Water Conservancy Project Management System Based on GIS

Abstract

The present invention relates to the field of water conservancy engineering technology, and discloses a water conservancy engineering management system based on GIS, including the following modules: data access and processing module, multi-dimensional factor comprehensive evaluation module, high-precision DEM generation and calibration module, flooding simulation module, scenario setting and uncertainty analysis module, decision support and early warning report module, machine learning and prediction module and real-time monitoring and updating module. The present invention introduces multiple factors, provides a more accurate risk index, supports dynamic adjustment of risk levels, makes flooding simulation and geological disaster prediction closer to actual conditions through high-precision DEM, and provides strong technical support for disaster prevention and emergency management through the automated early warning mechanism. The real-time monitoring and updating module ensures the rapid update of post-disaster data and provides detailed support for repair work.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy projects, and more particularly to a water conservancy project management system based on GIS. Background Art

[0002] Water conservancy projects are important infrastructure for ensuring national water resources security and social and economic development. However, under the influence of climate change and human activities, extreme hydrological events occur frequently, bringing unprecedented challenges to water conservancy projects. How to effectively use modern information technology to improve the management level of water conservancy projects has become an urgent problem to be solved.

[0003] Therefore, in terms of flood risk management, the present invention proposes a comprehensive management system that combines an innovative flood risk zoning method and inundation simulation technology optimization. Summary of the invention

[0004] The present invention provides a water conservancy project management system based on GIS to solve the technical problems in related technologies.

[0005] The present invention provides a water conservancy project management system based on GIS, comprising the following modules:

[0006] Data access and processing module: responsible for collecting and preprocessing various types of data, including historical hydrological data, topographic data, socio-economic data, human activity data, and historical records of geological disasters, and setting scoring criteria and weights for each factor; receiving data streams from IoT sensors in real time, and adjusting risk levels and initiating emergency response procedures in a timely manner for new changes or geological disaster warning information;

[0007] Multi-dimensional factor comprehensive assessment module: calculates the risk index of the region and quantifies the risk level faced by different regions;

[0008] High-precision DEM generation and calibration module: Create accurate terrain models, build high-precision digital elevation models through LiDAR point cloud data and drone images, and perform geometric correction through ground control points;

[0009] Inundation simulation module: Select appropriate hydrodynamic models according to the characteristics of the watershed, and apply the Bayesian optimization algorithm to find the optimal parameter combination to make the simulation results as close to the actual observed values ​​as possible;

[0010] Scenario Setting and Uncertainty Analysis Module: Evaluate the uncertainty and impact of different scenarios, assume that key input variables follow a specific probability distribution, and generate multiple possible scenarios through Monte Carlo random sampling. Perform a complete flooding simulation under each scenario, record the output results, and calculate confidence intervals to quantify the uncertainty;

[0011] Decision support and early warning reporting module: provides an intuitive user interface for browsing maps and querying statistical data. When the monitoring indicators exceed the preset threshold, the system will automatically trigger an early warning notification;

[0012] Machine learning and prediction module: Use machine learning algorithm to train models to predict the probability of geological disasters that may occur in the future;

[0013] Real-time monitoring and updating module: After a geological disaster occurs, quickly organize on-site investigation, update DEM to reflect the latest changes in river morphology, and record the damage to water conservancy facilities. If necessary, re-run the multi-dimensional factor comprehensive assessment model and update the risk zoning map.

[0014] Furthermore, in the multi-dimensional factor comprehensive evaluation module, the calculation formula of the regional risk index is as follows:

[0015] ;

[0016] in is the risk index, represents the historical hydrological factor score, represents the topographic factor score, represents the socioeconomic factor score, represents the score of human activity factor, represents the geological hazard factor score, , , , and Represent the weight of each factor.

[0017] Furthermore, the steps for creating the landform model are as follows:

[0018] LiDAR point cloud data acquisition: Use airborne or ground-based LiDAR systems to scan large areas and obtain high-density three-dimensional point cloud data;

[0019] Point cloud classification and filtering: Classify LiDAR point cloud data, distinguish ground points from non-ground points, apply filtering algorithms to remove noise points, and generate digital surface models;

[0020] Construct DEM based on point cloud: import the classified ground points into GIS software and use interpolation method to generate high-precision DEM to ensure that the spatial resolution of DEM is suitable for research needs;

[0021] Select ground control points: several ground control points with known coordinates are set up in the study area. These points should be evenly distributed and cover different terrain features. Accurate coordinates can be obtained through GPS measurement. Fixed markers on existing high-precision maps can also be used as GCPs.

[0022] Apply geometric correction algorithm: Select ground control points are applied to the DEM and an optimization algorithm is used to adjust the errors in the DEM.

[0023] Furthermore, the interpolation method of the creation step of the landform model is as follows:

[0024] ;

[0025] in It is the prediction point 's elevation; is the sample point The known elevation of is the weight coefficient, which is determined by the covariance function and reflects the influence of each sample point on the prediction point.

[0026] Furthermore, the watershed characteristics are described by the following formula:

[0027] ;

[0028] ;

[0029] in It's the depth of water. is the velocity vector, It's time. is the water surface height, is the height of the riverbed bottom, is the external force term, is the acceleration due to gravity.

[0030] Furthermore, the water flow dynamics model includes the following three dynamics models:

[0031] One-dimensional hydrodynamic model: It is suitable for narrow river channels where water flows mainly in a single direction. It takes into account the changes in water depth and flow velocity over time, but ignores the influence of lateral water flow.

[0032] Two-dimensional hydrodynamic model: suitable for areas where water flow changes significantly in the horizontal plane, capable of describing the lateral distribution of water flow, and more suitable for flood simulation in complex terrain;

[0033] Three-dimensional hydrodynamics model: Suitable for environments where the water flow also changes significantly in the vertical direction. It can capture the effects of turbulence and the influence of complex underwater structures on the water flow and provide the most detailed simulation results.

[0034] Furthermore, in the scenario setting and uncertainty analysis module, the degree of uncertainty is quantified and expressed using confidence intervals:

[0035] ;

[0036] in is the sample mean, is the standard deviation, is the sample size;

[0037] When considering the impact of geological disasters, different types of disaster scenarios are set up and simulation experiments are carried out to evaluate their impact on water conservancy projects.

[0038] Furthermore, the decision support and early warning report module includes an early warning report generation submodule, which automatically generates an early warning report based on a preset threshold value. The early warning report includes key indicators, charts, and suggested measures.

[0039] Furthermore, the following machine learning algorithm is used in the machine learning and prediction module, and its calculation formula is as follows:

[0040] ;

[0041] in Indicates category The probability of, that is, the proportion of samples belonging to category i in the current node, Represents the total number of categories, indicating all possible types of geological hazards, Represents Gini impurity, which is used to measure a data set An indicator of the degree of confusion and uncertainty;

[0042] ;

[0043] in Represents entropy, a measure of the data set The degree of confusion, Represents attributes, such as rainfall or river flow in historical hydrological data, According to the attribute After segmentation, each subset corresponds to a specific attribute value. and are the number of samples in the original dataset and the subset after segmentation, Represents information gain, which is used to measure an attribute For the dataset The classification contribution index of Is an attribute All possible values ​​of .

[0044] The beneficial effects of the present invention are:

[0045] The present invention introduces a multi-dimensional factor comprehensive evaluation module that combines multiple factors, provides a more accurate risk index, and supports dynamic adjustment of risk levels. The application of high-precision DEM and Bayesian optimization algorithm makes inundation simulation and geological disaster prediction closer to the actual situation. The application of automated early warning mechanism and machine learning algorithm provides strong technical support for disaster prevention and emergency management. The real-time monitoring and update module ensures the rapid updating of post-disaster data and provides detailed support for restoration work. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a structural block diagram of the GIS-based water conservancy project management system proposed by the present invention.

[0047] In the figure: 101, data access and processing module; 102, multi-dimensional factor comprehensive evaluation module; 103, high-precision DEM generation and calibration module; 104, flooding simulation module; 105, scenario setting and uncertainty analysis module; 106, decision support and early warning report module; 107, machine learning and prediction module; 108, real-time monitoring and update module. DETAILED DESCRIPTION

[0048] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0049] like Figure 1 As shown in the figure, the GIS-based water conservancy project management system includes the following modules:

[0050] Data access and processing module 101:

[0051] In this module, the data collected include historical hydrological data (such as rainfall, river flow, groundwater level, etc.), topographic data (such as high-resolution DEM), socio-economic data (such as population density, building distribution), human activity data (such as layout of water conservancy project facilities), and historical records of geological disasters;

[0052] In one embodiment of the present invention, the collected data is cleaned and standardized to ensure the quality and consistency of the data;

[0053] And according to the characteristics of each factor, set scoring criteria for each factor, for example:

[0054] Historical hydrological factor: scoring based on historical flood frequency, severity, and other factors;

[0055] Topographic factors: score based on factors such as altitude, slope, river network density, etc.

[0056] Socioeconomic factors: Scoring based on population density, economic value, infrastructure distribution, etc.

[0057] Human activity factor: Analyze the impact of human activities on the natural environment, such as land use change, engineering construction, etc.

[0058] Geological disaster factor: score based on the frequency and impact range of geological disasters such as earthquakes and landslides in the past;

[0059] Finally, the analytic hierarchy process (AHP) was used to determine the weights of each factor. , It represents the number of weight factors, and the weight distribution is obtained by statistically analyzing the existing data.

[0060] In one embodiment of the present invention, the data access and processing module also collects the following data:

[0061] Receive and analyze data streams from IoT sensors in real time, and use the Kalman filter algorithm to smooth the observed values ​​to ensure data accuracy;

[0062] The calculation formula of the Kalman filter algorithm is as follows:

[0063] ;

[0064] ;

[0065] in is the state estimate; is the predicted state; is the Kalman gain; is the observed value; is the observation matrix; and are the posterior and prior covariance matrices, respectively;

[0066] Adjust the risk level of relevant areas in a timely manner for emerging changes (such as new construction projects or changes in land use);

[0067] When receiving geological disaster warning information, immediately initiate emergency response procedures, reassess the risk status of the affected area, and push the latest results to relevant departments and the public;

[0068] Multidimensional Factor Comprehensive Assessment Module 102:

[0069] In this module, based on the comprehensive data collected, the risk index for each area is calculated using the following formula:

[0070] ;

[0071] in is the risk index; represents the historical hydrological factor score; represents the topographic factor score; represents the socioeconomic factor score; represents the score of human activity factor; Indicates the geological hazard factor score; represents the factor weight of the risk index, represents the factor weight of the historical hydrological factor, represents the factor weight of topographic factors, represents the factor weight of the human activity factor, Represents the factor weight of geological hazard factors.

[0072] High-precision DEM generation and calibration module 103:

[0073] In this module, high-precision DEM is constructed using LiDAR point cloud data and photos taken by drones, and geometric correction is performed through ground control points to ensure that its vertical error does not exceed ±0.5 meters.

[0074] In one embodiment of the present invention, the specific steps are as follows:

[0075] LiDAR point cloud data acquisition: Use airborne or ground-based LiDAR systems to scan large areas and obtain high-density three-dimensional point cloud data;

[0076] Point cloud classification and filtering: Classify LiDAR point cloud data, distinguish ground points from non-ground points (such as vegetation, buildings, etc.), apply filtering algorithms to remove noise points, and generate a digital surface model (DSM);

[0077] Construct DEM based on point cloud: import the classified ground points into GIS software and use interpolation methods (such as Kriging interpolation, nearest neighbor interpolation, etc.) to generate high-precision DEM;

[0078] Ensure that the spatial resolution of the DEM is appropriate for the research needs, usually 0.5 to 1 meter;

[0079] The interpolation method is as follows (Kriging interpolation):

[0080] ;

[0081] in It is the prediction point 's elevation; is the sample point The known elevation of is the weight coefficient, which is determined by the covariance function and reflects the influence of each sample point on the prediction point.

[0082] Select ground control points (GCPs): Several ground control points with known coordinates are set up in the study area. These points should be evenly distributed and cover different terrain features. Accurate coordinates can be obtained through GPS measurement. Fixed landmarks on existing high-precision maps can also be used as GCPs.

[0083] Apply geometric correction algorithms: Apply GCPs to DEM and use least squares or other optimization algorithms to adjust the errors in DEM to ensure its geometric accuracy.

[0084] The goal is to keep the vertical error within ±0.5 m to ensure the reliability of the DEM in flood simulation and other applications.

[0085] Submergence Simulation Module 104:

[0086] In this module, the appropriate hydrodynamic model (one-dimensional, two-dimensional or three-dimensional) is selected according to the characteristics of the watershed; for the selected hydrodynamic model, the Bayesian optimization algorithm is used to automatically find the optimal parameter combination so that the simulation results are as close as possible to the actual observed values;

[0087] The following formula is used to describe the water flow:

[0088] ;

[0089] ;

[0090] in It's the depth of water. is the velocity vector, It's time. is the water surface height, is the height of the riverbed bottom, is an external force (such as friction), is the acceleration due to gravity.

[0091] In one embodiment of the present invention, the following steps are specifically included:

[0092] Collect basic information: Collect basic data such as hydrology, meteorology, topography and geomorphology of the study area to understand the natural conditions and socio-economic characteristics of the basin.

[0093] Determine key parameters: Analyze factors such as water flow path, confluence velocity, river network density, etc. in the basin to identify key parameters that affect water movement.

[0094] The hydrodynamic model includes the following three dynamic models:

[0095] One-dimensional hydrodynamic model: Applicable to situations where the river channel is narrow and the water flows mainly in a single direction, such as plain rivers or artificial channels. The model takes into account the changes in water depth and flow velocity over time, but ignores the influence of lateral water flow.

[0096] Two-dimensional hydrodynamic model: It is suitable for areas where water flow changes significantly in the horizontal plane, such as lakes, reservoirs, urban floodplains, etc. The model can describe the lateral distribution of water flow and is more suitable for flood simulation under complex terrain.

[0097] Three-dimensional hydrodynamic model: Suitable for environments where the water flow also changes significantly in the vertical direction, such as mountain canyons or near-shore waters. The model can capture the effects of turbulence and complex underwater structures on the water flow, providing the most detailed simulation results.

[0098] Initial condition setting: Set reasonable boundary conditions according to upstream water, downstream drainage and lateral inflow (such as tributaries, groundwater recharge) to determine the initial states of water level, flow rate, etc. at the beginning of the simulation, usually based on historical observation data or existing monitoring data.

[0099] Parameter estimation: Physical parameters include roughness coefficient (Manning's n), riverbed friction coefficient, etc. These parameters can be obtained through field measurements or references; empirical parameters include empirical parameters that are difficult to measure directly (such as permeability and evaporation rate), which can be preliminarily estimated based on historical data and gradually adjusted in subsequent optimization.

[0100] Construct an objective function: Define an objective function to measure the performance of the model, such as minimizing the mean squared error (MSE) between the simulation results and the actual observations.

[0101] Use the Bayesian optimization algorithm to search for the optimal parameter combination, and build a surrogate model (such as Gaussian process regression) to predict the performance of untested parameter combinations, so as to efficiently find the global optimal solution. Continuously update the surrogate model and evaluate new parameter combinations until the convergence criterion is met or the predetermined maximum number of iterations is reached.

[0102] The optimized model was applied to known historical flood events and the consistency of the simulation results was compared with the actual observations.

[0103] Scenario Setting and Uncertainty Analysis Module 105:

[0104] In this module, for key input variables (e.g., rainfall, evaporation rate), it is assumed that they follow a certain probability distribution (e.g., normal distribution), and then a Monte Carlo random sampling method is used to generate multiple possible scenarios;

[0105] Perform a complete flooding simulation for each scenario and record the output results;

[0106] To quantify the degree of uncertainty, a confidence interval can be used:

[0107] ;

[0108] in is the sample mean, is the standard deviation, is the sample size.

[0109] It should be noted that when considering the impact of geological disasters, different types of disaster scenarios (such as mild, moderate, and severe landslides) should be set up and simulation experiments should be carried out to evaluate their impact on water conservancy projects.

[0110] In one embodiment of the present invention, the following steps are specifically included:

[0111] Identification of key input variables: Identify key input variables that have a significant impact on the inundation simulation results, such as rainfall, evaporation rate, soil moisture content, etc.

[0112] Data collection: Collect historical observation data and forecast data to ensure coverage of diverse situations in different seasons and years.

[0113] Probability distribution assumption: According to the historical data characteristics of the variable, select an appropriate probability distribution function. For example, rainfall usually follows a normal distribution or a lognormal distribution; evaporation rate may be more consistent with a gamma distribution, and use maximum likelihood estimation (MLE) or other statistical methods to estimate the parameters of the selected distribution, such as the mean μ and standard deviation σ.

[0114] Monte Carlo random sampling: Apply the Monte Carlo method to randomly draw a large number of samples (e.g., 10,000) from a chosen probability distribution to represent different scenarios.

[0115] Scenario construction: Each sample corresponds to a specific scenario, including a set of specific input variable values. These scenarios will be used in subsequent flooding simulation experiments.

[0116] In one embodiment of the present invention, the submergence simulation experiment is as follows:

[0117] Perform simulations: Perform a complete inundation simulation for each scenario combination of input variables and record output results, including water depth, flow rate, inundation extent, etc.

[0118] Result storage: All simulation results are collated and stored to provide basic data for subsequent analysis.

[0119] Confidence interval calculation: Calculate confidence intervals for output results to quantify the degree of uncertainty;

[0120] Sensitivity analysis: Analyze the impact of different input variables on simulation results and identify which variables are most influential, which helps improve models and management strategies.

[0121] Geological disaster scene setting: set geological disaster scenes of different types and severity, such as mild landslide, moderate landslide, severe landslide, etc.;

[0122] In each of the above disaster scenarios, the boundary conditions and initial conditions of the water conservancy project facilities are adjusted, the inundation simulation is re-run, its impact on the water conservancy project is evaluated, the combined effects of geological hazards and hydrological factors are comprehensively considered, the potential risks are comprehensively assessed, and corresponding response measures are formulated.

[0123] Decision support and early warning reporting module 106:

[0124] This module includes the following two submodules:

[0125] User interface design submodule: intuitively displays geographic information and simulation results, and provides convenient data query and statistical analysis functions.

[0126] Early warning report generation submodule: Automatically generate early warning reports based on preset thresholds. The early warning reports include standardized report formats such as key indicators, charts, and recommended measures.

[0127] Machine Learning and Prediction Module 107:

[0128] Collect and organize historical data for training models, train models based on historical data to predict the probability of future geological disasters, visually display the prediction results through charts and other means, and propose reasonable preventive measures based on the prediction results;

[0129] Machine learning algorithm selection: such as decision tree, random forest, neural network, etc.

[0130] In one embodiment of the present invention, the following algorithm example is given:

[0131] The algorithm used is decision tree:

[0132] Gini Impurity:

[0133] ;

[0134] in Indicates category The probability of, that is, the proportion of samples belonging to category i in the current node, Represents the total number of categories, indicating all possible types of geological hazards, Represents Gini Impurity, which is used to measure a data set An indicator of the degree of confusion or uncertainty in a transaction.

[0135] Information Gain:

[0136] ;

[0137] in Represents entropy, a measure of the data set The degree of confusion, Represents attributes, such as rainfall or river flow in historical hydrological data, According to the attribute After segmentation, each subset corresponds to a specific attribute value. and are the number of samples in the original dataset and the subset after segmentation, Represents information gain, which is used to measure an attribute For the dataset The classification contribution index of Is an attribute All possible values ​​of ;

[0138] The training steps of the above machine learning algorithm are as follows:

[0139] S01: Data Preprocessing

[0140] Collect and clean data: Collect historical hydrological, topographic, socio-economic, human activities and geological disaster data from various modules (such as data access and processing module, multi-dimensional factor comprehensive assessment module, etc.). Ensure data integrity and consistency, and remove noise and outliers.

[0141] Feature engineering: Select appropriate features based on business needs and standardize or normalize the features. For example, convert rainfall, river flow, population density, etc. into a format suitable for model input.

[0142] S02: Building a training set

[0143] Divide the dataset: Divide the dataset into a training set and a test set (usually in a ratio of 70:30 or 80:20), ensuring that the data distribution of the two sets is similar. Cross-validation can also be used to improve the generalization ability of the model.

[0144] Label definition: Clarify the target variable of each sample (such as whether a geological disaster occurred) as the label for supervised learning.

[0145] S03: Initialize the decision tree structure

[0146] Set the root node: use the entire training set As the initial node, all samples belong to this node.

[0147] S04: Feature selection and node splitting

[0148] Calculate information gain: For each attribute , calculate its information gain ,Information gain measures the degree to which the purity of a data set is improved after partitioning using a certain attribute.

[0149] Select the best feature: select the attribute with the largest information gain as the basis for splitting, create branches and assign samples to the corresponding child nodes.

[0150] S05: Recursively construct subtrees

[0151] Repeat step S04, and repeat the feature selection and node splitting process for each newly generated child node until the stopping condition is met (such as reaching the set maximum depth, the number of samples in the node is less than the threshold, all samples belong to the same category, etc.).

[0152] Real-time monitoring update module 108:

[0153] After a geological disaster occurs, quickly organize on-site surveys, use portable laser scanners or drones to quickly collect topographic data of the disaster area, and update the DEM to reflect the latest changes in river morphology;

[0154] At the same time, record the damage to water conservancy project facilities to provide a basis for subsequent repair work;

[0155] In view of the changes in river channel morphology caused by geological disasters, the model boundary conditions and initial conditions are adjusted to ensure that the simulation results are consistent with the actual situation;

[0156] If necessary, re-run the multi-dimensional factor comprehensive assessment model and update the risk zoning map;

[0157] Compare the updated inundation simulation results with the actual flood events to evaluate the accuracy of the model. If a large deviation is found, return to the relevant steps in the "Inundation Simulation Module" section to readjust the model parameters or improve the model structure until the simulation results are satisfactory.

[0158] At least one embodiment disclosed in the present invention provides a storage medium storing non-temporary computer-readable instructions for executing steps corresponding to one or more modules in the aforementioned GIS-based water conservancy project management system.

[0159] The computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

[0160] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. The GIS-based water conservancy project management system is characterized by: Includes the following modules: Data access and processing module: responsible for collecting and preprocessing various types of data, including historical hydrological data, topographic data, socio-economic data, human activity data, and historical records of geological disasters, and setting scoring standards and weights for each factor; receiving data streams from IoT sensors in real time, and adjusting risk levels and initiating emergency response procedures in a timely manner for emerging changes and geological disaster warning information; Multi-dimensional factor comprehensive assessment module: calculates the risk index of the region and quantifies the risk level faced by different regions; High-precision DEM generation and calibration module: Create accurate terrain models, build high-precision digital elevation models through LiDAR point cloud data and drone images, and perform geometric correction through ground control points; Inundation simulation module: selects the hydrodynamic model according to the basin characteristics, and applies the Bayesian optimization algorithm to find the optimal parameter combination to make the simulation results close to the actual observation values; Scenario Setting and Uncertainty Analysis Module: Evaluate the uncertainty and impact of different scenarios, assume that key input variables follow probability distributions, generate multiple scenarios through Monte Carlo random sampling, perform a complete flooding simulation under each scenario, record the output results, and calculate confidence intervals to quantify the uncertainty; Decision support and early warning reporting module: provides an intuitive user interface for browsing maps and querying statistical data. When monitoring indicators exceed preset thresholds, early warning notifications will be automatically triggered; Machine learning and prediction module: Use machine learning algorithm to train models to predict the probability of geological disasters that may occur in the future; Real-time monitoring and updating module: After a geological disaster occurs, DEM is updated based on on-site investigation to reflect the latest changes in river morphology, and the damage to water conservancy facilities is recorded. The multi-dimensional factor comprehensive assessment model is re-run to update the risk zoning map.

2. The GIS-based water conservancy project management system according to claim 1 is characterized in that: In the multi-dimensional factor comprehensive assessment module, the calculation formula for the regional risk index is as follows: ； in is the risk index, represents the historical hydrological factor score, represents the topographic factor score, represents the socioeconomic factor score, represents the score of human activity factor, represents the geological hazard factor score, represents the factor weight of the historical hydrological factor, represents the factor weight of topographic factors, represents the factor weight of the socioeconomic factor, represents the factor weight of the human activity factor, Represents the factor weight of geological hazard factors.

3. The GIS-based water conservancy project management system according to claim 2 is characterized in that: The steps to create a terrain model are as follows: LiDAR point cloud data acquisition: Use airborne or ground-based LiDAR systems to scan large areas and obtain high-density three-dimensional point cloud data; Point cloud classification and filtering: Classify LiDAR point cloud data, distinguish ground points from non-ground points, apply filtering algorithms to remove noise points, and generate digital surface models; Construct DEM based on point cloud: import the classified ground points into GIS software and use interpolation method to generate high-precision DEM; Select ground control points: several ground control points with known coordinates are laid out in the study area. The ground control points with known coordinates are evenly distributed and cover different terrain features. Accurate coordinates are obtained through GPS measurement, and fixed markers on existing high-precision maps are used as GCPs. Apply geometric correction algorithm: Select ground control points are applied to the DEM and an optimization algorithm is used to adjust the errors in the DEM.

4. The GIS-based water conservancy project management system according to claim 3 is characterized in that: The interpolation method of the creation step of the terrain model is as follows: ； in It is the prediction point 's elevation; is the sample point The known elevation of is the weight coefficient, which is determined by the covariance function and is used to evaluate the influence of each sample point on the prediction point.

5. The GIS-based water conservancy project management system according to claim 4 is characterized in that: The watershed characteristics are described by the following formula: ； ； in It's the depth of water. is the velocity vector, It's time. is the water surface height, is the height of the riverbed bottom, is the external force term, is the acceleration due to gravity.

6. The GIS-based water conservancy project management system according to claim 5 is characterized in that: The hydrodynamic model includes the following three dynamic models: One-dimensional hydrodynamic model: It is applicable to the case where the river channel is narrow and the water flows mainly in a single direction. It considers the changes of water depth and flow velocity over time and ignores the influence of lateral water flow. Two-dimensional hydrodynamic model: Applicable to areas where water flow changes significantly in the horizontal plane, describing the lateral distribution of water flow, and suitable for flood simulation in complex terrain; Three-dimensional hydrodynamics model: Applicable to environments where the water flow also changes significantly in the vertical direction, capturing the effects of turbulence and the influence of complex underwater structures on the water flow, and providing detailed simulation results.

7. The GIS-based water conservancy project management system according to claim 6 is characterized in that: In the scenario setting and uncertainty analysis module, the degree of uncertainty is quantified and expressed using confidence intervals: ； in is the sample mean, is the standard deviation, is the sample size; When considering the impact of geological disasters, different types of disaster scenarios are set up and simulation experiments are carried out to evaluate their impact on water conservancy projects.

8. The GIS-based water conservancy project management system according to claim 7 is characterized in that: The decision support and early warning report module includes an early warning report generation submodule. The early warning report generation submodule automatically generates an early warning report based on a preset threshold. The early warning report includes key indicators, charts, and recommended measures.

9. The GIS-based water conservancy project management system according to claim 8, characterized in that: The following machine learning algorithm is used in the machine learning and prediction module, and its calculation formula is as follows: ； in Indicates category The probability of, that is, the proportion of samples belonging to category i in the current node, Indicates the total number of categories, indicating all types of geological hazards, Represents Gini impurity, which is used to measure a data set An indicator of the degree of confusion and uncertainty; ； in Represents entropy, a measure of the data set The degree of confusion, Represents attributes, including rainfall and river flow in historical hydrological data, According to the attribute After segmentation, each subset corresponds to an attribute value. and are the number of samples in the original dataset and the subset after segmentation, Represents information gain, which is used to measure an attribute For the dataset The classification contribution index of Is an attribute All possible values ​​of .

10. A storage medium storing non-transitory computer-readable instructions, characterized in that: Used to execute the steps corresponding to the modules in the GIS-based water conservancy project management system as described in any one of claims 1-9.

Images