical water network information structure; Step 3: abstracting the physical water network into an information water network, constructing a data model of the physical water network element information and a virtual space environment of the physical water network; Step 4: binding detection buoys on the fish school to collect fish school movement trajectory data, collecting water flow condition data by using water flow sensors on the buoys and fixed monitoring equipment, and performing cleaning and standardization processing on the fish school movement trajectory data and the water flow condition data; Step 5: calculating the angle and proportional relationship analysis correlation between the fish school movement and the water flow speed and direction based on the cleaned and standardized fish school movement trajectory data and the water flow condition data, including calculating the angle between the fish school movement and the water flow speed and the proportion of the same direction of the fish school movement direction and the water flow direction; Step 6: optimizing the physical parameters of the conceptual model of the water network according to the irregular areas of the fish school movement trajectory and the fish school movement trajectory on the water network boundary, optimizing the ecological niche distribution according to the fish school movement trajectory and the water flow condition, analyzing the interaction between the fish school and the environment, quantifying the relationship and integrating the model; The specific method of optimizing the ecological niche distribution according to the fish school movement trajectory and the water flow condition, analyzing the interaction between the fish school and the environment, quantifying the relationship and integrating the model is as follows: Defining the ecological niche characteristics of the water network as food resource richness, water temperature suitability, water depth suitability and water flow speed suitability; Obtaining the fish school movement trajectory and the water flow condition based on the detection buoys bound on the fish school; Performing normalization processing on the ecological niche characteristics to obtain the parameter values of the food resource richness, the water temperature suitability, the water depth suitability and the water flow speed suitability; Determining the ecological niche comprehensive value by comprehensively considering each ecological niche parameter, wherein the expression of the ecological niche comprehensive value is: ； In the formula, E represents a niche comprehensive value, F, T, D, and V represent parameter values of food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively, F, T, D, and V represent weight coefficients of food resource abundance, water temperature suitability, water depth suitability, and water flow velocity suitability, respectively. Adjusting the ecological niche distribution in the water network model according to the ecological niche comprehensive value.

2. The digital-twin-based water network model construction method according to claim 1, characterized in that, The method for obtaining the physical water network element information in step 1 is as follows: Channel information acquisition: identifying the position and direction of the channel through GIS and satellite remote sensing images, determining the center line and boundary of the channel through field investigation, measuring the length of the channel along the center line in GIS to obtain the length, obtaining the width by averaging the values measured on site, obtaining the average width of different river sections by combining remote sensing image analysis for large rivers, estimating the depth according to the water level data and the shape of the river section, and determining the initial roughness coefficient according to the material of the channel by referring to the water conservancy engineering manual; Node information acquisition: determining the position information of the node by using GIS data and field investigation, analyzing the boundary condition attributes of the node, determining the connection relationship between the node and the surrounding channel, and determining the inflow and outflow conditions of the water flow at the node; Water flow information acquisition: obtaining the flow, flow rate and water level data of the water flow through the hydrological monitoring station; The method for obtaining the geographical environment information in step 1 is as follows: Geographical information acquisition, using GIS data to acquire natural geographical elements and human geographical elements in the area where the water network is located; Hydrological information acquisition, acquiring past hydrological parameters and historical hydrological data through hydrological monitoring stations; Geological information acquisition, referring to geological survey reports and geological maps to understand the stratum structure, rock type, and geological structure information in the area where the water network is located; Topographical information acquisition, using DEM data to acquire topographical information in the area where the water network is located, which includes elevation, slope, and slope direction; Environmental information acquisition, acquiring water quality parameters by periodically collecting water samples for analysis through water quality monitoring points, and acquiring the ecological environment status around the water network through ecological survey data.

3. The digital-twin-based water network model construction method according to claim 1, characterized in that, Step 2 determines the physical water network element information attribute items and the topological relationship between element information, establishes the physical water network information structure, and based on the physical water network information structure, constructs the conceptual model of the water network, including: The physical water network element information attribute items include: channel attribute items, node attribute items, and water flow attribute items; wherein the channel attribute items include: length, width, depth, roughness coefficient, and water quantity-water level relationship curve, the node attribute items include: location, size, artificial node function attribute, and natural node function attribute, and the water flow attribute items include: flow, flow velocity, flow direction, seasonal variation, and flood frequency; The topological relationship between element information includes: connectivity analysis, upstream and downstream relationship determination, and hierarchical relationship construction; Connectivity analysis analyzes the connectivity relationship between channels and channels, and between channels and nodes through GIS; determines the connectivity of the water network through SWAT; determines the upstream and downstream relationship of each element in the water network according to the topographical information and the natural flow direction of the water flow; at the intersection of the river, the upstream and downstream relationship is determined by analyzing the flow direction and flow distribution of the water flow; the hierarchical relationship is constructed by GIS according to the topographical data and the flow direction, and the hierarchical relationship of the water system is further determined in each watershed according to the length, flow, and watershed area parameters of the river, which includes: main river and tributary, primary tributary and secondary tributary; Based on the physical water network information structure, the channel, node, and water flow elements in the physical water network are used as object classes to construct the conceptual model of the water network using object-oriented modeling.

4. The digital-twin-based water network model construction method according to claim 1, characterized in that, Step 3: The method for abstracting the physical water network into an information water network and constructing the data model of the physical water network element information is as follows: The data model of the channel element is constructed by taking the channel centerline as the reference line, dividing the channel into several segments, assuming that the depth and width change linearly within each segment, determining the parameters of each segment based on the integrated existing data resources of the target area, and dynamically adjusting the roughness coefficient according to the channel material and the actual water flow situation, and the adjustment formula is: ； In the formula, represents the adjusted roughness coefficient, represents the initial roughness coefficient; k represents the adjustment coefficient, which is based on the channel material characteristics, combined with the historical data of similar water network engineering and the measured experience, and is first set to an initial range according to the material, and then is calibrated through model simulation and actual flow data comparison, and is obtained according to the flow velocity deviation feedback iteration; represents the actual flow velocity, represents the design flow velocity; The data model of the node element is constructed by taking the node center as the coordinate origin, determining the connection relationship through the coordinates of the connection points of the node and the channel, and the flow distribution model of the node is: ； wherein Qj represents the flow rate of the i-th branch channel, Q represents the total flow rate of the node, hj represents the water head at the inlet of the i-th branch channel, hi represents the water head at the inlet of the j-th branch channel, and m represents the number of branch channels. The data model of the water flow element records the changes of flow, flow velocity, and flow direction with time using time series data, and acquires continuous function expressions through interpolation and fitting of hydrological monitoring data.

5. The digital-twin-based water network model construction method according to claim 1, characterized in that, The method for constructing the physical water network virtual space environment in step 3 is as follows: The terrain information of the area where the water network is located is converted into a three-dimensional terrain model by using geographic information system (GIS) and computer graphics technology, and the terrain information includes but is not limited to elevation, slope and aspect; The terrain is constructed by using digital elevation model (DEM) data, and the terrain topography is generated by interpolating and rendering the DEM data; The physical water network element information is plotted in the three-dimensional terrain model according to its actual geographical position and shape, the three-dimensional entity model is generated according to the center line and width and depth data of the channel, the corresponding geometric shape is constructed according to the position and size information of the node, and the virtual space environment is rendered in combination with the environmental information, so as to obtain the physical water network virtual space environment.

6. The digital-twin-based water network model construction method according to claim 1, characterized in that, The method for cleaning and standardizing the fish school movement trajectory and water flow condition data in step 4 is as follows: Based on the previously collected fish school movement trajectory data, the average speed of each fish in different time periods and the position change amplitude of adjacent time points are calculated, and the speed and position change frequency distribution histogram is drawn by plotting the calculated speed and position change data, and the speed threshold and position change threshold are set; Based on the speed threshold and position change threshold, the real-time collected fish school movement trajectory data is judged, and when the fish school movement speed is greater than the speed threshold and the position change of adjacent time points is greater than the position change threshold, the abnormal data is determined and eliminated; The data is smoothed by a calculation formula, and the smoothing calculation formula is: ； In the formula, denotes the smoothed coordinates of the fish school at time t, and z denotes the moving average window size, and denotes the coordinates of the fish school at the u-th time. For the water flow condition data, the missing value processing is performed for the flow, flow rate and water level data, and if the data is missing at a certain time, the linear interpolation is performed to supplement the data before and after the time.

7. The digital-twin-based water network model construction method according to claim 1, characterized in that, The method for calculating the angle and proportional relationship analysis correlation of the fish school movement and the water flow speed and direction in step 5 is as follows: By calculating the angle and proportional relationship in different regions and time periods, the correlation between the fish school movement and the water flow is analyzed, and if the angle is small and the proportion is high, it indicates that the fish school movement in the region is greatly affected by the water flow, and accordingly it is judged that the water flow in the water network has a significant effect on the fish school behavior in the region.

8. The digital-twin-based water network model construction method according to claim 1, characterized in that, The method for optimizing the physical parameters of the conceptual model of the water network according to the irregular region of the fish school movement trajectory and the fish school movement trajectory at the boundary of the water network in step 6 is as follows: If the fish school movement trajectory frequently turns or gathers at the edge of the channel in a certain region, it indicates that the width of the channel is incorrect, and the width is adjusted by the distribution density of the fish school movement trajectory, and the adjustment formula is: ； In the formula, represents the adjusted channel width, represents the channel width before adjustment; represents an adjustment coefficient, which is based on the original design accuracy of the water network type and the activity characteristics of the target fish population, is initially set to an initial value by referring to the engineering experience of width optimization of similar water networks, has a value range of 0.1-0.5, is verified through small-range measured data, and is iteratively calibrated in combination with the fitting degree of the adjusted channel width and the fish population movement trajectory to obtain the adjustment coefficient; represents the distribution density of the fish population movement trajectory, represents a trajectory distribution density threshold, which is initially set to an initial threshold value based on the historical fish population trajectory density statistics of similar water networks, is combined with the water network type and the fish population habit, and is iteratively calibrated and optimized through a model. The nodes include confluence points and branch points, and the shape and water flow distribution parameters of the nodes are optimized by observing the fish school movement trajectory at the nodes: The branch water flow distribution parameters are respectively and , , The initial values are calculated based on node terrain, channel section data, historical flow meter engineering experience formula, wherein According to the proportion of fish entering two branches respectively and , wherein ; The adjusted water flow distribution parameters are: ； In the formula, denotes a node flow adjustment coefficient, is an adjusted water flow distribution parameter, is a pre-adjusted water flow distribution parameter; the node flow adjustment coefficient is based on the node type, branch channel hydraulic characteristics and the same type of water network node flow optimization experience, and is first set to an initial value range of 0.2-0.6, and then is obtained through model simulation and fitting degree feedback iteration calibration of the actual branch selection proportion of the fish population. The water flow velocity field is adjusted according to the relationship between the fish school movement trajectory and the water flow speed and direction.

9. A digital-twin-based water network model construction system, characterized by, The system for implementing the water network model construction method based on digital twinning according to any one of claims 1-8 comprises a regional data resource integration module, a water network model construction module, an optimized data acquisition and processing module, and a model optimization module; The regional data resource integration module is used to integrate the existing data resources of the target region to provide basic data support for model construction; The water network model construction module is electrically connected with the regional data resource integration module, and is configured to construct a water network model based on the integrated data resource; The optimized data collection and processing module is electrically connected with the water network model construction module, and is configured to collect fish school movement trajectory data and water flow condition data, provide dynamic data sources for model optimization, and clean and standardize the fish school movement trajectory data and the water flow condition data; The model optimization module is electrically connected with the optimized data collection and processing module, and is configured to optimize physical parameters and ecological niche distribution data of the water network model based on the processed fish school movement trajectory data and the water flow condition data.

Images