A water conservancy multi-model coupling parallel architecture design method

Abstract

The application provides a water conservancy multi-model coupling parallel architecture design method, which divides the whole water conservancy system into multiple basic forecast sections, and maps the forecast sections according to the topological relationship between the upstream and downstream thereof. Meanwhile, multiple computing units are divided by using the basin business characteristics and different models, and each unit corresponds to a computing node in addition to being mapped with the basic forecast section. The computing nodes are connected with each other through a high-speed network to form a distributed computing environment. Corresponding water conservancy model components, such as a hydrological model, an engineering scheduling model, a water power model and a water inflow forecast model, are arranged on each computing node, and the model components on different nodes can be combined and cooperated according to the requirement to realize comprehensive simulation on the whole water conservancy system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer vision, and particularly relates to a water conservancy multi-model coupling parallel architecture design method. BACKGROUND

[0002] With the continuous development of water conservancy projects and the increasing demand for fine management of water conservancy systems, traditional single-machine water conservancy models face problems such as low computing efficiency, difficulty in data sharing, poor model scalability, and the like when processing simulation and analysis of large-scale complex water conservancy systems. For example, when simulating complex problems such as flood evolution and water resource allocation in large river basins, single-machine models often cannot complete the calculation within an acceptable time, and cannot fully utilize multi-source data for comprehensive analysis. Therefore, a new architecture and method are needed to realize efficient integration and collaborative work of water conservancy models to improve the scientificity and accuracy of water conservancy project planning, design, and management.

[0003] A Chinese invention patent application with the application number CN119830225A discloses a model fusion method and device based on a topological structure and a computer device. The method includes: constructing a topological structure and determining the to-be-fused objects deployed in each node in the topological structure, wherein the topological structure includes multiple nodes; according to a model fusion algorithm, merging and iterating the to-be-fused objects deployed in multiple nodes within a preset node range in the topological structure to obtain a local fusion model after each merging and iteration of all nodes in the topological structure; and when the coefficients of the local fusion models after multiple merging and iterations of all nodes are consistent, determining that the local fusion models corresponding to each node complete merging to obtain multiple multi-task models in the topological structure. The disadvantage of this method is that the model fusion algorithm performs merging and iteration within the preset node range, and the determination of the node range lacks an adaptive mechanism. If the range is set unreasonably, the model convergence speed will be slow, and even fall into a local optimum, making it difficult to obtain a globally optimal fusion effect.

[0004] Chinese invention patent application CN119862008A discloses a multi-model scheduling architecture method based on LLM for multimodal input applications, particularly in the field of large-scale AI models. This method processes user input data by identifying and handling multiple modalities. The application implements the application and scheduling of multimodal models through text processing. The main processes include multimodal input identification and processing, text processing and related question output, resource type processing, and interactive experience optimization. Users can interact with the system in various natural ways without changing their behavioral habits, while simultaneously processing multiple types of input, improving interaction efficiency. However, this method has a drawback: by implementing multimodal model application and scheduling solely through text processing, the conversion and processing of non-textual modal information such as images, audio, and video may result in the loss of a significant amount of original information. This can lead to the model being unable to acquire complete and accurate multimodal features, affecting the accuracy and richness of the processing results. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a multi-model coupled parallel architecture design method for water conservancy. This method divides the entire water conservancy system into multiple basic forecast sections, which are mapped according to their upstream and downstream topological relationships. Simultaneously, multiple computing units are created using the operational characteristics of the watershed and different models. In addition to mapping to the basic forecast sections, each unit corresponds to a computing node. These computing nodes are interconnected via a high-speed network, forming a distributed computing environment. Corresponding water conservancy model components, such as hydrological models, engineering scheduling models, hydrodynamic models, and inflow forecasting models, are deployed on each computing node. Model components on different nodes can be combined and work collaboratively as needed to achieve a comprehensive simulation of the entire water conservancy system.

[0006] The purpose of this invention is to provide a method for designing a coupled parallel architecture for multi-model hydraulic engineering, including basic forecast section collection and mapping, and further including the following steps:

[0007] Step 1: The computing unit collects the mapping;

[0008] Step 2: Mapping the calculation unit to the basic forecast section;

[0009] Step 3: Taking the basic forecast section as the core, configure different models and parameter information required for different sections according to flood control needs or water resources to form a calculation scheme;

[0010] Step 4: Using the constructed calculation scheme, configure various data such as time periods and meteorological monitoring to form different scenario session plans;

[0011] Step 5: Analyze the constructed session plan, allocate various computing units to various nodes, and allocate and call physical resources according to the computing plan of the session;

[0012] Step 6: Visualize the results of the field calculation.

[0013] Preferably, the base forecast section collection mapping includes the following sub-steps:

[0014] Step 01: Collection and analysis of basin basic data;

[0015] Step 02: Initial selection of sections based on forecasting needs, including flood control forecasting sections, water resources management sections, and ecological flow guarantee sections;

[0016] Step 03: Assessment of section representation and feasibility;

[0017] Step 04: Section optimization and demonstration based on model simulation to determine the final base section, including at least section location and basic parameters.

[0018] In any of the above schemes, preferably, the step 1 includes the following sub-steps:

[0019] Step 11: Collection and preprocessing of calculation unit data;

[0020] Step 12: Analysis of calculation unit characteristics;

[0021] Step 13: Construction of calculation unit topological relationship;

[0022] Step 14: Calculation unit boundary mapping based on topological relationship;

[0023] Step 15: Formulate mapping rules.

[0024] In any of the above schemes, preferably, the step 11 includes obtaining calculation unit data from different models, using data cleaning algorithms to remove outliers and missing values, identifying and correcting outliers in flow data through box plots, and filling missing values in water level data using linear interpolation method;

[0025] Calculation unit data includes distributed hydrological model grid cell data, conceptual hydrological model sub-basin cell data, and river section or grid cell data of hydrodynamic model.

[0026] In any of the above schemes, preferably, the step 12 includes the following sub-steps:

[0027] Step 121: Calculate the geometric properties of each calculation unit;

[0028] Step 122: Analyze the resolution and coverage of the grid cell to determine its accuracy in describing the spatial characteristics of the basin;

[0029] Step 123: Determine the catchment range and boundary form of the sub-basin unit by using the drainage basin system extraction algorithm;

[0030] Step 124: Through spatial overlay analysis, compare the spatial overlap of different model calculation units to identify possible data conflict areas.

[0031] In any of the above schemes, preferably, the step 13 comprises the following sub-steps:

[0032] Step 131: Construct the river system topology, based on the drainage basin system, use the river network tracking algorithm of GIS to construct the upstream and downstream topology relationship of the river channel, assign a unique identifier to each river section unit, and establish the connection relationship between the units in the topology table;

[0033] Step 132: Construct the drainage basin unit topology, determine the confluence relationship between the sub-basin units according to the terrain slope and water flow direction; use the hydrological analysis tool of ArcGIS to generate the flow direction grid map of the drainage basin, and then divide the sub-basin boundary and establish the topology network between the sub-basins; in the distributed grid model, based on the water flow direction of adjacent grids, construct the local topology structure of the grid unit, and clearly define the water flow transmission path between the grids.

[0034] In any of the above schemes, preferably, the step 2 comprises the following sub-steps:

[0035] Step 21: Spatial overlay analysis;

[0036] Step 22: Distance calculation analysis;

[0037] Step 23: Topology relationship analysis.

[0038] In any of the above schemes, preferably, the construction of the calculation scheme comprises the following sub-steps:

[0039] Step 31: Create a calculation scheme and name it;

[0040] Step 32: By selecting the start and end basic forecast section, the system will automatically filter and determine the calculation area of the calculation scheme according to its topology relationship;

[0041] Step 33: In the calculation area, list the basic forecast sections and select a certain basic forecast section;

[0042] Step 34: For this basic forecast section, select each type of model mapped to this section;

[0043] Step 35: After determining the model of the basic forecast section, select the specific model parameters of the model;

[0044] Step 36: configure the hydrological model and parameters in the initial state of the base forecast section, select the reservoir model and parameters, and mark for post-processing;

[0045] Step 37: configure each base forecast section in turn according to the topological relationship, and save the calculation scheme.

[0046] In any of the above schemes, preferably, the step 32 comprises the following sub-steps:

[0047] Step 321: select the base forecast section;

[0048] Step 322: analyze the topological relationship;

[0049] Step 323: determine the calculation area.

[0050] In any of the above schemes, preferably, the step 322 comprises calling the pre-constructed watershed river topological relationship data, storing the upstream and downstream connection relationship of the river in the form of a directed graph, using breadth-first search in graph theory, starting from the starting section, searching along the topological relationship path until reaching the ending section, and including all river sections, sub-basins and other areas involved in the path in the calculation area.

[0051] In any of the above schemes, preferably, the step 5 comprises the following sub-steps:

[0052] Step 51: start the scheduling job flow;

[0053] Step 52: execute the parsing node task;

[0054] Step 53: information reading and calculation preparation;

[0055] Step 54: plan the calculation logic;

[0056] Step 55: determine the boundary information;

[0057] Step 56: start and execute the calculation unit;

[0058] Step 57: monitor and update the status.

[0059] In any of the above schemes, preferably, the step 53 comprises area instruction parsing, rainfall instruction parsing, and station and model parameter instruction parsing.

[0060] In any of the above schemes, preferably, the step 53 comprises hydrological calculation unit boundary determination, reservoir scheduling boundary determination, and hydrodynamic calculation unit boundary determination.

[0061] The application provides a water conservancy multi-model coupling parallel architecture design method, realizes efficient integration and collaborative work of water conservancy models, and improves the scientificity and accuracy of water conservancy engineering planning, design and management. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 A flow chart of a preferred embodiment of the water conservancy multi-model coupling parallel architecture design method according to the application.

[0063] Figure 2 A flow chart of an embodiment of a calculation scheme construction of the water conservancy multi-model coupling parallel architecture design method according to the application.

[0064] Figure 3 A flow chart of an embodiment of model parallel computing based on a field scheme of the water conservancy multi-model coupling parallel architecture design method according to the application. DETAILED DESCRIPTION

[0065] The application will be further described below in combination with the drawings and specific embodiments.

[0066] Embodiment one

[0067] As shown in the figure, a water conservancy multi-model coupling parallel architecture design method, step 1000 is executed, and the basic forecast section is collected and mapped. Different types of water conservancy models have different modeling granularity support due to different model characteristics (some models take small watershed river channels and nodes as granularity, and some models take regional outlet points as granularity), and in order to support dynamic adjustment and mixed construction of various types of hydrological models in different regions of the watershed, a unified basic section needs to be constructed. Figure 1 The basic section construction is divided into four stages: pre-data collection and analysis, section preliminary selection based on forecast requirements, section representativeness and feasibility evaluation, section optimization and final determination. The following sub-steps are included:

[0068] Step 1010 is executed, and the basin basic data collection and analysis includes the following sub-steps:

[0069] Step 1011 is executed, and the basin basic data is collected

[0070] Collect topographic and geomorphic data in the basin, including high-precision digital elevation model (DEM), divide the basin boundary, extract the water system through geographic information system software such as ArcGIS, and clearly define the basic information such as basin area and river network distribution. At the same time, obtain the underlying surface data such as soil type and vegetation cover in the basin, which will affect the precipitation and runoff process.

[0071]

[0072] ​Collect historical hydro-meteorological data, covering years of precipitation, evaporation, air temperature, wind speed, and other meteorological elements, as well as data on flow, water level, sediment content at various hydrological stations. Use Python, R, and other data analysis tools to statistically analyze historical data, draw precipitation-runoff relationship curves, water level-flow relationship curves, and master the hydrological characteristics and variation laws of the basin.

[0073] Perform step 1012, water conservancy project and monitoring network sorting

[0074] Sort out the water conservancy facilities in the basin in detail, including the location, scale, and operation and dispatching rules of reservoirs, sluices, dams, and pumping stations. For reservoirs, obtain key parameters such as reservoir capacity curve and flood discharge capacity curve; for sluices and dams, understand their control mode and flow capacity.

[0075] Determine the distribution of existing hydrological monitoring stations, including the location, monitoring period, and data quality of rainfall stations, water level stations, and flow stations. Analyze the coverage and representativeness of the monitoring stations to determine whether there are monitoring blind spots or weak links.

[0076] Perform step 1020, initial selection of sections based on forecasting needs, including flood control forecasting sections, water resources management sections, and ecological flow guarantee sections.

[0077] 1. Selection of flood control forecasting sections

[0078] In flood control key areas such as cities, densely populated areas, and around important infrastructure, select sections that can effectively reflect the flood evolution process. Usually, select straight river sections, regular sections, and sections without obvious water retention or diversion. Generally, set forecasting sections at appropriate distances upstream of the flood protection area to provide early warning. For example, set sections at a stable river section 10-20 kilometers upstream of a city to provide sufficient response time for city flood control.

[0079] At the same time, consider the correlation with water conservancy projects such as reservoirs and sluices, and set sections at appropriate locations downstream of reservoirs to monitor the impact of reservoir flood discharge on downstream river channels, and set sections upstream and downstream of sluices to monitor changes in water level and flow before and after sluice dispatching.

[0080] 2. Selection of water resources management sections

[0081] For water resources management needs, set forecasting sections at important water intake points in the basin, concentrated water user areas, and cross-border rivers. Set sections upstream of water intake points to forecast inflow and ensure water supply safety, and set sections at cross-border river sections to monitor water quantity and quality changes and provide data support for water resources allocation and coordination. For example, set sections at the intersection of cross-provincial rivers to monitor water resources in real time and avoid water resources disputes.

[0082] 3. Selection of ecological flow guarantee section

[0083] In the surrounding areas of ecological sensitive regions such as wetlands, fish spawning grounds, and rare aquatic habitats, select sections that can reflect ecological water demand. In combination with ecological protection goals and ecological flow standards, set sections in key river reaches to ensure that the river ecological base flow meets the requirements. For example, set a section upstream of a certain wetland protection area, predict the flow, and reasonably regulate water resources to maintain the stability of the wetland ecosystem.

[0084] Step 1030 is executed, and the section representation and feasibility evaluation includes the following sub-steps:

[0085] Step 1031 is executed, and the hydrological representation evaluation is performed

[0086] Using the collected historical hydrological data, mathematical statistical methods such as correlation analysis and frequency analysis are used to evaluate the correlation between the initial selected section and the upstream and downstream hydrological station data. Calculate the correlation coefficient of the section and the surrounding hydrological station flow and water level. The higher the correlation coefficient, the better the hydrological representation of the section. If the correlation coefficient is less than 0.7, the position of the section needs to be reconsidered or measures such as data interpolation extension are taken to improve the representation.

[0087] Analyze the flow pattern at the section to determine whether it is a stable uniform flow. For river reaches with sharp changes in bends, branches, and scouring and silting, be cautious in selection, or eliminate adverse effects through model correction and other methods.

[0088] Step 1032 is executed, and the engineering feasibility evaluation is performed

[0089] Consider the engineering conditions for setting the section, including the topographic conditions at the section, traffic convenience, construction difficulty, etc. Preferably, choose locations with flat terrain, convenient transportation, and easy construction and maintenance of monitoring facilities. Avoid setting sections in areas with complex geological conditions, high construction difficulty, and high maintenance cost.

[0090] Evaluate the mutual influence of the section and existing water conservancy projects and monitoring facilities to ensure that the section setting will not interfere with the operation of water conservancy projects and the collection of monitoring data. For example, the section cannot be set in the downstream scouring area of the dam of the reservoir to avoid affecting the stability of the monitoring equipment and the accuracy of the data.

[0091] Step 1040 is executed, and the final basic section is determined based on model simulation, which at least includes the section location and basic parameters.

[0092] 1. Section optimization based on model simulation

[0093] Hydrological forecasting models such as Xin'anjiang model, API, and spatiotemporal variable source model are used to simulate and analyze the preliminary sections. By adjusting the section location and parameters, the fitting degree of the prediction results and the measured data under different schemes is compared, and the section location and parameter combination with the best fitting effect are selected.

[0094] Sensitivity analysis is performed using the model to determine the degree of influence of section location changes on prediction results. For sections with high sensitivity, further optimization of their location is performed to ensure the reliability and stability of the prediction results.

[0095] 2. Multi-party collaboration to determine the final section

[0096] Experts in the fields of hydrology, water conservancy engineering, and ecological environment are organized to conduct research and discussion, and the opinions and suggestions of different professionals are combined to comprehensively evaluate the section setting scheme.

[0097] After multiple rounds of argumentation and optimization, the final basic section is determined, including section location, basic parameters, etc., providing a basis for subsequent hydrological forecasting work.

[0098] Based on the constructed unified basic section, spatial overlay with rivers is performed through geographic information system software such as ArcGIS, and the topological relationship of each type of basic section is matched and set based on the topological relationship of rivers within the basin.

[0099] Step 1100 is executed, and the calculation unit collects the mapping. The calculation units constructed by various models are collected, and the boundaries between calculation units are mapped and managed according to the characteristics and topological relationships of the calculation units of different models, including the following sub-steps:

[0100] Step 1110 is executed, and the calculation unit data is collected and preprocessed, including:

[0101] 1. Multi-source data collection

[0102] Calculation unit data is obtained from different models, including grid cell data of distributed hydrological models (such as spatiotemporal variable source), sub-basin unit data of conceptual hydrological models (such as Xin'anjiang model), and river section or grid cell data of hydrodynamic models. Data is collected through model output interfaces, database reading, etc., such as using the pandas library of Python to read attribute data such as area and terrain slope of sub-basins from CSV format output files, and using the GDAL library to read spatial data such as elevation and soil type of grid cells.

[0103] 2. Data cleaning and standardization

[0104] Data cleaning algorithms are used to remove outliers and missing values. Outliers in the flow data are identified and corrected using box plots. Missing values in the water level data are filled using linear interpolation. The data format is unified, and various types of calculation unit data are converted to a standard format supported by GIS (such as Shapefile, GeoJSON), and the attribute field naming is standardized to ensure the compatibility of different model data.

[0105] Step 1120 is executed, and the calculation unit characteristic analysis includes the following sub-steps:

[0106] Step 1121 is executed, and GIS spatial analysis tools are used to calculate the geometric properties of each calculation unit, such as area, perimeter, and centroid coordinates.

[0107] Step 1122 is executed, and the resolution and coverage of the grid cells are analyzed to determine their accuracy in describing the spatial characteristics of the watershed.

[0108] Step 1123 is executed, and for sub-basin units, the watershed drainage network extraction algorithm is used to determine their catchment area and boundary shape.

[0109] Step 1124 is executed, and spatial overlay analysis is used to compare the spatial overlap of different model calculation units to identify potential data conflict areas.

[0110] Step 1130 is executed, and the calculation unit topology relationship is constructed, including the following sub-steps:

[0111] Step 131: Constructing the topology of the river system

[0112] Based on the watershed drainage network, the river network tracing algorithm (D8 algorithm) of GIS is used to construct the upstream and downstream topology of the river. Each river section unit is assigned a unique identifier, and a topology table is established to record the connection between units, such as identifying the upstream and downstream relationship through the start node and end node numbers. For areas with bifurcated river channels, the directed acyclic graph (DAG) structure in graph theory is used to describe the complex water flow path.

[0113] Step 132: Constructing the topology of the watershed unit

[0114] For sub-basin units, the confluence relationship between units is determined based on terrain slope and water flow direction. The hydrological analysis tool of ArcGIS is used to generate a flow direction grid of the watershed, and then the sub-basin boundaries are divided, and the topology network between sub-basins is established. In the distributed grid model, based on the water flow direction of adjacent grids, the local topology structure of the grid unit is constructed, and the transmission path of water flow between grids is clearly defined.

[0115] Step 1140 is performed to map the boundaries of the calculation units based on the topological relationship; a direct mapping relationship of the boundaries of the calculation units is established according to the constructed topological network. The flow, water level and other calculation results of the upstream section are directly taken as the input boundary conditions of the downstream section between the upstream and downstream sections of the river; the outlet flow of the sub-basin is mapped to the corresponding river section unit for the connection relationship between the sub-basin and the river. Through the connection information in the topological table, efficient transmission of data between the calculation units is realized.

[0116] Step 1150 is performed to formulate the mapping rules; detailed boundary mapping rules are formulated to clearly define the data conversion and transmission mode between different types of calculation units. For example, it is specified that when the sub-basin overlaps the river section, the hydraulic parameters of the river section are used as the reference; for cross-border calculation units, data conflicts are handled according to the weight distribution principle. The mapping rules are written into a configuration file for subsequent modification and maintenance.

[0117] Step 1200 is performed to map the calculation units and the basic forecast section; the calculation units are mapped with the basic forecast section, and the calculation units are associated with the physical space position of the section, including the following sub-steps:

[0118] Step 1210 is performed to perform spatial overlay analysis; the spatial overlay function of GIS is used to overlay the calculation unit layer and the basic forecast section layer. Taking the grid calculation unit as an example, when the basic forecast section is located in a certain grid, the attribute information of the grid can be directly obtained through the overlay analysis; for the polygon sub-basin calculation unit, if the basic forecast section falls within the sub-basin range, the spatial intersection relationship can be determined by using the intersection analysis tool. In ArcGIS, the “Intersect” tool can be used to realize the intersection of layers and obtain the detailed attributes of the calculation unit where the section is located.

[0119] Step 1220 is performed to perform distance calculation analysis; when the basic forecast section cannot directly fall within the calculation unit, the distance calculation method is used to determine the association relationship. The distance between the basic forecast section and the centroid (or boundary) of each calculation unit is calculated, and the Euclidean distance formula is used to calculate the distance under the plane coordinates. According to the principle of the nearest distance, the basic forecast section is associated with the calculation unit with the nearest distance. In Python, the Geopy library can be used to conveniently calculate the spherical distance between two points to realize the distance analysis function.

[0120] Step 1230 is performed to analyze the topological relationship; the topological relationship between the calculation unit and the basic forecast section is constructed. For river-type calculation units and forecast sections, the topological connection is determined according to the upstream and downstream relationship of the river. The network analysis function of GIS is used to track the flow direction by taking the river calculation unit and the forecast section as network nodes, to determine the position of each section in the calculation unit network and establish the upstream and downstream topological association. By setting the river node and the connection relationship, a complete river topological network is constructed, and the topological mapping of the section and the calculation unit is realized.

[0121] Step 1300 is performed to configure different models and parameter information required under different sections according to flood control requirements or water resources, taking the basic forecast section as the core, to form a calculation scheme. As shown in the following table, the calculation scheme includes the following sub-steps: Figure 2

[0122] Step 1310 is performed to create and name the calculation scheme; the scheme creation function module is called through the user interface or API interface. The system background uses the unique identifier (UUID) generation technology to assign a globally unique ID to the newly created scheme, ensuring the uniqueness and identifiability of the scheme in the database. The user inputs the scheme name in the interactive interface, which needs to follow specific naming conventions, such as using the format of “basin name + business type + creation time” (e.g., “Yongding River Basin Flood Control Scheme_20250516”), which facilitates the classification management and retrieval of the scheme. After the scheme is created, the system establishes the corresponding scheme data table in the database and initializes the basic information fields of the scheme, such as scheme ID, name, creator, creation time, etc.

[0123] Step 1320 is performed to select the region; by selecting the start and end basic forecast sections, the system will automatically filter and determine the calculation region of the calculation scheme according to its topological relationship; including the following sub-steps:

[0124] Step 1321 is performed to select the basic forecast section; the user selects the start and end basic forecast sections in the geographic information system (GIS) visualization interface through mouse clicks or inputting section numbers, etc. The system uses the spatial query function of GIS to accurately locate the selected section in the basin water system layer according to the coordinate information of the section.

[0125] ​Step 1322: Analyze the topological relationship; the system calls the pre-constructed topological relationship data of the river basin, which is generated based on the D8 algorithm and stored in the form of a directed graph to represent the upstream and downstream connection relationship of the river. Using breadth-first search (BFS) in graph theory, starting from the beginning section, search along the topological relationship path until the end section is reached, and include all river segments, sub-basins, and other regions involved in the path in the calculation area. In this process, through the node connection information in the topological table, the branching and confluence of the river are identified to ensure that the calculation area completely covers the required range.

[0126] Step 1323: Determine the calculation area; use the spatial analysis function of GIS to perform geometric operations on the searched area to generate polygon vector data of the calculation area. Through the geometric properties such as area and perimeter of the calculation area, the range of the area is visualized and boundary checked, and if there are omissions or errors, the user can manually adjust the section selection and determine the calculation area again.

[0127] Step 1330: List the basic forecast sections within the calculation area and select a certain basic forecast section; based on the range of the calculation area, the system uses spatial query statements to filter all section data within the area from the basic forecast section database. These section data are displayed in the user interface in the form of a list or tree structure, and each section entry contains basic information such as section name, number, geographic location, and belonging basin. The user selects a basic forecast section from the list by clicking the mouse, keyboard selection, or search function, and the system obtains the unique identifier (ID) of the section for subsequent operation association.

[0128] Step 1340: Select various models mapped to the section for the basic forecast section, such as hydrological models, water resources models, and scheduling models.

[0129] 1. Model library call

[0130] The system connects the model library, which stores the metadata of various water conservancy models, including model name, function description, applicable scenario, input and output data format, etc. Through API interface or database query statement, the model list suitable for flood control business is retrieved, such as Xin'anjiang hydrological model, reservoir scheduling rule model, etc.

[0131] 2. Model recommendation and selection

[0132] The rule-based recommendation algorithm is adopted to recommend suitable models for users by combining information such as topography, climate conditions, and basin characteristics of the section area. For example, the Xin'anjiang model is preferentially recommended for sections in humid areas. Users can select the model to be mapped to the section by checking and dragging in the model list according to their own needs and recommendation information. The system records the ID and related configuration information of the selected model.

[0133] Step 1350 is performed to select specific model parameters of the model after determining the model of the base forecast section, including:

[0134] 1. Parameter template loading

[0135] The system loads the corresponding parameter template from the parameter template library according to the selected model. The parameter template contains information such as the name, type, unit, value range, and default value of all parameters of the model. For example, the parameter template of the Xin'anjiang model contains the basin storage capacity curve parameter (B value) and evaporation calculation parameter.

[0136] 2. Parameter selection and adjustment

[0137] Users can select and adjust parameters based on the following methods:

[0138] Experience value: Refer to existing research results under similar topography and climate conditions in the same type of basin to obtain initial parameter values from the parameter recommendation database. The recommended values in the database are sorted through literature research, historical project summary, and other methods.

[0139] Automatic calibration: Use historical measured data and optimization algorithms (such as genetic algorithm, particle swarm optimization algorithm) to automatically calibrate parameters. Minimize the error between measured flow and water level data and model simulation results as the objective function (such as root mean square error RMSE, Nash efficiency coefficient NSE), and automatically search for the optimal parameter combination through algorithm iteration calculation.

[0140] Manual adjustment: Users manually modify parameter values based on professional knowledge and practical experience. The system provides parameter sensitivity analysis tools to real-time display the changes in model output results by changing the value of a single parameter, helping users judge the impact of parameter adjustment on simulation results and assisting users in parameter optimization.

[0141] 3. Parameter verification

[0142] Use the reserved measured data of the verification period to verify the adjusted parameters. Calculate the error indicators of the simulation results and the measured data. If the error is within an acceptable range (such as NSE≥0.7), the parameter selection is confirmed; otherwise, return to adjust the parameters until the accuracy requirements are met.

[0143] Execute step 1360, there will be multiple models coupled at a certain section, for example, the flow results of the section hydrological model are used as the inflow of the reservoir scheduling model, which means that in addition to the normal setting of model parameters, post-processing configuration is also needed at the section. That is, in the initial state of the section, first configure the hydrological model and parameters, then select the reservoir model and parameters, and mark it as post-processing.

[0144] 1. Coupling relationship identification

[0145] When the user selects multiple models, the system automatically identifies the coupling relationship between the models according to the logical relationship and data flow between the models. Through the pre-defined model coupling rule library, it is determined whether the output data (such as flow) of the hydrological model can be used as the input data of the reservoir scheduling model. If the condition is met, it is determined that there is a coupling relationship between the two.

[0146] 2. Post-processing configuration

[0147] According to the coupling order, first complete the parameter configuration and verification of the hydrological model. Then select the reservoir scheduling model and set the parameters. In the parameter configuration interface, set a special post-processing mark field to mark the reservoir scheduling model as a post-processing model, and establish a data transmission mapping relationship to clearly define the correspondence between the flow data field output by the hydrological model and the inflow parameter of the reservoir scheduling model. The system records the execution order of the coupled models and the data transmission rules to ensure that data interaction and model operation are performed according to the correct process during calculation.

[0148] Execute step 1370, and complete the configuration of each basic forecasting section according to the topological relationship. Save the calculation scheme.

[0149] The system iterates through all the basic forecasting sections in the calculation area and checks whether the model configuration, parameter setting and coupling relationship of each section are complete and correct. Through data verification algorithms, it verifies whether the parameter values are within a reasonable range, whether the model input and output data formats are matched, etc. If there are incomplete configurations or errors, the system gives a prompt to guide the user to correct them. When all the sections are correctly configured, the system packages all the information of the calculation scheme (including scheme basic information, calculation area data, model configuration of each section, parameter values, coupling relationship, etc.) in JSON, XML or other formats and stores it in the database. It also generates a scheme version number to facilitate subsequent scheme management, calling and version tracing.

[0150] Execute step 1400, use the constructed calculation scheme to configure time periods, meteorological monitoring and other types of data to form different scenario schemes.

[0151] Initialization setting: create a scenario task, input the forecast start time, end time and time step, and complete the time dimension configuration.

[0152] Rainfall data loading: Retrieve the compiled hourly rainfall data from the data center, and distribute the areal rainfall data to each computing unit according to the watershed grid or sub-watershed division.

[0153] Calculation scheme invocation: Based on the watershed characteristics and flood control requirements, select a calculation scheme from the calculation scheme library and load the corresponding hydrological model, hydrodynamic model and parameter configuration file.

[0154] Step 1500 involves parsing the constructed session plan, allocating various computing units to different types of nodes, and then allocating and calling physical resources according to the session's computing plan. Figure 3 As shown, it includes the following sub-steps:

[0155] Execute step 1510 to initiate the scheduling job process. The system receives computation scheduling instructions via the RabbitMQ message queue or API interface. The instruction format follows a predefined protocol standard (such as JSON or XML format). When an instruction arrives, the system triggers the scheduling job process, first validating the instruction by matching key fields in the instruction (such as scheme ID, computation time range, job type, etc.) using regular expressions. If the instruction format is incorrect or key information is missing, an error message is returned. If the validation passes, the instruction is stored in the task queue, and a unique job ID is generated for subsequent process tracking and management.

[0156] Execute step 1520 to perform the node parsing task.

[0157] 1. Regional Command Parsing

[0158] Using the spatial query function of Geographic Information System (GIS), based on the topological relationship of the basic forecast cross sections and the start and end basic cross sections of the scheme area, the specific geographic boundary information is retrieved from the regional database by using SQL statements combined with spatial functions (such as ST_Contains).

[0159] 2. Analysis of Rainfall Commands

[0160] For rainfall instructions, regular expressions are used to extract key parameters such as rainfall time range and intensity. If the instruction contains a rainfall data file path (e.g., "hourly rainfall data in.csv format"), the data is loaded using a file reading function (e.g., the read_csv function in the pandas library in Python), and format conversion and quality checks are performed to remove outliers (e.g., negative rainfall values). For complex rainfall instructions (e.g., "72-hour rainfall prediction based on numerical weather prediction"), a meteorological data interface is called to obtain rainfall grid data from numerical model outputs, and interpolation algorithms (e.g., inverse distance weighting interpolation, Kriging interpolation) are used to convert the grid data into areal rainfall data applicable to each calculation unit in the watershed.

[0161] 3. Site and model parameter instruction parsing

[0162] Site instruction parsing extracts site numbers or names from instructions through string matching and database queries, and queries corresponding latitude and longitude coordinates, monitoring device types, and data collection frequencies in the site information database. For model and parameter instructions, a hierarchical parsing strategy is used to identify the model type (e.g., Xin'anjiang hydrological model, hydrodynamic model) and then parse its corresponding parameter list. During parameter parsing, parameter values are compared with predefined parameter templates to verify parameter types and value ranges. If there are errors, they are marked and recorded for subsequent correction.

[0163] Step 1530, information reading and calculation preparation

[0164] 1. Regional information acquisition

[0165] Based on the parsed regional spatial range, GIS spatial analysis tools (e.g., the SpatialAnalyst module of ArcGIS) are used to obtain basic geographic information such as terrain data (digital elevation model DEM), soil type data, and land use data within the region. Topological relationship data of rivers and sub-basins within the region are retrieved from the watershed topology database, including upstream and downstream connection relationships, sub-basin confluence paths, and other information, providing spatial data support for subsequent calculations.

[0166] 2. Rainfall event information acquisition

[0167] According to the rainfall instruction parsing results, rainfall event data within the corresponding time range is retrieved from the meteorological database. If it is historical rainfall data, the stored rainfall process file is directly read; if it is predicted rainfall data, the meteorological prediction model interface is called to obtain the latest rainfall prediction results. Data cleaning algorithms (e.g., outlier removal, missing value interpolation) are used to preprocess the rainfall data to ensure data integrity and accuracy.

[0168] 3. Site monitoring information acquisition

[0169] Through the Internet of Things data acquisition platform or database interface, real-time monitoring data such as water level, flow, and water quality of each monitoring site are acquired. Data compression techniques (such as Gzip, Zlib) are used to reduce data transmission volume and improve data reading efficiency. Real-time quality control is performed on the acquired data, and a threshold alarm mechanism (such as triggering an alarm when the water level exceeds the warning water level) is set to timely discover and handle abnormal data.

[0170] 4. Dispatching model and parameter information acquisition

[0171] Model code and configuration files participating in dispatching are loaded from the model library, and the jar / python package is used to realize rapid deployment of the model. According to the parameter instructions obtained through analysis, corresponding model parameter values are called from the parameter database. If the parameter values are not completely specified, default parameters or initial parameter values generated through a parameter recommendation algorithm (such as a parameter analogy algorithm based on similar basins) are used.

[0172] Step 1540, planning calculation logic

[0173] 1. Forecast section calculation unit association

[0174] Based on the geographic location of the forecast section, the GIS spatial overlay analysis function is used to determine the calculation unit associated with the section. For a distributed hydrological model, a mapping relationship between the section and the grid cell is established by judging whether the section coordinates are located within a certain grid cell. For a lumped parameter model, the sub-basin division and river topology relationship are used to determine the sub-basin where the section is located and the corresponding calculation unit. A data association table is established to record the correspondence between the forecast section and the calculation unit, as well as the data transmission direction and format requirements.

[0175] 2. Post-processing association and logic planning

[0176] For forecast sections with model coupling, the execution order and data transmission relationship of each model are determined according to the pre-configured post-processing rules. A directed acyclic graph (DAG) data structure is used to describe the calculation logic, with nodes representing calculation tasks (such as hydrological model calculation, reservoir dispatching model calculation) and edges representing data flow direction and task dependency relationship. For example, the output of the hydrological model calculation node is used as the input of the reservoir dispatching model calculation node. Through the topological sorting algorithm of the DAG, the execution sequence of the calculation tasks is generated to ensure that each model is calculated in the correct logical order.

[0177] Step 1550, boundary information determination

[0178] 1. Hydrological calculation unit boundary determination

[0179] The upstream boundary condition is usually measured flow data or the output of the upstream calculation unit, which is transmitted through a data interface (such as Web Service, API). The downstream boundary condition can be set as a fixed water level, a flow-water level relationship curve, or the boundary information fed back by the downstream calculation unit according to the actual situation of the river channel. By using hydrological principles (such as continuity equation, energy equation) and combining with the geometric parameters of the river section (such as section area, roughness), the hydraulic elements at the boundary, such as water level and flow velocity, are calculated to provide accurate boundary conditions for hydrological calculation.

[0180] 2. Reservoir scheduling boundary determination

[0181] The input boundary of reservoir scheduling includes inflow (provided by hydrological calculation unit), initial water level of reservoir (obtained from reservoir monitoring data), and downstream river channel safety discharge (determined according to flood control standard). The output boundary is the reservoir outflow, which needs to meet the constraints of downstream river channel flood carrying capacity and flood control scheduling rules. By establishing a reservoir scheduling model (such as rule scheduling model, optimization scheduling model), the outflow process is calculated according to the input boundary conditions and scheduling rules, which serves as the input boundary of the downstream calculation unit.

[0182] 3. Determination of boundary of hydrodynamic calculation unit

[0183] The upstream boundary of one-dimensional or two-dimensional hydrodynamic model can be set as flow process line or water level process line, and the downstream boundary is usually water level boundary condition. When using the method of characteristic line or finite difference method to discretely solve the hydrodynamic equation, the corresponding numerical boundary condition needs to be set according to the type of boundary. Through data interaction with the surrounding calculation units, the boundary information is updated in real time to ensure the accuracy and stability of hydrodynamic calculation.

[0184] Step 1560, start and execute calculation unit

[0185] Multi-threading or distributed computing framework is adopted to start each calculation unit in parallel. Before starting, each calculation unit checks whether the input boundary data is complete through the state query interface. If the input boundary data is not ready, it enters the waiting state; if the data is complete, it triggers the calculation task. During the calculation process, numerical calculation methods (such as finite element method, finite volume method) are used to solve the model equation, and the calculation progress and intermediate results are recorded in real time. After the calculation is completed, the output results are stored in the designated data storage location (such as database, file system), and the boundary state identifier of the calculation unit is updated to "completed" for subsequent calculation unit query and call.

[0186] Step 1570, monitor and update state

[0187] A real-time monitoring system is deployed to obtain the state information of each computing unit and the forecast section in real time through a message queue or WebSocket technology. For the computing unit, the computing progress, resource usage (CPU usage, memory occupancy), and computing result accuracy are monitored. For the forecast section, the changes in key elements such as water level and flow are monitored. When the computing unit completes the calculation or the state of the forecast section changes, the system automatically updates the boundary state and the forecast section state database, and notifies the relevant computing unit and the forecast section to perform the next calculation or data update through an event-driven mechanism, ensuring the continuity and accuracy of the entire calculation scheduling process.

[0188] In step 1600, the results of the game calculation are visualized and displayed.

[0189] Embodiment Two

[0190] The present application proposes a water conservancy multi-model coupling parallel architecture design method, which divides the entire water conservancy system into multiple basic forecast sections. The forecast sections are mapped according to their upstream and downstream topological relationships. Meanwhile, the basin business characteristics and different models are used to divide multiple computing units. In addition to being mapped with the basic forecast sections, each unit corresponds to a computing node. These computing nodes are connected to each other through a high-speed network to form a distributed computing environment. On each computing node, corresponding water conservancy model components such as hydrological models, engineering scheduling models, hydrodynamic models, and inflow forecasting models are deployed. The model components on different nodes can be combined and work together as needed to achieve comprehensive simulation of the entire water conservancy system.

[0191] (I) Basic forecast section mapping principle

[0192] Based on the in-depth analysis of key elements such as water flow movement, water resource distribution, and engineering regulation in the water conservancy system, the basic forecast section concept is innovatively introduced to finely divide the vast and complex entire water conservancy system. The basic forecast section is formed based on the basic forecast section of the hydrological and water resource model. The division of the forecast section is strictly selected according to the natural hydrological context, river trend, and key water conservancy facility layout of the water conservancy system.

[0193] A set of rigorous mapping mechanisms is established between the basic forecast sections according to the natural topological relationship between the upstream and downstream. This topological mapping is like an invisible but strong network that tightly connects each isolated section and clearly outlines the dynamic transmission path of water flow in the entire water conservancy system. For example, from the source to the mouth of the river section, the topological network realizes real-time linkage of water regime.

[0194] (II) Computing unit mapping principle

[0195] The numerous calculation units derived from the water conservancy model are not isolated, but are connected by strictly defined boundaries to form a topological relationship between upper and lower levels, realizing seamless connection of data and calculation process. This boundary coupling mechanism is like an invisible but strong link that connects each local unit and makes them work together to outline the overall operation of the water conservancy system. Specifically, after a calculation unit completes its internal established calculation process, the key results output, usually the outlet data at its boundary, such as the flow rate, flow velocity, and water level of the simulated water flow to the unit boundary, immediately become the inlet input data of the adjacent and downstream calculation unit, becoming the key initial condition for driving the next unit calculation process to start and accurately run. Following this pattern, along the direction of water flow or the business logic chain of the water conservancy system, each calculation unit cooperates closely in sequence, with the output of the previous unit continuously powering the subsequent unit, prompting the entire model system to operate like a precise mechanical clock, accurately and efficiently, gradually realizing the leap from local unit simulation to comprehensive simulation of the entire water conservancy system, providing solid technical support for scientific decision-making of water conservancy projects, sustainable management of water resources, and effective protection of the ecological environment.

[0196] (Three), mapping principle between calculation unit and basic forecasting section

[0197] In the water conservancy model system architecture, there is a specific mapping logical relationship between the basic forecasting section and the calculation units constructed by different water conservancy models, which is the key link to realize accurate simulation and efficient regulation of the water conservancy system.

[0198] From the mapping dimension of the basic forecasting section to the calculation unit, a basic forecasting section can be associated with specific calculation units under different water conservancy model frameworks. This is because different water conservancy models focus on different characteristics and process analysis of the water conservancy system. For example, a basic forecasting section located at a steep slope of the river mainstream, in the hydrodynamic model system, its sudden change data of flow rate and water level monitored and obtained will be mapped to the calculation unit responsible for simulating the energy conversion of river flow and the dynamic adjustment of river regime, to accurately deduce the motion trajectory of the water flow under this special terrain. In the hydrological model category, the precipitation and runoff monitoring values of this section will be associated with the calculation unit focusing on the analysis of the runoff generation and concentration mechanism in the basin, to help analyze the efficiency and process details of precipitation conversion to runoff in this area. This multi-mapping mode fully taps the multi-dimensional value of the basic forecasting section data, making it serve the core calculation needs of different water conservancy models in all directions.

[0199] On the other hand, from the perspective of the computing unit pointing to the basic forecasting section, there are cases where a single computing unit is not directly associated with a basic forecasting section. This is mainly due to the functional positioning of some computing units, which is not directly dependent on real-time monitoring section data, but focuses on the in-depth exploration of the inherent laws of the water conservancy system, the comprehensive analysis of historical data, or the prediction of future trends. Such computing units provide high-level wisdom support for the long-term layout and macro decision-making of the water conservancy system through complex operations on various non-section data, complementing those computing units that are closely driven by basic forecasting section data, and together weaving a rigorous and efficient operation network of the water conservancy model system.

[0200] (Four), Computing unit and node adaptation principle

[0201] A one-to-one corresponding close relationship is established between the computing unit and the computing node, ensuring the accurate allocation and efficient use of computing resources. Each computing node is equipped with hardware resources and software environments that are in perfect harmony with the characteristics of the corresponding computing unit. For example, for a computing unit responsible for flood evolution simulation, its corresponding computing node is equipped with high-performance multi-core processors and large-capacity memory to meet the demanding requirements of hydrodynamic models for massive grid data complex numerical operations in a short time. For a computing unit that focuses on long-term water resource allocation planning, its computing node optimizes the storage system to ensure the proper storage and efficient retrieval of water resource supply and demand data spanning decades, and cooperates with the engineering scheduling model to achieve optimal allocation of water resources.

[0202] (Five), Model distributed collaborative computing task principle

[0203] The complex computing process of the water conservancy model is decomposed into multiple relatively independent sub-tasks, which can be executed simultaneously on different computing nodes. Taking the flood evolution model as an example, the entire flood evolution process can be divided into flood generation, river flood propagation, and floodplain inundation calculation sub-tasks. Each sub-task has clear input and output and independent computing logic, and can start execution without waiting for the intermediate results of other sub-tasks, as long as the initial data required is ready.

[0204] The parallel framework is responsible for scheduling and coordinating these decomposed tasks to ensure they are executed in the correct order and dependency relationship on appropriate computing nodes. The parallel framework will arrange the start and execution order of tasks reasonably based on the dependency relationship between tasks (such as reservoir discharge decision-making requiring upstream inflow prediction results) and the availability of computing resources, dynamically allocate computing resources to each task, and enable multiple tasks to advance in parallel on different nodes, maximizing the use of computing resources and shortening the overall model computation time.

[0205] Embodiment three

[0206] Take the Daqing River as an example, the Xin'anjiang hydrological model, reservoir regulation model, and two-dimensional hydrodynamic model need to be integrated.

[0207] (1) Collection and mapping of basic forecast sections

[0208] There are 23 Xin'anjiang hydrological model forecast sections, 7 reservoir forecast sections, and 1 hydrodynamic section in the basin.

[0209] After summarizing and arranging, 23 basic forecast sections are formed, and the upstream and downstream relationship mapping is performed.

[0210] (2) Collection and mapping of calculation units

[0211] There are 13 Xin'anjiang hydrological model calculation units, 7 reservoir calculation units, and 1 hydrodynamic calculation unit in the basin. At the same time, the boundary mapping of the calculation units of each type of model is performed.

[0212] (3) Mapping of calculation units and basic forecast sections

[0213] The forecast section and basic section mapping of the Xin'anjiang hydrological model, reservoir regulation model, and hydrodynamic model in the basin is generated.

[0214] (4) Construction of calculation scheme

[0215] Check the basic forecast sections of the entire basin (Gongnongliang, Ziyaxinhefengchaoliang (new), Erdaogezheshang (Haihe mainstream)) as the final basic forecast section, and construct the flood control forecasting and dispatching integrated business of the basin to form the calculation scheme (named: Daqing River hydrology and hydrodynamic reservoir demonstration scheme).

[0216] (5) Set the scene

[0217] Based on the constructed calculation scheme (named: Daqing River hydrology and hydrodynamic reservoir demonstration scheme), the forecast time is selected as 2023-7-29 8:00, the preheating period is 24 hours, the prediction period is 306 hours, and the calculation length is 1 hour. The rainfall information is selected as the full period of measured rainfall, and the specific measured + predicted rainfall scene scheme (named: Daqing River hydrology and hydrodynamic coupling scene) is formed.

[0218] (6) Parallel computing

[0219] According to the constructed measured + predicted rainfall scene scheme (named: Daqing River hydrology and hydrodynamic coupling scene), the calculation units are distributed to the nodes, and the physical resource allocation is called through the calculation scheme (named: Daqing River hydrology and hydrodynamic reservoir demonstration scheme).

[0220] According to the calculation scheme (named: Daqing River hydrology and hydrodynamic reservoir demonstration scheme), the basic section, model, parameter, and topology information of the scheme are obtained.

[0221] First, take each basic forecast section in the area as the core calculation point (total 23), associated with the corresponding calculation unit (13 calculation units); associated with the reservoir dispatch post-processing of the basic forecast section (7 reservoir calculation units) and the hydrodynamic post-processing (1 hydrodynamic calculation unit); plan the logical structure of the calculation.

[0222] Then start the calculation of the calculation unit (hydrology, reservoir dispatch, hydrodynamic, etc. 21 units). After that, query whether the input boundary of the unit is completed, if the input boundary is completed, start the calculation of the calculation unit; update the boundary of the calculation unit after the calculation is completed, indicating that the calculation unit has completed the calculation. Other calculation units are calculated in turn according to this logic.

[0223] The calculation unit calculates at the same time, and the boundary state monitoring and the forecast section state monitoring are carried out at the same time. To update the boundary state and the forecast section state, to ensure the next calculation unit and the next forecast section to carry out the calculation.

[0224] (7) Achievement display

[0225] The calculation results of the field scheme (named: Daqing River hydrological dispatch hydrodynamic coupling field) will be divided into hydrology, reservoir, and hydrodynamic results.

[0226] In order to better understand the present application, the above is described in detail in combination with the specific embodiments of the present application, but it is not a limitation of the present application. Any simple modification according to the technical essence of the present application to the above embodiments is still within the scope of the technical solutions of the present application. The focus of each embodiment in the specification is the difference from other embodiments, and the same or similar parts can be referred to each other. For system embodiments, since they basically correspond to method embodiments, the description is relatively simple, and the relevant parts refer to the part of the method embodiment.

Claims

1. A water conservancy multi-model coupling parallel architecture design method comprising a basic forecast section collection mapping, characterized in that, Further comprising the following steps: Step 1: The calculation unit collects the mapping, comprising the following sub-steps: Step 11: Calculation unit data collection and preprocessing, the calculation unit data comprising the grid cell data of the distributed hydrological model, the sub-basin cell data of the conceptual hydrological model, and the river section or grid cell data of the hydrodynamic model; Step 12: Analysis of the characteristics of the calculation unit; Step 13: Construction of the topological relationship of the calculation unit, comprising the following sub-steps: Step 131: Construction of the water system topology; Step 132: Construction of the basin unit topology; Step 14: Mapping of the calculation unit boundary based on the topological relationship, establishing the direct mapping relationship of the calculation unit boundary according to the constructed topological network; on the upstream and downstream sections of the river, the calculation results of the flow and water level of the upstream section are directly taken as the input boundary conditions of the downstream section; for the connection relationship between the sub-basin and the river, the outlet flow of the sub-basin is mapped to the corresponding river section unit; Step 15: Formulate mapping rules to clearly define the data conversion and transmission methods between different types of calculation units; Step 2: Mapping of the calculation unit and the basic forecasting section; Step 3: Taking the basic forecasting section as the core, configuring different models and parameter information required under different sections according to flood control requirements or water resources to form a calculation scheme, the construction of the calculation scheme comprising the following sub-steps: Step 31: Create a calculation scheme and name it; Step 32: By selecting the start and end basic forecasting sections, the system will automatically filter and determine the calculation area of the calculation scheme according to its topological relationship; Step 33: In the calculation area, list the basic forecasting sections and select a certain basic forecasting section; Step 34: For the basic forecasting section, select the various models mapped to the section; Step 35: After determining the model of the basic forecasting section, select the specific model parameters of the model; Step 36: In the initial state of the basic forecasting section, configure the hydrological model and parameters, select the reservoir model and parameters, and mark it for post-processing; Step 37: According to the topological relationship, configure each basic forecasting section in turn, and save the calculation scheme; Step 4: Using the constructed calculation scheme, configure the time period, meteorological monitoring data, and form the scene scheme of different scenarios; Step 5: According to the constructed scene scheme, analyze and allocate various calculation units to various nodes, and call the physical resource allocation through the scene calculation scheme; Step 6: Visualize the results of the scene calculation.

2. The method of claim 1, wherein, The basic forecasting section collection mapping comprises the following sub-steps: Step 01: Collection and analysis of basin basic data; Step 02: Preliminary selection of sections based on forecasting requirements, the sections including flood control forecasting sections, water resources management sections, and ecological flow guarantee sections; Step 03: Evaluation of section representation and feasibility; Step 04: Section optimization and demonstration based on model simulation to determine the final basic section, the basic section at least including the section location and basic parameters.

3. The method of claim 2, wherein, The step 11 comprises obtaining the calculation unit data from different models, removing abnormal values and missing values by using a data cleaning algorithm, identifying and correcting outliers in the flow data by using a box plot, and filling in missing values in the water level data by using linear interpolation.

4. The method of claim 3, wherein, The step 12 comprises the following sub-steps: Step 121: calculating the geometric properties of each calculation unit; Step 122: analyzing the resolution and coverage of the grid unit to determine the accuracy of its description of the spatial characteristics of the basin; Step 123: determining the catchment area and boundary shape of the sub-basin unit by using a basin drainage extraction algorithm; Step 124: comparing the spatial overlap of different model calculation units by spatial overlay analysis to identify possible data conflict areas.

5. The method of claim 4, wherein, The step 13 comprises the following sub-steps: Step 131: constructing the river system topology, using the river system as the basis, using the river network tracking algorithm of GIS to construct the upstream and downstream topology of the river, assigning a unique identifier to each river section unit, and establishing a topology table to record the connection between units; Step 132: constructing the basin unit topology, determining the confluence relationship between sub-basin units according to the terrain slope and water flow direction; using the hydrological analysis tool of ArcGIS to generate a flow direction grid of the basin, and then dividing the sub-basin boundaries and establishing the topology network between sub-basins; in the distributed grid model, based on the water flow direction of adjacent grids, the local topology structure of the grid unit is constructed to clearly define the transmission path of water flow between grids.

6. The method of claim 5, wherein, The step 2 comprises the following sub-steps: Step 21: spatial overlay analysis; Step 22: distance calculation analysis; Step 23: topology relationship analysis.

7. The method of claim 6, wherein, The step 32 comprises the following sub-steps: Step 321: selecting the basic prediction section; Step 322: analyzing the topology relationship; Step 323: determining the calculation area.

8. The method of claim 7, wherein, The step 322 comprises calling the pre-constructed river channel topology relationship data of the basin, storing the upstream and downstream connection relationship of the river channel in the form of a directed graph, using breadth-first search in graph theory to search along the topology relationship path from the starting section until the end section, and including all river sections and sub-basins involved in the path in the calculation area.

9. The method of claim 8, wherein, The step 5 comprises the following sub-steps: Step 51: starting the scheduling job flow; Step 52: executing the parsing node task; Step 53: information reading and calculation preparation; Step 54: planning the calculation logic; Step 55: determining the boundary information; Step 56: starting and executing the calculation unit; Step 57: monitoring and updating the status.

Images