Drainage basin key node optimization scheduling system based on digital twinborn technology

Abstract

The invention relates to the technical field of optimal scheduling, and discloses a drainage basin key node optimal scheduling system based on a digital twinborn technology. The data acquisition module is used for dividing a target drainage basin into local drainage basins and acquiring comprehensive drainage basin data; the model splicing module is used for splicing the sub-models into a drainage basin simulation model and identifying key nodes; the node identification module is used for identifying abnormal nodes; the mode analysis module is used for identifying the relation between the associated water area and the water direction demand and formulating an optimal scheduling mode; the scheme formulating module is used for formulating an optimal scheduling scheme; according to the method, the discrete sub-models can be spliced into the drainage basin simulation model orderly and smoothly, so that the drainage basin simulation model can achieve an integral virtual-real combination simulation effect on the target drainage basin, and an integral-local-integral virtual-real combination simulation effect of the target drainage basin is achieved; and an accurate and reliable basis is provided for optimization scheduling operation of subsequent key nodes.

Description

Technical Field

[0001] This invention relates to the field of optimization scheduling technology, and more specifically, to a watershed key node optimization scheduling system based on digital twin technology. Background Technology

[0002] A watershed is the core geographical unit of water cycle and water resource management. Water conservancy projects such as reservoirs, dams, and pumping stations within it serve as key nodes, together forming a complex water resource regulation system. Scientific and precise analysis, identification, and optimized scheduling of these key nodes through digital twin technology are crucial to ensuring the efficient use of water resources in the watershed.

[0003] The patent application with publication number CN120562783A discloses a method for dynamic scheduling optimization and simulation of watershed water resources based on digital twins. The method includes: acquiring at least one water resource and its available scheduling quantity within the watershed, and constructing a scheduling network for the water resources within the watershed; acquiring at least one demand area within the watershed, and simulating the regional water demand and water usage of that demand area; simulating the water loss coefficient based on the environmental characteristics and rainfall / evaporation conditions of the water resource's location; acquiring the current scheduling scheme, simulating at least one point to be optimized within the current scheduling scheme, and forming at least one preliminary optimization scheme; calculating the demand satisfaction, scheduling excess, scheduling completion time, and optimization coefficient of the preliminary optimization scheme; selecting the preliminary optimization scheme with the largest optimization coefficient as the scheduling optimization scheme, simulating the scheduling optimization scheme, and obtaining the actual scheduling result after optimization.

[0004] Existing optimization scheduling systems typically employ a holistic modeling approach for the target watershed when conducting simulation analysis using virtual models. This holistic modeling approach faces problems such as excessive data acquisition burden and huge computational resource requirements, resulting in low model accuracy and difficulty in performing detailed simulations of key areas at multiple levels. Consequently, it fails to achieve a combined virtual and real simulation effect of the target watershed from the whole to the part and then from the part to the whole, reducing the accuracy and reliability of optimization scheduling operations for key nodes.

[0005] In view of this, the present invention proposes a watershed key node optimization scheduling system based on digital twin technology to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a watershed key node optimization scheduling system based on digital twin technology, applied to a watershed management platform, comprising:

[0007] The data acquisition module is used to draw watershed dividing lines on the hydrological plan of the target watershed, divide the target watershed into local watersheds based on the watershed dividing lines, and collect comprehensive watershed data of the local watersheds, which includes hydrological data, topographic data and meteorological data.

[0008] The model stitching module is used to combine comprehensive watershed data with digital twin technology to construct multi-level sub-models. Based on the model stitching criteria, the sub-models are stitched together in sequence to form a watershed simulation model, and key nodes are identified from the watershed simulation model.

[0009] The node identification module is used to collect time-limit parameters of the target watershed, determine the monitoring cycle of the watershed simulation model, simulate the node status of key nodes in the next monitoring cycle, and identify abnormal nodes from the key nodes based on the node status.

[0010] The pattern analysis module is used to identify the associated waters of abnormal nodes in the target watershed, determine the water direction demand relationship between the associated waters and the abnormal nodes, and formulate an optimized scheduling mode for the abnormal nodes based on the water direction demand relationship.

[0011] The scheme formulation module is used to combine the comprehensive learning data of the abnormal nodes in the next monitoring cycle, input the comprehensive learning data into the scheme learning model, and formulate the optimized scheduling scheme for the abnormal nodes.

[0012] Furthermore, the method for dividing the target watershed is as follows:

[0013] The hydrological plan map of the target watershed is retrieved by querying the geographic information system, highlighting the boundary line of the hydrological plan map, and identifying the main water area and A branch water areas in the hydrological plan map;

[0014] Mark the intersection points of the main water area and A branch water areas one by one. Using the intersection points as the dividing benchmark, draw local auxiliary lines passing through the intersection points on the hydrological plan and divide the main water area into continuous sub-water areas through the local auxiliary lines.

[0015] Using a branch water area and a sub-water area as the standard, the local auxiliary lines are extended to both sides to the boundary line of the hydrological plan, thus generating the watershed dividing line;

[0016] The target watershed located between two adjacent watershed dividing lines, and the target watershed located between the watershed dividing line and the boundary line of the adjacent location, are denoted as local watersheds, resulting in B local watersheds.

[0017] Furthermore, the sub-model is constructed as follows:

[0018] A basic model with three inner and outer layers is constructed using digital twin technology, and the three layers are respectively named the data layer, model layer and functional layer in an outward manner.

[0019] Three circularly distributed data units are set up on the data layer. Hydrological data, topographic data and meteorological data are imported into the three data units one by one to generate hydrological units, topographic units and meteorological units.

[0020] The basic terrain of the local watershed is simulated in the model layer, and the basic terrain is rendered and colored using rendering technology. Attribute information and logical relationships are added to the basic terrain.

[0021] In the functional layer, model constraints are set, and model permissions are configured for the model constraints, which causes the basic model to be converted into sub-models, resulting in B sub-models.

[0022] Furthermore, the model splicing criterion is: to splice the splicing surfaces at the same level sequentially from the inside out.

[0023] Furthermore, the method for stitching together the watershed simulation model is as follows:

[0024] A1: Based on the water flow direction of the main water area, the sub-models of the B local watersheds are arranged in order to generate a model queue;

[0025] A2: The first sub-model in the current model queue is designated as the reference model, the first sub-model in the remaining model queue is designated as the superimposed model, and the superimposed model is rotated and tilted until the superimposed model and the reference model are aligned. The contact interface between the superimposed model and the reference model is designated as the splicing surface.

[0026] A3: Import the aligned superimposed model and the reference model into the 3D space, and mark C model points with the same spatial coordinates on the functional layer, model layer and data layer of the superimposed model and the reference model respectively, and denot them as inner alignment point, middle alignment point and outer alignment point;

[0027] A4: Spatially align the C inner alignment points, C middle alignment points, and C outer alignment points on the two splicing surfaces sequentially to generate a local splicing model;

[0028] A5: Import the partially spliced ​​model into the model queue, and repeat steps A2-A4 until all sub-models are spliced. Then, synchronize the model constraints and model permissions on the last partially spliced ​​model to splice the watershed simulation model.

[0029] Furthermore, when identifying key nodes, the intersection of the main waterway and the branch waterway in the watershed simulation model is used as the base point. After moving the base point one standard length in the direction of the main waterway, the location of the moved base point is recorded as the key node, thus obtaining D key nodes.

[0030] Furthermore, the parameters include watershed monitoring duration, standard monitoring duration, and water area anomaly duration;

[0031] The method for determining the monitoring cycle is as follows:

[0032] The difference between the monitoring duration of D watersheds at B key nodes and the standard monitoring duration is calculated. The absolute values ​​of the D differences are summed and averaged, and then compared with the standard monitoring duration to calculate the proportion of B exceeding the limit.

[0033] Remove the maximum and minimum values ​​of the abnormal duration of the water area at each of the B key nodes, and then... The average of the duration of each abnormal water area is calculated by summing the durations of the abnormal water areas;

[0034] Based on the average abnormal value and the proportion of exceeding limits, the duration of the test for B key nodes is calculated. The duration of the test for B nodes is accumulated and averaged to calculate the interval duration. Using the interval duration as the standard and the current time as the end point, the historical time period is divided into F consecutive monitoring cycles.

[0035] Furthermore, node states include normal simulation states and abnormal simulation states; the simulation method for node states is as follows:

[0036] Configure simulation parameters that meet the model constraints on the watershed simulation model, and assign simulation permissions consistent with the model permissions;

[0037] Based on the water flow of B key nodes in the watershed simulation model during F monitoring cycles, the water flow of B key nodes in the next monitoring cycle is simulated respectively. The difference between the water flow of B key nodes in the next monitoring cycle and the water flow of the previous monitoring cycle is calculated.

[0038] When the absolute value of the flow difference is greater than the standard flow threshold, the node state is recorded as an abnormal simulation state; when the absolute value of the flow difference is less than or equal to the standard flow threshold, the node state is recorded as a normal simulation state.

[0039] Furthermore, the associated waters include both the inflow waters and the outflow waters;

[0040] The main waterway and branch waterway located between any two adjacent key nodes are denoted as the inter-point waterway, the inter-point waterway located upstream of the abnormal node is denoted as the inlet waterway, and the inter-point waterway located downstream of the abnormal node is denoted as the outlet waterway.

[0041] Water demand relationships include water shortage diversion relationships and saturation diversion relationships; the method for determining water demand relationships is as follows:

[0042] When the water flow difference of the abnormal node is positive, the water demand relationship is a saturated draw-out relationship; when the water flow difference of the abnormal node is negative, the water demand relationship is a water shortage draw-out relationship.

[0043] Furthermore, the optimized scheduling modes include node introduction mode and node deduction mode; the method for formulating the optimized scheduling mode is as follows:

[0044] When the water demand relationship of an abnormal node is a saturated outflow relationship, a node outflow mode is determined; when the water demand relationship of an abnormal node is a water shortage outflow relationship, a node inflow mode is determined.

[0045] The water flow, water flow difference, associated water areas, water direction demand relationship, and optimized scheduling mode of the abnormal nodes in the next monitoring cycle are summarized one by one to generate comprehensive learning data.

[0046] The technical advantages of this invention, a watershed key node optimization scheduling system based on digital twin technology, are as follows:

[0047] (1): This invention divides the overall target watershed into fragmented local watersheds and constructs virtual models corresponding to the local watersheds. This can achieve the effect of data collection and model construction of the target watershed by breaking it down into parts, thereby reducing the burden of data collection. At the same time, combined with the model splicing criteria, the discrete sub-models can be spliced ​​into a watershed simulation model in an orderly and smooth manner, so that the watershed simulation model can play an overall virtual-real simulation effect on the target watershed. This achieves the virtual-real simulation effect of the overall-local-overall of the target watershed, providing an accurate and reliable foundation for the subsequent optimization and scheduling operations of key nodes.

[0048] (2) By determining the duration of the monitoring period, the present invention enables the watershed simulation model to accurately simulate the water flow changes of key nodes in the target watershed during the monitoring period in advance, thereby realizing the effective simulation conversion between the physical dimension of the target watershed and the virtual dimension of the simulation model. This allows for effective simulation analysis of the dynamic changes in water flow in the target watershed in the future, providing a reliable data foundation for the optimized scheduling of key nodes in the target watershed and improving the dynamic balance of water resources in the target watershed. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the architecture of a watershed key node optimization scheduling system based on digital twin technology, provided in Embodiment 1 of the present invention.

[0050] Figure 2This is a flowchart illustrating a method for optimizing the scheduling of key nodes in a watershed based on digital twin technology, as provided in Embodiment 2 of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1: Please refer to Figure 1 As shown in this embodiment, a watershed key node optimization scheduling system based on digital twin technology is applied to a watershed management platform and includes:

[0053] The data acquisition module draws watershed dividing lines in the target watershed, divides the target watershed into local watersheds, and collects comprehensive watershed data from the local watersheds.

[0054] The target watershed refers to the river basin through which the water resources undergoing this critical node optimization and scheduling operation flow, and also serves as the largest water area involved in subsequent critical node marking and optimization scheduling;

[0055] In this embodiment, the target watershed is not a single body of water, but a combination of multiple waterways that intersect with each other. In this case, the target watershed will contain more than one confluence of waterways, sluice gates, pumping stations, and other nodes that can promote the flow of water resources in the target watershed.

[0056] Since the target watershed covers a large area and contains a large amount of relevant data, collecting relevant data from the target watershed in a holistic manner would increase the burden of data collection and reduce the accuracy of data collection. Therefore, it is necessary to divide the target watershed into multiple independent and small local watersheds to achieve the effect of data collection from the whole watershed to the parts.

[0057] Specifically, the method for dividing the target watershed is as follows:

[0058] The hydrological plan map of the target watershed is retrieved by querying the geographic information system, highlighting the boundary line of the hydrological plan map, and identifying the main water area and A branch water areas in the hydrological plan map;

[0059] Mark the intersection points of the main water area and each of the A branch water areas one by one. Using the intersection points as the dividing reference, draw local auxiliary lines passing through the intersection points on the hydrological plan and divide the main water area into continuous sub-water areas through the local auxiliary lines. The local auxiliary lines here are relatively short and will not touch any boundary lines or local auxiliary lines at other locations.

[0060] Using a branch waterway and a sub-waterway as the standard, local auxiliary lines are extended to both sides to the boundary line of the hydrological plan, generating watershed dividing lines. The direction and standard of each local auxiliary line extending to both sides are not consistent, but the underlying logic is consistent, that is, using a branch waterway and a sub-waterway as the standard; thus ensuring that each two adjacent watershed dividing lines contain a branch waterway and a sub-waterway.

[0061] The hydrological plan view located between two adjacent watershed dividing lines, and between the watershed dividing line and the boundary line of the adjacent location, is recorded as a local map, and the target watershed located in the local map is recorded as a local watershed, thus obtaining B local watersheds.

[0062] It should be noted that the area and shape of each local watershed may be different, so each local watershed can be represented separately for the water area it contains. However, when all local watersheds are combined in sequence, a complete target watershed can be formed.

[0063] Comprehensive watershed data is used to represent the specific water resource flow in the main and tributary waters of a local watershed, enabling comprehensive watershed data to serve as a prerequisite for subsequent optimized scheduling of water resources in the target watershed.

[0064] Specifically, integrated watershed data includes hydrological data, topographic data, and meteorological data;

[0065] Hydrological data is used to represent the water resource flow, direction and other related data of the main and tributary waters in a local watershed, and can be used as data to represent the water resources themselves in the local watershed.

[0066] In this embodiment, hydrological data includes, but is not limited to, water flow direction, water resource flow rate, and water flow velocity; hydrological data is acquired after detection by pressure sensors and flow velocity sensors.

[0067] Topographic data refers to data such as the depth and curvature of the main and tributary waterways in a local watershed, which can be used as data representing the secondary dimensions of waterways in a local watershed.

[0068] In this embodiment, the terrain data includes, but is not limited to, riverbed depth, river width, and river curvature. When collecting terrain data, it is necessary to first mark multiple terrain points in the local watershed at preset intervals, and then obtain the data by real-time detection of each terrain point using drone aerial photography equipment and depth sensors in the water area.

[0069] Meteorological data is used to represent the changes in the main and tributary waterways in a local watershed under rainfall conditions; it can be used as data to represent the water environment dimension of a local watershed.

[0070] In this embodiment, meteorological data includes, but is not limited to, rainfall amount and rainfall rate. When collecting meteorological data, it is necessary to first determine the collection time of the local watershed, so that data such as rainfall amount and rainfall rate can be collected at different collection times. For example, the collection times are 4:00, 8:00, 12:00, 16:00, 20:00 and 24:00 every day.

[0071] The model stitching module uses comprehensive watershed data as a benchmark and combines digital twin technology to construct sub-models of local watersheds. Based on the model stitching criteria, the sub-models are stitched together into a watershed simulation model, and key nodes are identified from the watershed simulation model.

[0072] After obtaining the comprehensive watershed data, each comprehensive watershed data can only represent the comprehensive situation in each local watershed at the physical level. In order to facilitate virtual simulation and other operations on local watersheds, it is necessary to combine the comprehensive watershed data with digital twin technology to construct a sub-model that can represent local watersheds at the virtual digital level.

[0073] In this embodiment, each local watershed corresponds to a sub-model, and each sub-model has specific integrated watershed data. Due to the positional continuity and data independence between two adjacent local watersheds, the positional continuity and data independence are also maintained between two adjacent sub-models.

[0074] Specifically, the method for constructing sub-models is as follows:

[0075] A basic model with three inner and outer layers is constructed using digital twin technology, and the three layers are respectively named the data layer, model layer and functional layer in an outward manner.

[0076] Three ring-shaped data units are set up on the data layer of the basic model. Hydrological data, topographic data and meteorological data are imported into the three data units one by one, so that the three data units are converted into hydrological units, topographic units and meteorological units respectively.

[0077] The basic terrain of the local watershed is simulated in the model layer of the basic model. The basic terrain is rendered and colored using rendering technology, and attribute information and logical relationships are added to the basic terrain.

[0078] In the functional layer of the basic model, model constraints are set, and model permissions are configured for these constraints. This causes the basic model to be converted into sub-models, resulting in B sub-models. Model constraints are parameters used to control and constrain the logic level of the sub-models, preventing negative phenomena such as confusion and errors at the logic level. Simulation permissions are the maximum simulation limit for the sub-model during simulation, preventing the sub-model from simulating beyond its permissions.

[0079] It should be noted that attribute information is used to represent information at the level of the geographical topography in the local watershed, such as the design water level and design flow velocity, while logical relationships are used to represent information at the level of the geographical topography in the local watershed, such as the direction of water confluence and the relationship between gates and water level opening. This ensures that the constructed sub-model can effectively match the actual topography and water conditions of the local watershed, thereby improving the accuracy of the virtual simulation of the sub-model.

[0080] A watershed simulation model is a model used to simulate the overall situation of water flow and topography in a target watershed, making it a direct analytical object for subsequent optimization and scheduling of key nodes in the target watershed.

[0081] When splicing sub-models into a watershed simulation model, it is necessary to do so under the constraints of model splicing criteria to ensure the orderly splicing of sub-models and improve the accuracy and reliability of the splicing results of the watershed simulation model.

[0082] Specifically, the model splicing principle is as follows: splice the splicing surfaces at the same level in sequence from the inside out; this ensures that two adjacent sub-models maintain the splicing order from the inside out when splicing the models, and also ensures that the splicing surfaces of the sub-models remain seamless and smooth, thereby improving the smooth transition effect of the splicing surfaces of two adjacent sub-models.

[0083] The method for stitching together the watershed simulation model is as follows:

[0084] A1: Based on the water flow direction of the main water area, the sub-models of the B local watersheds are arranged in order to generate a model queue;

[0085] A2: The first sub-model in the current model queue is designated as the reference model, the first sub-model in the remaining model queue is designated as the superimposed model, and the superimposed model is rotated and tilted until the superimposed model and the reference model are aligned. The contact interface between the superimposed model and the reference model is designated as the splicing surface.

[0086] A3: Import the aligned superimposed model and the reference model into the 3D space, and mark C model points with the same spatial coordinates on the functional layer, model layer and data layer of the superimposed model and the reference model respectively, and denot them as inner alignment point, middle alignment point and outer alignment point;

[0087] A4: First, spatially align the C inner alignment points on the splicing surface of the overlay model and the reference model. Then, spatially align the C middle alignment points on the splicing surface of the overlay model and the reference model. Finally, spatially align the C outer alignment points on the splicing surface of the overlay model and the reference model to generate a local splicing model.

[0088] A5: Import the partially spliced ​​model into the model queue, and repeat steps A2-A4 until all sub-models are spliced. Then, synchronize the model constraints and model permissions on the last partially spliced ​​model to splice the watershed simulation model.

[0089] Key nodes refer to specific objects in the watershed simulation model that can serve as locations and facilities for water flow monitoring and optimized scheduling, so that relevant water flow data at key nodes can affect subsequent optimized scheduling.

[0090] Specifically, when identifying critical nodes, the intersection of the main waterway and the tributary waterway in the watershed simulation model is used as the base point. After moving the base point a standard length in the direction of the main waterway flow, the location of the moved base point is recorded as the critical node, resulting in D critical nodes. The standard length is a pre-set value used to represent the distance between the critical node and the intersection point, ensuring that the critical node maintains a reasonable distance from the intersection point.

[0091] In this embodiment, key nodes need to be identified and determined from the watershed simulation model. The number of key nodes may not be unique, and the positions between any two key nodes may also be different, so as to monitor the water flow of the entire watershed of the target watershed from multiple points.

[0092] The node identification module collects time-limit parameters in the target watershed, determines the monitoring cycle of the watershed simulation model, simulates the node status of key nodes in the next monitoring cycle, and identifies abnormal nodes from the key nodes.

[0093] The time limit parameter refers to the comprehensive parameter contained in the target watershed that can affect the simulation time of two adjacent simulations of the watershed simulation model. In other words, the time limit parameter can be used to determine the time interval between two adjacent simulations of key nodes of the watershed simulation model.

[0094] In this embodiment, the implemented parameters include watershed monitoring duration, standard monitoring duration, and water area anomaly duration.

[0095] Watershed monitoring duration refers to the actual duration of each water resource monitoring session at key nodes in a target watershed over historical periods.

[0096] Standard monitoring duration refers to the standard duration span of each water resource monitoring session at key nodes in the target watershed over historical periods.

[0097] Abnormal water level duration refers to the duration span of each instance of excessively high water levels at key nodes in the target watershed during historical periods. In this embodiment, the watershed monitoring duration, standard monitoring duration, and abnormal water level duration are all obtained by querying the watershed management platform.

[0098] After collecting the time limit parameters, each monitoring cycle in the watershed simulation model can be determined based on the time limit parameters, so that the monitoring cycle can serve as the time period for evaluating the real-time status of key nodes.

[0099] Specifically, the method for determining the monitoring cycle is as follows:

[0100] The monitoring duration of each of the B key nodes and the D watershed monitoring durations are subtracted from the standard monitoring duration. The absolute values ​​of the D differences are summed and averaged, and then compared with the standard monitoring duration to calculate the proportion of the B exceedances.

[0101] The formula for calculating the percentage of over-limit is:

[0102] ;

[0103] In the formula, For the proportion exceeding the limit, For the first Monitoring duration for each river basin Standard monitoring duration;

[0104] Remove the maximum and minimum values ​​of the abnormal duration of the water area at each of the B key nodes, and then... The average of the duration of each abnormal water area is calculated by summing the durations of the abnormal water areas;

[0105] The formula for calculating the outlier mean is:

[0106] ;

[0107] In the formula, This is an outlier mean. For the first The duration of abnormal water conditions;

[0108] Based on the average abnormal value and the proportion of exceeding limits, the duration of the test for B key nodes is calculated. The duration of the test for B nodes is accumulated and averaged to calculate the interval duration. Using the interval duration as the standard and the current time as the end point, the historical time period is divided into F consecutive monitoring cycles.

[0109] The formula for calculating the interval duration is:

[0110] ;

[0111] In the formula, For interval duration, For the first The average anomalies of key nodes, For the first The percentage of key nodes exceeding the limit.

[0112] Once the monitoring period is obtained, the state of key nodes in the next monitoring period can be evaluated based on the actual water flow of the target basin within the historical time period. This allows for advance simulation and prediction of water flow changes at key nodes in the future, and the node state is used as the simulation result for key nodes.

[0113] In this embodiment, the node status includes normal simulation status and abnormal simulation status; different simulation statuses indicate that the key node corresponds to different simulation results in the next monitoring cycle. When the node status is normal simulation status, it means that the water flow change of the key node simulated by the watershed simulation model in the next monitoring cycle remains normal; otherwise, it is the opposite.

[0114] It should be noted that the next monitoring cycle refers to the first monitoring cycle after the current monitoring cycle, making the next monitoring cycle a time period that starts from the current time and ends at a time after an interval.

[0115] Specifically, when simulating the node status of key nodes in the next monitoring cycle, firstly, simulation parameters that meet the model constraints are configured on the watershed simulation model, and simulation permissions consistent with the model permissions are granted; then, based on the water flow of B key nodes in the watershed simulation model over F monitoring cycles, the water flow of B key nodes in the next monitoring cycle is simulated respectively; next, the water flow difference is calculated by subtracting the water flow of B key nodes in the next monitoring cycle from the water flow of the previous monitoring cycle; finally, when the absolute value of the water flow difference is greater than the standard water flow threshold, the node status of the key node is recorded as an abnormal simulation state; when the absolute value of the water flow difference is less than or equal to the standard water flow threshold, the node status of the key node is recorded as a normal simulation state.

[0116] In this embodiment, the standard water flow threshold refers to the maximum value of the water flow difference when the node state of the critical node is recorded as an abnormal simulation state, thus serving as a direct numerical basis for judging the node state of the critical node in the next monitoring cycle.

[0117] An abnormal node is a key node whose node state is in an abnormal simulation state, and it is used as the object for subsequent water resource optimization scheduling. In this embodiment, the key node whose node state is in an abnormal simulation state is recorded as an abnormal node, and H abnormal nodes are obtained.

[0118] It should be noted that the number of abnormal nodes may be one or more, and the maximum number of abnormal nodes is B. Abnormal nodes may be in any state of being adjacent or not adjacent.

[0119] The pattern analysis module identifies the associated waters of abnormal nodes in the target watershed, determines the water direction demand relationship between the associated waters and the abnormal nodes one by one, and formulates an optimized scheduling mode for the abnormal nodes based on the water direction demand relationship.

[0120] Associated water areas refer to the main water areas and tributary water areas in the target water basin that are geographically connected to the abnormal node. This allows associated water areas to serve as the direct objects for optimized scheduling operations of the abnormal node, thereby enabling the water resources water basins where water flow is introduced and drawn out by the abnormal node in the next monitoring cycle.

[0121] In this embodiment, the associated water area includes the inflow water area and the outflow water area; the inflow water area refers to the water area located upstream of the abnormal node, where water resources flow towards the abnormal node, and the outflow water area refers to the water area located downstream of the abnormal node, where water resources flow out of the abnormal node.

[0122] Specifically, when identifying associated waterways, the locations of D key nodes are marked one by one in the target watershed. The main waterway and branch waterway located between any two adjacent key nodes are recorded as inter-point waterways. The inter-point waterways located upstream of the abnormal node are recorded as inflow waterways, and the inter-point waterways located downstream of the abnormal node are recorded as outflow waterways.

[0123] In this embodiment, when two abnormal nodes are geographically adjacent, the associated water areas corresponding to the two abnormal nodes will overlap. Therefore, the associated water area can be used as the water area for water flow optimization scheduling of two adjacent abnormal nodes at the same time.

[0124] The water demand relationship is used to represent whether the water resource flow between associated water areas and abnormal nodes meets the water flow demand.

[0125] Specifically, the water demand relationship includes the water shortage diversion relationship and the saturation diversion relationship; the water shortage diversion relationship means that the abnormal node will experience water shortage in the next monitoring cycle, and the associated water area needs to increase the water flow to the abnormal node; the saturation diversion relationship means that the abnormal node will experience saturation in the next monitoring cycle, and the associated water area needs to reduce the water flow to the abnormal node.

[0126] The method for determining the water demand relationship is as follows:

[0127] The positive and negative relationships of the water flow differences at H abnormal nodes were identified one by one;

[0128] When the water flow difference of the abnormal node is positive, it means that the water flow of the abnormal node is large in the next monitoring cycle, and the water flow of the abnormal node needs to be reduced. At this time, the water demand relationship between the associated water area and the abnormal node is a saturated output relationship.

[0129] When the water flow difference of an abnormal node is negative, it indicates that the water flow of the abnormal node is small in the next monitoring cycle, and the water flow of the abnormal node needs to be increased. At this time, the water demand relationship between the associated water area and the abnormal node is a water shortage diversion relationship.

[0130] It should be noted that, since the abnormal node is judged based on the absolute value of the water flow difference, and an abnormal node will only appear if the absolute value of the water flow difference is greater than 0, the water flow difference of the abnormal node will not be equal to 0.

[0131] Optimized scheduling mode refers to the working mode of optimizing water flow scheduling for abnormal nodes. Optimized scheduling mode includes node introduction mode and node deduction mode.

[0132] Specifically, the method for formulating an optimized scheduling mode is as follows:

[0133] When the water demand relationship of an abnormal node is a saturated outflow relationship, the abnormal node needs to reduce the water flow. At this time, a node outflow mode is determined.

[0134] When the water demand relationship of an abnormal node is a water shortage diversion relationship, the abnormal node needs to increase the water flow, and a node introduction mode is then formulated.

[0135] The scheme formulation module combines the comprehensive learning data of the abnormal node in the next monitoring cycle and inputs the comprehensive learning data into the scheme learning model to formulate an optimized scheduling scheme for the abnormal node.

[0136] Comprehensive learning data refers to comprehensive data that can influence the optimization and scheduling of water flow at abnormal nodes in the next monitoring cycle. It comes from the combination of different parameter data of the aforementioned abnormal nodes.

[0137] In this embodiment, when combining comprehensive learning data, the water flow, water flow difference, associated water areas, water direction demand relationship and optimized scheduling mode of H abnormal nodes in the next monitoring period are summarized one by one to generate H comprehensive learning data.

[0138] The scheme learning model is based on reinforcement learning technology. It can automatically identify and formulate water flow optimization scheduling schemes for abnormal nodes based on comprehensive learning data, so that the scheme learning model can generate optimization scheduling schemes that best match the actual situation of abnormal nodes.

[0139] In this embodiment, the scheme learning model is not obtained directly. It requires training data from a large amount of historical comprehensive learning data of different types and corresponding optimized scheduling schemes. The model is obtained through repeated and multiple iterations of optimization training to ensure that the scheme learning model can meet the requirements for the automatic formulation of optimized scheduling schemes for subsequent abnormal nodes. The specific training process of the scheme learning model is an existing technology in this field and will not be described in detail here.

[0140] In this embodiment, the optimized scheduling scheme includes specific measures for optimizing the scheduling of water flow in the main waterway and branch waterway where the abnormal node is located in the target watershed. These specific measures include, but are not limited to, introducing water resources from the inlet waterway into the waterway where the abnormal node is located, and diverting water resources from the waterway where the abnormal node is located to the outlet waterway.

[0141] When carrying out water resource optimization and scheduling operations in the inflow and outflow water areas, it is necessary to use facilities such as pumping stations and gates located in each inflow and outflow water area to achieve dynamic and proactive optimization and scheduling of water resources in the target watershed. This avoids water shortages or saturation in some water areas of the target watershed in the future, achieving a dynamic balance of water resources in the target watershed, and thus improving the rationality, reliability, and stability of water resource optimization and scheduling in the target watershed.

[0142] Example 2: Please refer to Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. This embodiment provides a method for optimizing the scheduling of key nodes in a watershed based on digital twin technology, applied to a watershed management platform. It is implemented using a watershed key node optimization scheduling system based on digital twin technology, including:

[0143] S01: Draw the watershed dividing line on the hydrological plan of the target watershed, and divide the target watershed into local watersheds based on the watershed dividing line, and collect comprehensive watershed data of the local watersheds;

[0144] S02: By combining comprehensive watershed data with digital twin technology, a multi-level sub-model is constructed. Based on the model splicing criteria, the sub-models are sequentially spliced ​​into a watershed simulation model, and key nodes are identified from the watershed simulation model.

[0145] S03: Collect time-limit parameters of the target watershed, determine the monitoring cycle of the watershed simulation model, simulate the node status of key nodes in the next monitoring cycle, and identify abnormal nodes from the key nodes based on the node status.

[0146] S04: Identify the associated waters of the abnormal nodes in the target watershed, determine the water direction demand relationship between the associated waters and the abnormal nodes, and formulate an optimized scheduling mode for the abnormal nodes based on the water direction demand relationship.

[0147] S05: Combine the comprehensive learning data of the abnormal node in the next monitoring cycle, input the comprehensive learning data into the scheme learning model, and formulate an optimized scheduling scheme for the abnormal node.

[0148] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A watershed key node optimization scheduling system based on digital twin technology, applied to a watershed management platform, characterized in that, include: The data acquisition module is used to draw watershed dividing lines on the hydrological plan of the target watershed, divide the target watershed into local watersheds based on the watershed dividing lines, and collect comprehensive watershed data of the local watersheds, which includes hydrological data, topographic data and meteorological data. The model stitching module is used to combine comprehensive watershed data with digital twin technology to construct multi-level sub-models. Based on the model stitching criteria, the sub-models are stitched together in sequence to form a watershed simulation model, and key nodes are identified from the watershed simulation model. The node identification module is used to collect time-limit parameters of the target watershed, determine the monitoring cycle of the watershed simulation model, simulate the node status of key nodes in the next monitoring cycle, and identify abnormal nodes from the key nodes based on the node status. The pattern analysis module is used to identify the associated waters of abnormal nodes in the target watershed, determine the water direction demand relationship between the associated waters and the abnormal nodes, and formulate an optimized scheduling mode for the abnormal nodes based on the water direction demand relationship. The scheme formulation module is used to combine the comprehensive learning data of the abnormal nodes in the next monitoring cycle, input the comprehensive learning data into the scheme learning model, and formulate the optimized scheduling scheme for the abnormal nodes.

2. The watershed key node optimization scheduling system based on digital twin technology according to claim 1, characterized in that, The method for dividing the target watershed is as follows: The hydrological plan map of the target watershed is retrieved by querying the geographic information system, highlighting the boundary line of the hydrological plan map, and identifying the main water area and A branch water areas in the hydrological plan map; Mark the intersection points of the main water area and A branch water areas one by one. Using the intersection points as the dividing benchmark, draw local auxiliary lines passing through the intersection points on the hydrological plan and divide the main water area into continuous sub-water areas through the local auxiliary lines. Using a branch water area and a sub-water area as the standard, the local auxiliary lines are extended to both sides to the boundary line of the hydrological plan, thus generating the watershed dividing line; The target watershed located between two adjacent watershed dividing lines, and the target watershed located between the watershed dividing line and the boundary line of the adjacent location, are denoted as local watersheds, resulting in B local watersheds.

3. A watershed key node optimization scheduling system based on digital twin technology according to claim 2, characterized in that, The sub-model is constructed as follows: A basic model with three inner and outer layers is constructed using digital twin technology, and the three layers are respectively named the data layer, model layer and functional layer in an outward manner. Three circularly distributed data units are set up on the data layer. Hydrological data, topographic data and meteorological data are imported into the three data units one by one to generate hydrological units, topographic units and meteorological units. The basic terrain of the local watershed is simulated in the model layer, and the basic terrain is rendered and colored using rendering technology. Attribute information and logical relationships are added to the basic terrain. In the functional layer, model constraints are set, and model permissions are configured for the model constraints, which causes the basic model to be converted into sub-models, resulting in B sub-models.

4. A watershed key node optimization scheduling system based on digital twin technology according to claim 3, characterized in that, The model splicing principle is to splice the splicing surfaces at the same level sequentially from the inside out.

5. A watershed key node optimization scheduling system based on digital twin technology according to claim 4, characterized in that, The method for stitching together the watershed simulation model is as follows: A1: Based on the water flow direction of the main water area, the sub-models of the B local watersheds are arranged in order to generate a model queue; A2: The first sub-model in the current model queue is designated as the reference model, the first sub-model in the remaining model queue is designated as the superimposed model, and the superimposed model is rotated and tilted until the superimposed model and the reference model are aligned. The contact interface between the superimposed model and the reference model is designated as the splicing surface. A3: Import the aligned superimposed model and the reference model into the 3D space, and mark C model points with the same spatial coordinates on the functional layer, model layer and data layer of the superimposed model and the reference model respectively, and denot them as inner alignment point, middle alignment point and outer alignment point; A4: Spatially align the C inner alignment points, C middle alignment points, and C outer alignment points on the two splicing surfaces sequentially to generate a local splicing model; A5: Import the partially spliced ​​model into the model queue, and repeat steps A2-A4 until all sub-models are spliced. Then, synchronize the model constraints and model permissions on the last partially spliced ​​model to splice the watershed simulation model.

6. A watershed key node optimization scheduling system based on digital twin technology according to claim 5, characterized in that, When identifying key nodes, the intersection of the main waterway and the branch waterway in the watershed simulation model is used as the base point. The base point is moved one standard length in the direction of the main waterway flow, and the location of the moved base point is recorded as the key node, thus obtaining D key nodes.

7. A watershed key node optimization scheduling system based on digital twin technology according to claim 6, characterized in that, The parameters to be implemented include watershed monitoring duration, standard monitoring duration, and water area anomaly duration; The method for determining the monitoring cycle is as follows: The difference between the monitoring duration of D watersheds at B key nodes and the standard monitoring duration is calculated. The absolute values ​​of the D differences are summed and averaged, and then compared with the standard monitoring duration to calculate the proportion of B exceeding the limit. Remove the maximum and minimum values ​​of the abnormal duration of the water area at each of the B key nodes, and then... The average of the duration of each abnormal water area is calculated by summing the durations of the abnormal water areas; Based on the average abnormal value and the proportion of exceeding limits, the duration of the test for B key nodes is calculated. The duration of the test for B nodes is accumulated and averaged to calculate the interval duration. Using the interval duration as the standard and the current time as the end point, the historical time period is divided into F consecutive monitoring cycles.

8. A watershed key node optimization scheduling system based on digital twin technology according to claim 7, characterized in that, Node states include normal simulation states and abnormal simulation states; the simulation method for node states is as follows: Configure simulation parameters that meet the model constraints on the watershed simulation model, and assign simulation permissions consistent with the model permissions; Based on the water flow of B key nodes in the watershed simulation model during F monitoring cycles, the water flow of B key nodes in the next monitoring cycle is simulated respectively. The difference between the water flow of B key nodes in the next monitoring cycle and the water flow of the previous monitoring cycle is calculated. When the absolute value of the flow difference is greater than the standard flow threshold, the node state is recorded as an abnormal simulation state; when the absolute value of the flow difference is less than or equal to the standard flow threshold, the node state is recorded as a normal simulation state.

9. A watershed key node optimization scheduling system based on digital twin technology according to claim 8, characterized in that, Related waters include inflow waters and outflow waters; The main waterway and branch waterway located between any two adjacent key nodes are denoted as the inter-point waterway, the inter-point waterway located upstream of the abnormal node is denoted as the inlet waterway, and the inter-point waterway located downstream of the abnormal node is denoted as the outlet waterway. Water demand relationships include water shortage diversion relationships and saturation diversion relationships; the method for determining water demand relationships is as follows: When the water flow difference of an abnormal node is positive, the water demand relationship is a saturated draw-out relationship; when the water flow difference of an abnormal node is negative, the water demand relationship is a water shortage draw-out relationship.

10. A watershed key node optimization scheduling system based on digital twin technology according to claim 9, wherein the optimization scheduling mode includes a node introduction mode and a node deduction mode; the method for formulating the optimization scheduling mode is as follows: When the water demand relationship of an abnormal node is a saturated outflow relationship, a node outflow mode is determined; when the water demand relationship of an abnormal node is a water shortage outflow relationship, a node inflow mode is determined. The water flow, water flow difference, associated water areas, water direction demand relationship, and optimized scheduling mode of the abnormal nodes in the next monitoring cycle are summarized one by one to generate comprehensive learning data.

Images